Optimal Soft Starting Of Voltage-controller-fed Im Drive Based On Voltage Across Thyristor

IEEE TRANSACTIONS ON POWER ELECTRONICS, VOL. 12, NO. 6, NOVEMBER 1997

1041

Optimal Soft Starting of Voltage-Controller-Fed IM Drive Based on Voltage Across Thyristor Venkata V. Sastry, Senior Member, IEEE, M. Rajendra Prasad, and T. V. Sivakumar, Student Member, IEEE

Abstract—AC voltage controllers are used as induction motor starters in fan or pump drives and the crane hoist drives. This paper presents a method of identifying the end of soft start of an ac voltage-controller-fed induction motor (IM) drive based on the voltage across the nonconducting thyristor through a dynamic simulation of the whole drive system. A two point current minimization technique is adopted to operate the drive system at the required optimal voltage under all operating conditions. This minimizes the motor losses. Graphic modeling of the whole drive system is done in a modular format using Design Star and dynamic simulation is done using SABER. The dynamic simulation results of the whole drive system are supported with experimental data.

The main sections of the paper are organized as follows. In Section II, the soft-start and the optimization techniques followed are explained. The simulation of “whole drive system” is discussed in Section III. The various modules developed in SABER simulator are also briefly explained. In Section IV, the simulation results are presented and discussed. In Section V, the experimental setup used for its validation is discussed. In Section VI, the experimental results of the proposed drive system are compared with that of the dynamic simulation results. II. SOFT START

I. INTRODUCTION

AND

OPTIMIZATION

A. Optimal Soft Starting

A

C VOLTAGE-controller-based soft starters offer many advantages over conventional starters like smooth acceleration, ease in implementation of current control, open circuit transition to line voltage, and also energy savings at lightly loaded conditions. Energy savings by voltage control is achieved by reducing the applied voltage if the load torque requirement can be met with less than rated flux. By this way the core loss and stator copper losses can be reduced [1]. When current minimization and power factor maximization concepts were used in combination, the energy saving was found to be good [2], however, the dynamics of the system was poor. When the angle load pattern concept was employed, it gave a better dynamic response [3], [4], but the energy saving was not satisfactory. To be able to obtain energy efficient operation and also obtain a good dynamic response, the slip is to be controlled. This could be achieved using the voltage across the nonconducting thyristor as a feedback parameter [5] as it is an indirect index of back electromotive force (emf) and, hence, speed. Therefore, for identifying the end of soft start, the voltage across the nonconducting thyristor itself could be used. In this paper, simulation and experimental results of such a voltage-controlled drive system as a soft starter are presented. The optimum point at which the losses are minimum was found out by a two-point current minimization technique [6].

During soft start, initially the thyristors are fired at an equal to . When the motor is at standstill, the voltage across the nonconducting thyristor is measured and stored as VREF. Then, is decremental by 0.50 /cycle (ALPSTP1) until reaches 4 . This ensures an initial rise in motor current. The fundamental component of the line current drawn is rectified and sampled once in every 30 ms. If the current is less than the current limit CURLIM for a given motor, then is decremented by 0.10 /cycle (ALPSTP2) until the identification of the end of soft start. If current exceeds the current limit CURLIM, then will not be decremented until the motor accelerates or the current limit is not seen. The end of optimal soft start is identified with the fall of voltage across the nonconducting thyristor to a value below 75% of the value when the motor is at standstill. The voltage across the nonconducting thyristor is the difference between the source voltage and the back emf. The magnitude and the phase lag of back emf (with regard to the source voltage) keeps changing as the motor accelerates. As also keeps changing, it is very difficult to express the voltage across the nonconducting thyristor through an expression involving back emf and source voltage. Hence, empiricism is needed in choosing the value 75%. However, the value 75% is established through experimental tests conducted on various medium range motors and the latter is verified through dynamic simulation as presented in this paper.

Manuscript received June 7, 1996; revised October 24, 1996. This work was supported in part by Prof. D. Schroder of T. U. Munich and by the Volkswagen Foundation, Germany. Recommended by Associate Editor, D. A. Torrey. The authors are with the Department of Electrical Engineering, Indian Institute of Technology, Madras 600 036, India. Publisher Item Identifier S 0885-8993(97)06408-9.

