Research On Electromagnetic Parameters Identification Method Of Sensorless Control Permanent Magnet Synchronous Machine

Accurately obtaining the rotor position and speed information is necessary to achieve highprecision and high-performance control of permanent magnet synchronous machines(PMSM).However,the system’s cost and space will increase using mechanical sensors(photoelectric encoders,incremental encoders,rotary encoders,Hall sensors),even reduce system operation reliability.Simultaneously,the electromagnetic parameters of PMSM are easily affected by temperature rise,electromagnetic interference,or other factors due to the strongly coupled,multi-parameter,non-linear of the PMSM,affecting the performance and stability of the servo system designed offline.Therefore,position sensorless control and online parameter identification have become research hotspots in PMSM control.The main investigations of the thesis are as follows:To solve the problems that the current PMSM initial rotor position estimation methods have low adaptability to the PMSMs with different saliency properties and high sensitivity to electrical parameters variation.An initial rotor position estimation method based on the dynamical equations of the PMSM is proposed to solve the problems above.Firstly,a series of vibration signals of the PMSM is generated by applying a high-frequency voltage to the rotating virtual d-axis of the rotor.Then,the relationship between the vibration signals and the rotating virtual d-axis position is established.Next,the initial d-axis rotor position of the PMSM is estimated by analyzing the vibration response.At last,the rotor magnetic polarity is identified on account of the saturation saliency effect.The theoretical analysis and simulation,as well as experiments,are carried out to verify the effectiveness and accuracy of the innovation method.The results demonstrate that the proposed method has the advantages of easy to implement,high adaptability,and strong robustness.The maximum initial rotor position estimation error of the proposed method is 1.7 electrical degrees below other methods.The basic theory of sliding mode variable structure control is introduced,and then the sliding mode back-EMF observer and phase-locked loop(PLL)rotor position estimation module are designed.Finally,the rotor position online estimation module’s effectiveness is verified in the simulations and experiments from multiple perspectives such as system control performance and parameter change robustness.To overcome the problem that the deficient-rank of the PMSM electromagnetic parameters online identification equations,a full-rank online identification method is proposed to identify the full electromagnetic parameters of the SPMSM,which is controlled with self-adjusting parameters sliding mode observer(SMO).First,the rotor permanent magnet flux linkage value is estimated by calculating the q-axis back electromotive force estimation value and the rotational speed estimation value while online estimating the SPMSM rotor position.Then,the permanent magnet flux linkage estimated value is used as an input value of the extension Kalman filter method(EKF)parameters identification algorithm to establish the full-rank parameters identification equations and estimate the stator resistance and stator.Finally,the SMO parameters are adjusted according to the stator resistance and stator inductance estimated value for improving all electromagnetic parameters identification accuracy of the SPMSM.Moreover,the higher identification accuracy and better parameter change tracking performance of the method are shown in the simulation and experimental results.The proposed full-rank identification method is based on the multiplexing of the SMO observer function.That is,SMO is responsible for rotor position observation and flux identification at the same time.The fullrank equation can be constructed without signal injection and the time-division full-rank algorithm,which reduces the motor control system’s complexity.