A Thesis entitled Optimal Gear Shifting Strategy for a Seven-speed Automatic Transmission Used on a Hydraulic Hybrid Vehicle by Yaoying

Wang Submitted to the Graduate Faculty as partial fulfillment of the requirements for the Master of Science Degree in Mechanical Engineering

__________________________________________ Dr. Walter W. Olson, Committee Chair

__________________________________________ Dr. Yong Gan, Committee Member

__________________________________________ Dr. Mohammad Elahinia, Committee Member

__________________________________________ Dr. Patricia Komuniecki, Dean College of Graduate Studies

The University of Toledo March 2012

Copyright 2012, Yaoying Wang This document is copyrighted material. Under copyright law, no parts of this document may be reproduced without the expressed permission of the author.

An Abstract of Optimal Gear Shifting Strategy for a Seven-speed Automatic Transmission Used on a Hydraulic Hybrid Vehicle by Yaoying Wang Submitted to the Graduate Faculty as partial fulfillment of the requirements for The Master of Science in Mechanical Engineering The University of Toledo May 2011

Hydraulic technology can be used to capture and transfer high levels of energy extremely quickly and have a longer operating life compared with similarly sized electric systems. The hydraulic hybrid vehicle (HHV) includes two power sources that propel the vehicle: a fuel-efficient diesel combustion engine and hydraulic components. This technology replaces a conventional drive train with a hydraulic one, which eliminates the need for a mechanical transmission and driveline. To explore an optimal gear shifting strategy with best fuel economy for a seven-speed automatic transmission used on a hydraulic hybrid vehicle, a strategy is designed with a highest possible gear criterion as long as the torque requirement can be satisfied, except for braking process and torque demanding situations. The optimization strategy takes several other criteria into consideration, such as high motor displacement criterion, to improve efficiency and fuel economy. Then the optimization strategy is developed on the basis of these criteria from two main aspects of the existing SIMULINK truck model. One approach is based on the hydraulic
iii

motor working conditions, such as motor displacement, and the other is based on the drivers intention, which is interpreted as the driver pedal position. This controller is able to recognize the drivers intention to change the speed and incorporate it into gear shifting decision making. Then a SIMULINK controller model is developed based on the optimal gear shifting strategy and criteria and validated both in fuel economy and power performance by analyzing the simulation results in the Federal Urban Driving Cycle.

iv

Acknowledgments
First, I would like to express my sincerest gratitude to my advisor, Dr. Walter W. Olson, for his guidance on my graduate study and research. It is very lucky and a great honor to be his student and have the opportunity to work with him in such an interesting research area of hydraulic power. His academic enthusiasm and persistence do inspire my motivation for further study and future research work. Without his tremendous support and patience, I would not be able to complete this thesis. Special thanks to my committee members Dr. Yong Gan and Dr. Mohammad Elahinia, who give constructive advices and comments on my thesis. I would specifically like to give a thank you to the help and friendship of my partners, Andrew Sulzer and James Sweetman, who spare their precious time to review and correct my writing while they are busy working on their theses. Many thanks to the financial and technical support from Southwest Research Institute (SwRI), Bapiraju Surampudi, Joe B. Redfield and Glenn R. Wendel. Most thankful for my family and friends who always give their unconditional love and support to me wherever I am and whatever I do. Thank you all for accompanying me in my heart.
v

Table of Contents

Acknowledgments..................................................................................................... v Table of Contents ..................................................................................................... vi List of Tables ............................................................................................................. ix List of Figures ............................................................................................................ x Nomenclature ...........................................................................................................xii Chapter 1 Introduction .......................................................................................... 1
1.1 1.2 1.3 Background of Research ................................................................................. 1 Problem Statement ........................................................................................... 3 Work Outline ................................................................................................... 3

Chapter 2 Literature Review ............................................................................... 5

2.1 2.1.1. 2.1.2 Hydraulic Hybrid Vehicles ............................................................................ 5 Parallel ..................................................................................................... 7 Series ......................................................................................................... 8
vi

2.2

Optimal Transmission Strategies ..................................................................... 9

Chapter 3 Control Criteria and Shifting Strategy ........................................ 18

3.1 3.2 Control Criteria.............................................................................................. 19 Shifting Strategy ............................................................................................ 23 Shifting Based on the Driver Pedal ......................................................... 23

3.2.1

3.2.2 Shifting Based on Hydraulic Motor Conditions .......................................... 25 3.3 Control Algorithm ......................................................................................... 25

Chapter 4 Simulation Design ............................................................................ 27

4.1 4.2 Transmission Model ...................................................................................... 27 Controller System Model .............................................................................. 29 Driver Shifting Controller ....................................................................... 30 Speed Computation ................................................................................. 31 Braking Controller .................................................................................. 33 Priority Selection..................................................................................... 34 Overall Speed Violation of the Hydraulic Motor .................................... 35 Dwell Time Controller ............................................................................ 36 Gear Memory .......................................................................................... 36 Gear Ratio Matching ............................................................................... 37
vii

4.2.1 4.2.2 4.2.3 4.2.4 4.2.5 4.2.6 4.2.7 4.2.8

Chapter 5 Simulation Results............................................................................ 39

5.1 5.2 5.3 5.4 5.5 Fuel Consumption ......................................................................................... 40 Gear Shifting Schedule .................................................................................. 42 Tracking Performance ................................................................................... 44 Hydraulic Motor Speed ................................................................................. 47 Summary ....................................................................................................... 48

Chapter 6 Summary and Conclusion .............................................................. 50

6.1 6.2 6.3 Summary ....................................................................................................... 50 Conclusion ..................................................................................................... 51 Future Work ................................................................................................... 51

References ................................................................................................................. 53

viii

List of Tables

Table 5.1 Simulation results comparison between controller and original model ....... 49

ix

List of Figures

Figure 1-1 Configuration block diagram of the hydraulic hybrid vehicle ..................... 2 Figure 2-1 Parallel configuration [5] ............................................................................. 7 Figure 2-2 Series configuration [5] ................................................................................ 8 Figure 3-1 Plots of efficiency vs. motor displacement of each gear ............................ 21 Figure 3-2 Plots of motor displacement vs. acceleration of each gear ........................ 21 Figure 4-1 Original model with default shifting schedule ........................................... 28 Figure 4-2 Block diagrams with the optimal shifting controller in orange .................. 29 Figure 4-3 All block diagrams of the optimal controller top layer .............................. 30 Figure 4-4 Driver shifting controller block diagrams in orange .................................. 31 Figure 4-5 Speed computation block diagrams in grey ............................................... 33 Figure 4-6 Braking controller block diagrams in orange ............................................. 34 Figure 4-7 Hydraulic motor speed limit controller block diagrams in red .................. 35 Figure 4-8 Dwell time controller block diagrams in magenta ..................................... 36 Figure 4-9 Gear memory by using unit delay .............................................................. 37 Figure 4-10 Gear ratio matching controller ................................................................. 38 Figure 5-1 EPA Federal Urban Driving Schedule (FUDS) .......................................... 40 Figure 5-2 Fuel consumption simulation results comparison: fuel consumptionKg vs. time (seconds)......................................................................................................... 41 Figure 5-3 Controller model simulation results: gear ratio vs. time (seconds) ............ 42 Figure 5-4 Original model simulation results: gear ratio vs. time (seconds) ............... 43
x

Figure 5-5 Controller model simulation results: actual and desired vehicle speed (m/s) vs. time (second) .......................................................................................................... 44 Figure 5-6 Original model simulation results: actual and desired vehicle speed (m/s) vs. time (second) .......................................................................................................... 45 Figure 5-7 Simulation results comparison between original and controller model: vehicle speed error (m/s) vs. time (second) ................................................................. 45 Figure 5-8 Controller model simulation results: hydraulic motor speed (radian/second) vs. time (second) ................................................................................ 47 Figure 5-9 Original model simulation results: hydraulic motor speed (radian/second) vs. time (second) .......................................................................................................... 48

xi

Nomenclature

CVT EPA FUDS HHV IC

Continuously variable transmission Environmental Protection Agency Federal Urban Driving Schedule Hydraulic hybrid vehicle Internal combustion Final drive ratio Transmission gear ratio Minimum transmission gear ratio Radius of the vehicle tire

