Research On Sensorless Control Of Synchronous Reluctance Motor

Synchronous Reluctance Motor(SRM)is a kind of brushless machine which has no permanent magnet or excitation winding on its rotor.It depends on the saliency of the rotor to generate electromagnetic torque,so the mechanical structure of SRM is simple,with a cheap manufacturing cost and good speed adjusting performance.In order to achieve sensorless control,this thesis focuses on analysis of the existing methods for position estimation of SRM,and build a sensorless control scheme based on improved High Frequency Signal Injection(HFSI)and Active-Flux observer with an adapted fusion position observer.The main work is as follows:Firstly this thesis describes the basic structure and control system of the synchronous reluctance motor.The principle of the reluctance torque generated by the synchronous reluctance motor due to the salient is discussed.The mathematic model of the synchronous reluctance motor in the two-phase synchronous rotation axis system is presented.Then the general structure of motor vector control system and several commonly used control strategies for SRM is also discussed.Secondly,based on the mathematical model of the synchronous reluctance motor,the basic principle of the flux observer is introduced,and the application of the Active-Flux observer in angle estimation is analyzed.The stability and dynamic performance of the velocity phase-locked loop in the Active-Flux observer is discussed.The applicable speed range of the Active-Flux observer is ensured.Thirdly,to solve the problem that the flux observer cannot be applied to low-speed position estimation,a high-frequency signal injection method is proposed.Based on the fundamental mathematical model of the synchronous reluctance motor,the mathematical model under the high-frequency excitation signal is deduced.The principle of position estimation by the high-frequency signal injection method is analyzed,which indicates a problem that the high-frequency feedback current signal need phase compensation exists.The q-axis feedback current signal needs phase compensation to ensure the signal-to-noise ratio of the effective signal.An improved feedback signal processing method is proposed.By introducing a feedback d-axis current signal into signal processing progress,the signal compensation is avoided and the original method is improved in tracking the error signal.Estimation error of the original method is reduced and the stability is improved.Fourthly,to solve the problem of method switching and position observation fusion in the process of speed-up and speed-down of two methods,an improved velocity Phase Locked Loop(PLL)is designed.Both the error angle signal of the Active-Flux observer and the error angle signal of the HFSI method are integrated according to a strategy with speed as a parameter.The fusion error signal enter the proportional-integral(PI)controller to get the final velocity estimate.Compared with the original method,the improved method has a smoother tracking effect in the speed switching process,simplifies the system structure,and has better dynamic performance.In the end of this thesis,the proposed improved HFSI method and improved fusion observer are simulated and experimented.The position estimation method and the vector control system of the synchronous reluctance motor were simulated in the MATLAB/Simulink software.Using the aforementioned control scheme and estimation model,a SRM vector control hardware experiment system was built on the dSPACE DS1103 platform.Through simulation and experiment,the work data and experimental waveforms of improved HFSI method and Active-Flux observer proposed in this paper are obtained.Compared with the original method,the effectiveness and superiority of the improved method above are verified.