Design of a Digital Controlled Solar Water Pump

Drive System for a Nano-ltration System

Chuanlin Jin, Wei Jiang
Smart Energy Laboratory
School of Energy and Power Engineering
Yangzhou University
196 HuaYangXiLu Rd, Room N521
Yangzhou, China, 225127
Email: jiangwei@yzu.edu.cn
AbstractStand-alone solar power system provides simple
yet exible solution to certain applications which demand high
mobility. This paper introduces the design of a single solar
panel powered induction machine-driven water pump system,
which is used in an on-board nano-ltration system. The system
conguration is evaluated at the rst stage to provide high ratio
voltage step-up; complementary PWM and space vector PWM
scheme is chosen for dc-dc stage and dc-ac stage respectively; V/F
based control scheme is designed and integrated with maximum
power point tracking algorithm for the best use of solar panel.
The control system is implemented digitally in a TMS320F2812
DSC. Meaningful mode of operation is tested and satisfactory
experimental results are collected.
I. INTRODUCTION
Solar is one of the most environment-friendly renewable
energy sources. The process of solar energy generation re-
quired no extra energy input and yields no harmful emissions.
Therefore, solar energy is considered as one of the most
important primary energy sources in the future.
Even today, some remote areas cannot be covered by the
utility grid; however, most of those areas are of abundant
sun illumination. Therefore, stand-alone power system will
be quite suitable for those regions. Generally, there are two
solutions for the stand-alone power system, one is micro-grids
based, in which the renewable energy sources are connected to
a common dc or ac bus via power electronic converters. The
electric/electronic appliances and equipments within the grid
area can be fed with high quality electric power using power
electronic converters. Micro-grids based solution requires huge
transmission and distribution infrastructure investment, which
is indeed a micro version of utility grid. The other solution
is non-grid based, in which one load is fed by one or more
sources. This solution offers high mobility and exibility
for different applications, for example solar powered water
pumping for irrigation.
Water treatment is of very importance for daily life. In such
appliances, collected rain water is pumped via an induction
machine driven water pump into a nano-ltration chamber,
which delivers clean water for drinking. Due to the nature of
this application, the water treatment appliance has to be highly
mobile, so that only one photovoltaic panel and one fractional
horse-power induction machine is used in the system.
This paper investigates the design and implementation of
the energy conversion stage of a water pump system, which
is powered by one photovoltaic panel without any energy
storage and loaded by one fractional horse-power induction
machine (FHPIM) of 150W. The system conguration is
evaluated at the rst stage to provide high ratio voltage
step-up; and complementary PWM and space vector PWM
scheme is chosen for dc-dc stage and dc-ac stage respectively;
V/F based control scheme is designed and integrated with
maximum power point tracking algorithm for the best use of
solar panel. The control system is implemented digitally in
a TMS320F2812 DSC, meaningful scenarios of operation is
tested to prove the feasibility of the design.
II. THEORY OF OPERATION
As can be seen from Fig.1, an fractional horse-power
induction motor drive system consists of ve interrelated parts:
the photovoltaic array, the dc-dc converter, the dc-ac inverter,
the control module and the pump.
DCDC
Couvt
Th
lhu:
`uvt
MllT \l S\l"M
M l
So'u luu'
l|lTM
lum
Fig. 1. System Structure
TABLE I
SOLAR PANEL SPECIFICATIONS
Maximum power 205W
Maximum power voltage 26.6V
Maximum power current 7.71A
Open circuit voltage 33.2V
Short circuit current 8.36A
The power ow from the solar panel has to be maximized
and boosted to supply the DC bus of IM drive system. Given
the solar panel specication in Table I, the range of input
IEEE PEDS 2011, Singapore, 5 - 8 December 2011
978-1-4577-0001-9/11/$26.00 2011 IEEE
982
voltage can be selected. In order to run a 220V AC induction
motor, a DC bus of 350V is required considering the safety
margin, and the motor has to be connected in delta. The front-
end converter has to boost the input voltage with a large ratio.
An isolated dc-dc topology is used as our front-end converter
as shown in Figure 2(a): the input solar panel voltage is
inverted into high frequency ac using a single phase full-bridge
inverter; a high-frequency (HF) transformer is used to step-up
the voltage to a compatible level; on the secondary of the HF,
a half bridge rectier is used to rectify the HF ac voltage as
well as double the dc voltage. As shown in the gure, S
1
and S
2
are switched above 50% duty cycle at 50kHz, S
3
and
S
4
are switched below 50% duty with 180 degree phase-shift
with respect to S
1
and S
2
, D
1
and D
2
together with two series
connected capacitor form a voltage doubler for dc output. The
minimum voltage gain for this converter can be found as,
V
o
V
i
= 2nD (1)
which forms a proportional gain as the small signal transfer
function. The control design of this converter is simple due to
its zero order properties.