SwRI

Southwest Research Institute Output torque corresponding to Output torque corresponding to Vehicle speed Motor speed Torque convertor efficiency Minimum torque convertor efficiency

xii

Chapter 1

Introduction

1.1

Background of Research
In many applications especially those where high power densities are

required, hydraulic hybrid systems can offer a more efficient alternative to those driven by electric motors. Hydraulic technology can be used to capture and transfer high levels of energy extremely quickly compared with similarly sized electric systems, which generally require long periods over which batteries have to be charged. Hydraulic systems are also likely to have a longer operating life than battery-powered devices. Similar to an electric hybrid vehicle, which includes a gas or diesel engine and battery, the hydraulic hybrid vehicle (HHV) includes a diesel engine and a hydraulic power system that, in laboratory testing, has achieved significant fuel economy over traditional UPS vehicles. Hydraulic hybrid technology includes two power sources that propel the vehicle: a fuel-efficient diesel combustion engine and hydraulic components. This technology replaces a conventional drive train with a
1

hydraulic one, which eliminates the need for a mechanical transmission and driveline. These vehicles can store energy from the hydraulic system, even after the vehicle is turned off. This storage allows the vehicle to start with this energy, instead of relying on the engine to propel the vehicle [1]. To improve fuel economy of a military vehicle or family of vehicles, the hydraulic hybrid technology is applied and the ways it can benefit ancillary vehicle functions and enhance the mission usefulness of the vehicle are investigated as the purpose of this project. Figure 1-1 below is the configuration of the hydraulic hybrid vehicle developed in this project.

Figure 1-1 Configuration block diagram of the hydraulic hybrid vehicle

1.2

Problem Statement
The objective of this thesis is to explore an optimal gear shifting strategy for

the best fuel economy with a seven-speed automatic transmission used on a hydraulic hybrid vehicle. A simulation model of this automatic transmission was developed in SIMULINK by engineers at Southwest Research Institution that takes the seven-speed gear ratios as a shifting schedule input to the transmission model and exports the vehicle speed, vehicle distance, driver pedal position, current gear ratio, fuel consumption, simulation time, motor pressure, motor displacement, motor speed and efficiency as outputs to a MATLAB workspace. Though shown as a seven- speed model, the strategy discussed is for the range from 2nd to 7th gear as a shift directly from 1st to 2nd gear is not allowed while the vehicle is moving. In addition, since the transmission is no longer driven by the IC engine, the instantaneous fuel consumption cannot be computed from outputs of the transmission model. Part of the work of this thesis is to research and identify better operational measures by which fuel economy can be optimized.

1.3

Work Outline
Chapter 2 is the literature review which represents the development of

hydraulic hybrid vehicles along with illustrations of two main configurations: parallel and series. It also introduces some different approaches to transmission optimization.
3

Chapter 3 represents the shifting criteria and analyzes the shifting strategy for the transmission. Chapter 4 shows the simulation block diagrams built in

MATLAB/SIMULINK according to the shifting strategy. Chapter 5 discusses the simulation results and compares the results to the original shifting model simulation. Chapter 6 is the conclusion of the optimal shifting strategy simulation of the transmission and represents the future work.

Chapter 2

Literature Review

2.1

Hydraulic Hybrid Vehicles

A hybrid vehicle combines two or more sources of power. Currently, hybrid

technology is widely recognized as the most effective measure to solve the energy problem. Heavy vehicles such as city buses have the characteristics of high stop-and-go duty cycles and high power ow braking energy, which needs to nd an efficient way to store and reuse the braking energy [2]. The additional power source can be electrical, chemical, hydraulic, ywheel operated or any other form of power storage and production [3]. Within the many hybridization options, battery and fuel cell have the characteristics of high energy density and are well suited for light vehicles. However, the high internal resistances and handling the wasted battery are major obstacles for commercialization. Both fuel cell and battery hybrid vehicles can only marginally recover the braking energy. Moreover, high frequency charging and discharging will lead to overheating and battery destruction. An ultra- capacitor (UC)
5

has higher power density than a battery because their operation does not employ a chemical reaction, but the high cost and relatively lower reliability constrain their applications [2,5,4]. As an important method of hybrid technology, hydraulic hybrid vehicles (HHV) now attract the attention of worldwide research institutions and commercial industrial companies. The first-ever hydraulic-hybrid diesel urban-delivery vehicle reportedly improves fuel economy by 60 to 70% and reduces carbon-dioxide emissions more than 40% in initial laboratory testing. EPA estimates that the technology has the potential to save more than 1,000 gallons/yr for each urban-delivery vehicle [11]. In a hydraulic hybrid vehicle, the hydraulic power assists the conventional internal combustion engine by providing additional torque to the driveshaft [6]. The hydraulic hybrid system with accumulators and hydraulic pump/motors have the potential for improving fuel economy by operating the engine in the optimum efficiency range and making use of the regenerative braking during deceleration. When braking, a hydraulic hybrid can recover and reuse braking energy that is normally wasted. As the vehicle stops, energy from the wheels pumps fluid from a low-pressure reservoir into the high-pressure accumulator. When the vehicle subsequently accelerates, the stored energy propels the vehicle. According to EPA officials, this process recovers and reuses more than 70% of the energy normally wasted during braking, and it also reduces brake wear by about 75%, substantially increasing the savings [11].
6

There are two main configurations for hydraulic hybrid vehicles, parallel and series. Both of them have the main components which consist of a high-pressure accumulator storing energy by using hydraulic fluid to compress nitrogen gas stored inside each accumulator, a low-pressure reservoir storing hydraulic fluid after it has been used by the pump/motor, a pump/motor converting high-pressure hydraulic fluid into rotating power for the wheels and transmitting braking energy back to the high-pressure accumulator, and an engine pump transmitting pressurized hydraulic fluid to the pump/motor, the high-pressure accumulator, or both [1].

2.1.1.

Parallel
Shan [7] reports parallel hydraulic hybrid vehicles are easier to implement,

but efficiency gains are limited by the solid link between the wheels and engine.

Figure 2-1 Parallel configuration [5]

The rear pump/motor acts as the drivetrain, such that in the motor mode, it drives the axle using high pressure uid and, as the vehicle brakes, the pump/motor directs the uid to the high pressure accumulator by switching into the pump mode. The high pressure accumulator is usually designed to satisfy internal pressure loads up to 35MPa (5000 psi), whereas the low pressure accumulator holds the internal pressure of up to 1.4MPa (200 psi) [8, 9].

2.1.2

Series
Series hydraulic hybrid vehicles allow engine speed to be decoupled from

vehicle speed. This permits a control strategy where the engine and other hydraulic components operate only near maximum efficiency.

Figure 2-2 Series configuration [5]

In this configuration, the conventional driveline is completely removed; it is

8

connected through hydraulic pipes and the drive pump/motors are used to transfer power and propel the vehicle. The drive pump/motor converts high pressure hydraulic fluid into rotating power in the motor mode and transmits regenerative braking energy back to the high pressure accumulator in the pump mode. Since the direct link between the engine and the driveline components is removed, the engine is separated from the road and higher efficiencies are anticipated. [7].

