Proportional Wavelet Sliding Mode Controller for Torque Ripple Reduction in BLDC Motor

Abstract

Permanent magnetic synchronous machines with trapezoidal back-electromotive force and six-step operation, also known as brushless direct current motors, have high torque undulations because needed current waveform can’t be provided. Proper vector control techniques used to reduce chattering, are high dependent on machine parameters and are complex to implement on real applications. Sliding Mode and Pseudo Sliding Mode are controllers with high dynamical performance and simple implementation, but increase torque undulations in machine control applications. On the other hand, wavelet multiresolution controllers have the feature of chattering reduction, but are complex to tune and analyze, due to the number of parallel detail and approximation coefficients used. This works proposes a simplified wavelet controller, replacing approximation coefficients with a proportional controller. This controller was used to design a proper sliding manifold for a sliding mode speed control, on a BLDC motor, without using machine parameters. A genetic algorithm was used to determine an optimum mother wavelet for a fixed decomposition level, leaving just one parameter to tune in future works. In results, an improvement to the proposed controller was suggested, for a controller with high dynamical and steady state performance, without torque undulations and easier to tune than wavelet multiresolution controllers.