A Novel Self-tuning Speed Control Scheme for the Permanent Magnet Synchronous Motor

For the permanent magnet synchronous motor (PMSM), it is challenging to control the speed under the varying conditions, such as wide-speed range and sudden load. Therefore, a novel self-tuning speed controller is proposed for the current source inverter fed three-phase PMSM. This scheme includes two parts: the single-neuron based controller and the parameter self-tuning algorithm in real time. The controller takes the error as the input, and combines the proportional item and the modified active function to meet the actual requirements. Based on the stochastic gradient descent algorithm and the PMSM mathematical model, the real-time online self-tuning of the parameters including the weight and bias is achieved, and the other parameters like the proportional coefficient and learning rate can be easily obtained. With the proposed speed control scheme, the dynamic response and adaptive ability can be improved. Especially under the condition of the wide speed range, the settle time changes little, and the overshoot of the system decreases with the increase of speed. And the parameters of this scheme can be self-tuned in online. The comparison experiments have been executed to verify the proposed ST controller.