FPGA‐based Hardware‐in‐the‐loop Real‐time Simulation Implementation for High‐speed Train Electrical Traction System
Xizheng Guo,Yiguo Tang,Mingkang Wu,Ziyu Zhang,Jiaqi Yuan
DOI: https://doi.org/10.1049/iet-epa.2019.0655
2020-01-01
IET Electric Power Applications
Abstract:IET Electric Power ApplicationsVolume 14, Issue 5 p. 850-858 Research ArticleFree Access FPGA-based hardware-in-the-loop real-time simulation implementation for high-speed train electrical traction system Xizheng Guo, Corresponding Author Xizheng Guo xzhguo@bjtu.edu.cn School of Electrical Engineering, Beijing Jiaotong University, Beijing, 100044 People's Republic of ChinaSearch for more papers by this authorYiguo Tang, Yiguo Tang School of Electrical Engineering, Beijing Jiaotong University, Beijing, 100044 People's Republic of ChinaSearch for more papers by this authorMingkang Wu, Mingkang Wu School of Electrical Engineering, Beijing Jiaotong University, Beijing, 100044 People's Republic of China Electric Power Research Institute, China Southern Power Grid, Guangzhou, 510663 People's Republic of ChinaSearch for more papers by this authorZiyu Zhang, Ziyu Zhang School of Electrical Engineering, Beijing Jiaotong University, Beijing, 100044 People's Republic of ChinaSearch for more papers by this authorJiaqi Yuan, Jiaqi Yuan School of Electrical Engineering, Beijing Jiaotong University, Beijing, 100044 People's Republic of ChinaSearch for more papers by this author Xizheng Guo, Corresponding Author Xizheng Guo xzhguo@bjtu.edu.cn School of Electrical Engineering, Beijing Jiaotong University, Beijing, 100044 People's Republic of ChinaSearch for more papers by this authorYiguo Tang, Yiguo Tang School of Electrical Engineering, Beijing Jiaotong University, Beijing, 100044 People's Republic of ChinaSearch for more papers by this authorMingkang Wu, Mingkang Wu School of Electrical Engineering, Beijing Jiaotong University, Beijing, 100044 People's Republic of China Electric Power Research Institute, China Southern Power Grid, Guangzhou, 510663 People's Republic of ChinaSearch for more papers by this authorZiyu Zhang, Ziyu Zhang School of Electrical Engineering, Beijing Jiaotong University, Beijing, 100044 People's Republic of ChinaSearch for more papers by this authorJiaqi Yuan, Jiaqi Yuan School of Electrical Engineering, Beijing Jiaotong University, Beijing, 100044 People's Republic of ChinaSearch for more papers by this author First published: 09 March 2020 https://doi.org/10.1049/iet-epa.2019.0655Citations: 3AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onFacebookTwitterLinkedInRedditWechat Abstract The hardware-in-the-loop (HIL) real-time simulation for high-speed train electrical traction system aims to reduce the design cost and speed up control verification process of algorithms in the developmental stage of the traction control unit. In this study, based on the dSPACE real-time simulator, the multiple-simulator, multiple-simulation step of HIL real-time simulation system is first built. Second, for the associated discrete circuit modelling method, an optimisation method is proposed to minimise the switching loss and improve the simulation accuracy by selecting the optimal discrete-time switch admittance parameter, GS. To decrease the computational burden for field-programmable gate array (FPGA), a decoupling method without simulation latency is presented to reduce the matrix dimension of the system model. Finally, the real-time simulation models of electrical traction system are realised by CPU + FPGA-based simulator, in which the power electronics converter models are computed in FPGA with a fixed 100 ns time-step. The validity and reliability of the real-time simulation system is verified by the HIL simulation and experimental results, which indicate that the real-time HIL simulation at the nanosecond level improves the accuracy essentially. 1 Introduction The electrical traction system of the high-speed train includes two subsystems: the inverter fed induction motor drive system; the auxiliary power supply system provides AC 380 V/50 Hz supply to the train auxiliary loads. In the initial development process of control unit for high-speed train electrical traction system and auxiliary power supply system, it is improper to validate the designed traction control unit (TCU) and auxiliary control unit (ACU) in real system without eliminating the software and hardware errors, and also some fault and maximum conditions cannot be conducted in laboratory. Therefore, the hardware-in-the-loop (HIL) real-time simulation technology provides an alternative means for these problems and can reduce cost, time and risk substantially [1-3]. For the HIL real-time