You are on page 1of 9

IEEE TRANSACTIONS ON EDUCATION, VOL. 45, NO.

3, AUGUST 2002

253

Teaching Nonlinear Modeling, Simulation, and Control of Electronic Power Converters Using MATLAB/SIMULINK
Vitor Ferno Pires, Member, IEEE, and Jos Fernando A. Silva, Senior Member, IEEE
AbstractThis paper describes an efficient method to teach analysis and simulation of power electronic converters to undergraduate students, using system level nonlinear state-space models. System-level modeling of power electronic converters reproduces only the ideal switching behavior of the semiconductors and is a useful concept for the numerical simulation of power converters, since simulations present no convergence problems and require little computational time. Switched state-space models, programmed in the MATLAB/SIMULINK software package, can be advantageously used to simulate power converters at the system level and also to design and study their controllers. Switched state-space nonlinear models should be obtained using a theoretical framework suitable for the enhanced control of variable structure power systems. Since the method is inherently nonlinear, no approximated linear models are needed; and since state-space models are used, modern control techniques (sliding mode, neural networks, fuzzy logic) for power converters can easily be used. This paper summarizes the proposed methodology and gives some examples. Index TermsElectronic power converters, power converter control, power converter modeling, power converter simulation.

I. INTRODUCTION OMPUTER simulation plays an important role in the design, analysis, and evaluation of power electronic converters and their controllers. Designing and developing power electronic circuits without suitable computer simulation is extremely laborious, error-prone, time-consuming, and expensive. Therefore, it is essential to teach, at the undergraduate level, power converter modeling and simulation, together with the dynamic behavior of the converter, using a theoretical framework suited for controller design and development. Nowadays, a variety of software tools, such as SPICE [1][3], EMTP [4], SABER [5], CASPOC [6], SIMPLORER [7], [8], SPECTRE [9], etc., is available to simulate electrical and electronic circuits. The most used simulators are SPICE or PSPICE [10], [11], user-friendly programs designed to perform analysis of low power analog electronic circuits. Several power electronics professors [12][14] have used SPICE to simulate the behavior of power electronics converters.
Manuscript received July 23, 2001; revised February 11, 2002. V. F. Pires is with Centro de Automtica Universidade Tcnica de Lisboa, Escola Superior Tecnologia Setbal, Inst. Politcnico de Setbal, Setbal 2914-508, Portugal (e-mail: vpires@est.ips.pt). J. F. A. Silva is with Centro de Automtica Universidade Tcnica de Lisboa, Instituto Superior Tcnico, Universidade Tcnica de Lisboa, Lisboa Codex 1096, Portugal. Publisher Item Identifier S 0018-9359(02)05061-6.

SPICE simulation has, therefore, become a part of several power electronics courses. SPICE is fitted with physical models of semiconductors and switching devices. Physical modeling produces well-detailed results, but leads to high calculation effort and often introduces convergence problems during transient simulation (SPICE has 86% success rate [15]). To overcome these convergence problems, the power semiconductors are usually fitted with suitably designed snubbers. Students and teachers [13], [14], [16] are, therefore, committed to solving the convergence problems and to reducing the computing time and memory needed, both time consuming tasks. With suitable snubber circuits and device values, SPICE is able to perform the time-domain, steady-state behavior and transient analysis needed to study most power electronic circuits. However, some built-in capabilities of SPICE, such as dc operating point and small signal ac frequency analysis, are not usable for power converters. SPICE ac frequency analysis does not provide the dynamic transfer functions of power converters, because of the switching operation of the power semiconductors. Therefore, two simulation packages are often needed [17]. Furthermore, because of the huge amount of memory and CPU time required to converge problems, SPICE cannot easily handle the usually long, transient analysis behavior of complex switching converter topologies [17], [18]. Since most physical details are not needed in the study of power electronics, some well-known authors [18] have used macro models of the semiconductor switching devices to improve convergence and increase simulation speed. The authors have been using SPICE as an instructional tool to teach power converters at an introductory level. The success rate in examinations shows that student ability to understand the behavior of several power electronics converters have been increased. Despite this success, the continual use of SPICE simulations leads to a loss of student commitment and ability to do the mathematical modeling needed for more advanced courses, such as control of power converters. Students concentrate on the skills needed to obtain convergence of the solutions with closed-loop control and tend to use only well established converter controllers. They often use trial and error methods to design high-performance controllers for power converters. Therefore, undergraduate students should also learn some nonlinear mathematical modeling and converter control. Additionally, students should take advantage of the mathematical models they have derived to perform the simulation of the power electronic converters with closed-loop control.

