A fuzzy CMAC for AC servo system

A fuzzy cerebellar model articulation controller (FCMAC) is presented for controlling AC servo system to improve both global stability and performance. The controller introduces fuzzy theory into traditional cerebellar model articulation controller (CMAC) to fuzzify the state quantization way and the relationships between input states and hypercubes. This reflects the characteristics of human cerebellum learning and can reduce memory size greatly. The proposed FCMAC is applied on AC servo control and compared with a traditional CMAC. The simulation results show that the proposed control is effective and yields superior performance.