2.2

Optimal Transmission Strategies

As it is important part of the vehicle powertrain to convert torque and rotation, the optimization of the transmission can increase the performance of the vehicle efficiency and fuel economy significantly. This depends on different working conditions and rules, and various optimal control methods are applied to different kinds of transmissions. A. Haj-Fraj [10] introduces an optimal control approach for gear shift operations in automatic transmissions as a multistage decision process by making use of dynamic programming method. Starting from a veried model of a typical powertrain, Haj-Fraj considers three optimization parameters as a performance measurement for evaluating the gear shift process: the control data for the clutch pressure in the gear box, the engine load-reduction and the evaluation of the gear shift duration. As the passengers comfort is a subjective issue and varies from driver
9

to driver, it can generally be stated that the smoother the acceleration change is and the smaller the peaks of the jerk are, the more comfortable the gear shifting is. To engage the clutch of the target gear with a very smooth and slow rising pressure in order to get a very slow transition in the dynamical behavior of the powertrain in order to get a smooth acceleration but a very long gear shift process. A control law is derived analytically in an explicit form by minimizing the performance measure over each process stage and the optimization finally is solved with a sequential quadratic programming algorithm. The results have shown that the passengers comfort and the duration of the gear shift operation represent contrary issues. This means that an improvement in one criterion leads to a deterioration in the other one. Therefore the optimization problem represents be leads to find a reasonable solution that can define a compromise between the two issues. The shift schedule proposed by Gong Jie [13] for the ground vehicle automatic transmission by studying the function of the torque converter and transmission shift schedule can keep the torque converter working in the high efficiency range under all the working conditions except in the low efficiency range on the left when the transmission worked at the lowest shift, and in the low efficiency range on the right when the transmission worked at the highest shift. In order to evaluate the economic performance of the torque converter, the efficiency of the torque converter is required no less than an ideal value 75% in engineering machinery, in use ( =

= 80% in the normal automobile). The shift

10

schedule aims to control and set the common working point of engine and torque converter so that the torque converter efficiency so that is no less than the ideal value

. In other words, the points satisfying the condition are used as the shift points.

The gear ratio

and corresponding torque

are analyzed as the key

factors to determine the shift points. After the experimental test-bed results, the torque converter efficiency was controlled to be over 75% in the high efficiency range by gear shift according to the actual load acting on the drivetrain. This method focuses most on the relationship between the fixed gear ratio and the continually changing actual gear ratio along with its torque value, which is much easier to control compared to the dynamic controller stated in Haj-Frajs. Toshinichi Minowa [15] investigates a powertrain control model for an automatic transmission providing efficient control for both the engine and the transmission which leads to better fuel economy and acceleration feeling by optimizing the shift timing and throttle valve opening. The gear shift timing consists of the vehicle speed and the throttle valve opening simulating the engine load. To achieve the optimal fuel consumption, gear shift timing is calculated by comparing the efficiency of the torque transmitted to the wheels at each gear shift ratio, using the fuel flow rate characteristic of the engine and torque convertor. The controller is developed based on the concept that the optimal gear shift timing is selected with the driven horse-power required for running, and the engine torque is controlled by the throttle valve opening so as to maintain the driven shaft torque demanded by the
11

driver which can be calculated from accelerator pedal angle and vehicle speed. To avoid the busy shift and fluctuations of the driven shaft torque, hysteresis are applied to the fuel flow rate calculated every gear shift to prevent gear command signal changing several times. The test pattern model used to verify the optimization shift timing results is very simple and straight forward, so the simulation results are relatively better with known driving circle. M. Pachter [17] presents and analyzes differential game model of an automatic transmission for a road vehicle control which leads to a synthesis of the optimal feedback strategy for the selection of the transmission ratio of the automatic gearbox. The vehicle is considered as object, which is governed by two controlling agents: the setting by the driver of the throttle and the gear-change controller. He remarks that most road vehicles are equipped with manual or automatic gearboxes which have a finite number of discrete transmission ratios. But for CVT where the transmission ratio is a continuous variable, his analysis hinges on the fact that the dynamic function assumes a discrete set of values. Therefore, in order to analyze the case of a continuous secondary control variable, one should discretize the transmission ratio or alternatively, use an adaptation of our theory to a discrete-time model for the dynamical system. Quan Zheng [33] introduces a new coordinated engine-transmission control approach for the neutral idle input clutch application phase aiming at improving fuel economy during urban driving and reducing engine vibrations transmitted to the
12

passenger compartment. It involves ramping up engine speed smoothly and simultaneously engaging the forward clutch smoothly. Eective coordination can be realized by controlling the engine and turbine speeds closed loop so that they follow specied trajectories. Therefore, the controller design objective is to realize optimal trajectory tracking by synthesizing the control inputs so that the engine and turbine speeds track the specied desired trajectories closely. Due to the Multi-Input Multi-Output nature of the control problem, an optimal Linear Quadratic Regulator with Explicit Model Following is used to allow the system dynamic response to track two desired trajectories for engine and turbine speeds. The results show that the proposed control strategy can achieve satisfactory performance. There are several issues that remain to be explored in this paper. The rst issue is state estimation. In the current formulation, the states are assumed to be available for feedback implementation. However, state estimation is needed to obtain clutch pressure and the derivative of clutch pressure, and dynamic estimation of engine indicated torque is also necessary. The second issue is better inclusion of the vehicle dynamics in the design. One limitation of the current implementation is that the state related to vehicle acceleration is ignored in the process of model simplication. This limits accurate studies of vehicle vibrations under dierent control strategies. In Alarico Macors [34] work, the design of a hydro-mechanical transmission is defined as an optimization problem in which the objective function is the average efficiency of transmission, that is to minimize the total loss of the
13

transmission, while the design variables are the displacements of the two hydraulic machines and gear ratios of ordinary and planetary gears. Since the object function doesn't have an analytical formulation, the optimization problem is solved by a direct-search algorithm based on the swarm method, which showed a good speed convergence and the ability to overcome local minima. And the use of evolutionary algorithms is also able to reduce the importance of the initial research point and the trapping in local minima far from absolute minimum. However the advantage of the continuous speed variation of the power-split drives is counterbalanced by a reduced efficiency, caused by the double energy conversion taking place in the hydrostatic transmission. Therefore, the design of the power split drive still need further study. The proposed procedure does not depend on experience and previous knowledge because no assumption had to be made on the component's sizing; the optimality of the output is based on the implemented search algorithm while the quality of the classical designs depends strongly on the designer's experience. B.Mashadi [18] designs a gear-shifting strategy of an automated manual transmission by taking into consideration the eects of these parameters, with the application of a fuzzy control method. The controller structure is formed in two layers. In the rst layer, two fuzzy inference modules are used to determine the necessary outputs. In the second layer a fuzzy inference module makes the decision of shifting by upshift, downshift, or maintain commands. The behavior of the fuzzy controller is examined by making use of ADVISOR software. It is shown that at
14

dierent driving conditions the controllers make correct decisions for gear shifting accounting for the dynamic requirements of the vehicle. Both the engine state and the drivers intention also eliminate unnecessary shifts that are present when the intention is overlooked. Both strategies for the vehicle to reach the maximum speed starting from rest allow the gear shift to be made consecutively. Considerable dierences are observed between the two strategies in the deceleration phase. The engine-state strategy is less sensitive to downshift, taking even unnecessary upshift decisions. The state intention strategy, however, interprets the drivers intention correctly for decreasing speed and utilizes engine brake torque to reduce the vehicle speed in a shorter time. Magnus Pettersson and Lars Nielsen [21] uses engine control during the gear shift for a manual transmission without using the clutch during the shift event. To minimize the total time needed for a gear shift will excite the driveline resonances which may lead to problems with disengaging the old gear and synchronizing speeds for engaging the new gear. Internal driveline torque control is a novel idea introduced in their work for handling resonances and increasing shift quality. By estimating the transmitted torque and controlling it to zero by engine control, the gear can systematically be disengaged with minimized driver disturbances and faster speed synchronization. Two main advantages of the control system are: fast shifts to neutral gear, despite disturbances and driveline oscillations at the start of the gear shift; the control scheme is simple and robust against variations among different gears and
15

damping of driveline resonances can be obtained with an observer in combination with a PID feedback structure. Different from the manual or automatic transmission with fixed gear ratios, continuously variable transmission (CVT) with its continuous ratio offers the potential to substantially improve the part-load fuel efficiency of spark-ignited engines. The control of CVT has traditionally been designed using static arguments, like by identifying the best efficiency points in the engine map for each constant power requirement and by following that curve using some heuristics as much as possible also in transients. R. Pffner [36] presents the solution of the fuel-optimal control problem for transient conditions using the numerical optimization package DIRCOL. Based on this optimal solution a simplied but causal control strategy is proposed which offers almost the same benets. The resulting engine operation trajectory is pictured in a engine torque vs. engine speed figure and the optimal solution therefore consists of bringing the system to the corresponding best efficiency curve as fast as possible, and to keep the system on this curve as long as possible. During the periods when the system moves towards or away from this curve the gear ratio has to be changed with maximal possible speed while the optimal engine power trajectory is more or less constant over time. Michiel Pesgens [35] develops a transmission ratio controller for a hydraulically actuated metal push-belt continuously variable transmission (CVT), which consists of an anti-windup PID feedback part with linearizing weighting and a
16