0018-9359/02$17.00 2002 IEEE

254

IEEE TRANSACTIONS ON EDUCATION, VOL. 45, NO. 3, AUGUST 2002

To reach this goal, students should be able to analyze power semiconductor circuits using classic network theory without going through complicated mathematics. The derived models should be suitable for implementing a computer simulation of the power converter. This accomplishment can be a powerful motivation for the somewhat tedious mathematical modeling work, because students tend to prefer the automatic use of the computer, encouraged by its well-known relevance as a powerful tool for research and development. MATLAB/SIMULINK software [19] is widely used for the simulation of almost all types of dynamic systems. This software package is also valuable for teaching and learning since it provides a series of standard routines and software toolboxes, such as a control toolbox, system identification blocks, nonlinear control design block set, and neural networks block set, which enable students to perform system simulation, identification, and control. The latest versions of MATLAB/SIMULINK include a Power System Blockset [20]. This toolbox features electrical models of power semiconductors and the most commonly used power devices (machines, transformers, power lines, voltage sources), and allows simulation of power systems and power electronics. This package is valuable for simulating well-known topologies several of which are included as demonstrations, but it tends to generate too many algebraic loops on more complex or novel power topologies. These algebraic loops are difficult to handle (because they are inherent to the modeling method) and are time consuming, often preventing simulation convergence. Furthermore, this toolbox does not easily allow open-loop or closed-loop simulation of series associations of power rectifiers, nor does it study the steady and the transient-states in cases of unbalanced or distorted and/or polluted power supply. Considering the approach of [18] with PSPICE and SIMPLORER, the authors think that a system-level simulation, considering only the ideal switching and functional behavior of power semiconductors, would be desirable for MATLAB/SIMULINK. The system-level simulation is fast enough and free of algebraic loops and convergence problems (SIMULINK has built-in integration methods suited to deal with stiff systems). Therefore, it could avoid the problems of the Power System Blockset mentioned above. Additionally, the system-level derived models to implement in SIMULINK can be used for closed-loop controller design, since they are switched state-space models. This advantage is lost when using the Power System Blockset or SIMPLORER. Considering the increasing capabilities of MATLAB/SIMULINK for the simulation of dynamic systems [21], it is advantageous to adapt the ideal models of semiconductors and simulation methods presented here for this software since only one software package is needed. The simulation time is short (a few seconds); an excellent graphical interface is available with parametric identification of the system and the ability to choose the numerical integration method and toolboxes for closed-loop control. In addition, the SIMULINK package offers the benefits of a hierarchical structure and uses MATLAB as its mathematical engine. If required, the modeling method here proposed could be adapted to other programs. Since the goal is to teach nonlinear

mathematical modeling and control and the simulation of power converters, this paper shows, in Section II, how to write system-level models of power electronics circuits. In Section III, examples of pulse width modulation (PWM) ac/dc and dc/ac power electronic converters are given. The simulation models described are quite suitable to study power electronics converters in drives or other applications whose simulation times are not too long, since only the ideal behavior of the power switches is considered. This work was initially developed for research in the area of new topologies for power electronics. However, further developments allowed its use as a valuable teaching aid. Therefore, this work presents a new way to teach undergraduate students the dynamic behavior of power electronics circuits without cutting down the analytic skills needed to learn and synthesize power converter controllers. The new method can also be used as verification of analytical methods, allowing students to check their mathematical work quickly and use it for power converter behavior and controller development. II. MODELING PROCEDURE To obtain a nonlinear model easily for power electronic circuits, students need only to apply Kirchhoffs circuit laws. To avoid the use of complex mathematics, the electrical and semiconductor devices must be represented as ideal components (zero ON voltages, zero OFF currents, zero switching times). Therefore, auxiliary binary variables can be used to determine the state of the switches. Then, students must ensure that the equations obtained by the use of Kirchhoffs laws should include all the permissible states due to power semiconductor devices being ON or OFF. The steps to obtain a system-level modeling and simulation of power electronic converters are listed below. The two last steps obtain the controller(s) and perform closed-loop simulations. 1) Determine the state variables of the power circuit in order to write its switched state-space model [22]. 2) Assign integer variables to the power semiconductor (or to each switching cell) ON and OFF states (generally, if the circuit includes power switches, it needs 2 topologies minus those forbidden by topological restrictions). 3) Determine the conditions governing the states of the power semiconductors or the switching cell. 4) Assume the main operating modes of the converter (continuous or discontinuous conduction or both) or the modes needed to describe all the possible circuit operational modes. Then, apply Kirchhoffs laws and combine all the required stages into a switched state-space model, which is the desired system-level model. 5) Write this model in the integral form, or transform the differential form to include the semiconductors logical variables in the control vector: the converter will be represented by a set of nonlinear differential equations. 6) Implement the derived equations with SIMULINK blocks (open-loop system simulation is then possible to check the obtained model). 7) Use the obtained switched space-state model to design linear or nonlinear controllers for the power converter.