B. Optimization After identifying the end of an optimal soft start, a delay of 30 cycles is provided to allow the flux to settle. Then the process of optimization is activated by incrementing by 0.10 /cycle. A two-point current minimization technique is

0885–8993/97$10.00  1997 IEEE

1042

IEEE TRANSACTIONS ON POWER ELECTRONICS, VOL. 12, NO. 6, NOVEMBER 1997

Fig. 1. The concept behind two-point current minimization.

used for optimization. This technique can be understood from Fig. 1. This figure shows the variation of current with respect to . At any sampling instant , the new value of is calculated based on the difference between two successive current samples and (1) where is the value of at any time instant , is the value of at time instant , and is the step change in per cycle. As long as the difference is positive, the real time optimization process continues. Once the difference becomes negative, incrementing of is stopped and this value of is known as (c.f. Fig. 1). This condition is identified as the end of optimization routine. During simulation of the whole system, the end of soft start and optimization for achieving current minimization are identified by the module ACCON. This module is explained in Section III. The flow for soft start and optimization routine is shown in Fig. 2. III. SIMULATION OF AC VOLTAGE-CONTROLLER-FED IM DRIVE One traditional way of simulating an IM drive is to use an equivalent circuit model for IM. However, this is not the best approach when it is needed to study the performance of the drive during transients. The best approach is to describe the IM model through differential equations in - model. However, when it is needed to simulate the IM drive with switching networks (like ac voltage controller), it is not wise enough to describe the whole drive system through equations for various modes of operation. It is better if the switched network is represented as a circuit. Also, when it is needed to simulate a large system, it is better to follow a modular approach while modeling the whole system. The use of modularity in modeling offers many advantages such as the following: 1) ease of partitioning of the design tasks; 2) ease of reuse in other designs; 3) ease of debugging where ever there is a problem in the subsystem. Hence, the modeling of the whole drive system is done in MAST language of SABER simulator to conform to modular approach. The block diagram of the whole drive system formulated according to Design Star (graphic interface) is as shown in Fig. 3 [7], [8].

Modeling of the Whole Drive System: The whole drive system consists of various subsystems: 1) electrical; 2) electromechanical; 3) mechanical; and 4) control. The approach for modeling of various modules identified under each subsystem is explained in this section. 1) Electrical System: This system consists of three phase sources and a three-phase ac voltage-controller module ACV3PH. The module ACV3PH describes the SCR module connected back-to-back to form a three phase ac voltagecontroller circuit. The SCR is modeled as a low/high resistance model for on/off state of the thyristor [7]. 2) Induction Motor Model: This system consists of an IM module INDMAC. The module INDMAC defines the induction machine model in a stationary reference frame. The differential equations used to describe the dynamics of the induction motor are given in Appendix A. 3) Mechanical Load: This basically defines a mechanical load on the shaft of the induction motor. The equation describing the load on the IM is given by (2) where is the speed in rad/s. The value of is so chosen that at rated speed a load of 2.2 N-m is applied to the motor. This load is 0.155 times the full load. 4) Control System: This system consists of various modules such as IFILTER, VAKMES, ACCON, and FIR to implement the soft starting based on the voltage across nonconducting thyristor and optimization. The description of various submodules used are as follows. IFILTER: This submodule consists of an active filter and a diode bridge rectifier with a capacitive filter. This circuit is shown in Fig. 4. The analogue filter extracts the fundamental component of ac current. This current is then rectified by a diode bridge rectifier and filtered using a capacitive filter. This dc component of the fundamental current is fed to the module ACCON. VAKMES: This module is basically a “discrete” block. This module samples regularly the voltage across the thyristor with a time lapse of 0.3 ms from the instant the current through the thyristor is zero. The time lapse of 0.3 ms is so chosen that at the instant of sampling, neither of the thyristors in that phase conducts. This information is fed to the module ACCON. The netlist for this submodule is given in Appendix B. ACCON: This module is basically a decision making module in a sense that this block calculates the value of , identifies the end of soft start, and does real-time optimization based on the inputs it gets from modules VAKMES and IFILTER. An external clock is used (c.f. Fig. 3) to sample the rectified fundamental component of the line current. The new value of during soft starting is calculated as given by (3). During optimal soft starting, as long as the current is less than current limit CURLIM then (3a)

SASTRY et al.: OPTIMAL SOFT STARTING OF VOLTAGE-CONTROLLER-FED IM DRIVE

1043

Fig. 2. Flow chart for soft start and optimization—module ACCON.

where (3b) for all other values of

is applied to the corresponding thyristor at delay angle (c.f. Fig. 5). The interaction between various software modules is shown in Fig. 6.

