Adaptive Control Of Permanent Magnet Synchronous Motors For Vehicles

New energy vehicles is the direction of today's vehicle technology.Electric vehicles is an important part of new energy vehicles.PMSM has become preferred motor for EVS owing to its small volume,high efficiency and power density.However,there are 3 problems of self-adaption control of PMSM.First,there are four adjustment parameters in cascade control of speed loop and current loop in PMSM,making traditional parameters adjustment and advanced parameters self-adjustment rather complicated and have high computation cost.In addition,there is no evident physical meaning in controller parameters of speed loop and current loop.Second,rotor position is fed back using position sensor,which,however,could increase the cost of control system.Failure of sensor further reduce the reliability and accuracy of position feed back decreases due to noise of the sensor.In auto industry,cost,reliability,safety and other performance are of great significance.Last but not least,precise parameters of the motor are needed in order to guarantee performance of the PMSM control system.Parameters of the PMSM are affected by temperature and magnetic saturation,making it difficult to design control system of good performance and degrading dynamic performance of the system.Therefore,parameters identification of PMSM is an urgent need.Problems above could be settled as follows:(1)Setting formula of PMSM speed loop and current loop was derived using serial PI controller.Proportion and integral coefficient of serial PI controller of PMSM speed loop and current loop were set by two control parameters(band width of current loop and damping coefficient of the speed loop),making it convenient to set current of PMSM motor and parameters of speed controller.In addition,set control parameters have obvious physical meaning.(2)PMSM sensorless position and speed identification on-line were modified using Extended Kalman Filter.Also,state information like speed and position of the motor was identified.Results show that EKF has good dynamic tracking ability and noise immunity.(3)Parameters of PMSM were identified using recursive least squares(RLS)method with forgetting factor.Those parameters include inductance,resistance and rotor flux.Results show that this method of parameters identification features high identification accuracy,short time and good anti-interference performance.Error of inductance is 0.5%,with error of rotor flux and resistance at 0.3%and 0.17%,respectively.Finally,EKF state estimate and RLS method with forgetting factor were integrated with serial PI controller.Parameters of PI controller were simplified,realizing single parameter adaptive control of current loop and speed loop,reducing cost of on-board PMSM,improving robustness and control quality for sensorless PMSM.In the end,scientificity of this method is verified under urban UDDS driving mode.