PIRES AND SILVA: TEACHING NONLINEAR MODELING, SIMULATION, AND CONTROL OF ELECTRONIC POWER CONVERTERS

255

Fig. 1.

Buck dc/dc regulator (first quadrant chopper).

8) Perform closed-loop simulations and evaluate converter performance. III. EXAMPLES A. Buck DC/DC Regulator Consider the buck regulator of Fig. 1. Students should be able to preview the continuous and discontinuous conduction modes of operation. Using the procedure outlined in Section II, the modeling steps are the following. and Step1) The state variables of the Fig. 1 chopper are . can be represented Step2) The states of the power switch defined as by a switching variable if if is driven ON AND is OFF. (1)

(a)

(b) Fig. 2. (a) Block diagram to model the i current, v and v voltages of the circuit of Fig. 1; (b) Block diagram showing the control method, the converter load and output devices.

Step3) To model operation in the continuous and discontinuous conduction modes, the states of the power diode can be described by the binary variable if if Therefore, the AND AND voltage (Fig. 1) is . (2)

(3) Step4) Applying Kirchhoffs laws to Fig. 1, students can write (4) and (5) , the inductor current and Step5) For a resistor load output voltage can be obtained by integrating (4) and (5) over time (6) and (7) Step6) Using MATLAB/SIMULINK, the subsystem describing the behavior of the current (6) and voltage

(7) is shown in Fig. 5(a). The voltage and are obtained using respectively (3) and (1). Step7) A simple control method [Fig. 2(b)] is used to set the is driven on by the input current. The power switch clock signal (clocked turn-on) and driven off when the current exceeds a defined value (peak-current is a step from 3 A to turn-off). This defined value 0.4 ms. 9 A at Step8) The algorithm for solving the differential equations and the step size should be chosen before running any simulation. This example considers a 10-kHz switching frequency and 100 time steps per period. The RungeKutta 45 algorithm, with a step size of 10 s, has been selected. Simulations of the dc/dc Buck 5 , 0, 2 mH, converter were made using 0, 5 mF, 50 V, 15 V, 0 A. Students can easily change the component values and verify what happens to the voltages and currents. Fig. 3(a) shows the current, which has a step from 3 A to 9 A at 0.4 ms, the current and the Clk clock pulses. Fig. 3(b) shows the and voltages. From 0 to 0.4 ms operation is in the equals when the discontinuous conduction mode, and current is zero. From 0.4 ms to 1 ms operation is zero when is in the continuous conduction mode, and switch Q1 is off. The converter simulations and graphics display took only 10 s on a 166-MHz Pentium. B. AC/DC Single Phase Converter Single-phase pulse-width modulated (PWM) voltagesourced, switch-mode rectifiers (Fig. 4) are used in industrial applications, such as uninterruptible power supply (UPS) systems and ac/dc power supplies for telecommunications equipment. They are capable of bidirectional power flow, operation with near unity power factor, and input current with

256

IEEE TRANSACTIONS ON EDUCATION, VOL. 45, NO. 3, AUGUST 2002

can be represented by the switching variables ) defined as ( if OR is ON AND AND are OFF AND are OFF AND OR is ON.