!
?
3
+
S!
S?
S3
S+
\d
D!
D?
\o!
\o?
\o

(a)
S1
S2
S4
S3
(b)
Fig. 2. dc-dc converter (a) schematics, (b) PWM logic
Photovoltaic array has output characteristics of nonlinear-
ities obviously, which are affected by external environment
(sunshine intensity, temperature and load) and itself tech-
nical index (output resistance). Only at one voltage level,
photovoltaic array can output maximum power, and then the
photovoltaic array reaches the highest point of power output
voltage curve, called the Maximum Power Point. Since the
conversion efciency of photovoltaic array is rather low, and
for effective use, Maximum Power Point Tracking (MPPT)
appears very important. In this study, the Perturbation and
Observation method (P&O) is used, which is indicated in Fig.
3. As indicated in Figure 3, by observing the power variation
and voltage difference, the location of current operation point
can be determined by the algorithm hence the decision of next
operational point can be made.
0 5 10 15 20 25 30 35
0
50
100
150
200
250
Solar cell array power curve
Solar cell array voltage (Volt)
S
o
l
a
r
c
e
l
l
a
r
r
a
y
p
o
w
e
r
(
W
a
t
t
)P1
P3
P2
P4
Fig. 3. Power curve of a solar panel
Variable voltage and variable frequency (VVVF) control is
selected as the means for power regulation. The control block
of the system is shown in Fig. 4. The estimated solar panel
V
MPP
is used as the reference for the VVVF control loop,
the resulted error is fed into a PI controller to generate an
excitation frequency; the output stator voltage of IM can be
obtained by look-up V/F table; the generated voltage reference
value is used in SVPWM generator for six gate pulses.
PI regulator
SVPWM
U
f
Reference Usp*
Usp
f
* Us
Fs
Frequency Slope function PWM1
PWM2
PWM3
PWM4
PWM5
PWM6
u '
+
-
Fig. 4. VVVF control of induction motor
III. POWER STAGE DESIGN
Given the specication in Table I, the maximum voltage
and current for the primary side MOSFET can be calculated.
Suppose a maximum duty of 80% for the primary side inverter
switch, so we can calculate that the max-current through the
MOSFET is 10.45A through I
max
= I
sc
/0.8. On this basis,
IRF540, whose I
dnom
is 33A and V
ds
is 100V, is chosen
as the inverter switch. The dc system should at least offer
a voltage gain of 16 in order to take care of very low solar
panel input (20V ). Therefore, using the half bridge rectier
secondary, the ratio is chosen as n = 8, therefore we set the
turn ratio 13 : 100. At the secondary side, as the voltage level
is 350V , and power is about 150W, the maximum current
is 0.43A, therefore fast recovery diode ES1J is sufcient for
this application, which average current is 1A. In dc-ac inverter
circuit, IPM FSBS5CH60 is chosen as the power stage for the
three phase inverter. It is an advanced smart power module
(SPM) that Fairchild has newly developed and designed to
provide very compact and high performance ac motor drives
mainly targeting low power inverter-driven application like air
conditioner and washing machine. The supply voltage of IPM
FSBS5CH60 is 500V, which can satisfy our dc link voltage
of 350V, and 5A output current is sufcient for the solar
983
pump application. It combines optimized circuit protection and
drive matched to low-loss IGBTs. System reliability is further
enhanced by the integrated under-voltage lock-out and short-
circuit protection. The high speed built-in HVIC provides
optocoupler-less single-supply IGBT gate driving capability
that further reduce the overall size of the inverter system
design. Each phase current of inverter can be monitored
separately due to the divided negative dc terminals.
IV. PERIPHERAL CIRCUIT DESIGN
The isolated sampling circuit is adopted as the voltage and
current sensing circuit. The advantages of this sampling circuit
are: 5V power supply only; the main circuit and control circuit
are isolated by HCPL7800 isolation amplier, thereby the
digital control circuit can be protected.
For the auxiliary power supply, the solar panel output volt-
age is converted to 10V rst to supply the control peripherals
using a LM2596 simple switcher. The 10V is output is then
converted to 15V for the gate drive and other supporting
circuit. The reason for this design is that the solar panel voltage
will uctuate during the test, if solar panel voltage drops
down below 15V, the power stage will lose power supply.