set point feedforward which is generated by the hierarchical (coordinated) controller. Physical constraints on the system, especially with respect to the hydraulic pressures, are accounted for using a feedforward part to eliminate their undesired effects on the ratio. The total ratio controller guarantees that at least one of the pressure setpoints is always minimal with respect to its constraints, while the other is raised above the minimum level to enable shifting. This approach has potential for improving the efficiency of the CVT, compared to non-model based ratio controllers with experimental results showing that adequate tracking is obtained together with good robustness against actuator saturation. The largest deviations from the ratio setpoint are caused by actuator pressure saturation. It is further revealed that all feedforward and compensator terms in the controller have a benecial effect on minimizing the tracking error. Most optimal approaches stated above depend on the working condition of engine and torque converter. For this case, in this thesis, as the transmission is no longer driven by the IC engine, the instantaneous fuel consumption cannot be computed from outputs of the transmission model. Therefore the optimal gear shifting decision cannot be made based on the best fuel economy of a specific gear under a certain driving condition. A new method based on the current transmission model is designed and introduced in the following chapters.

17

Chapter 3

Control Criteria and Shifting Strategy

The function of a vehicle transmission is to adapt the traction available from the drive unit to suit the vehicle, the surface, the driver and the environment. It has a decisive effect on the reliability, fuel consumption, ease of use, road safety and transportation performance [37]. The optimization of the transmission can significantly increase the performance of the vehicle efficiency and fuel economy. As introduced in chapter 1, a simulation model of a truck was developed in SIMULINK by engineers at Southwest Research Institution based on the configuration of the hydraulic hybrid vehicle shown in Figure 1-1. It is a seven-speed transmission which can provide more efficient fuel performance than a normal four-speed one. The strategy discussed is for the range between 2nd and 7th gear as a shift directly from 1st to 2nd gear is not allowed while the vehicle is moving. In the transmission model, a shifting command that consists of one of the seven-speed gear ratios is taken as an input, and it outputs the vehicle speed, vehicle

18

distance, driver pedal position, current gear ratio, fuel consumption, simulation time, motor pressure, motor displacement, motor speed and efficiency as outputs to a MATLAB workspace. In this chapter, an optimal control strategy for the gear shifting schedule of the automatic transmission is developed depending on the proper application of the decision-making algorithm for best fuel economy. Since the transmission is no longer driven by the IC engine, the instantaneous fuel consumption cannot be computed from outputs of the transmission model. The control strategy developed in this chapter is to research and identify better operational measures by which fuel economy can be optimized.

3.1

Control Criteria
The design of a controller requires certain performance criteria to be

established as objectives. The overall objective of this controller was to provide a gear shifting schedule that would minimize fuel consumption for a hydraulic hybrid truck. Unlike conventional vehicles, this trucks drive wheels are not directly attached to the internal combustion engine. Therefore the torque and wheel speeds are controlled independently from motor speed. While the internal combustion engine must provide power to a storage system, the rate at which the power is consumed does not need to match the rate of power production. Therefore, the internal combustion engine can continuously operate at the point of its best
19

efficiency. When sufficient power has been stored, the internal combustion engine can be turned off until the stored power has dropped to some replenishment level. At that time, it would again recharge the storage system. Thus the objective of minimizing fuel consumption will not depend on matching engine torque and speed to immediate driving conditions. As it will be explained later, meeting this objective will depend on promoting higher displacements in a variable displacement hydraulic motor. First of all, the analysis of the truck model shows that higher efficiency of the motor is obtained in a higher gear of the transmission. Therefore, the transmission should be in the highest possible gear as long as the torque demand can be met while driving. The analysis of data produced by the SwRI s-function black box model of the hydraulic motor in Figure 3-1 shows that the highest efficiency for the motor is near full displacement. Therefore, the controller should strive to keep the hydraulic motor working near full displacement. Figure 3-2 shows that high acceleration requires higher torque, thus requiring a lower gear. However, taking this plot into account, the transmission should upshift to get higher motor displacement as long as the torque requirement can be satisfied. So, the highest possible gear criterion is made for this controller.

20

Figure 3-1 Plots of efficiency vs. motor displacement of each gear

Figure 3-2 Plots of motor displacement vs. acceleration of each gear

Another criterion that the controller must meet is to provide the wheel speed and torque necessary to meet the EPA Federal Urban Driving Cycle. This
21

cycle, meant for comparing automobile fuel efficiency consists of 1369 seconds of velocities simulating large city stop-and-go traffic. For a large truck, this is a very demanding schedule to meet as the accelerations required are at the torque limits of the propulsion system. However, the driving cycle cannot be used directly. Initially, the demands for torque and the conditions for braking must be interpreted from the actions of the acceleration pedal which will be discussed in the strategy section. There are also certain safety standards imposed. Based on the limitations of the truck components, shifting from a lower gear to a higher gear, or shifting from a higher gear to a lower gear requires that the transmission be shifted into each gear in between. For example, a shift from 4th gear to 2nd gear, as might be required driving a braking event, requires that shift sequence is from 4th gear to 3rd gear and then to 2nd gear. This is required in order to sequence the transmission internal clutches so the transmission is not destroyed or so that the gear shift can be successfully made. The vehicle must start from a stop in 2nd gear. 1st gear is not used for normal driving conditions. Another transmission requirement is that each gear during an upshift requires a dwell time in that gear for conditions to stabilize as the transmission is under driving force. This requirement is relaxed during downshift so that maximum use can be made of regenerating braking. Besides that, the gear shifting time is also taken into consideration as the gear shifting needs certain time to complete. The hydraulic motor has a rotational speed limitation of 3000 RPM. At speeds that exceed this, the motors rotating parts produce enough inertia such that
22

the motor could explode or otherwise be damaged.

3.2

Shifting Strategy
The optimal shifting strategy is mainly developed from two different

aspects of the truck model. One approach is based on the drivers intention which is given as pedal position output of the transmission model. Based on the concept that the driving cycle is unknown to the controller, drivers intention is the only input can be taken into account as a prediction of the driving cycle. The intention of the driver pressing the pedal indicates the torque or the acceleration is required from driver and releasing the pedal indicates braking. In addition to interpreting drivers behavior, another aspect that is taken into consideration is the hydraulic motor displacement, which can guarantee the hydraulic motor working within the range where higher efficiency can be achieved.

3.2.1

Shifting Based on the Driver Pedal

The intention of the driver is interpreted from the position of the driver

pedal. In the most basic form, the depression of the pedal is interpreted as the driver is demanding torque to accelerate while releasing of the pedal is interpreted as the driver intends to decelerate. There are two parameters that can be considered here under the drivers

23

control, one is the pedal position, and the other is the pedal rate. The pedal position is represented with a numerical value from 0 to 1 with regard to the full pedal position as 1. The pedal rate is defined as the rate of increase of pedal position during one sample time T (T=5 milliseconds). When the driver uses a high rate of increase in pedal position, it is interpreted as an urgent torque demand. By observing the relationship between pedal rate and acceleration requirement, the average value of the pedal rate was found to be 0.01 per sample time from averaging all of the rates of increase over the FUDS driving cycle. So when driver presses the pedal down with rate double of the average value per sampling time, a downshift command is applied to increase torque. In addition, an upshift decision is made when pedal rate is negative which is interpreted as the driver wants to decelerate by retarding the pedal based on the highest possible gear criterion. The release of the pedal indicates that the driver wants to decelerate the vehicle. The vehicle can decelerate in two ways: coast down and active braking. However, braking is neither an input nor an output of the truck model. Therefore, a surrogate had to be developed. If the pedal position is less than 0.05, it is considered as a potential sign for braking, so a downshift decision is made for braking mode. Otherwise, in the case of pedal position greater than 5% along with pedal rate within the range from 0 to 0.005 will keep on the current gear.