(8)

0. Step3) Students should note that the diodes impose Therefore, the conditions governing the states of the power semiconductors considered in (8), are related to switch drive signals and to diode bias voltages and currents. Therefore, according to (8)
(a)

is driven ON AND if OR if OR and is driven OFF AND if OR if then

is driven OFF then is driven ON (9)

Step4) Applying Kirchhoffs laws to Fig. 4, students can easily write (10) and (11) and
(b) Fig. 3. Main waveforms of the buck converter; (a) i the Clk clock pulses; (b) the v and v voltages. and i currents and

(12) where the switching function is given by (13). This definition assumes continuous conduction mode in the boost mode, and two-level operation, typical of voltage source PWM reversible rectifiers if if AND AND . (13)

(11) and current (12), Considering the voltage again applying Kirchhoffs laws, students can easily obtain the switched, state-space equations of the PWM rectifier (14)
Fig. 4. Single-phase voltage source switch-mode rectifier.

and (15)

low harmonic content, and they can behave as power factor preregulators. Therefore, it is important for most students of electrical engineering to study the behavior of these power converters alone or with their controllers. Using the procedure outlined in Section II, the modeling steps are: Step1) The state variables of the rectifier of Fig. 4 are chosen to be and . Step2) To avoid shorting the output capacitor (and destroying the power semiconductors), the two switches of each bridge leg must be in complementary states (topological restriction). Therefore, the semiconductors are grouped into two switching cells, whose states

Step5) For a resistor load and output voltage (14) and (15)

, the converter input current can be obtained by integrating

(16) and (17) Step6) Using MATLAB/SIMULINK, the subsystem describing the behavior of the boost inductor current (16) is shown in detail in Fig. 5. This subsystem Single-Phase

PIRES AND SILVA: TEACHING NONLINEAR MODELING, SIMULATION, AND CONTROL OF ELECTRONIC POWER CONVERTERS

257

Fig. 5. Block diagram describing the behavior of the input current of the single phase ac/dc converter.

Fig. 7. Block diagram of the input current controller.

Fig. 6. Block diagram for the model of the circuit of Fig. 4.

ac/dc Converter, with the inductor ( , ) in the source line is then included in the implementation of equation (17) as shown in Fig. 6. The bounded integrator ensures a positive . The figure also presents a closed-loop controller for the input current ( ) (current controller), which is explained in the next step. Step7) Sliding mode [23], [24] is used to control the input current . Equation (14), written in the controllability canonical form, shows the first order dependence of on the switching function . Therefore, a sliding surface , able to enforce the control goal and also provide robustness and system-order reduction, is (18) ) and also Using sliding mode stability ( considering (18) (14), the control law (19) is derived. It defines , which is the switching function to obtain a sinusoidal wave in phase with the input voltage. This action achieves a near unity power factor. As power switches (e.g., MOS power transistors, or insulated gate bipolar transistors) present upper bounds for their switching frequency, a positive are designed to achieve constant and the current ripple 2 switching frequencies suitable for the power semiconductors used S1 S4 are driven ON and S2 S3 are driven OFF and S2 S3 are driven ON (19) and S1 S4 are driven OFF The approach considered here is similar to current mode control, but a PWM modulator is not needed. The semiconductors are driven ON or OFF just when needed at variable switching frequency.

(a)

(b) Fig. 8. (a): Reference for the input current (i scaled input voltage (V =15). ). (b) Input current (i ) and

The current controller subsystem (Fig. 7) outputs the switching function . The block product generates the . A step is used to change the phase sinusoidal reference of this reference [Fig. 8(a)], enabling power factors close to 1 before the step and close to 1 (for inverter operation) after the step. Block sum_3 implements (18), and block together with Relay3 (with hysteresis width 2 ) implement (19). The derived input current controller is encased in the rectifier model (Fig. 6). The use of subsystems is encouraged, enabling alternative controller testing by simple replacement of the existing controller with a new block, whose output must be the switching variable . The study of the rectifier with three-level 1 0 1 . operation is possible just by allowing Step8) Simulations will allow students to verify that this controller draws input current with sinusoidal shape, also presenting near unity power factor. Operation in the inverter mode is also studied. Prior to running the simula-