simulation, the real-time simulator, in which the real-time simulation model of power electronics converter (PEC) is implemented, and the real TCU/ACU constitute the closed-loop testing system, as shown in Fig. 1. Fig. 1Open in figure viewerPowerPoint High-speed train electrical traction system (a) Real test system, (b) HIL simulation The establishment of the real-time simulation model of the traction drive system and the auxiliary power supply system is an important part of the whole HIL system. Traditional CPU-based HIL simulators, where the time step is usually in 40–100 μs range, can hardly take into accounts the entire switching event from pulse width modulation (PWM) in modern power electronics systems, so the multiple switching events within one fixed simulation time step h may produce erroneous simulation results [4-7]. The high parallelism provided by field-programmable gate array (FPGA) can compute the real-time simulation model from microsecond to nanosecond range, which is suitable for PEC with high-switching frequency [8-10]. On the other hand, the modelling method of PEC in FPGA is a challenging task. The state–space modelling method analyses all possible PEC network equations under different switching combinations step by step. However, the possible circuit topology will increase exponentially with the increasing number of switches, which brings great challenges for the model implementation in FPGA [11-13]. The voltage-controlled current source method is highly accurate, but it requires additional iterative calculations which lead to higher computation effort, so it is not suitable for FPGA-based real-time simulation [14-16]. Due to the fixed network equations, the associate discrete circuit (ADC) method improves the simulation efficiency [17-19]. However, this method introduces artificial transients when the switching state changes, which results in additional losses compared to the ideal switching modelling method. Selecting the optimal switch admittance parameter can partially solve this problem. The common method for selecting the optimal switch admittance parameter is to consider a priori switch admittance parameters and then compare them with the benchmark test results to find the corresponding optimal value. However, the efficiency of this trial and error method is very low [20, 21]. This paper is organised as follows: Section 2 presents the structure of multi-simulator real-time simulation system for the high-speed train electrical traction system. Section 3 proposes the optimal discrete-time switch admittance parameter selection method, which is verified by the offline simulation of the three-phase inverter. This method combined with the decoupling techniques is applied to the high-speed train electrical traction system. FPGA implementation process and HIL real-time simulation and experimental results are described and discussed in Section 4. Finally, Section 5 gives the conclusion. 2 Structure of multi-machine real-time simulation for high-speed train electrical traction system Fig. 2 shows the circuit topology of the high-speed train electrical traction system. The primary winding of the traction transformer connects the single-phase 25 kV/50 Hz AC power grid through the pantograph. The two secondary windings of the transformer connect the double four-quadrant converter (4QC), which is controlled by interleaved pulses to reduce the harmonics injection into the power grid and provide DC power supply to the two traction inverters. The main parameters are shown in Table 1. Table 1. Parameters of high-speed train electric traction system Parameters Unit Value AC power grid, us kV 25 traction transformer secondary side-rated voltage, es1, es2 kV 1.77 intermediate DC circuit-rated voltage, udc kV 3.6 DC–DC converter-rated output voltage, udc_aux V 600 DC bus support capacitor, Cout mF 9 traction transformer secondary side leak inductance, Ls mH 2.56 braking resistor, REO Ω 6.5–9 traction induction motor-rated power, P kW 568 4QC and traction inverter switching frequency Hz 450 Fig. 2Open in figure viewerPowerPoint Circuit topology of the high-speed train electrical traction system (a) Main circuit of electrical traction system, (b) Division model of traction system The auxiliary DC–DC converter working on soft-switching mode and converting the high voltage (DC 3600 V) to a common level (DC 600 V) output is also called as half-bridge zero-current switching (ZCS) PWM DC–DC converter. The auxiliary inverter with LC filter outputs AC 380 V/50 Hz for the auxiliary load supply. The main parameters are shown in Table 2. Table 2. Parameters of auxiliary converter Parameters Unit Value SPS solver none