24

3.2.2 Shifting Based on Hydraulic Motor Conditions

Another aspect that can be considered for minimizing the fuel consumption is to maximize the hydraulic motor efficiency. According to the relationship between efficiency and motor displacement, and efficiency and motor speed, the higher motor displacement and speed, the higher efficiency. So the hydraulic motor needs to work in the high range of displacement and speed to get high efficiency. As shown in Figure 3-1, the highest efficiency, 95%, of the motor occurs at a high motor displacement range from 0.7 to 0.95. A downshift will be applied when motor displacement value is more than 0.95 to prevent over driving the displacement.

3.3

Control Algorithm
Based on the analysis of the control criteria and strategy, the main algorithm

of the controller can be determined by taking all the input and output parameters of the transmission model into account. First of all, the controller will always upshift the transmission to the highest possible gear provided the driving power requirements can be met based on the highest possible gear criterion. There are some exceptions for downshift, pedal position is less than 5%, which indicates braking mode, and pedal rate more than 0.005 per sampling time which indicates more torque requirement from the driver. Otherwise, when the pedal rate is less than 0.005 per sampling time the controller will stay in the current gear. Secondly, the maximum motor displacement limitation
25

is set to 95% and downshift will be applied if it exceeds that value. When the current motor speed reaches 3000 RPM (314 radian/second), an upshift is always applied to reduce the motor speed. In addition, any command derived from the drivers intention will also be checked with the current vehicle speed to protect the motor speed from exceeding the maximum value. All the shifting commands must be checked with the 2nd to 7th gear range controller and then combined with the 2-second dwell time other than when braking. Finally, as the gear shifting needs certain time to accomplish, the shifting in process (SIP) signal is presented with either 1 which shows the shifting is still in process or 0 which means the shifting is already completed. The shifting command only can be applied when the shifting in process (SIP) signal is 0. In other words, every proposed shifting command must be checked with SIP signal before it can be delivered to the final output.

26

Chapter 4

Simulation Design

SIMULINK provides a platform to simulate the performance of the optimal shifting controller by running the transmission model in a certain driving cycle.

4.1

Transmission Model
The transmission model shown in Figure 4-1 is a discrete-time based

SIMULINK model which runs a 1369-second simulation driving cycle with a 0.005-second sampling time. The model simulates the shifting of a 6-speed transmission under conditions imposed by the EPA Federal Urban Driving Cycle for a hydraulic hybrid truck. The truck dynamics are contained within a black box s-function provided by Southwest Research Institute. This s-function uses the gear ratio as the input, and then provides the vehicle speed, the driver pedal position, the hydraulic motor speed, the motor displacement and the hydraulic pressure among
27

other factors.

Out1

DesVehSpd ActVehSpd ActVehSpd (m/s) DesVehSpd (m/s)

Out2

Out3

VehDistance VehDistance (m)

Out4

SIP Shift In Progress (SIP)

Out5

DriverPedal DriverPedal (0-1)

Out6

CurrentGearRatio CurrentGearRatio

-1 Automatic Shifting or Des Gear Ratio

In1 Out7

FuelConsumed FuelConsumed (Kg)

Out8

tt Simulation Time (tt) seconds

Pressure (N/m2)

Pressure Pressure (N/m2) MotorDisp MotorDisp (0-1) MotorSpeed MotorSpeed (rad/sec) PtimesEff Press*Eff

Motor Disp (0-1)

Motor Speed (rad/s)

P*ef f

VehicleWithSeriesHydraulicPowerTrainAndFTPDriveCycle+Driver1

Figure 4-1 Original model with default shifting schedule

A simulation model of the controller was developed in SIMULINK based on the original truck model provided by Southwest Research Institute, presented in Figure 4-2. The highest level of this controller-truck model consists of three components: the controller system, the truck s-function and the outputs. The s-function black box is the same as the original truck model which uses the gear ratio from the controller system block and outputs 12 parameters including vehicle speed, vehicle distance, driver pedal position, shifting in process signal (SIP), simulation time, current gear ratio, fuel consumption, simulation time, motor
28

pressure, motor displacement, motor speed and efficiency as matrices into MATLAB workspace.

Figure 4-2 Block diagrams with the optimal shifting controller in orange

4.2

Controller System Model

The controller is a five-input and one-output system which uses vehicle

speed, driver pedal position, shifting in process signal (SIP), motor displacement and motor speed. The output is the gear ratio for the transmission. The controller contains the following components which interpret the control algorithm discussed in chapter 3.

29

Algorithm of the Controller: 1.Highest Gear Optimization. 2.Max Motor Speed (314rad/s) first priority, upshift when exceed. 3.Max Displacement limit 0.95, downshift when exceed. 4.Always upshift unless: 1). Pedal position <0.05 indicates braking, then downshift. 2). Pedal position increasing means driver is demanding torque, then stay on the current gear. 3). Pedal position increasing >=0.005 /sample time, then downshift. 5.Dwell time 5 sec, may cause problems when braking time to a stop is so short that the gear cannot shift to the 2nd gear in time, so no dwell time is applied when vehicle speed < =0.3m/s.

Note: gear number 1 corresponds to 2nd Gear gear number 6 corresponsed to 7th Gear

Current Gear Number (1 to 6)

1 z Gear Memory

1 SIP Value is 1 if shift is in progress, zero other used to inhibit shifting

>0 SIP Check No shifting if SIP

Gear Sent

4.18 2nd Gear Ratio1 2.24 Motor Speed Limit Shift 4 Motor Speed If 1 A Motor Speed Violoation is Occurring! Mandates upshift if motor speed exceeds set value Highest Priority: Not subject to Dwell Time
Motor Speed Gear Shift Command: 1 = upshift 0 = no action Current Gear Number (1 to 6) Gear Range Limited Gear

1 2 3 4 5 6 1 z used to break alg loop

3rd Gear Ratio1 1.69 4thGear Ratio1 1.20 5thGear Ratio1 0.90 Limit Gears to Range 2nd - 7th >0 Give priority to an upshift 6thGear Ratio1 0.78 7thGear Ratio1

1 Gear Ratio

Add

*
Convert to Gear Ration

3 Actual Velocity
Gear Delay Current Veh Velocity Prop Gear after limit Prop Gear after Dwell Time

DriverPedal 2 Driver Pedal 5 Displacement

Pedal Position Driver Shift Command Displacement

Computed Gear Motor Speed

-1,0,1 values Create gear number to go to

Prop Gear

Range Limited Gear

Proposed Gear

Gear Command Given to Transmission

Dwell Time Controller Limit Gears to Range 1 to 6 (gears 2nd to 7th) Computed Motor Speed >= 314 Speed Check If Prop gear violates motor speed use current gear (no shift) Shift Cmd after spd ck

Driver Command Shifting

Driver Command

Proposed Gear

Veh Speed Prop gear command without dwell time Prop gear command after dwell time Pedal Position Prop gear

Prop Gear Number (1 to 6)

Give priority to Braking VehSpeed<0.3m/s without dwell time

Figure 4-3 All block diagrams of the optimal controller top layer

4.2.1

Driver Shifting Controller

The shifting decisions made by the controller will be presented in signals of

1, -1 and 0, which represent upshift, downshift, and on action respectively. The torque requirements are interpreted from the driver pedal position and the pedal position rate. The pedal position is represented with a numerical value from 0 to 1 and the pedal rate is defined as the rate of increase of pedal position during one sample time T (T=0.005 second). The controller always implements an upshift command unless the torque requirement cannot be satisfied by the current gear. As mentioned before, a high rate of increase in pedal position with a value of 0.005 per sampling time is interpreted as
30

an urgent torque demand from the driver, a downshift command is applied to increase torque in this case. In addition, pedal position greater than 0.05 with the pedal rate staying in the range from 0 to 0.005 indicating the driver is demanding torque gently and will keep the shifting command in the current gear. This is shown
Always upshift unless: 1). Pedal position <0.05 indicates braking, then downshift. 2). Pedal position increasing means driver is demanding torque, then stay on the current gear. 3). Pedal position increasing >=0.005 /sample time, then downshift. 4).Max Displacement limit 0.95, downshift when exceed.

in Figure 4-4.

