Research On Control Strategy Of Permanent Magnet Synchronous Motor With Uncertain System Parameters

The vector control has become the preferred strategy for the high-performance control of permanent magnet synchronous motor since it is with the advantages of fast torque response and wide speed range.However,the system parameters of motor control system are uncertain.Firstly,system parameters are usually unknown.Secondly,with the variations of temperature,magnetic saturation,and work conditions,each parameter still varies during the operation of the system.In this dissertation,a series of problems of the vector control of permanent magnet synchronous motor caused by parametric uncertainty are studied in depth,and some control strategies considering parametric uncertainty are proposed.Specifically,the research is mainly carried out from the following aspects:(1)The mathematical model of permanent magnet synchronous motor with respect to three-phase stationary coordinate,the vector control and the principle of coordinate transformation are researched.Then,the mathematical model of permanent magnet synchronous motor with respect to two-phase rotating coordinate is established.Furthermore,the factors that lead to the uncertainty of each system parameter are analyzed,and the influence of the parametric uncertainty on the control performance is studied.(2)Aiming at the problem of unknown system parameters,a multi-parameter identification process based on the optimization of efficiency is designed.All system parameters that the inner torque loop relies on can be accurately identified.Meanwhile,the time of multi-parameter identification is reduced and the efficiency is improved through the reuse of excitation signals and the data processing based on the Dixon criterion.Furthermore,based on the results of multi-parameter identification,natural angular frequency and phase margin,the current controller is designed to improve the design efficiency of the controller.The stability analysis shows that the system is globally asymptotically stable under the control of the designed controller.Experimental results show that the efficiency of multi-parameter identification is significantly improved through the designed multi-parameter identification process,and the designed controller is with good static and dynamic performance.(3)Aiming at the problem of parametric variations,adaptive control strategies considering parametric variations are proposed.On one hand,in the inner torque loop,this dissertation takes the decoupling control of permanent magnet synchronous motor as an example.An adaptive decoupling control strategy considering parametric variations is proposed.Firstly,based on the Lyapunov stability theory,an adaptive decoupling controller with which the system is asymptotically stable is proposed.Then,the radial basis function neural networks is used to optimize all parameters of decoupling controller to optimize decoupling performance in the whole ranges of speed and torque.Experimental results show that the torque and current are with fast response and small fluctuation,that is to say,the system with the proposed strategy is with good decoupling performance under the condition of parametric variations.On the other hand,in order to reduce the effects of parametric variations on the speed control performance,an adaptive speed control strategy considering parametric variations is proposed.Firstly,based on the Lyapunov stability theory,an adaptive speed controller with which the system is asymptotically stable is proposed.Then,the radial basis function neural networks is used to optimize all parameters of speed controller to optimize speed control performance in the whole range of speed.Experimental results show that the system with the proposed strategy is with fast speed response and small speed fluctuation,that is to say,the effects of parametric variations on the speed control performance are effectively suppressed.(4)Aiming at the uncertainty of load torque,control strategies considering the uncertainty of load torque are proposed.On one hand,when the load torque is unmeasurable,the nonlinear torque error exists because of the variations of other system parameters.In order to reduce the nonlinear torque error,a control strategy based on the calibration of torque error is proposed.Firstly,based on the general polynomial regression model,a calibration algorithm that is not affected by parametric variations is designed,and the close-loop control of torque is carried out through the designed calibration algorithm.Then,the optimal regression model is found through the designed comparison criteria.Experimental results show that the accurate torque control in the whole ranges of speed and torque is achieved through the proposed control strategy.On the other hand,in order to reduce the effects of random fluctuations or sudden changes of load torque on the speed control performance,a control strategy based on the disturbance observer is proposed.Firstly,according to the principle of disturbance observer based on the system inverse,a disturbance observer is designed to achieve the accurate and fast observation of load torque.Then,the observed value is used as a feedforward compensation term to improve the anti-disturbance performance.The effectiveness of the proposed control strategy is verified through the simulation.