Decreasing control signal vibrations in the anti-noise model-free sliding mode control algorithm

摘要

Thanks to the model-free sliding mode control (MFSMC) algorithm, we were able to overcome errors resulting from the modelling process. When this algorithm was tested in a noise-prone environment, it was observed that the system was not able to track the desired path. In previous research, an anti-noise model-free sliding mode control (AMFSMC) algorithm was proposed. It improved the performance of MFSMC algorithm and made it noise-resistant. The control signal of that AMFSMC algorithm has a lot of chattering, which limits the use of the algorithm in practice. In this research, an improved anti-noise model-free sliding mode control (IAMFSMC) algorithm was proposed to solve the chattering problem. The control law of AMFSMC was modified using three techniques: boundary layer, quasi-sliding mode control, and saturation function. The IAMFSMC algorithm was applied to a second-order linear system. The system performance using the three techniques was compared to that when the AMFSMC algorithm was applied. Simulation results showed that using the saturation function gave good tracking response, despite the presence of noise, with chattering-free and smooth control signal.