Enhanced Neural Network Model Predictive control based on TLBO algorithm: Applied to speed control of PMSM

This paper proposes an enhanced neural network model predictive control based on the teaching learning-based optimization algorithm, for the speed control of the permanent magnet synchronous motor. The objectives of this strategy are firstly, to solve the online computational problem and the nonconvex solution of model predictive control, using neural network identification and online metaheuristic optimization, secondly, to improve computational efficiency, real-time applicability and tracking accuracy of the conventional neural network model predictive control based on the teaching learning-based optimization algorithm using the proposed contribution of adding a sequence of reduction weighting coefficients to its cost function, over the prediction horizon. To assess the efficiency of the proposed control algorithm, a comparative study with the conventional neural network model predictive controller based on the teaching learning algorithm, is done. The mean square error, mean absolute error, root mean square error and computing time of the speed response of the PMSM are calculated and analyzed for the both controllers.The obtained results demonstrate that the proposed control algorithm gives better control performance than the other controller.