2.

LITERATURE SURVEY
2.1 Introduction
High performance control and estimation techniques for induction motor drives are very fascinating and challenging subjects of R & D, and recently, they received wide attention in the literature. Although the technology started more than two decades ago, it is going through rapid advancement in the recent years. High performance control is now finding increasing acceptance in industrial drives for applications, such as steel mills, paper mills, servos, machine tools, robotics, elevators, and electric vehicles. The invention of vector or field-oriented control, and the demonstration that ac motor can be controlled like a separately excited dc motor, brought renaissance in the high performance control of induction motor drives. The advent of microprocessors in the 1970s made the vector control increasingly acceptable from the 1980's. In fact, with vector control, induction motor drive outperforms the dc drive because of higher transient current capability, increased speed range and lower rotor inertia. High performance adaptive and optimal control techniques can be easily applied on vector-controlled drives because of simple dc machine-like transient model. The advent of modern digital signal processors, ASIC chips, powerful personal computers, user-friendly simulation and CAD tools, artificial intelligence (AI) techniques, and advancement of control and estimation theories are continuously extending the frontier of high performance ac drives.

2.2 Vector or Field-Orientated Control

Vector control is the foundation of modern high performance drives. It is also known as decoupling, orthogonal, or transvector control. Vector control techniques can be classified as indirect or feed forward method, and direct or feedback method depending on the method of unit vector generation for vector transformation. In the indirect vector control, synchronously rotating vector components of stator current i qs and i ds are controlled independently to control the torque and rotor flux, respectively. The unit vector signal, that transforms the synchronously rotating stator voltages into stationary frame signals generated from the speed signal and slip signal. The flux can also be simply controlled by open loop. The drive can easily be operated from zero speed to constant power field-weakening region. It is the most popular vector control method in industry. However, the machine parameter variation affects the slip gain, and correspondingly, both static and dynamic performances of the drive are affected. The on-line tuning for parameter variation is more difficult [3]. In the direct vector control, all the basic elements of the control are essentially the same as indirect vector control except the unit vector signal, which is derived from the flux vector that can be estimated either by voltage model, current model, or by close loop observer. It is also possible to estimate the speed signal from the voltage and current signals. The flux vector estimation with voltage model does not work near zero speed, but with current model it can be easily extended to zero [3].

2.3 Direct Torque and Flux Control (DTFC)

The DTFC method is basically a performance-enhanced scalar control method and is popularly known as direct torque control (DTC). It can be shown that the developed torque of a machine is proportional to the product of synchronously rotating stator flux, rotor flux and the angle between them. Basically, DTC has torque and stator flux control in the outer loops. The speed loop can be added on the torque loop with the speed encoder or estimated speed. The machine voltages and currents are sensed to estimate the torque and flux vector that gives information about stator flux location. The control loop errors generate the digital signals through the respective hysteresis-band comparators. A three-dimensional look-up table then selects the most appropriate voltage vector to satisfy the flux and torque demands. Since the feedback signals are being estimated from the machine terminal, the low-speed limitation and parameter variation problem are similar to direct vector control. The drive has fast transient response, and has the simplicity of implementation due to absence of close loop current control, traditional PWM algorithm and vector transformation. However, the inherent limitations of limit cycle control, such as pulsating torque, pulsating flux and additional harmonic loss exist. Recently, a large number of papers are appearing in literature to improve the DTC control [6].

2.4 Adaptive and Optimal Control

A classical control design based on linear plant model and time-invariant parameters can hardly be accepted for high performance control. A vector-controlled drive system can give fast response, but its poles and zeros can vary due to plant parameter variation. The electrical parameters of a machine may vary by saturation, temperature and skin effect, and the mechanical parameters are determined by the coupling load. A high gain negative feedback loop can linearize a plant (within the stability constraint) and attenuate parameter variation and external

