Sliding Neural Artificial Controller of Induction Motor Combined with NPC Five Level Inverter

Abstract

Sliding mode is a robust approach used in nonlinear control; however, it can lead to high frequency chattering caused by the discontinuous control action. This phenomenon can become a significant issue, particularly when the system's state is near the sliding surface. To minimize the chattering phenomenon, we have replaced the discontinuous control with a neural artificial network. This has led to the development of a new controller, referred to as the sliding neural artificial controller, which is applied in speed control. This approach appears to effectively address these challenges while ensuring optimal control performance. Induction motors play a crucial role in the industry. To achieve speed variation, multi-level inverters are utilized due to their significant advantages. They enable operation in medium and high voltage as well as high-power applications while delivering improved voltage waveforms with low total harmonic distortion, making them ideal for electric machine applications. This article examines the Speed Sliding Neural Artificial (SM-ANN) controller employed to regulate the speed of an induction motor in combination with a Five Neutral Point Clamped inverter. The study focuses on evaluating the performance based on the physical parameters of the squirrel cage motor, such as speed, electromagnetic torque, and current, in comparison to the conventional speed sliding mode controller. Simulation results indicate significant improvements achieved by the SM-ANN controller, particularly in reducing ripple in the motor parameters. These simulations, conducted using Matlab/Simulink, effectively demonstrate the robustness and efficiency of the SM-ANN controller.