> 0.005 1 Pedal Position 1 z Unit Delay Difference of Driver Pedal position If pedal depression velocity is greated than .005/Ts, the driver is urgently demaning torque 1st priority

-1 or 0 values

>0 1 Up Shift

0 or 1 values

If Driver is depressing pedal: Torque Demanded : Do not up shift otherwise allow upshift 2nd priority

>= 0

-1, 0, 1 values

Give priority of choice to pedal vel cmd

-1,0,1 values

2 Displacement

> 0.95

-1 or 0 values

>= 0

If Hys Motor Disp is greater than 0.95: downshift 0 No Shift -1 Down shift Assume lowest gear slowing Driver Pedal Retard: If pedel depress < 5% assume braking modes downshift >= 0.05

Give priority of choice to disp cmd

-1,0,1 values

-1 or 0 values

>= 0

1 Driver Shift Command

Final Gear Command Prioity given to braking (down shift)

Figure 4-4 Driver shifting controller block diagrams in orange

4.2.2

Speed Computation
The maximum hydraulic motor rotational speed is 3000 RPM which can be
31

converted into radians per second as:

The proposed gear command made by the driver shifting controller also had to be considered in the proposed corresponding hydraulic motor rotation speed based on the proposed gear ratio, , and checked with the maximum hydraulic

motor speed limitation based on the criteria:

The ratio for the final drive is

and the radius of the vehicle tire is

This is implemented as shown in Figure 4-5.

32

1 Current Veh Velocity 13.4483 Velocity factor (Nf/re) 1 Computed Gear Motor Speed Compute Motor Speed from gear ratio

2 Proposed Gear 4.18 2nd Gear Ratio 2.24 3rd Gear Ratio 1.69 4thGear Ratio 1.20 5thGear Ratio 0.90 6thGear Ratio 0.78 7thGear Ratio

1 2 3 4 5 6

*
Proposed Gear Ratio

Figure 4-5 Speed computation block diagrams in grey

4.2.3

Braking Controller
In order to get to the lowest gear when the vehicle stops, a downshift

command will be made during the braking process. As there is no braking pedal signal in the transmission model, the driver pedal position of less than 5% is predicted braking. A downshift decision is made in the pedal controller to provide sufficient time to get to the 2nd gear before the vehicle stops. A figure for the braking controller is shown below.

33

wn shift lowest gear 2 owing1 Displacement 0 No Shift1

> 0.95

-1 or 0 v alues

If Hys Motor Disp is greater than 0.95: downshift

Give priority of choice to disp cmd

0 No Shift 1 Pedal Position -1 Down shift Assume lowest gear slowing Driver Pedal Retard: If pedel depress < 5% assume braking modes downshift Final Gear Command Prioity given to braking (down shift) >= 0.05 -1 or 0 values >= 0
-1,0,1 v alues

Driver Shift Command

Figure 4-6 Braking controller block diagrams in orange

4.2.4

Priority Selection
In the driver pedal shifting controller, the first priority is given to the

urgent torque demand from the driver with a pedal rate greater than 0.005/T. Otherwise the algorithm maintains the current gear for torque demand for the increasing rate of the pedal less than 0.005/T, which indicates moderate torque demand from the driver. The upshift command with decreasing pedal rate is secondary to increasing pedal rate. Within the pedal controller, which consists of the driver pedal shifting controller, braking controller, and hydraulic motor efficiency controller, the first priority is given to the braking process. The second priority is the hydraulic efficiency controller, with 95% limitation, which can keep the hydraulic motor working within the high efficiency range. For the overall controller of the transmission, first priority is given to the shifting in process (SIP) signal controller as no new action can be initiated while a previous shift command is being executed.
34

The hydraulic motor speed controller is the second priority as long as the violation of the maximum hydraulic motor speed limitation is not violated.

4.2.5

Overall Speed Violation of the Hydraulic Motor

The maximum hydraulic motor rotational speed limitation of 3000 RPM

(134 radians/second), must be strictly obeyed regardless of current hydraulic motor speed condition or the proposed hydraulic motor speed after calculation. An upshift is applied when the current hydraulic motor speed exceeds the maximum limitation to reduce the motor speed. Furthermore, no command can be made by the shift controller if the corresponding proposed hydraulic motor speed exceeds the maximum limitation. Figure 4-7 shows the hydraulic motor speed limit controller.

Purpose: to enforce an upshift on motor speed violation Next Gear Up 1 1 Motor Speed 0 Motor Speed Check >= 314 1 Gear Shift Command: 1 = upshift 0 = no action

Hold Gear Constant

Figure 4-7 Hydraulic motor speed limit controller block diagrams in red

35

4.2.6

Dwell Time Controller

A two-second dwell time used to stabilize the gear during upshift is applied

to the transmission by using the dwell time controller, except for braking processes. The controller shown in Figure 4-8 uses a delay loop which lasts for two seconds to be the timer of the dwell time. It is initiated by a gear change signal to reset the dwell loop.

1 Gear Delay 3 Gear Command Given to Transmission

12:34 Digital Clock -1,0,1 values Check Gear Shift ~= 0 Reset for Shift time When Gear Shift

Prop Gear after limit 12:34 Digital Clock1 Del T >= 2 Dwell Time Setting in Seconds >0 Switch2 1 Prop Gear after Dwell Time

Gear delay Loop 1 z Last Shift Time Memory

Figure 4-8 Dwell time controller block diagrams in magenta

4.2.7

Gear Memory
Every new gear command is made on the basis of the previous gear
36

command. The last gear command of the controller is stored by using a unit delay block. As shown in Figure 4-9 every proposed shifting command has to be combined with the last gear command and stored to the next step sampling time. For example, a downshift command with a value of -1 will be combined with the last gear command of 4 (5th gear ) and output a gear command as 3.

Current Gear Number (1 to 6)

1 proposed gear command Create gear number Proposed Gear1 to go to1 z Gear Memory Gear Sent >0 SIP Check No shifting if SIP

4.18 2nd Gear Ratio1 2.24 3rd Gear Ratio1 1.69 4thGear Ratio1 1.20
Range Limited Gear

1 2 3 4 5 6 1 z used to break alg loop

1 Gear Ratio

dd

Gear

5thGear Ratio1 0.90 6thGear Ratio1 0.78 7thGear Ratio1

mit Gears to Range 2nd - 7th >0 Give priority to an upshift

*
Convert to Gear Ration

imit

Prop Gear af ter Dwell Time

Giv en to Transmission

Figure 4-9 Gear memory by using unit delay

Dwell Time Controller

Prop gear command af ter dwell time

4.2.8

Gear Ratio Matching

Prop Gear Number (1 to 6)

Pedal position

Prop Gear af ter braking check

Prop gear command without dwell time

Braking Check

Every gear command must be within the 2nd to 7th gear range according to

the criteria in Chapter 3. The gear range controller checks every gear command before delivering it to the transmission by setting the lowest gear to be 1 (2nd gear)
37

and highest gear to be 6 (7th gear). The the gear ratios of 0.78, 0.9, 1.2, 1.69, 2.24 and 4.18 are matched with the gear numbers from the highest gear 6(7th gear) to the lowest gear 1(2nd gear). In doing this, the gear range controller can match each gear

command with its corresponding gear ratio and then output the gear ratio to the truck
1

model. For instance, the gear range controller takes in a gear command number 4 z
Gear Memory

(5 gear), finds its gear ratio 1.2 and then outputs it to the truck model. Figure 4-10 >0 shows theshifting ifratio matching controller. No gear SIP
SIP Check

th

eate gear number Proposed Gear1 to go to1

Propsed gear Gear Sent 4.18 2nd Gear Ratio1 2.24 3rd Gear Ratio1 1.69 4thGear Ratio1 1.20 5thGear Ratio1 0.90 6thGear Ratio1 0.78 7thGear Ratio1 1 2 3 4 5 6 1 z used to break alg loop

1 Gear Ratio

*
Convert to Gear Ration

Figure 4-10 Gear ratio matching controller

38
Prop Gear Number (1 to 6)

Chapter 5

Simulation Results

In this chapter, the simulation results are presented and discussed by comparing them with the default gear shift schedule. Results include fuel consumption, tracking performance, and motor speed limit. The simulation runs in a 1369-second velocity-base EPA Federal Urban Driving Schedule (FUDS) with a discrete 0.005-second sampling time.