ODE4(Runge–Kutta) simulation step ns 100 half bridge IGBT switching frequency kHz 4 resonant IGBT switching frequency kHz 8 input voltage V 1800 transformer ratio — 0.6 transformer leakage inductance (converted to the secondary side), Lk μH 1.27 input filter, C1–C4 μF 1200 resonant capacitor, Ca μF 4.5 DC filter inductance, Lo mH 1.0 low voltage DC support capacitor, Co mF 0.98 output filter inductance, Lauc mH 0.21 In order to simulate the real working environment for the TCU and ACU, the whole system model is divided into three subsystems: traction drive system, auxiliary converter system, traction power supply system starting sequential logic, as shown in Fig. 2b, which are implemented in two simulators, respectively. The system employs dSPACE as the real-time simulator, which mainly includes: (i) the DS1006 processor board, which is based on the AMD quad-core processor. It is responsible for data communication between different simulators which send measurement data to the workstation and calculate complex, precisely detailed simulation models; (ii) the FPGA board DS5203 of Xilinx Virtex-5, which offers 6 DAC and 16 digital high-resolution I/O channels. The voltage and current signals from the PEC models are sent to control units through the DAC channels, and the control PWM signal is identified by the high resolution I/O channels; (iii) other boards, such as digital I/O board DS4004 that simulates the relay status, the resistive sensor simulation board DS2401 that makes it possible to simulate resistive sensors. The division of the real-time simulation model is shown in Fig. 3. The functions of two real-time simulators are defined as follows: simulator 1 is aimed at implementing MASTER and SLAVE_D system, which is operated in the different core of the multi-core processor. Simulator 2 is responsible for auxiliary converter system, which is set up to the SLAVE system. The real-time simulation data exchanging of different simulators is via the fibre interface, which can achieve 1.25 Gbit/s communication rate. Fig. 3Open in figure viewerPowerPoint Structure of multi-simulator real-time simulation system based on dSPACE According to the different dynamic characteristics of the system, the simulation model is carried out under different simulation time steps. For the traction system as the example, traction power supply system starting sequential logic, motor motion equation, traction transformer model and so on, which are not strict with the simulation time step, are processed in the CPU board with a low-speed subsystem (simulation time step h1 is 60 μs). The simulation model of PEC is calculated in FPGA board with the high-speed (simulation time step h2 is 100 ns). The data exchange of the two different time-step systems is through the peripheral high-speed bus, as shown in Fig. 4. Therefore, the whole real-time simulation model for high-speed train electrical traction system is built in the multi-simulator and multi-step mechanism, which fully utilises the FPGA hardware resources of the real-time simulators. Fig. 4Open in figure viewerPowerPoint Implementation of dual time-scale real-time simulation system 3 Improved ADC modelling method 3.1 Associated discrete circuit modelling method The power electronics switch in PEC can be equivalent to a small inductance L when the switch is turned ON and a small capacitor C with damping resistor R when the switch is turned OFF, as shown in Fig. 5. This method guarantees a constant system matrix by modelling each switch as a conductance (denoted as GS) in parallel with a current source (denoted as jS), and their values change with the discretisation method. Using the backward Euler method, the companion circuit is as shown below (1) (2) Throughout this paper, the letter n + 1 denotes the next simulation time step and n denotes the current simulation time step, h denotes the simulation time step. The system matrix can be obtained by the modified nodal approach (MNA) [22]. (3) where H is the system matrix, x is the node voltage and the branch current vector and b is the input vector of known current injections including. Since H is a constant matrix, H−1 can be pre-calculated and stored in the memory, which reduces computational complexity and increases simulation speed. Fig. 5Open in figure viewerPowerPoint ADC switch model The switching status of the power electronics switch has two states: ON or OFF. However, for the active switching devices such as IGBT which contains anti-parallel diode, when the external switching logic signal, denoted as S, is controlled as OFF, the switching status is determined by the voltage and branch current. A new state machine is proposed in Fig. 6, in