Commutation of permanent magnet A.C. servo motors with incremental encoders via fuzzy reasoning

Abstract

A control method which drives the permanent magnet A.C. servo motor without the detection of the rotor position by the absolute position transducer such as an absolute encoder or a resolver is described. Only an incremental encoder is coupled to the motor shaft in order to obtain information about electrical commutation, motor speed, and motor position. A fuzzy algorithm is developed in order to estimate the absolute rotor position which is essential to electrical commutation. The center of gravity method defuzzifies the output variables of the fuzzy rules to generate the current command for estimation. A pulse interval measurement method is discussed, which enables the measurement of the motor speed and the motor position with the same resolution from a standstill to the rated speed. A digital current control method is proposed, which makes the permanent magnet A.C. servo motor fed by impressed currents. The entire control algorithm is implemented on the control system consisting of a digital signal processor (DSP) ADSP2101. Experimental results show that the proposed control scheme can be effectively used in controlling the permanent magnet A.C. servo motor with only an incremental encoder as a position transducer with high dynamic performance.