39

30

Federal Urban Driving Schedule (FUDS)

25

Vehicle Speed(m/s)

20

15

10

200

400

600

800

1000

1200

1400

Time(sec)

Figure 5-1 EPA Federal Urban Driving Schedule (FUDS)

5.1

Fuel Consumption
Figure 5-2 shows that the total fuel consumption of the controller simulation

is 3.7367 Kg, which is 5.22% less than the fuel consumption of the original default shifting simulation, 3.9427 Kg. The number of gear changes in the controller simulation is 315, compared to 197 in the default simulation.

40

4 3.5
Fuel Consumed(Kg)

Fuel Consumption

3 2.5 2 1.5 1 0.5 0

Controller Model FuelConsumption Original Model Fuelconsumption

200

400

600

800

1000

1200

1400

Time(sec)

Figure 5-2 Fuel consumption simulation results comparison: fuel consumption Kgvs. time (seconds)

As the default shifting schedule of the original model does not include the 2-second dwell time criterion, the fuel consumption of the original model with 2-second dwell time will be a little higher than 3.9427 Kg without dwell time.

41

5.2

Gear Shifting Schedule

GearRatio of Controller

4.5

3.5

GearRatio

2.5

1.5

0.5

200

400

600

800

1000

1200

1400

Time(sec)

Figure 5-3 Controller model simulation results: gear ratio vs. time (seconds)

The gear shifting schedule of the controller simulation shows no unexpected oscillations of the gear selection when a 2-second dwell time applied, as illustrated in Figure 5-3. All the downshifts are either due to the urge demand of torque from the driver or violations of the hydraulic motor speed and displacement limitations as expected in the strategy. Shifting with less than 2 seconds dwell time only occurs when the hydraulic motor speed reaches the maximum limit 314 radian/second (3000 RPM), or during braking downshift process, as introduced in the previous chapters.
42

GearRatio of Original Shifting 4.5

3.5

3
GearRatio

2.5

1.5

0.5

200

400

600 Time(sec)

800

1000

1200

1400

Figure 5-4 Original model simulation results: gear ratio vs. time (seconds)

The default gear shifting schedule of the original model, shown in Figure 5-4, presents relatively simple gear changes and unexpected oscillation at approximately 700 seconds, which may be due to the lack of dwell time.

43

5.3

Tracking Performance
VehicleSpeed of Controller

30

Actual Vehicle Speed Desired Vehicle Speed

25

20

Vehicle Speed(m/s)

15

10

200

400

600

800

1000

1200

1400

Time(sec)

Figure 5-5 Controller model simulation results: actual and desired vehicle speed (m/s) vs. time (second)

44

VehicleSpeed of Original Shifting 30 Actual Vehicle Speed Desired Vehicle Speed

25

20
Vehicle Speed(m/s)

15

10

200

400

600 Time(sec)

800

1000

1200

1400

Figure 5-6 Original model simulation results: actual and desired vehicle speed (m/s) vs. time (second)

VehicleSpeed Error
10

Vehicle Speed Error(m/s)

Original Model Vehicle Speed Error Controller Model Vehicle Speed Error

-2

-4 0

200

400

600

800

1000

1200

1400

Time(sec)

Figure 5-7 Simulation results comparison between original and controller model: vehicle speed error (m/s) vs. time (second)

45

The combined time-based plots of actual vehicle speed (m/s) and desired vehicle speed (m/s) are depicted in Figure 5-5 and Figure 5-6. The results of figure-20 indicate that the truck controlled by the optimal gear shifting controller can track the desired driving cycle well. One exception occurs between 200 seconds and 300 seconds with a 8m/s peak value difference during the high-acceleration and high-speed period around 17m/s to 25m/s. This difference can be seen clearly from Figure 5-7 which shows the error of the actual vehicle speed from both original and controller model when compared to the desired vehicle speed. The vehicle speed error stays no more than 3m/s except for the period between 200 seconds and 300 seconds as mentioned before. In fact, the FUDS is used for light-duty vehicle test. Therefore, the most aggressive acceleration rates might not be realistic for heavy-duty truck [7], which can account for the peak error of vehicle speed displayed between 200 seconds and 300 seconds. The total actual simulation driving distance of the truck is 11428m which is 4.68% less than the desired total distance of FUDS 11989m. The tracking performance of the original model, as shown in figure-21, has a total driving distance of 11592m, a 3.31% difference between actual driving distance and desired driving distance. Though the controller model drives 1.37% less distance than the original one, the difference is less than the designated maximum difference of 5%. Therefore, the simulation results of the controller still show good performance on tracking the desired driving cycle.

46

5.4
350 300
Motor Speed(rad/s)

Hydraulic Motor Speed

MotorSpeed of Controller Model

250 200 150 100 50 0

200

400

600

800

1000

1200

1400

Time(sec)

Figure 5-8 Controller model simulation results: hydraulic motor speed (radian/second) vs. time (second)

47

300

MotorSpeed of Original Shifting

250
Motor Speed(rad/s)

200

150

100

50

200

400

600

800

1000

1200

1400

Time(sec)

Figure 5-9 Original model simulation results: hydraulic motor speed (radian/second) vs. time (second)

The hydraulic motor speed is well below the maximum limit of 14 radian/second (3000RPM) , as presented in Figure 5-8. When compared to the results of original model in Figure 5-9, it is obvious that the hydraulic motor speed of the controller model is higher, or in other words, keeps in the high-speed range more often, which eventually results in higher motor efficiency and better fuel economy.

5.5

Summary
This chapter presents and analyzes the results of the controller model
48

simulation and compares them with the results of the original model. Table-1 shows the comparison between the original model with default shifting schedule and the model with shifting controller in term of fuel consumption, number of gear changes, total driving distance, and the difference between actual and desired distance.

Table 5.1 Simulation results comparison between controller and original model

Fuel consumption (Kg) controller original 3.7367 3.9427

Total driving distance (m) 11428 11592

Driving distance difference between actual and desired (%) 4.68 3.31

Shifting change times 315 197

The controller simulation showed improved fuel consumption and acceptable tracking performance. However, the original model has slightly better tracking performance.

49

Chapter 6

Summary and Conclusion

6.1

Summary
This paper proposes a fuel-economy optimization gear shifting strategy for

a seven-speed automatic transmission used on a hydraulic hybrid vehicle in order to maximize fuel economy. This strategy is designed with a highest possible gear criterion as long as the torque requirement can be satisfied, except for braking process and torque demanding situations. The optimization strategy takes several other criteria into consideration, such as high motor displacement criterion, to improve efficiency and fuel economy as well. Then the optimization strategy is developed on the basis of these criteria from two main aspects of the existing SIMULINK truck model. One approach is based on the hydraulic motor working conditions, such as motor displacement, and the other is based on the drivers intention, which is interpreted as the driver pedal position. This controller is able to recognize the drivers intention to
50

change the speed and incorporate it into gear shifting decision making. This paper then develops a SIMULINK controller model based on the optimal gear shifting strategy and criteria and validates the model both in fuel economy and power performance by analyzing the simulation results in the Federal Urban Driving Cycle.

6.2

Conclusion
The simulation results show that the SIMULINK optimal gear shifting

controller model is able to increase the fuel economy by 5.22% with a 3.7367 Kg fuel consumption compared to the original default shifting schedule. The controller model is also able to keep the hydraulic motor speed below the 3000RPM maximum speed limitation when driving. Moreover, by keeping the hydraulic motor speed higher, the hydraulic motor efficiency can stay in high range more often to get better fuel economy. The controller model performs well in the tracking with 4.68% distance difference between the actual and desired total driving distance which is less than the 5% designed standard.