(3c) In the present case, ALPSTP1 is 0.50 and ALPSTP2 is 0.10 . If current is more than current limit CURLIM, then (4) The values of ALPSTP1 and ALPSTP2 are chosen based on various parameters like rating of the motor, machine parameter values, load on the motor, and chosen soft-start time. During optimization the new value of is calculated as given in (1). The netlist for this module (in MAST language) is given in Appendix C. FIR: This module generates the necessary gate pulses of 120 pulse width according to the value of it gets from the module ACCON. The module generates six sawtooth waveforms (one for each thyristor), each wave phase shifted by 60 . The firing pulse at for each thyristor is generated by comparing with a corresponding sawtooth waveform. Whenever sawtooth wave is greater than , a positive pulse

IV. SIMULATION RESULTS The simulation results of ac voltage-controller-fed IM drive are shown in Fig. 7. Fig. 7(a) shows the variation of rectified component of the fundamental current and the flags used for identifying end of soft start and optimization. From Fig. 7(a), the different stages of soft start and optimization can be seen. During the period “A,” soft start is carried out. This is clearly seen with an initial rise and fall of current and as well as fall of voltage across the nonconducting thyristor [c.f. Fig. 7(b)]. During this period it can also be seen from Fig. 7(c) that alpha is decremented initially at 0.50 /cycle and then at 0.10 /cycle. At time , the voltage across the thyristor falls below 75% of VREF and, hence, it is identified as end of soft start. The flag EOST is set to “one” to indicate the end of soft start. Period “B” indicates the 30 cycles of delay time provided to allow the flux to settle down. During this period alpha remains constant [vide Fig. 7(c)]. The end of the delay is

1044

IEEE TRANSACTIONS ON POWER ELECTRONICS, VOL. 12, NO. 6, NOVEMBER 1997

Fig. 3. The graphic modeling of the drive system using Design Star.

Fig. 4. The analogue filter circuit—module IFILTER.

and (f) shows the effectiveness of optimization. In Fig. 7(e), the rectified component of the fundamental current with and without optimization are presented. It can be seen that the current with optimization is less than the current without optimization. In Fig. 7(f), the motor terminal voltage applied under steady-state condition with and without optimization are presented. From this figure, it can also be seen that the voltage applied to the motor after optimization is less than that without optimization, and the voltage applied to the motor need not be full supply voltage.

V. EXPERIMENTAL SETUP

Fig. 5. The firing module.

indicated by setting the flag EOST to “two.” During the period “C,” optimization is carried out. This is very clearly seen with a decrease in current and as well with an increment in alpha [c.f. Fig. 7(c)]. At time , the end of optimization is identified and it is indicated by setting the flag FOPT to “one.” Fig. 7(d) shows the rectified component of the motor terminal voltage and the supply voltage. From this figure, it can be clearly seen that the full voltage need not be applied to the motor at the end of soft starting. Hence, this method of soft starting is identified as “optimal” soft starting. Fig. 7(e)

The whole drive system was developed based on an 8-b single-chip controller 8751, which consists of built-in timers, ports, and memory. The whole dedicated system used for the control of induction machine drive is shown in Fig. 8. A 3hp motor is used as a test machine. The nameplate data and parameters of the test machine are given in Appendix D. Two ports of the microcontroller are configured as input ports. One port is used to read the analogue to digital converter. The other port is used to read the user set parameters like current limit (CURLIM), (through dual in-line package switches) and also end of conversion signal from ADC. The pins T0 and T1 are used for ADC channel selection. Voltage across the nonconducting thyristor was stepped down using a potential divider and read through an ADC. The voltage across the nonconducting thyristor measurement circuit is shown in Fig. 9. Line current was measured through a current transformer. Fundamental component of the current was rectified, filtered, and read through the same ADC. The

SASTRY et al.: OPTIMAL SOFT STARTING OF VOLTAGE-CONTROLLER-FED IM DRIVE

1045

Fig. 6. The interaction between various software modules under SABER simulator.

interrupt INT0 is used to inhibit the firing pulses under abnormal conditions. The interrupt INT1 is used for R-phase zero-crossing detection.