which the switch state transition can be easily implemented in FPGA. Fig. 6Open in figure viewerPowerPoint Switch state transition (a) Power diode, (b) IGBT 3.2 Selection of switch admittance parameter GS value for minimum switching loss The switch representation of ADC modelling method introduces artificial transients, which causes additional loss compared to the ideal switching modelling method and reduces the simulation accuracy. In this section, taking one bridge arm of the auxiliary three-phase inverter as an example to illustrate the optimum GS selection method, as shown in Fig. 7. In Fig. 7, sw1 and sw2 indicate the upper and lower switching devices of the single-phase bridge arm. At the n step, sw1 is in OFF state, the switch current and voltage are (4) (5) The switch is equivalent to a capacitor, and the energy stored on the capacitor is (6) The energy stored on the capacitor exists in the form of a parallel current source (7) In the n + 1 step, the switch sw1 is turned ON and is equivalent to an inductor, the current value of the parallel current source is (8) From (6)–(8), we can see that the when the switch sw1 is turned ON, the energy stored in the capacitor is ignored and is replaced by an inductor without any energy. Fig. 7Open in figure viewerPowerPoint Commutation process of one inverter arm (a) sw1 OFF, sw2 ON, (b) sw1 ON, sw2 OFF Similarly, it is also assumed that at the n step, the switch sw1 is in ON state, the switch current and voltage are (9) (10) The switch is equivalent to an inductor, and the energy stored on the inductor is (11) The energy stored on the inductor exists in the form of parallel current source (12) In the n + 1 step, the switch sw1 is turned OFF and is equivalent to a capacitor, the current value of the parallel current source is (13) From (11)–(13), we can see that when the switch sw1 is turned OFF, the energy stored in the inductor is ignored and is replaced by the capacitor without any energy. So, energy is always ignored when the switch state changes. The switch state of sw2 is opposite to the switch sw1, and the losses are the same. The total switching losses can be calculated as shown in (14), (15), and m, k are the total switching times of turning ON and OFF during the entire simulation process (14) (15) To minimise the switching loss, the optimal GS can be obtained as follows: (16) For the most PECs, the total switching times in the ON and OFF states are equal, and the switch voltage and current have the following relationship (17) Therefore, the optimal GS can be expressed as (18) It can be seen from (18) that the optimal GS obtained with the minimum loss as the objective function is related to the specific working point voltage and current of the switch. With the parameters of Table 1, the optimal GS can be solved before modelling as (19) In order to verify the optimal parameter with the minimum switching loss, the GS is changed from 0.1 to 1 and complete the offline simulation. The total switching loss is calculated by MATLAB, and the results are as shown in Fig. 8. It can be seen that the minimum total switching loss will be obtained at the point GS = 0.2614, which can prove the correctness of (18) and (19). Fig. 8Open in figure viewerPowerPoint Switching loss varies with GS To further verify the optimal GS selection method, the SimPower System (SPS)/Matlab is employed as the reference model. Three typical simulation results are compared, which are the optimal GS, the smaller GS (0.1) and the larger GS (0.5). The relative error of the inverter current with different GS is shown in Fig. 9. Fig. 9Open in figure viewerPowerPoint Relative error at different GS (a) A-phase current of inverter side, (b) A-phase current of the load side It can be seen that the relative error of A-phase current is not >1% with optimal GS. The relative errors of the smaller and larger GS are larger than the optimal value of GS. 3.3 Decoupling modelling method without latency Reduce the dimension of the system matrix can decrease the computation time in FPGA, which can reduce the computational burden for FPGA significantly. A decoupling method without simulation latency is proposed in this section. As is shown in Fig. 10, the one-port network is often connected by an inductor or capacitor, in which it can be used to separate the circuit into subsystems by the Forward Euler. For the inductor, the subsystem input vectors are the branch current and output vectors are the nodal voltage across the inductor. For the capacitor, the subsystem input vectors are the nodal voltage and output vectors are the branch current across it. The input vectors are updated at every time step (20) From (20), the