Improved Control Technique Based on Neural Network for AC-Chopper of Railway Substations

Improving the efficiency of the unbalance compen-sators in high-speed railway (HSR), mainly depends on the development of an intelligent control to limit the unbalance factors imposed by the Moroccan energy provider (ONEE). In this paper, a neural network model (ANN) to control the AC-chopper converter has been proposed. The aim is to achieve a stable operation under real conditions of variation of power consumption by substation. The efficiency of this control mode compared to a control method based on look-up table (LUT) with PID corrector is presented. The different stages of the system as well as the control strategy have been modeled and implemented in MATLAB/Simulink.