Application analysis of several improved cerebellar model articulation controllers for electric load simulator

With the deep research on Cerebellar Model Articulation Controller (CMAC), several improved CMAC methods are proposed from different aspects for electric load simulator. This paper analyzes these improved CMAC methods, i.e. BOWA-CMAC, anti-overlearning CMAC, smoothed CMAC, novel smoothed CMAC and NFCMAC respectively, and the implementation approach of these methods are discussed in detail. Compounding them with PD controller, the corresponding hybrid CMAC-PD controllers can be obtained. An electric load simulator is introduced to test the stability, accuracy, smoothness and loading performance of different controllers under different loading frequencies. The experimental results show the control performance of these improved CMAC methods.