whole system model can be further decoupled into several sub-systems and each sub-circuit is connected to voltage or current sources with values from the previous step to solve for the state variables individually. Taking the auxiliary converter as an example, the detailed discrete decoupling model of the auxiliary converter is shown in Fig. 11. Fig. 10Open in figure viewerPowerPoint Separation at inductor and capacitor (a) One-port network, (b) System partitioning Fig. 11Open in figure viewerPowerPoint Original auxiliary converter with partitioning into five subsystem For the subsystem 1, the transformer primary winding current i1pri is modelled as one of the input vectors, which is the state variable i2sec in subsystem 2. For the subsystem 2, the input v2sec is the nodal voltage of subsystem1 v1pri. Considering the transformer ratio m, their relationship is defined as follows: (21) For subsystems 3, 4, decoupling the system at Co and Laux, and subsystem 5 is the induction machine. Using the ADC switch model mentioned before, Fig. 11 can be transformed into Fig. 12, the corresponding modified nodal equation by the MNA method is shown below (22) (23) (24) (25) Fig. 13 is the block diagram of the auxiliary converter model when it is implemented in FPGA, the five subsystems are connected with state variables. In each subsystem, addition, shift and logic selection operation can be finished in one clock cycle, so the simulation latency is avoided by the decoupling method. The further simulation results are shown in Section 4. Fig. 12Open in figure viewerPowerPoint Discrete decoupling ADC model of the auxiliary converter Fig. 13Open in figure viewerPowerPoint Block diagram of the auxiliary converter model implemented in FPGA 4 FPGA implementation of the electrical traction system model The mathematical model established in Section 3 can be solved easily by floating-point operation in Matlab/Simulink offline simulation, which can obtain high-accuracy results. However, if the models are calculated in the FPGA-based real-time simulator, it is impossible to apply the floating point operations due to the limited FPGA hardware resources and strict latency requirement. For the implementation of the electrical traction system model, the fixed-point operations are employed to speed up task execution and reduce FPGA hardware resource consumption. The method can be divided into three steps: (i) Defining the base value for the voltage, current, Vbase, Ibase, modifying system matrix to per-calculated model. (ii) Selecting the Q-format for fixed-point operation. (iii) FPGA implementation. On the other hand, traditional FPGA-based real-time simulation requires designers to master the hardware description language (HDL). The method proposed by Vivado High-Level Synthesis (VHLS) allows designers to describe the simulation model using a high-level description language (such as C, C + +) and generate HDL code automatically [23]. The pipeline and parallel processing are also automatically executed to meet the timing constraints. Using VHLS tools can significantly reduce design complexity, and the design flow chart is as shown in Fig. 14. Fig. 14Open in figure viewerPowerPoint VHLS development process When the design meets all the requirements, it can be packaged as IP and imported into the real-time simulation system to complete the HIL simulation. The overall model including the traction and auxiliary system consumes FPGA available resources as shown in Table 3. Table 3. Summary of FPGA resource usage FPGA board Resources Occupancy rate, % Available traction drive system DSP 74 640 slices 78 14,720 flip flop 61 58,880 suxiliary converter system DSP 51 640 slices 60 14,720 flip flop 53 58,880 In order to verify the validity and applicability of the proposed modelling method, the HIL real-time simulation platform for high-speed train electrical traction system is built in Fig. 15, and the parameters of the electrical traction system are listed in Table 1. Fig. 15Open in figure viewerPowerPoint Photograph of hardware in the loop real-time simulation platform for high-speed train electrical traction system The TCU and ACU have connected the dSPACE simulator through the signal processing unit. The control development package, provided by the dSPACE simulator, performs experimental conditions setting and related signal monitoring, and completes the experiments of high-speed traction system in various conditions. Fig. 16a presents HIL simulation results of the 4QC no-load starting-up process, in which is1, is2 are the traction transformer secondary winding current, is is the equivalent primary winding, udc is the DC side voltage. The duration time of