An Artificial Neural Network-Based Model Predictive Control for Three-Phase Flying Capacitor Multilevel Inverter

Abstract

Model predictive control (MPC) has been used widely in power electronics due to its simple concept, fast dynamic response, and good reference tracking. However, it suffers from parametric uncertainties, since it directly relies on the mathematical model of the system to predict the optimal switching states to be used at the next sampling time. As a result, uncertain parameters lead to an ill-designed MPC. Thus, this paper offers a model-free control strategy on the basis of artificial neural networks (ANNs), for mitigating the effects of parameter mismatching while having a little negative impact on the inverter's performance. This method includes two related stages. First, MPC is used as an expert to control the studied converter in order to provide a dataset, while, in the second stage, the obtained dataset is utilized to train the proposed ANN. The case study herein is based on a four-level three-cell flying capacitor inverter. In this study, MAT-LAB/Simulink is used to simulate the performance of the proposed method, taking into account various operating conditions. Afterward, the simulation results are reported in comparison with the conventional MPC scheme, demonstrating the superior performance of the proposed control strategy in terms of robustness against parameters mismatch and low total harmonic distortion (THD), especially when changes occur in the system parameters, compared to the conventional MPC. Furthermore, the experimental validation of the proposed method is provided based on the Hardware-in-the-Loop (HIL) simulation using the C2000TM-microcontroller-LaunchPadXL TMS320F28379D kit, demonstrating the applicability of the ANN-based control strategy to be implemented on a DSP controller.INDEX TERMS Model predictive control, Artificial neural network, Multilevel inverter, Total harmonics distortion, Hardware-in-the-Loop (HIL) simulation. I. INTRODUCTIONP OWER converters have been used in a wide range of applications in recent decades due to their improved performance and efficiency [1]. In this context, substantial attention has been paid to multilevel inverters (MLIs) that introduced with certain advantages, such as: (i) reducing the total harmonic distortion, (ii) reducing common-mode voltage, and (iii) reducing the dv/dt (that is, the rate of voltage change over time in switching instant) stress that leads to a reduction in electromagnetic emissions. There are a number of well-established and conventional MLI topologies that have been implemented and refined over time, such as: flying capacitor (FC), cascaded H-Bridge (CHB), neutral

Low Computational Burden Predictive Direct Power Control of Quasi Z-Source Inverter for Grid-Tied PV Applications

“…However, the control strategies are carried out in an external target microcontroller kit (in our research, the C2000TM microcontroller-LaunchPadXL TMS320F28379D kit). The HIL emulator necessitates collaboration between the Host-PC and the target LAUNCHXL-F28379D, accomplished through a virtual serial COM port [40,41]. In this case, the Host-PC sends measured signals (or input characteristics) such as V pv , e abc , and i abc to the LAUNCHXL-F28379D kit.…”

A Comparison of the Response and Voltage Regulation Performance of the Single-Channel DC/DC Boost Converter Circuit with Artificial Neural Networks, Fuzzy Logic and PID Controllers

“…As part of the neural network design of the single-channel boost converter, for this research, a mathematical model by the MATLAB/Simulink program was used to create a feed forward neural network control block. This control block uses coaching by entering the input conditions of dc voltage, DC voltage output and output current, defined as duty cycle value [19][20][21]. However, for better understanding, it is shown as a model used in the coaching of neural networks as shown in Figure 7.…”

Face Recognition System in Intelligent Building Based on Neural Network

“…All nodes in the neural network can choose their own weights and connection strength to determine whether the connection parameters are correct. Different connection methods will have multiple obvious end patterns corresponding to specific points, that is, the same node receives the same result when it receives the same signal and inputs it as a training set to the output process, which is called "neural network" [17][18]. Figure 2…”