258

IEEE TRANSACTIONS ON EDUCATION, VOL. 45, NO. 3, AUGUST 2002

tions, the algorithm for solving the differential equations and the step size must be chosen. For this example, considering switching frequencies around 5 kHz and about 20 time steps per period, the DormandPrince (ode5) algorithm, with a step size of 10 s, has been selected. Simulations of the single-phase ac/dc converter were 100 , 10 000 F, 10 made using 0.01 , 220 V, 0 220 V mH, 0 A. and 0 Students can easily change the component values and verify what happens to the voltages and currents. Fig. 8(a) shows the , which is designed to be in phase with the input voltage (50-Hz sinusoid) during the first 30 ms, but to be in opposite phase during the next 30 ms. This situation allows testing the current controller, as the rectifier operates in the rectifier region and then in the inverter region, if the current controller is fast enough. Fig. 8(b) confirms that the controller performs with the required speed as the input current follows the reference, except for the unavoidable ripple. The rectifier input current is almost sinusoidal in shape and is in phase, or is 180 out of phase, with the voltage source, providing power factors close to 1 or 1, respectively [Fig. 8(b)]. Students can note the rectifier or inverter operation at unity power factor and can easily confirm these results in laboratory classes. The converter simulations and graphics display took 12 s on a 166-MHz Pentium, which is fast enough to be used in teaching classes. Recent computers, running at 1 GHz or more, can cut down this time by at least a factor of 5. Proposed Exercises: I) Use a sinusoidal PWM modulator with a triangular carrier to perform open-loop simulations. Design an average current mode controller and compare the performance obtained with that presented in Fig. 8(b). II) Derive a three-level control by rewriting equation (13). Use two relay blocks (Fig. 7) to implement the new equation. Repeat simulations and compare performance. C. AC/DC Three-Phase Converter Three-phase PWM voltage-source converters (Fig. 9) are suited to high power industrial applications and capable of bidirectional power flow, near unity power factor operation. They can behave as power factor compensators, and the AC currents can be almost sinusoidal waveforms with low harmonic content. The converter model can be obtained using the concepts already presented in the previous example. and DC voltage Step 1) Consider the AC currents (Fig. 9) as the state variables. Step 2), 3) The switching function , which represents the states of the switches, can be defined according to (20). This definition prevents internal shorts in the voltage source switch-mode converter, ensuring that the two switches of each bridge leg are always in complementary states (topological restriction) if if is ON or is ON is ON or is ON (20)

Fig. 9.

Three-phase voltage source switch-mode converter.

Step 4) Students can verify that, from Kirchhoffs laws, this power converter can be described by the following state-space 1 2 3): equations ( (21a) (21b)

with

(22)

are functions of the output voltage ( ), where the voltages and the state of the switches (variable ), according to (23) (23) Now applying steps 5) through 6), the SIMULINK block diagram is obtained (Fig. 10). The block diagram includes the AC current controller and two input voltages (only two of the three-phase variables are independent in this three-wire converter). Step 7) To obtain sinusoidal ac currents, a sliding mode controller can be used. For simplicity, one switching function for each converter leg should be defined according to (24), as in (19) if if (24) 100 , 10 000 Simulations were made using F, 10 mH, 0.01 , and 220 V, with integration settings and initial conditions similar to those of the previous example. Again, the ac current reference [Fig. 11(a)] is designed to show rectifier and inverter operation (converter reversibility). Students can verify that ac currents track the reference [Fig. 11(b)], and are in phase, or 180 out of phase, relative to the corresponding input voltage [Fig. 11(b)]. Power factors close to 1 or 1 are achieved. It is also verified that the power flows [Fig. 11(c) and (d)] to the ac source voltages, when

PIRES AND SILVA: TEACHING NONLINEAR MODELING, SIMULATION, AND CONTROL OF ELECTRONIC POWER CONVERTERS

259

Fig. 10.

Block diagram for the model of the three-phase ac/dc converter.