the starting-up process is about 0.7 s, the DC side voltage increased from 2600 to 3600 V. Fig. 16b shows the HIL simulation results of two 4QCs current and voltages, in which uAB1, uAB2 are the 4QC output voltage, us1, us2 are the transformer secondary voltage. It indicates that the phase difference of uAB1, uAB2 is one modulation cycle, which can reduce the current harmonics to the power grid. Fig. 16c shows the current FFT analysis of the two 4QCs, we can see that the 15th, 17th, 19th, 21st current harmonics are eliminated effectively in the synthesis current is. Fig. 17 is the comparison of current is1 at full load condition between real-time simulation and experiment results. There is no significant difference between them. Fig. 16Open in figure viewerPowerPoint Real-time simulation voltage and current waveforms of 4QC (a) Current and voltage of the 4QC, (b) Phase current, AC side voltage and power supply voltage, (c) FFT analysis results of current Fig. 17Open in figure viewerPowerPoint HIL simulation and experimental results of the current is1 at full-load condition Fig. 18 compares the real-time simulation and experiment result of traction induction motor stator current ia with the transition between different synchronous modulation method, which is widely applied in the high-power electrical traction system. In this paper, the central 60° synchronous PWM modulation method is employed due to the switching frequency limitation of the traction inverter. Fig. 19 describes the real-time simulation results of stator flux locus with different modulation method [24, 25]. These results are basically the same, which proves the effectiveness of the HIL simulation. Fig. 18Open in figure viewerPowerPoint Traction induction motor stator current ia1 (200A/div) transition between different modulation methods (a) Central 60˚ modulation with N = 5, (b) Central modulation with N = 5 to with N = 3, (c) Central modulation with N = 3 to the square waveform Fig. 19Open in figure viewerPowerPoint Real-time simulation results of stator flux locus with the different modulation method (a) Asynchronous modulation, (b) Square waveform modulation Fig. 20 depicts the simulation and experimental results of half-bridge voltage output and current in the auxiliary DC–DC converter. It can be seen clearly that the primary switches (S+, S−) and the auxiliary switch Sa could be turned off at ZCS. Since the reverse recovery characteristic of the power diode is not considered in the modelling method, the details of the waveform are a little different. Fig. 21 shows auxiliary DC–DC converter output voltage and current waveforms with the load variation. For the HIL real-time simulation, the algorithm in ACU has achieved good steady-state performance and dynamic response. Fig. 20Open in figure viewerPowerPoint Half-bridge voltage and current output of auxiliary DC–DC converter (a) HIL simulation results, (b) Experimental results Fig. 21Open in figure viewerPowerPoint Voltage and current waveforms of the auxiliary DC–DC converter 5 Conclusion This paper presents a multi-simulator multi-simulation step HIL real-time simulation platform for the high-speed train electrical traction system. An optimisation method of discrete-time switch admittance parameter is proposed for the ADC modelling method. The proposed modelling method is verified by comparing the SPS simulation results. The decoupling method without simulation latency is implemented in FPGA with a very small simulation time step 100 ns, which can improve the real-time simulation accuracy. The validity and reliability of the real-time simulation system is verified by the HIL simulation and experimental results, which indicates that the research of modelling and FPGA implementation method is meaningful for the development of the electrical traction system in the initial develop and test stages. 6 Acknowledgments This work was supported by the Fundamental Research Funds for the Central Universities under Grant no. 2019JBM062. 