disturbance. In adaptive control, the controller parameters (and sometimes the structure) vary to adapt continuously the variation of plant parameters to give the desired stability, dead-beat response and robustness. A control system can be defined as optimal where a performance parameter, such as time of execution, efficiency, or energy consumption, etc., is optimized. In a general sense, the term "optimal" means doing a job in the best of all possible ways. The adaptive control can be classified as either explicit or implicit. A simple example of explicit or direct adaptive control is the gain scheduling control of an inertia (J) varying speed control system provided the parameter is known apriori, or can be identified on real-time basis. In a more complex self-tuning control (STC), the system poles can be assigned, or poles, zeroes and gain may remain unique irrespective of parameter variations. In such a control system, a plant parameter estimation algorithm solves the plant model on-line by observer method. A tuning algorithm then adjusts the control parameters based on the estimation of plant parameters [13]. In recent literature, H-infinity control has been proposed and found to be more effective than pole allocation STC. The examples of implicit or indirect adaptive control methods are model referencing adaptive control (MRAC) and sliding mode control (SMC). Such a control gives robust performance of the drive, i.e., the response is not affected by any parameter variation (such as J), or load disturbance effect. In MRAC, the plant response is forced to track the response of a reference model irrespective of plant parameter variation or load disturbance. Evidently, the desired robustness of the control system is obtained at the cost of optimal response speed [13]. Sliding mode control (SMC) is also known as variable structure control system (VSS). Like MRAC, it gives robust performance against plant parameter variation and load disturbance

effect, but it is somewhat easy to implement. The control can be easily applied to a vectorcontrolled induction motor drive. In a sliding mode control, the reference model is stored in the form of predefined phase plane trajectory, and the drive system response is forced to follow or slide along the trajectory by a switching control algorithm. The structure or topology of the control is varied intentionally between the positive and negative feedback control modes so that the average response of the system is stable although in individual structure it may be unstable [14].

2.5 Intelligent Control

Intelligent control is based on artificial intelligence (AI), which can be defined as computer emulation of human thinking process. The AI techniques are generally classified as expert system (ES), fuzzy logic (FL), artificial neural network(ANN) and genetic algorithm(GA). Expert system, based on Boolean algebra, uses hard or precise computation, whereas fuzzy logic, neural network and genetic algorithm are based on soft or approximates computation. With a control based on AI, a system is said to be intelligent, autonomous, adaptive, selforganizing or learning control. The conventional control design is based on mathematical model of the plant. Often the plant model is unknown, or ill defined, or the system may be nonlinear, complex, and multivariable with parameter variation problem. An intelligent control system can identify the model, if necessary, and give predicted performance even with wide range of parameter variation. Of course, if a model is available, it can be used for simulation study where the control can be optimized by iteration [15]. Fuzzy logic deals with problems that have vagueness or uncertainty, and uses member ship functions with values between 0 and 1 to solve the problem. Fuzzy control can give robust

adaptive response of a drive with nonlinearity, parameter variation and load disturbance effect. The fuzzy controller will then have two input signals, i.e., the loop error and the error rate of change. The control output is the increment of current signal, which is integrated to generate the current command. Basically, it is a nonlinear-PI control where the gain components are adaptive in nature. A rule matrix relates the loop error, the error rate of change and the increment of current signal variables. The input signals are fuzzified, the corresponding control rules are evaluated from the membership functions and rule table, composed and finally defuzzified to derive the control signal [16]. Artificial neural network or neural network is the most generic form of AI compared to expert system and fuzzy logic. Basically, it is interconnection of artificial neurons with usually nonlinear transfer function at the output. A neural network can be feed forward or feedback (or recurrent) type. A feed forward multi-layer network consists of input layer, output layer and one or more hidden layers of neurons. The network is trained to generate multiple input-output matching patterns. The training can be off-line, on-line or combination of both. Neural network has been used in various control and signal processing applications in power electronics and drives. Some of these applications are: one or multi-dimensional nonlinear function

generation, PWM control, zero phase shift harmonic filtering of waves, waveform FFT signature analysis, on-line diagnostics, direct or inverse nonlinear model emulation of machine, model referencing adaptive control (MRAC), inverse dynamics control of drives, etc. For neural control, apriori mathematical model of the plant may not be necessary. In fact, plant data can emulate the direct or inverse model for high performance control [15].

