Research On Vector Control System Of Six Phase Permanent Magnet Synchronous Motor Based On Improved Adaptive Dynamic Programming

With the rapid development of intelligent control theory,the ability of fast learning liberates the complicated human debugging,and the research on intelligent control system used in permanent magnet synchronous motor has been a hot topic in the field.In this paper,a new control algorithm is proposed,which is of high efficiency,sensitive response and stable dynamic performance.The algorithm is applied to the vector control system of six phase motor.The simulation results show that the controller designed by this algorithm can improve the response speed and accuracy of the control system,and proved that the improved algorithm has stronger adaptability and robustness.The main research contents are as follows:(1)Analysis the structural of six phase permanent magnet synchronous motor which the phase between the windings is 30 electrical degrees,the flux equation,voltage equation,torque equation and equation of motion,and the conversion between the coordinates of the model in the natural stationary coordinate system,the alpha beta coordinate system and the d-q coordinate system.A new coordinate transformation method is used for vector decoupling,and the model of motor under the condition of open-phase is developed to be used in vector control system.(2)Construct the space vector control system of the six phase permanent magnet motor based on id=0 principle.The two-level six-phase space vector modulation strategy based on two vectors and the three-level six-phase space vector modulation strategy based on five vectors are studied respectively.Reduce the number of space vector according to the different effect of the vectors in α-β and X-Y plane,five vector which can eliminate the harmonic of the X-Y plane is selected as the basic synthetic vector to compose reference voltage vector and to eliminate the number of 5,7,17,19,...harmonic voltage.Optimized the vector sequence in order to solve the problem of neutral point voltage offset of three level inverter.(3)The fitness variance factor is introduced to optimize the update of the weight coefficient of the particle swarm optimization algorithm,which can accelerate the convergence speed and avoid premature convergence.On the basis the weight coefficient between 3 networks,the PID coefficient is increased,which optimizes the structure of PID neural network and makes the function fitting ability more perfect.The direct heuristic dynamic programming was improved according to the optimization of PID neural network to build network action and critic networks and the improved PSO algorithm to renew the network weights.The improved algorithm avoids the computation of the first derivative,reduces the computational complexity,and optimizes the convergence and robustness.(4)Use the improved adaptive dynamic programming algorithm to design the speed and current controller in the vector control system of the six phase permanent magnet synchronous motor.The adaptive controller according to the error between feedback and thetarget value update control parameters and output control signal,and the three-level six-phase inverter output AC to adjust the motor.Compared with the traditional adaptive dynamic programming controller,the experimental results show that the improved algorithm has better sensitivity,dynamic characteristics and anti-interference ability.