Deformable Fuzzy Neural Network Control Algorithm of PMSM
To enhance the deficiencies of conventional vector control technology of permanent magnet syn-chronous motor,which pertains to the incapacity to dynamically modify controller parameters when the mo-tor work in complex conditions,resulting in subpar control system performance.A combined intelligent con-trol strategy based on Gaussian radial basis function neural network and fuzzy control has been proposed.Based on the speed error and the rate of change of the error,a incremental compensation-type two-dimen-sional deformable fuzzy neural network PID controller is constructed.The Jacobian matrix of the permanent magnet synchronous motor isobtained through the RBF neural network parameter identifier.The structure information of the deformable fuzzy neural network is determined through the variable structure algorithm.Simulation results in MATLAB/Simulink show that the control system improves the response speed and re-duces overshoot during motor startup and when the target velocity changes.The velocity alteration appears negligible upon encountering an abrupt alteration in the load conditions,promptly reverting to the prescribed value and thus optimizing the functionality of the vector control system.
NETL
NSTL
万方数据
