An Exploratory Study In The Adapted-Fuzzy CMAC Position Servo Control System Of PMSM

As an executor in AC servo system, Permanent Magnet Synchronous Motor (PMSM) is efficient, economical, and easy to maintenance. Thus, it is widely used in devices that require accurate location, such as Numerically Controlled Machine Tool and robot. However, the control performance of PMSM drive is influenced by many uncertainties, for example, parameter variations, external load disturbances, un-modeled targets and nonlinear dynamics. To achieve high-performance PMSM position servo system, advanced control scheme, which is static precision, transient responses and self-adaptation and anti-disturbances, is needed in practice. Based on the analysis on the features and control principles of PMSM, this thesis established its math model using id=0 vector control method. Three close loops control system, which includes current loop, velocity loop and position loop, was built up. For position loop, the main feedback loop, we chose a new adapted-fuzzy CMAC intelligent controller, which includes self-learn ability of neural net and quick response merit of fuzzy controller. Finally, the author simulates PMSM position servo system by the simulink of matlab, an advanced computer simulation tool. Results from simulations and experiments show it improved robust and real-time response property of PMSM position servo system by using this new controller. In summary, this thesis provides a new intelligent controller to improve the PMSM position servo system.