Statistical Performance Verification of the FS-MPC Algorithm Applied to the Matrix Converter

Abstract

In recent publications statistical model checking (SMC) has been proposed as a method for verifying the performance of finite-set model predictive control (FS-MPC) algorithms applied to power electronics converters. SMC is a powerful method originating from statistics, which can provide statistical evidence of a system's performance in a stochastic environment. In this paper, SMC is applied to a direct matrix converter, which operates in a grid with different harmonic distortion levels and voltage sags. Using the proposed method it is possible to evaluate not only the performance of the control algorithm in terms of the output current distortion but also evaluate the effects of the weighting factor selection and grid distortions on the device utilization. The obtained results show that high grid distortions and voltage sags increase the number of switching cycles. This information can be of great importance to identify the most stressed devices and how the control algorithm can be adapted to extend the lifetime of the devices and thereby the system during different grid conditions at the very early stage of the converter system design.

Supervised imitation learning of FS-MPC algorithm for multilevel converters

“…How high is the robustness of the trained network to model uncertainties? In [14] it was demonstrated that FS-MPC algorithm shows a high robustness to model parameter mismatch, the question is if this robustness is also inherited by the trained ANN controller. In the following section these questions will be answered by performing a HIL verification of the trained ANN controller.…”