Study On Parameters Identification Of PMSM Based On Affine Projection

Due to the advantage of high performance and easy control,the permanent magnet synchronous motors have been widely used in many fields.In the actual system,the motor parameters will change according to different operating conditions,which will affect the performance of control system,so it will be a great significance to obtain the motor parameters accurately.Because the recognition model is not full rank,when we are trying to identify the parameters of the motor,each of them cannot be identified independently.The nonlinear factors of the inverter will make the reference voltage of stator inconsistent with the actual value,which will lower the identification accuracy of the motor parameters.In this paper,we will carry out some relevant research on how to solve the above-mentioned problems which will affect the identification of motor parameters.First of all,mathematical model of permanent magnet synchronous motor in natural coordinate system is established,and mathematical model in the two-phase rotating coordinate system is derived based on the coordinate transformation technology.On this basis,the vector control principle of the motor and the realization of SVPWM are explained,and a vector control model is built on MATLAB,the model is used to simulate the identification of motor parameters online.Secondly,in order to improve the accuracy of identification which will be reduced by nonlinear factors of the inverter,a discrete mathematical model with nonlinear voltage is established.To solve the problem of non-full rank of identification equation,by injecting a weak magnetic current into the d-axis which can add a new stator voltage equation,then all the parameters can be decoupled.The identification system of motor parameter is built base on the Adaline neural network,the Least Mean Square Algorithm with Variable Step-Size and the Affine Projection Algorithm with Variable Step-Size will be used to update the weight of network separately,these two algorithms are used to solve the contradiction between the steady-state error and the convergence speed in the identification algorithm,then the experiment of identification is simulated to verify the effectiveness of the affine projection algorithm proposed in this paper.Finally,the hardware platform and software flow of the permanent magnet synchronous motor control system are explained,andthe proposed algorithm with Variable Step-Size is experimental verified.The result of experiment shows that the proposed algorithm can effectively reduce the steady state error and improve the convergence speed of parameter identification result.