Modeling?Analysis And Advanced Control Method Applications Of Permanent Magnet Synchronous Motor AC Servo Systems

With the rapidly development in power electronic technology, microelectronic technology, speed regulation system and modern control theory, the inevitable trend is that AC speed-regulation system replaces DC speed-regulation system. Friction torque affects the precision and performance of the servo system, especially at low speed or back and forth movement. Considering economic and technical reasons, it is very hard to eliminate the effect brought by friction torque through processing craft. Therefore, it has meaningful and good application prospects to eliminate the effect brought by friction torque using the control method.This paper does the research on how to suppress the friction effect when permanent magnet synchronous motor is working on position mode. Based on analysis of model for PMSM servo system, the system friction is modeled by the Stribeck model, whose parameters are identified off-line by piecewise linearization recursive least square algorithm and genetic algorithm respectively, and the feedforward compensation is presented using the estimated values of the identified model. Simulation and experiment results show that the proposed method can reduce the influence of the friction on the low velocity performance of servo system.Considering the practical application, the friction parameters will change with the variable working conditions. The compensation of the friction model based on the nominal working condition is a deviation. Since friction compensation model error does not belong to the slow and constant disturbance, which is more suitable to treat it as polynomial form. Therefore, a GPI observer is designed to estimate the error. Simulation results demonstrate the compound feedforward algorithm can still be better to reduce the friction effect of the servo system under the change of friction parameters.In practical application, the motor parameters may change due to the variation of temperature, magnetic saturation degree or other working conditions. The fixed control parameters are often difficult to achieve the desired performance. Therefore, identifying the parameters that represent current working condition, and then adjusting the controller parameters accordingly are necessary. As a result, a self-tuning algorithm for current loop parameters based on inductance identification is proposed. Simulation and experiment results show that the proposed algorithm can identify the inductance well and the current loop tracking performance is good when using the above algorithm. So it has a strong applicability and generality.