VI. EXPERIMENTAL RESULTS The experimental results are as shown in Fig. 10. Fig. 10(a) shows the variation of the rectified component of the fundamental current during various stages of soft start and optimization. Fig. 10(b) shows the variation of the rectified component of the voltage across the nonconducting thyristor. Fig. 10(c) shows the variation of speed with time. From Fig. 10(a), it can be seen that when the motor is accelerating, the current is also increasing. Once the motor accelerates to the rated speed, the current falls [period “A” in Fig. 10(a)]. The voltage across the nonconducting thyristor also falls [c.f. Fig. 10(b)]. Then the process of optimization is activated. The period of optimization is associated with a fall in the current [period “B” in Fig. 10(a)] and a rise in the voltage across the nonconducting thyristor. Fig. 10(d) shows the variation of the rectified component of fundamental current without optimization. Comparing Fig. 10(a) with Fig. 10(d), it can be clearly seen that the current drawn by the motor with real-time optimization is less than that without optimization. Fig. 10(e) shows the variation of the rectified component of the source voltage and the motor terminal voltage during various stages of soft starting and optimization. From this figure, it is seen that the full voltage has not been applied to the motor at the end of soft starting. Hence, this method of soft starting is called “optimal” soft starting. From Figs. 7 and 10, it can be observed that simulation and experimental results are the same except for the difference in their time scales. This has been intentional to minimize the simulation time with reduced moment of inertia on the load. The motor terminal voltage rise, just before the end of soft start during simulation [c.f. Fig. 7(f)] is not so fast as it is observed experimentally [c.f. Fig. 10(e)]. This is mainly due to the fact that the induction machine model does not include saturation effect.

VII. CONCLUSIONS In this paper, a novel method of identifying the end of soft start of a voltage-controller-fed IM drive using a “new” parameter, namely the voltage across the nonconducting thyristor, is presented. This method is speed feedback less detection of the end of acceleration of the drive system. Dynamic simulation and experimental results for the whole drive are presented. Simulation of the whole drive system is done in Design Star-SABER environment. The various modules are developed in MAST language in a modular format. The newness of this paper lies in identifying the end of soft start based on the voltage across the nonconducting thyristor through a dynamic simulation of the whole drive system with modularity in mind using a versatile simulator. These ideas are also verified experimentally using an 8b single-chip microcontroller-based voltage-controller-fed IM drive system.

APPENDIX A THE DYNAMIC - MODEL OF INDUCTION MACHINE IN A STATIONARY REFERENCE FRAME

where stands for derivative, stands for flux linkages, stands for number of pole pairs, and stands for speed in rad/s. The relation between flux linkages and currents are as follows:

1046

IEEE TRANSACTIONS ON POWER ELECTRONICS, VOL. 12, NO. 6, NOVEMBER 1997

(a)

(b)

(c)

(d)

(e)

(f)

Fig. 7. Simulation results for IM drive during soft-start and optimization intervals. (a) Variation of the rectified component of the fundamental current. (b) Variation of speed and voltage across the nonconducting thyristor. (c) Variation of alpha. (d) Variation of motor voltage and source voltage through various stages of soft starting and optimization. (e) Variation of rectified component of the fundamental current with and without optimization. (f) Variation of motor terminal voltage with and without optimization.

and

are calculated from the stator phase voltages as follows:

as follows:

The mechanical equation describing system dynamics is as follows: The torque

in

is defined in terms of the - currents

SASTRY et al.: OPTIMAL SOFT STARTING OF VOLTAGE-CONTROLLER-FED IM DRIVE

1047

Fig. 8. 8-b microcontroller-based experimental setup for IM drive system.

Fig. 9. Circuit for measuring voltage across the nonconducting thyristor.

