Research On Predictive Current Control Strategy Of PMSM

With the development of science and technology,the speed regulation system of permanent magnet synchronous motor(PMSM)has been applied to many fields,such as electric vehicles,aerospace,robots and so on,which consequently raises higher performance requirements.For the speed-regulating system with fixed motor body,improving the control strategy can greatly improve the speed-regulating performance of the system.Among the existing control strategies,predictive control is the most popular one,but the application of predictive control in the field of power transmission is not mature yet.Considering the above background,this paper studies the application of predictive control in the speed-regulating system of PMSM.To begin with,this paper makes a classified discussion of the PMSM's physical structure,and builds a mathematical model of the interior permanent magnet synchronous motor(IPMSM)to analyze its operating area and power.The traditional vector control system and its control strategy of PMSM are introduced,and moreover,this paper employs the voltage feedforward decoupling to accelerate the speed response,which improves the dynamic performance of the system.In addition,the PI regulator of current inner loop is replaced by the predictive controller to accelerate the current response.The current prediction controllers studied in this paper are divided into two types:model predictive current(MPC)and deadbeat predictive current control.The dual-vector MPC among them are adopted to increase the number of voltage actions within a period,which effectively reduces the current ripple.However,there exists the problem of enormous calculation.Therefore,the value function based on voltage error judgment is used to replace the traditional value function based on current deviation,which reduces the complexity and calculation of algorithm and improves the stability of system.Meanwhile,aiming at the relatively high requirement of model accuracy for the predictive control method of deadbeat current,this paper optimizes the predictive expected voltage to reduce the dependence of the prediction model on motor parameters.In order to further improve the anti-parameter disturbance performance of the system,this paper proposes the predictive control method of PMSM deadbeat current based on the Newton iteration method and adds the disturbance feedforward compensation.By combining the value function of model predictive control,the deadbeat predictive control is transformed into a quadratic optimization problem under voltage constraint.Based on the Newton-Raphson method,the selected expected voltage of the next moment is close to the ideal voltage value,and considering the time delay compensation,on the basis of which a disturbance compensator is added to enhance the overall system robustness.Finally,the experimental platform is built by taking advantage of the laboratory resources,the hardware composition and basic working principle of control loop are introduced as well.By step response,the parameters setting of two improved current controllers are conducted,and it is proved that the improved deadbeat predictive control with disturbance compensation has better anti-parameter disturbance performance.On this basis,the speed loop is closed to verify and analyze the steady-state response and dynamic response performance of the system under two control strategies.Experimental results show that the improved deadbeat control can effectively improve the system's ability to resist parameter disturbance and has great control performance.The improved dual-vector MPC control obviously improves the current control performance,reduces the current ripple and speeds up the current response,which is verified experimentally.