Therefore V
in
10V 15V structure is chosen for reliability
considerations. For safety consideration, we use the method
of opto-isolation to isolate the drive circuit and the control
circuit.
V. ALGORITHM IMPLEMENTATION
The control design includes a PI controller for dc-dc con-
verter voltage control, a PI controller for frequency generation
whose reference is fed from a MPPT routine. The dc-dc
converter can be treated as a zero order system and the
frequency generation can be treated as a rst order system.
The cross-over frequency for the dc-dc converter is selected
as 1kHz and 10Hz for the frequency generation loop. The
resulted PI controller in s-domain can be discretized into
digital form using Equation 2.
u(n) = K
p
e(n) +
K
p
T
T
i
n

k=0
e(k) (2)
Space vector PWM (SVPWM) is applied for VVVF control.
Ten PWM signals are needed to drive the power stage as
indicated in the hardware design. Four of them are used to
gate the MOSFET of full bridge dc-dc converter and the other
six are used to the IPM. Figure 5 shows the ow chart for
SVPWM and the corresponding tasks can be processed in the
interrupt service routine,
1) Determine V
ref
, and angle ;
2) Determine time duration (T
1
, T
2
, T
0
);
3) Determine the switching time of each IGBT (S
1
to S
6
)
To implement the VVVF control, two variables have to be
adjusted: the output ac voltage V and the frequency f. And
the following relation holds,
m

=
V
f
= C (3)
Sample of DC bus voltage
and bus current, and motor
voltage and motor current.
Set time base
Through the PI
regulator, the inverter
output frequency f is
obtained.
Calculate
modulation wave
cycle
According to U/f
function, calculate the
motor voltage U.
Search the sector, where is the U.
Calculate the action
time T
Put action time T into registers
Interrupt return
SVPWM
Fig. 5. Flowchart of SVPWM interrupt subroutine
where m is the amplitude modulation index and m = V/V
dc
,
V
dc
is the input voltage of inverter. is the angle of phase
voltage vectors rotation increment, which is modied during
each PWM underow interrupt service routine during DSC
operation. The vector rotation increment and rotation angel is
given as
=
2f
T
PWM
(4)
= + (5)
where is the angle of voltage vector in the reference
frame, f is the output voltage frequency. In order to generate
a constant ux for VVVF, modication of m and is
sufcient. If the speed of motor need increasing to match the
maximum power point, has change quickly; therefore letting
= + 1 will increase the motor speed. Conversely, if
= 1, the motor speed will be reduced. At the mean
time, m is also modied in SVPWM service routine; so that
V/F is keep constant while changing the motor speed.
If f is increased, the output power of motor will be
increased, so the output power of photovoltaic panel will also
be greater.Conversely, if f is decreased, output power will be
decreased too, and the output power of photovoltaic panel will
drop. Therefore, the output power of the IM should be adjusted
constantly in order to keep the output power of photovoltaic
panel staying in the MPP.
As indicated in Fig. 6, if V/F is kept constant in SVPWM
algorithm, the output power can be adjusted by changing the
984
Motor input voltage V
and frequency f changes
Motor speed n
changes
Motor output power
p changes
Photovoltaic array
output power changes
Fig. 6. Principle of power control
excitation frequency. So photovoltaic arrays load matching
can be realized by changing the speed of the pump. Namely
through speed reference regulation, we can adjust the photo-
voltaic arrays output load, and make the output power follow
the maximum power point of the solar panel all the time.
VI. TESTING RESULT
The induction machine test bed is indicated in the Fig.7, the
test is conducted in an indoor lab environment. As can be seen
in Fig.7, the system includes a regulated power supply, which
output voltage is from 20V to 40V adjustable during the test,
an embedded F2812 control board, a power board for the solar
pump converter and a FHPIM of 120W, three phase 220V AC
input. The system is later redesigned to a better form factor
with everything inside a box as presented in Fig. 8.
The input voltage is initially set at 24V DC and 350V dc
voltage can be obtained on the system dc-link. The test is
conducted with constant V/F ratio for 220V AC/50Hz, dif-
ferent amplitude modulation index is selected for test purpose
as shown in the gure below. Under four different speed tests,
the system presents excellent electrical and thermal stability.