The drive with nonlinear load dynamics accepts the torque control signal and gives position as the output. The test data from the nonlinear plant can be used to train a neural network to generate an inverse model. Simulation data can be used for training if the model is known. The inverse model is used in series with the plant to cancel the dynamics and nonlinearity of the plant and get open loop deadbeat control performance. The plant may have parameter variation problem. Therefore, a feedback loop is added for correction. The error signals can also be used for on-line training of the network. The same control principle can be extended to a multiple degree-of-freedom robotic manipulator with internal coupling to get robust performance. Model referencing adaptive control can also be applied to an inverse dynamics based control system to get improved performance [2]. 1.3 Problem Introduction In the scalar control methods of voltage and current fed drives the voltage or current and the frequency are the basic control variables of the induction motor. In a voltage fed drive, both torque and air gap flux are functions of voltage and frequency. This coupling effect is responsible for the sluggish response of the induction motor. If the torque is increased by incrementing the frequency, the flux tends to decrease. But it is compensated by the sluggish flux control loop feeding in additional voltage. This transient dipping of flux reduces the torque sensitivity with slip and therefore lengthens the response time. This limitation can be overcome by applying vector or field oriented control methods, in which AC machine is controlled like a separately excited DC machine. In a DC machine, the torque is given by Te=Kt Ia If. The control variables Ia and If can be considered as orthogonal or decoupled vectors. In normal operation, the field current If is set to

maintain the rated field flux and torque is changed by changing Ia. Due to decoupling the torque sensitivity remains maximum in both transient and steady state operations. For an induction machine, the machine operation is considered in a synchronously rotating reference frame where the sinusoidal variables appear as dc quantities. The direct axis component Id and the quadrature axis component Iq are similar to If and Ia. The decoupled torque across the air gap is given by Te=Kt' iqs ids. The variables iqs and ids are mutually decoupled and can be independently varied without affecting the orthogonal component. For normal operation, as in a dc machine, the current ids remains constant and the torque is varied by varying the i qs component. Thus both magnitude and phase of a vector variable are controlled. The phase variables are converted to d-q component by 3- to 2- transformation. In addition to fast transient response due to decoupling control, the conventional stability problem of an induction motor, that is by crossing the breakdown torque point, does not exist here. The control can easily be designed to have four-quadrant operation. Therefore the vector controlled induction motor drives can be used for high performance applications, where traditionally; DC machines have been used.

References
1. P.C.Krause, O. Wasynczuk and S.D. Sudhoff, Analysis of Electric Machinery, IEEE Press, 1994. 2. D.W. Novotny and T.A.Lipo, Vector Control and Dynamics of AC Drives, Clarendon Press Oxford, 1996. 3. B.K. Bose, Power Electronics and AC Drives, Prentice-Hall, 1986. 4. Yen-Shin Lai and Ye-Then Chang, Design and Implementation of Vector-Controlled Induction Motor Drives Using Random Switching Technique with Constant Sampling Frequency, IEEE Transactions on Power Electronics, Vol. 16, No. 3, pp. 400-409, May 2001.