6.3

Future Work
As there is no braking pedal in the truck model, the braking intention can

51

only be predicted from the driver behavior trend on the accelerating pedal. And due to many influence factors in the truck model, the positive correlation between the hydraulic motor speed and the displacement can only be concluded approximately. Therefore, a more efficient schedule for high speed driving period can be developed based on a braking pedal input and accurate relationship between the hydraulic motor speed and the displacement, which can track the driving cycle or road conditions better. Furthermore, with instantaneous fuel consumption output from the truck model, the optimal shifting strategy can be improved with more efficient choices. The FUDS only simulates the driving test on flat roads and roads with slopes are left to be developed in the future. The optimal gear shifting strategy for a seven-speed automatic transmission developed in this thesis is part of the research work for a future hydraulic hybrid truck. The work done by this thesis is presented in simulation, which still needs to be applied on the hydraulic hybrid truck transmission and adjusted according to the real conditions of the truck in the future.

52

References

[1] R. Hall, UPS and J. J. Kargul, U. S. EPA, Hydraulic Hybrid Promises Big Savings for UPS, Hydraulics & Pneumatics, Oct 2006, p 42 [2] YJ. Kim, Integrated Modeling and Hardware-In-The-Loop Study for Systematic Evaluation Of Hydraulic Hybrid Propulsion Options. Doctor of Philosophy in the University of Michigan 2008. [3] A. Mohaghegh-Motlagh and M. H. Elahinia, Dynamic Force Prole in Hydraulic Hybrid Vehicles--A Numerical Investigation, Vehicle System Dynamics, Vol. 48, No. 4, April 2010, p 405428 [4] AW. Stienecker, T. Stuart, C. Ashtiani, An Ultracapacitor Circuit for Reducing Sulfation In Lead Acid Batteries For Mild Hybrid Electric Vehicles. Journal of Power Sources 2006 [5] A. Mohaghegh-Motlagh, M. Abuhaiba, M. Elahinia, and W. Olson, Hydraulic Hybrid Vehicles, K. Myer, ed.,Wiley, 2007. ISBN: 978-0-471-79369-4. [6] EPA, Draft Report on the Environment, 2003. [7] M. Shan Ph.D. Dissertation Modeling and Control Strategy for Series Hydraulic Hybrid Vehicle, The University of Toledo, December 2009 [8] A. Mohaghegh-Motlagh and M.H. Elahinia, M. Abuhaiba, and W. Olson (2007) Application of Smart Materials for Noise And Vibration of Hydraulic Systems, ASME DETC/CIE, September 47, Las Vegas, NV. [9] C. Schroeder, A. Mohaghegh-Motlagh, T.M. Nguyen, and M.H. Elahinia, Modeling of A Hydraulic Hybrid P/M For Evaluation of Vibration Characteristics, ASME International Mechanical Engineering Congress, November, 2007. Seattle,WA, p 11-15 [10] A. Haj-Fraj, F. Pfeier, Optimal Control of Gear Shift Operations In Automatic Transmissions, Journal of the Franklin Institute 338 (2001) p 371390
53

[11] Hydraulic Hybrids Boost Fuel Economy, Hydraulics and Pneumatics, Machine Design, October 2006 [12] B. Kwon, H. Kim, Dynamic Analysis of Shift Quality for Clutch to Clutch Controlled AT, KSME International Journal, Vol 14, No. 12, (2000), p 1348-1357 [13] J. Gong, D. Zhao, Y. Chen, N. Chen, Study on Shift Schedule Saving Energy of Automatic Transmission of Ground Vehicles, Journal of Zhejiang University SCIENCE 2004 5(7), p 878-883 [14] H. Sun, L. Yang, J. Jing, Hydraulic/Electric Synergy System (HESS) Design for Heavy Hybrid Vehicles, Energy 35 (2010), p 5328-5335 [15] T. Minowa, H. Kimura, N. Ozaki, M. Ibamoto, Improvement of Fuel Consumption for A Vehicle with An Automatic Transmission Using Driven Power Control With A Powertrain Model, JSAE Review 17(1996), p375-380 [16] H. Sun, Multi-Objective Optimization for Hydraulic Hybrid Vehicle Based on Adaptive Simulated Annealing Genetic Algorithm, Engineering Applications of Articial Intelligence 23 (2010), p 2733 [17] M. Pachter, Optimal Control Problem for an Automatic Transmission, Journal of Optimization Theory And Applications, Vol. 42, No. 1, Jan 1984 [18] B. Mashadi, A. Kazemkhani, and R. Baghaei-Lakeh, an Automatic Gear-Shifting Strategy for Manual Transmission, Proc.I.Mech.E Vol. 221 Part I: J. Systems and Control Engineering, p 757-768 [19] H. Lee, S. Sul, H. Cho, and J. Lee, Advanced Gear-Shifting and Clutching Strategy for A Parallel-Hybrid Vehicle, IEEE Industry Applications Magazine, November/December 2000 [21] M. Pettersson and L. Nielsen, Gear Shifting by Engine Control, IEEE Transactions On Control Systems Technology, Vol. 8, No. 3, May 2000, p 495-507 [22] B. Wu, C. Lin, Z. Hlip, H. Peng and D. Assanis, Optimal Power Management for a Hydraulic Hybrid Delivery Truck, Vehicle System Dynamics 2004. Vol 42. Nos. 1-2., p 23-40 [23] A. Delaney, Hydraulic Hybrids, Engineering & Technology 22 November - 5 December 2008 [24] R. Klop and M. Ivantysynova , Investigation of Noise Sources on A Series Hybrid Transmission, International Journal of Fluid Power 12 (2011) No. 3, p 17-30
54

[25] H. Ohnishi, J. Ishii, M. Kayano, H. Katayama, A Study on Road Slope Estimation for Automatic Transmission Control, JSAE Review 21 (2000), p 235-240 [26] E. Galvagno, M. Velardocchia, A. Vigliani, Analysis and Simulation of A Torque Assist Automated Manual Transmission, Mechanical Systems and Signal Processing 25 (2011), p 18771886 [27] Y. Li, F. Yu, C. Wu, Application of Neural Network to Downhill Shift Strategy for Automatic Transmission, J. Shanghai Jiaotong Univ. (Sci.), 2010, 15(4), p 498-503 [28] Y. Yan, G. Liu and J. Chen, Integrated Modeling and Optimization of A Parallel Hydraulic Hybrid Bus, International Journal of Automotive Technology, (2010) Vol. 11, No. 1, p 97104 [29] J. Pu, C. Yin, J. Zhang, Fuel Optimal Control of Parallel Hybrid Electric Vehicles, Front. Mech. Eng. China 2008, 3(3), p 337342 [30] G. Lucente, M. Montanari, C. Rossi, Modeling of an Automated Manual Transmission System, Mechatronics, 17 (2007), p 7391 [31] H. Sun, J. Jing, Research on the System Conguration and Energy Control Strategy for Parallel Hydraulic Hybrid Loader, Automation in Construction 19 (2010), p 213220 [32] J.R. Duncan and E.L. Wegscheid, Determinants of Off-Road Vehicle Next Term Previous Term Transmission Next Term Shift Quality, Applied Ergonomics 1985, 16.3, p 173-178 [33] Q. Zheng, K. Srinivasan, G. Rizzoni, Optimal Engine Transmission Control of Neutral-Idle Clutch Application, JSAE Review 22 (2001), p 463472 [34] A. Macor, A. Rossetti, Optimization of Hydro-Mechanical Power Split Transmissions, Mechanism and Machine Theory 46 (2011), p 19011919 [35] M. Pesgens, B. Vroemen, B. Stouten, F. Veldpaus and M. Steinbuch, Control of A Hydraulically Actuated Continuously Variable Transmission, Vehicle System Dynamics Vol. 44, No. 5, May 2006, p 387406 [36] R. Pffner, L. Guzzella, C.H. Onder, Fuel-Optimal Control of CVT Powertrains, Control Engineering Practice 11 (2003), p 329336

55