The system is tested under different excitation frequency; two
of the tests are indicated in Fig. 9 and Fig. 10. As can be seen
in the gure, CH1 shows the transformer primary voltage of
24V (yellow curve, 1V/div, 10X attenuated) which is supplied
by a dc power supply. The CH2 (blue curve, 1V/div, 10X
attenuated) standing for the input voltage from the dc power
supply, which emulates the solar panel output. The dc link
voltage (CH3, purple curve, 100V/div) is the output voltage
from the dc-dc converter and the input of dc-ac inverter. CH4
(green, 100V/div) stands for the inverter output which is
regulated in open loop as the modulation index m is under
change. The experimental results shows that as the solar panel
voltage is regulated, the speed of IM changes. Fig. 9 shows the
input/output waveforms of the power stage at the modulation
index of 0.8 and 50Hz of excitation frequency, and Fig. 10
shows the input/output waveforms for m = 0.5, f = 31.25Hz.
VII. CONCLUSION
This paper presents the digital converter design for a solar
powered water pump system in a nano-ltration system. The
system conguration, PWM algorithm, control system struc-
ture and power stage hardware design are presented in detail.
The control system is implemented in a TMS320F2812 DSC
and tested in several meaningful modes of operation. The test
results indicated the effectiveness of the system design and
digital control implementation.
Fig. 7. Test bed with 120W IM
Fig. 8. Converter prototype
ACKNOWLEDGMENT
The authors would like to thank the nancial support from
the Deans Ofce of School of Energy and Power Engineering,
Provost Ofce of Yangzhou University, and Undergraduate
Student Innovation Funding grant.
REFERENCES
[1] M.A. Vitorino, M.B.R. Correa, C.B. Jacobina and A.M.N. Lima, Sen-
sorless induction motor drive for photovoltaic pumping applications,
23
th
Annual IEEE Applied Power Electronics Conference and Exposi-
tion, pp. 11391143, Feb. 2008.
[2] R. Marouani and F. Bacha, Sensorless induction motor drive for pho-
tovoltaic pumping applications, Advanced Electromechanical Motion
Systems & Electric Drives Joint Symposium, 2009. ELECTROMOTION
2009, pp. 16 , Jul. 2009.
[3] Dezheng Wu and S.D. Pekarek and B. Fahimi, A Field Reconstruction
Technique for Efcient Modeling of the Fields and Forces Within
Induction Machine, IEEE Trans. Energy. Conv., vol. 24, no. 2, pp. 366
- 374, Jun. 2009.
[4] B. K. Bose, P. M. Szczesny, and R. L. Steigerwald, Microcomputer
control of a residential photovoltaic power conditioning system, IEEE
Trans. Ind. Appl., vol. 21, no. 5, pp. 11821191, Sep. 1985.
[5] C. Hua, J. Lin, and C. Shen, Implementation of a dsp-controlled pho-
tovoltaic system with peak power tracking, IEEE Trans. Ind. Electron.,
vol. 45, no. 1, pp. 99107, Feb. 1998.
985
(a)
(b)
Fig. 9. Test waveforms under m=0.8 f=50.00Hz with 120W IM, (a) system
input; (b) system output
(a)
(b)
Fig. 10. Test waveforms under m=0.5 f=31.25Hz with 120W IM, (a) system
input; (b) system output
[6] K. H. Hussein, I. Muta, T. Hoshino, and M. Osakada, Maximum pho-
tovoltaic power tracking: an algorithm for rapidly changing atmospheric
conditions, IEE Proc. Generation, Transmission and Distribution, vol.
142, no. 1, pp. 13502360, Jan. 1995.
[7] E. I. Ortiz-Rivera and F. Z. Peng, Analytical model for a photovoltaic
module using the electrical characteristics provided by the manufacturer
data sheet, 36
th
IEEE Power Electronics Specialists Conference, pp.
20872091, Jun. 2005.
[8] N. Femia, G. Petrone, G. Spagnuolo, and M. Vitelli, Optimization of
perturb and observe maximum power point tracking method, IEEE
Trans. Power Electron., vol. 20, no. 4, pp. 963 973, Jul. 2005.
[9] T. Esram and P. L. Chapman, Comparison of photovoltaic array max-
imum power point tracking techniques, IEEE Trans. Energy Convers.,
vol. 22, no. 2, pp. 439449, Jun. 2007.
[10] M. A. S. Masoum, H. Dehbonei, and E. F. Fuchs, Theoretical and
experimental analyses of photovoltaic systems with voltage and current-
based maximum power-point tracking, IEEE Trans. Energy Convers.,
vol. 17, no. 4, pp. 514522, Dec. 2002.
[11] N. Femia, G. Petrone, G. Spagnuolo, and M. Vitelli, A pure realization
of loss-free resistor, IEEE Trans. Circuits Syst., vol. 51, no. 8, pp.
16391647, Aug. 2004.
986