Fig. 11. (a) Reference waveform for the ac current i . For t 0.05 s the converter will operate in the inverter mode. (b) Current i and voltage source V =15. (c) Three-phase ac currents showing the transition from rectifier to inverter at t 0.05 s. (d) Output voltage V , showing that, in the inverter mode (t 0.05 s), the capacitor is delivering energy to the three-phase ac system. (e) Converter leg voltage u . (f) Line-to-line output voltage V .

the current reference is reversed, since the capacitor voltage decreases. The simulation time was about 35 s using a 166-MHz Pentium, which is fast enough for the classroom. These examples highlight a simple method to integrate the teaching of power converter modeling, simulation, and control. More exercises can be proposed to derive other controllers. Changing the controller is straightforward: replace the controller block with the new design, provided that it outputs 1 or 0 to drive the switches ON or OFF. For instance, a sinusoidal PWM modulator can be used to control the output voltage, and the capacitor value can be increased to store enough energy to allow almost constant dc voltage supply in the inverter mode. Then, for a dc voltage of 400 V the converter leg

is plotted in Fig. 11(e) and the line-to-line voltage , is shown in Fig. 11(f). The presence of two and three voltage levels in these voltages is clearly illustrated. Validation of the simulations in the laboratory is a very important step since it enables students to gain or enhance insight in the simulation method and, above all, it increases experience in handling the problems of real power converters. Proposed Exercises: III) Repeat exercise I) of the previous example for this converter. Compare the result with that presented in Figs. 11(e) and 8(f). IV) Evaluate the ac current ripple and compare it with the design value. Explain the differences.

voltage

260

IEEE TRANSACTIONS ON EDUCATION, VOL. 45, NO. 3, AUGUST 2002

Fig. 12.

Association of two three-phase converters connected back to back.

Fig. 13.

SIMULINK block diagram for the association of two three-phase converters connected back to back.

V)

Design a space vector current controller, applying the Concordia transformation to the ac currents and plotplane. Evalting the converter leg voltages in the uate and compare the obtained current ripple.

D. Extension to Associations of Electronic Power Converters One of the main areas of power electronics is the study of converters to supply ac machines from the mains. Two three-phase converters connected back-to-back (Fig. 12) enable operation at variable torque and/or speed and unity power factor. This association is advantageous since, removing the high frequency ac current harmonics, the mains seems to supply a pure resistor. This scheme can be also used in UPS. For these applications, the line-side converter operates in rectifier mode (ac/dc converter), while the load-side converter operates in inverter mode (dc/ac converter), supplied by the dc link capacitor voltage. Fig. 13 shows the block diagram used in the MATLAB/SIMULINK program to simulate the circuit presented in Fig. 12. In this example both converters feature current controllers. Fig. 14 shows the output currents of the dc/ac converter. Students can then be encouraged to design the dc voltage controller and ac output voltages, as well as torque and speed controllers.

Fig. 14.

Three-phase output currents of the dc/ac converter.

IV. CONCLUSION This paper has outlined and illustrated a method to obtain nonlinear, switched, state-space models of power converters, suited for simulation and control design. As the methodology uses state-space models, electronic power converters, associations of power converters, and electromechanical devices or drives, with elaborate control systems, can be effectively simulated. Associations of electromechanical systems and/or electronic power converters, which are often complex and difficult to analyze and to converge, due to stiffness, can also be studied. Simulation times are a few tens of seconds,