APPENDIX B THE NETLIST FOR MODULE VAKMES #Template for vak measurement #ma1,ma2 voltage across thyristor points #mo1 sampled output #pc,pref pc and pref are current sampling nodes #neg ground node #vsamp sampled value of the voltage across nonconducting thyristor #fsamp Flag to indicate the instant of sampling voltage across nonconducting thyristor template vakmes ma1 ma2 mo1 pc pref neg nstp1,nstp2 electrical ma1,ma2,mo1,pc,pref,neg number nstp1 0.3m,nstp2 0.05m

state nu flag 0,fsamp,vsamp var i iout branch vm1 v(ma1,ma2) branch vmo v(mo1,neg) branch vpc v(pc,pref) when(time step done) if(abs(vpc) 0.01 & flag 0) schedule event (time nstp1, fsamp, 1) schedule event (time nstp1 nstp2, fsamp, 0) flag 1 if(fsamp

1) vsamp

if(abs(vpc)

abs(vm1)

0.01 ) flag 0

equations i(ma1 ma2) 0 i(mo1 neg) 0 i(pc pref) 0 iout i(mo1 neg) iout:v(mo1)-v(neg) vsamp

1048

IEEE TRANSACTIONS ON POWER ELECTRONICS, VOL. 12, NO. 6, NOVEMBER 1997

(a)

(b)

(c)

(d)

(e) Fig. 10. Experimental results for IM drive during soft-start and optimization intervals. (a) Variation of the rectified component of the fundamental current with optimization. (b) Variation of the voltage across the nonconducting thyristor. (c) Variation of speed with optimization. (d) Variation of the rectified component of the fundamental current without optimization. (e) Variation of motor voltage and source voltage through various stages of soft starting and optimization.

SASTRY et al.: OPTIMAL SOFT STARTING OF VOLTAGE-CONTROLLER-FED IM DRIVE

APPENDIX C THE NETLIST FOR MODULE ACCON

1049

#SAMPLING OF CURRENT when(time step done) newcur vc2 if(vc3 0.0) oldcur newcur

#NETLIST FOR ACCON template accon cin1 cin2 csam cout1 neg imax, alpmax, alpmin, alpstp, fac, fac1, tmin, tset,del,dif,dif1

when(dc domain) alpold alpmax alp alpold

electrical cin1,cin2,csam,cout1,neg number imax 7.50,alpmax 112,alpmin 90, alpstp 0.50,fac 0.75,fac1 1.00,tmin 0.07, tset 0.60,del 20m,dif 0.05,dif1 0.10 number fact,alpstp1 state nu eost 0,alp,alpold,tst1 state nu newcur 0, oldcur 0,flagalp 0,volt 0 state nu fopt 0,fcur 0,flim 0,ldch 0 state nu flag 0,told 0,fsamp,del1,vref var i iout1 val nu dalp parameters fact (1.0-fac)*fac1 alpstp1 alpstp/5.00 branch vc1 v(cin1,neg) branch vc2 v(cin2,neg) branch vc3 v(csam,neg)

#DECEREMENT OF ALPHA SETTING when(ti me step done) del1 del

#INITIAL SECTION when(time

tmin & fsamp ) volt abs(vc1) alp alpold dalp alpold alp eost 0

#SOFT-START SECTION me step done & (time 0) eost

when(ti if(abs(vc1) else if((newcur if((newcur if((newcur

abs(fac*volt)) eost 0 eost 1 imax) & eost 0 & fcur fcur 1 imax) & eost 0) flagalp 0 imax) & eost 0) flagalp 1

if(flagalp 0) if( (abs(vc1) abs((1-fact)*volt))) if(fsamp 1 & flim alp alpold alpold alp flim 1

tmin) &

0)

0) dalp

#DELAY (FOR ALPHA) BLOCK if(alp when(time step done) if(time told del1 & flag 0) schedule event(time 2u, fsamp, 1) schedule event(time 25u, fsamp,0)

alpmin) alp alpmin

else alp alpold alpold alp

flag 1 told

time

if(flagalp if(eost

if(fsamp if(fsamp

1) flag 0 0) flim 0

1) alp

alpold

0) vref tst1

alp time

1050

IEEE TRANSACTIONS ON POWER ELECTRONICS, VOL. 12, NO. 6, NOVEMBER 1997

#END OF SOFT-START SECTION #DELAY BLOCK

when(time step done & eost if(time (tst1 tset)) eost 2 else eost 1 alpold vref

APPENDIX D INDUCTION MOTOR PARAMETERS AND NAMEPLATE DATA

1)