5. Robert D. Lorenz, Thomas A. Lipo and Donald W. Novotny, Motion Control with Induction Motors, IEEE Proceedings, Vol. 82, No. 8, pp. 1215-1240, August 1994. 6. S. Wade, M.W. Dunnigan and B.W. Williams, Simulation of Induction Machine Vector Control and Parameter Identification, IEE Conference Publication No. 399, pp. 42-47, October 1994. 7. C.M. Liaw, F.J. Lin and Y.S. Kung, Design and Implementation of a High Performance Induction Motor Servo Drive, IEE Proceedings-B, Vol. 140, No. 4, pp. 241-248, July 1993. 8. Xingyi Xu and Donald W. Novotny, Selection of the Flux Reference for Induction Machine Drives in the Field Weakening Region, IEEE Transactions on Industry Applications, Vol. 28, No. 6, pp. 1353-1358, November/December 1992. 9. Joseph Vithayathil, Field Oriented Control (Vector Control) of 3 phase Squirrel Cage Induction Motors, JIETE, 37(1), pp 57-73, 1991. 10. Robert D. Lorenz and Donald B. Lawson, Flux and Torque Decoupling Control for FieldWeakened Operation of Field-Oriented Induction Machines, IEEE Transactions on Industry Applications, Vol. 26, No. 2, pp. 290-295, March/April 1990. 11. Zaher Daboussi and Ned Mohan, Digital Simulation of Field-Oriented Control of Induction Motor Drives Using EMTP, IEEE Transactions on Energy Conversion, Vol. 3, No. 3, pp. 667-673, September 1988. 12. Robert D. Lorenz, Controllers for High-Performance Applications, IEEE Transactions on Industry Applications, Vol. 22, No. 2, pp. 293-297, March/April 1986. 13. A. Brickwedde, Microprocessor Based Adaptive Speed and Position Control for Electrical Drives, Conf. Rec. IEEE IAS Annu. Meet. Pp. 411-417, 1984. 14. B.K. Bose, Sliding Mode Control of Induction Motor, Conf. Rec. IEEE Annu. Meet., pp. 479-486, 1985. 15. B.K. Bose, Expert System, Fuzzy Logic, and Neural Network Applications in Power Electronics and Motion Control, Proceedings of the IEEE, Vol. 82, No. 8, Aug. 1994. 16. P. Guillemin, Fuzzy Logic Applied to Motor Control, IEEE Trans. on Industry Applications, Vol. 32, No. 1, Jan./Feb. 1996. 17. David M. Brod and Donald W. Novotny, Current Control of VSI-PWM Inverters, IEEE Transactions on Industry Applications, Vol. 21, No. 4, pp. 562-570, May/June 1985.

18. C.C. Chan and Huaqian Wang, An Effective Method for Rotor Resistance Identification for High-Performance Induction Motor Vector Control, IEEE Transactions on Industrial Electronics, Vol. 37, No. 6, pp. 477-482, December 1990. 19. P.C. Krause and C.H. Thomas, Simulation of Symmetrical Induction Machinery, IEEE Transactions on Power Apparatus and Systems, Vol. 84, No.11, pp. 1038-1053, November 1965. 20. S. Ertem and Y. Baghzouz, Simulation of Induction Machinery for Power System Studies, IEEE Transactions on Energy Conversion, Vol. 4, No. 1, pp. 88-94, March 1989. 21. Pekik Argo Dahono and Qamaruzzaman, A New Reduced Order Model of Induction Motors, IEEE Catalogue No. 95TH8130, pp. 651-655, 1995. 22. F.D. Rodriguez and O. Wasynczuk, A Refined Method of Deriving Reduced Order Models of Induction Machines, IEEE Transactions on Energy Conversion, Vol. 2, No. 1, pp. 31-37, March 1987. 23. O. Wasynezuk, Yi-Mia Diao and P.C.Krause, Theory and Comparison of Reduced Order Models of Induction Machines, IEEE Transactions on Power Apparatus and Systems, Vol. 104, No. 3, pp. 598-606, March 1985. 24. T.L. Skvarenina and P.C. Krause, Accuracy of A Reduced Order Model of Induction Machines in Dynamic Stability Studies, IEEE Transactions on Power Apparatus and Systems, Vol. 98, No. 4, pp. 1192-1197,July/Aug. 1979. 25. P.C. Krause, F. Nozari , T.L. Skvarenina and D.W. Olive, The Theory of Neglecting Stator Transients, IEEE Transactions on Power Apparatus and Systems, Vol. 98, No. 1, pp. 141148,Jan./Feb. 1979. 26. N. Gunaratnam and D.W. Novotny, The Effects of Neglecting Stator Transients in Induction Machine Modeling, IEEE Transactions on Power Apparatus and Systems, Vol. 99, No. 6, pp. 2050-2058, Nov./Dec. 1980. 27. Gill G. Richards and Owen T. Tan, Simplified Models for Induction Machine Transients under Balanced and Unbalanced Conditions, IEEE Transactions on Industry Applications, Vol. 17, No. 1, pp. 15-21, Jan./Feb. 1981. 28. D.W. Novotny and J.H. Wouterse, Induction Machine Transfer functions and Dynamic Response by Means of Complex Time Variables, IEEE Transactions on Power Apparatus and Systems, Vol. 95, No. 4, pp. 1325-1335, July/Aug. 1976.