Research On Dynamic Disturbance Control Method Of Asynchronous Motor

Vector control strategies make asynchronous motor can achieve the control performance comparable with DC motor attracted attentions by motor driver designer. However, Due to various working conditions and time-varying parameters, the stability of the motor and control system will be impacted greatly by disturbance if using traditional PID, which restricts the development of the vector control technology in many application fields especially in electric vehicle with high dynamics and energy-efficient demands.Therefore, the key problems in the vector control system of asynchronous motor are, disturbance identification with high performance of self renew ability, closed-loop controller strategy designed to restrain the disturbance when working conditions were changed and parameters were time-varied and keep good speed performance, maximum efficiency control based on the optimal values of rotor flux or excitation current.The thesis aims at the new vector controller strategy which can restrain disturbance and has efficiency optimization function. The main work and innovation point are as follows:The transient disturbance while motor running should be divided into two types, one comes from motor parameters variation because of influences of temperature and magnetic field, the other comes from real time interference of rotor speed, rotor flux, load torque and so on because of detection or the change of working conditions. Because the existing parameter identification methods are confined to some fixed mathematic models, can’t renew itself and has lower real time adaptive ability. This thesis propose Chained Extended Kalman Filtering (CEKF) to identify the rotor resistance, mutual inductance, rotor flux without speed sensor, which provide the basis of anti-interference and efficiency optimization. The simulation and experiment results show the effectiveness of the methods.The predictive functional control (PFC) is studied to improve the disturbance rejection ability of traditional vector control based on PID. Based on the basic theories of predictive functional control, the predictive functional control with second order characteristic model is easily focused. The vector control based on predictive functional control is proposed by selecting basis function, designing predictive model, determining the error correction equation, choosing the reference trajectory and rolling optimization, and is improve by the identify results. Meanwhile, the reduced-order disturbance observer was designed to carry on dynamic compensation while parameters variation and the random external loads were considered as an ensemble disturbance. This method can overcome the shortcoming of general predictive control method under strong disturbance and improve the anti-disturbance characteristics of asynchronous motor control system while maintaining the comprehensive performance. The simulation and experiment results show the effectiveness of the methods.The loss model of asynchronous motor is established. By training both the sample data of the loss model and the identification results in the previous section, the strategy of support vector machine (SVM) was put forward to find the directly or indirectly nonlinear fitting relationship between these variables. The new loss model based on SVM with anti- disturbance function is established, and is used in real time maximum efficiency control strategy which can decide the given excitation current according to input message in real-time. The simulation results show the effectiveness of the methods.