#OPTIMISATION BLOCK when(time step done & eost

2

if(newcur imax) flagalp 0 else flagalp 1 if(flagalp 0 & fopt 0) if( (newcur oldcur) dif) if(fsamp 1 & flim 0) alp alpold dalp alpold alp flim 1

alp

if (alp alpmax

alpmax)

else fopt 1 alpold alp

of(flagalp 1 & fopt 0) alp if(fopt 1) alp alpold

values if(alp (alpmax-4.0)) dalp if(alp (alpmax-4.0)) dalp alpstp1

equations i(cin1 neg) 0 i(cin2 neg) 0 i(cout1 neg) iout1 iout1:v(cout1)-v(neg) alp

alpold

alpstp

REFERENCES [1] N. Mohan, “Improvement in energy efficiency of induction motors by means of voltage control,” IEEE Trans. Power Apparatus Syst., vol. PAS-99, pp. 1466–1471, July/Aug. 1980. [2] T. M. Rowan and T. A. Lipo, “A quantitative analysis of induction motor performance improvement by SCR voltage control,” IEEE Trans. Ind. Applicat., vol. IA-19, pp. 545–553, July/Aug. 1983. [3] G. Bhuvaneswari, “An intelligent solid-state induction motor controller,” Ph.D. dissertation, Indian Instit. Technol., Madras, India, Jan. 1992. [4] R. S. J. Iyengar, “Novel techniques for energy efficient voltage controlled induction motor drives,” Ph.D. dissertation, Indian Instit. Technol., Madras, India, May 1995. [5] V. V. Sastry, P. S. Ganesh, and V. Madhavi, “High-performance induction motor controller using thyristor voltage feedback,” in Proc. Int. Conf. Power Electron., Drives, Energy Syst. Ind. Growth, New Delhi, India, Jan. 1996, pp. 44–47. [6] V. V. Sastry, P. Oomman, and K. Sundarraman, “Line voltage adaption for lightly loaded induction motors,” in Proc. Int. Conf. Power Conversion Ind. Contr., Singapore, Oct. 1986, pp. 87–91. [7] V. V. Sastry, D. Metzner, and D. Schroeder, “An integrated approach to understanding of power electronic circuits and systems,” IMACSTC1 Int. Conf. Math. Modeling, Montreal, ON, Canada, 1993, pp. 663– 667. [8] Saber Ref. Manual, Analogy Inc., Beaverton, OR, 1992.

Venkata V. Sastry (M’69–SM’82) was born in Vallaragudaba (A.P.), India, on August 4, 1941. He received the B.E. (Hons.) degree from Andhra University, Visakhapatnam, India, in 1963 and the M.Tech. and Ph.D. degrees from the Indian Institute of Technology, Kharagpur, India, in 1965 and 1968, respectively. Since 1968, he has been with the Indian Institute of Technology, Madras, in various capacities. Currently, he is a Professor of electrical engineering and the Head of the Computer Center. He visited T.U. Braunschweig and T.U. Munich, F.R. Germany, as an Alexander Van Humbolt scholar during 1972 to 1973 and again in 1992. He visited the National University of Singapore for two years from 1982 to 1984 and the University of Washington, Seattle, WA, for six months in 1993. He has published several papers and has six Indian patents. His research interests are in power electronic systems, electrical machines, modeling and simulation, real-time computer control, and computer graphics. Dr. Sastry is a Fellow of the National Academy of Engineers, New Delhi, India. He has been the recipient of six awards from NRDC, VASVIK Foundation, and IETE (India).

SASTRY et al.: OPTIMAL SOFT STARTING OF VOLTAGE-CONTROLLER-FED IM DRIVE

M. Rajendra Prasad was born in Vidyanagar (A.P.), India. He received the B.E. degree from Sri Venkateswara University, Tirupati, India, in 1993. He is currently pursuing the M.S. degree from the Indian Institute of Technology, Madras, India. Since 1994 he has been with the Indian Institute of Technology, Madras, India, as a Research Scholar. His research interests are in power electronic systems and hardware systems design for the real-time control applications.

1051

T. V. Sivakumar (S’97) was born in Karaimadai (T.N.), India. He received the B.E. degree from Bharathiar University, Coimbatore, India, and the M.Tech. degree from Anna University, Madras, India, in 1993 and 1995, respectively. He is currently pursuing the Ph.D. degree from the Indian Institute of Technology, Madras. Since 1995, he has been with the Indian Institute of Technology, Madras as a Research Scholar. His research interests are in power electronic systems, modeling, and simulation and real-time computer control.