PIRES AND SILVA: TEACHING NONLINEAR MODELING, SIMULATION, AND CONTROL OF ELECTRONIC POWER CONVERTERS

261

and no convergence problems were found. The somewhat tedious work needed to obtain the switched, state-space model of the converter, which is a theoretical frame necessary for controller design, can be advantageously used to perform the simulation. Therefore, this methodology is an effective tool to teach undergraduate students to simulate electronic power converters dynamics and their control design. Using the MATLAB/SIMULINK software package, students or power converter control engineers are provided with a powerful simulation and control tool for power converters. Several ac/dc and dc/ac converters have been successfully simulated. The examples given illustrate the considerable potential of the presented techniques as teaching aids. REFERENCES
[1] Berkeley University. (2001, Sept. 6) SPICE MANUALS [Online]. Available: from bwrc.eecs.berkeley.edu/Classes/IcBook/SPICE/MANUALS [2] University of Pennsylvania, Department of Electrical Engineering. (2001, Sept. 6) SPICEA Brief Overview, SPICE Users Guide [Online]. Available: http://www.seas.upenn.edu/~jan/spice/spice.overview.html [3] Intusoft. (2001, Sept. 5) Better design through simulation. [Online]. Available: http://www.intusoft.com/. [4] ATP. (2001, Sept. 5) Alternative transients program. [Online]. Available: http://www.emtp.org/. [5] SABER. (2001, Sept. 5) Saber mixed-signal simulator. [Online]. Available: http://www.emtp.org/http://www.analogy.com/products/simulation/simulation.htm#Saber. [6] CASPOC. (2001, September) CASPOC 2001: A simulation odyssey. [Online]. Available: http://www.caspoc.com/ [7] SIMEC Corporation. (2001, September) Simplorer simulation system. [Online]. Available: http://www.simplorer.com. [8] Ansoft. (2001, September 6) SIMPLORER. [Online]. Available: www.simec.com. [9] Cadence. (2001, September 6) SPECTRE circuit simulator. [Online]. Available: http://www.cadence.com/datasheets/spectre_cir_sim.html [10] MicroSim Corporation, PSpice Reference Manual, 1991. [11] Pspice.com. (2001, September 6) Design community in action. [Online]. Available: http://www.pspice.com/. [12] M. H. Rashid, SPICE for Circuits and Electronics Using PSpice. Englewood Cliffs, NJ: Prentice-Hall, 1990. [13] N. Mohan, Power electronics: Computer simulation, analysis, and education using PSpice, Minnesota Power Electronics Research & Education, 1992. [14] M. H. Rashid, Power Electronics Circuits, Devices, and Applications. Englewood Cliffs, NJ: Prentice-Hall, 1993. [15] Veribest. (2000, September) VERIBEST circuit simulator. [Online]. Available: http://www.veribest.com/sales/ads/analog/convergence.html [16] D. W. Hart, Circuit simulation as an aid in teaching the principles of power electronics, IEEE Trans. Educ., vol. 36, pp. 1016, Feb. 1993.

[17] [18] [19] [20] [21] [22] [23] [24]

, A software tool for learning the dynamic behavior of power electronics circuits, IEEE Trans. Educ., vol. 39, pp. 5055, Feb. 1996. O. Apeldoorn, S. Schroder, and R. De Doncker, A new method for Power electronics system-simulation with Pspice, in Proc. ISIE97 Conf., Guimares, Portugal, July 1997, pp. 217222. Math Works Inc., Matlab Users Guide, 1997. Mathworks. (2000, Sept.) Power system blockset 2.1 release notes. [Online]. Available: http://www.mathworks.com/access/helpdesk/help/techdoc/relnotes/powersys.shtml D. Graovac and V. Katic, Modeling and simulation of AC/DC converter in nonstandard supply conditions, in ELECTRIMACS96, Sept. 1996, pp. 547550. J. F. Silva, Control methods for power converters, in Power Electronics Handbook, M. H. Rashid, Ed. New York: Academic, 2001, ch. 19, to be published. V. Utkin, Sliding mode control design principles and applications to electric drives, IEEE Trans. Ind. Electron., vol. 40, pp. 2326, Feb. 1993. W. Gao and J. Hung, Variable structure control: A survey, IEEE Trans. Ind. Electron., vol. 40, pp. 222, Feb. 1993.

Vitor Ferno Pires (M96) received the B.S. degree in electrical engineering from Institute Superior of Engineering, Lisbon, Portugal, and the M.S. and Ph.D. degrees in electrical and computer engineering from Technical University of Lisbon (UTL), Portugal, in 1988, 1995, and 2000, respectively. Since 1991, he has been a member of the teaching staff at Electrical Engineering Department, Superior Technical School of SetbalPolytechnic Institute of Setbal, Setbal, Portugal. Currently, he is a Professor, teaching power electronics and control of power converters. He is also a Researcher at Centro de Automtica, UTL. His present research interests include the areas of low-distortion rectifier topologies, converter control, modeling, and simulation.

Jos Fernando A. Silva (M91SM00) born in Mono Portugal in 1956, received the Dipl.Ing. in electrical engineering and the Ph.D. degree in electrical and computer engineering (power electronics and control), from Institute Superior of Engineering (IST), Technical University of Lisbon (UTL), Lisbon, Portugal, in 1980 and 1990, respectively. He is currently an Associate Professor of Power Electronics at IST, teaching power electronics and control of power converters and a Researcher at Centro de Automtica, UTL. His main research interests include power semiconductor devices, modeling and simulation, new converter topologies, and sliding mode control of power converters. He has authored or coauthored one book, one book chapter, and over one hundred papers.

You might also like