| Permanent magnet synchronous wind generator(PMSG)is widely used in the field of power generation for its high power generation efficiency and strong low voltage ride through ability.Vector control maneuver is one of the main control strategies of PMSG,and its high performance control depends on the accurate PMSG parameters.However,the PMSG resistance,inductance and permanent magnet flux linkage parameters are changing at different temperatures and working conditions,which reduces control performance.of PMSG.Therefore,on-line identification of PMSG parameters is of great significance.In this thesis,PMSG was taken as the research object,and the improved extended Kalman filter(EKF)method was deigned to achieve highprecision parameter identification,the correctness of the proposed method was verified by simulation and experiment.The main research contents of this thesis were as follows:(1)On the basis of mathematical models in each coordinate system of PMSG,the space vector pulse width modulation(SVPWM)and the machine side converter vector control principle of PMSG were briefly described.(2)EKF was difficult to select the system noise matrix Q and measurement noise matrix R,which resulted poor identification accuracy,an EKF optimized by adaptive mutation particle swarm optimization(AMPSO)for PMSG parameter identification was proposed to solve this problem.The appropriate fitness function was established through the state estimation error,which served as the guiding standard for AMPSO to optimize the EKF optimal Q and R matrices,and the self-learning radius and extreme value mutation operation were introduced into the particle swarm optimization(PSO)to ensure the accuracy of Q and R matrix.The simulation results showed that the recognition accuracy and speed of the proposed method are better than the EKF and PSO-EKF.(3)Aiming at the problem that the PMSG parameter identification of AMPSO optimized EKF only identifies two parameters,a PMSG parameter identification of EKF optimized by mean particle swarm optimization(EDMPSO)with extreme perturbation was proposed.To reduce the large amount of high-order operations,a two-thread reduced-order identification model was established to realize multi-parameter identification of resistance,inductance and permanent magnet flux linkage,and an appropriate fitness function of PSO was designed.The operation of the extreme disturbance and the average best position were introduced into the PSO to make it jump out of the local optimum,and the adaptive correction of Q and R matrices of EKF was implemented.Simulation showed that the method had a good convergence speed and identification accuracy,and its generalization ability better than traditional methods.(4)The RT-LAB experimental platform was established to verify the two proposed methods,the PMSG parameter identification of two-thread EDMPSO-EKF can realize multi-parameter identification,whose identification accuracy and convergence speed were better than the PMSG parameter identification method of AMPSO optimized EKF,and its generalization ability was better. |