7 References 1Hasanzadeh, A., Edrington, C.S., Stroupe, N., et al: 'Real-time emulation of a high-speed microturbine permanent-magnet synchronous generator using multiplatform hardware-in-the-loop realization', IEEE Trans. Ind. Electron., 2014, 61, (6), pp. 3109– 3118 2Xu, F., Dinavahi, V., Xu, X.: 'Hybrid analytical model of switched reluctance machine for real-time hardware-in-the-loop simulation', IET Electr. Power Appl., 2017, 11, (6), pp. 1114– 1123 3Venkadesan, A., Himavathi, S., Sedhuraman, K., et al: 'Design and field programmable gate array implementation of cascade neural network based flux estimator for speed estimation in induction motor drives', IET Electr. Power Appl., 2017, 11, (1), pp. 121– 131 4Ould-Bachir, T., Blanchette, H.F., Al-Haddad, K.: 'A network tearing technique for FPGA-based real-time simulation of power converters', IEEE Trans. Ind. Electron., 2015, 62, (6), pp. 3409– 3418 5Liu, C., Guo, X., Ma, R., et al: 'A system-level FPGA-based hardware-in-the-loop test of high-speed train', IEEE Trans. Transp. Electrification, 2018, 4, (4), pp. 912– 921 6Shen, Z., Dinavahi, V.: 'Dynamic variable time-stepping schemes for real-time FPGA-based nonlinear electromagnetic transient emulation', IEEE Trans. Ind. Electron., 2017, 64, (5), pp. 4006– 4016 7Liang, T., Dinavahi, V.: 'Real-time device-level simulation of MMC-based MVDC traction power system on MPSoC', IEEE Trans. Transp. Electrification, 2018, 4, (2), pp. 626– 641 8Hadizadeh, A., Hashemi, M., Labbaf, M., et al: 'A matrix-inversion technique for FPGA-based real-time EMT simulation of power converters', IEEE Trans. Ind. Electron., 2019, 66, (2), pp. 1224– 1234 9Liang, T., Dinavahi, V.: 'Real-time system-on-chip emulation of electrothermal models for power electronic devices via hammerstein configuration', IEEE J. Emerging Sel. Topics Power Electron., 2018, 6, (1), pp. 203– 218 10Guo, X., Yuan, J., Tang, Y., et al: 'Hardware in the loop real-time simulation for the associated discrete circuit modeling optimization method of power converters', Energies, 2018, 11, (11), pp. 1– 14 11Mahseredjian, J., Dinavahi, V., Martinez, J.A.: 'Simulation tools for electromagnetic transients in power systems: overview and challenges', IEEE Trans. Power Deliv., 2009, 24, (3), pp. 1657– 1669 12Tant, J., Driesen, J.: ' Accurate second-order interpolation for power electronic circuit simulation'. 2017 IEEE 18th Workshop on Control and Modeling for Power Electronics (COMPEL), Stanford, CA, 2017, pp. 1– 8 13Zhu, J., Teng, G., Qin, Y., et al: ' A fully FPGA-based real-time simulator for the cascaded STATCOM'. 2016 IEEE Energy Conversion Congress and Exposition (ECCE), Milwaukee, WI, 2016, pp. 1– 6 14Kiffe, A., Riediger, W., Schulte, T.: ' Advanced preprocessing and correction-methods for automated generation of FPGA-based simulation of power electronics'. 2013 15th European Conf. on Power Electronics and Applications (EPE), Lille, 2013, pp. 1– 10 15Kiffe, A., Schulte, T.: ' FPGA-based hardware-in-the-loop simulation of a rectifier with power factor correction'. 2015 17th European Conf. on Power Electronics and Applications (EPE'15 ECCE-Europe), Geneva, 2015, pp. 1– 8 16Zhang, F., Yang, X., Xue, W., et al: ' A fast IGBT model considering the dynamic performance of both IGBT and antiparallel diode'. 2018 IEEE Applied Power Electronics Conf. and Exposition (APEC), San Antonio, TX, 2018, pp. 276– 279 17Hui, S.Y.R., Christopoulos, C.: 'Computer simulation of a converter-fed DC drive using the transmission-line modeling technique', IEEE Trans. Power Electron., 1991, 6, (4), pp. 636– 644 18Dagbagi, M., Hemdani, A., Idkhajine, L., et al: 'ADC-based embedded real-time simulator of a power converter implemented in a low-cost FPGA: application to a fault-tolerant control of a grid-connected voltage-source rectifier', IEEE Trans. Ind. Electron., 2016, 63, (2), pp. 1179– 1190 19Larijani, M.R., Zolghadri, M.-R., Shahbazi, M.: ' Design and implementation of an FPGA-based real-time simulator for H-bridge converter'. 2016 7th Power Electronics and Drive Systems Technologies Conf. (PEDSTC), Tehran, 2016, pp. 504– 510 20Matar, M., Iravani, R.: ' An FPGA-based real-time digital simulator for power electronic systems'. Proc. Int. Conf. on Power Systems Transients (IPST'07), Lyon, France, 2007, pp. 4– 7 21Razzaghi, R., Foti, C., Paolone, M., et al: 'A method for the assessment of the optimal parameter of discrete-time switch model', Electr. Power Syst. Res., 2014, 2, (8), pp. 80– 86 22Wedepohl, L.M., Jackson, L.: 'Modified nodal analysis: an Essential addition to electrical circuit theory and analysis', Eng. Sci. Educ. J., 2002, 11, pp. 84– 92 23Montano, F., Ould-Bachir, T., David, J.P.: 'An evaluation of a high-level synthesis approach to the FPGA-based sub-microsecond real-time simulation of power converters', IEEE Trans. Ind. Electron., 2017, 65, (1), pp. 636– 644 24Narayanan, G., Ranganathan, V.T.: ' Two novel synchronized bus-clamping PWM strategies based on space vector approach for high power drives', IEEE Trans. Power Electron., 2002, 17, (1), pp. 84– 93 25Narayanan, G., Ranganathan, V.T.: 'Overmodulation algorithm for space vector modulated inverters and its application to low switching frequency PWM techniques', IEEE Proc. Electr. Power Appl., 2001, 148, (6), pp. 521– 536 Citing Literature Volume14, Issue5May 2020Pages 850-858 FiguresReferencesRelatedInformation