Application of machine learning controller in matrix converter based on model predictive control algorithm

Finite control set model predictive (FCS-MPC) algorithms are famous in power converter for its easy implementation of constraints with cost function than classical algortihms. However computation complexity increases when swicthing state is high converters such as matrix converter, multilevel and this impose a serious drawback to compute multi-step prediction horizon MPC algorithm which further the computation. To overcome above said difficulty, machine learning based artificial neural network (ANN) controller proposed. The training data ANN derived from trained offline an accuracy 70.3%. proposed shows similar better performance terms total harmonic distortion (THD), peak overshoot during dynamic change reference current load parameter less execution time. Further, tested OPAL-RT using hardware in-loop (HIL) simulation showed that it outperforms controller.