Research Of Position Estimation Algorithm Based On Sliding Mode Observer For Permanent Magnet Synchronous Motors

The essential condition of high efficiency,high power and torque density,and wide speed range in permanent magnet synchronous machine(PMSM)system indicates the accurate rotor position detection.The position sensorless algorithm is universally employed in industrial application due to its advantages of low cost and high adaptability to extreme work conditions.Regarding the issues of position detection error and harmonic influence in current position sensorless algorithm,this paper conducts thorough researches and analyses on the methods of position estimating of PMSM using sliding-mode position observer(SMO)principle.The main work is as follows:First of all,for the principles of SMO and vector control,the mathematical model of PMSM in synchronous coordinate system and stationary coordinate system is respectively derived,based on which,the principle and structure of vector control model are discussed,and the SMO is described.Meanwhile,this paper illustrates the principles of SMO and the factors responsible for high-frequency buffeting,and analyze the architecture of the traditional SMO using the saturation function.In addition,the effect of the cutoff frequency of the low-pass filter(LPF)on the attenuation and phase delay of the disturbance signal in the back electromotive force is also discussed in the paper.In terms of the position estimated error in SMO,the paper outlines the impact caused by motor parameters variation on the accuracy of rotor position estimation on the basis of error mechanism,and thus gives the suggestions on motor parameter selection.Besides,for the purpose of accessing the relationship between inverter voltage error and rotor position estimation error,the influences of dead time effect and voltage drops of switch devices are introduced.Furthermore,a compensation algorithm which utilizes the current signal and the electrical magnetic force(EMF)amplitude is proposed to reduce the negative effect on position estimation made by voltage amplitude and phase error.With regard to the impact on back EMF estimation made by harmonic current and harmonic voltage,the paper derives the mathematical relationship between the estimated position ripples and system harmonics,from which,an improved SMO using resonance type approach law is proposed to avoid the delay of back EMF phase and the limited suppression capacity to harmonics in conventional SMO using LPF.Eventually,a position sensorless control system on the platform of the interior PMSM is established.In MATLAB,the comparative simulations and analysis of the PMSM vector control system with the improved SMO and traditional SMO are separately carried out to verify the correctness of the theoretical analysis.With reference to the established simulation system and selected parameters,a vector control experiment system for PMSM with improved SMO was constructed on dSPACE DS1103 platform.Experimental results under various working conditions show that the improved SMO can not only effectively eliminate the harmonics and high-frequency buffeting signals in the back EMF,but also prevent the phase compensation and reduce the position estimation error,and therefore improve the steady-state performance and dynamic response of the system.The reasonability and effectiveness of the proposed improved SMO are verified by the corresponding experimental results above.