Research On Estimation Algorithm Of Magnet Pole Position And Speed Of Permanent Magnet Linear Synchronous Motor In Novel Feeding System

The linear motor dirct drive system is the developing direction of highe velocity and high precision numerical control machine tool due to its many advantages, such as high speed, high acceleration, hige rigidity, and high position accuracy. A linear optical grating position sensor attached to the mover of the permanent magnet linear synchronous motor (PMLSM) provides the required position and speed information. However, the position sensor will result in adding cost and size, reduceing the operation stabliligy, and limiting the application of drive system. Thus, combining the PMLSM feeding system with the position sensorless control technology would overcome the drawbacks introduced by position sensor.The research theme of this dissertation is magnet pole position and speed estimation algorithms of permanent magnet linear synchronous motor (PMLSM) in the novel feeding system. Combined with the project of National Natural Science Foundation of China—"Research on high speed and high precision direct dirve system without position sensor" (No. 50475101), estimator of position and speed of the PMLSM dirve system is established, and the estimation algorithm, optimization and control as well as some key technologies of the PMLSM system are studied systematically by combining theoretical analysis, computer simulation and experiment.Firstly, the background and significance of the research are introduced, the development trend and the current research situations of the PMLSM system and position sensorless control technology are introduced, and the position and speed estimation techniques of PMSM are expatiated. Based on the reference frames and trandform criteria of PMLSM, mathematic models of PMLSM under three reference frames are given respectively. The PMLSM based on position sensorless control technology is analyzed and researched in detail. And the state equations of position sensorless control system of PMLSM are obtained.Secondary, operating PMLSM often subjects to parameters perturbation, load perturbation, and other uncertainty. So a model of PMLSM that can be only used in state estimation is built based on the "infinite mass" hypothetical condition. The PMLSM feeding system has strong non-linearity. Combining with the non-linear estimation theory and filter technique, the extended Kalman filter (EKF) is adopted to estimate the magnet pole position and speed of PMLSM. The estimation performance of position and speed estimator based on EKF is lucubrated by simulation experiments.EKF applies the Kalman filter to nonlinear systems by simply linearsing all nonlinear models. A number of serious limitations arise from its use of first order linearization. Based on unscented transformation, the unscented Kalman filter (UKF) can overcome the limitations of EKF. The UKF has been applied successfully in many nonlinear estimation fields, and has higher estimation performance than the EKF's. The UKF is introducred in the states estimation of PMLSM. The UKF estimation algorithm and PMLSM sensorless control system using UKF are simulated. And, a comparison of estimation performance of UKF and EKF for estimating position and speed is carried out.In the state estimation field of nonlinear and non-Guass distribution systems, particle filtering (PF) has more obviously advantagies than other filteing methord. An exploratory study on the application of particle filtering techniques to the state estimation of PMLSM feeding system is carried out. The Monte-Carlo particle filtering algorithm based on the sequential importance sampling (SIS) is given and analyzed in detail. Then, the position and speed estimator of PMLSM based on PF is designed. However, its effectiveness for improving states estimation performance for PMLSM has been unexplored. The PF algorithm for estimation speed and position of PMLSM is simulated. And, a comparison of estimation performance of PF, UKF and EKF for estimating position and speed is carried out by simulation experiments.Lastly, the experimential facility of PMLSM feeding system is introduced in detail. The measurement circuits system is developed to obtain three phase voltages and currents of PMLSM. Under different operating modes, magnet pole position and speed estimation experiments using EKF and UKF are made. The experiments results demonstrate that the speed and position estimator is effectiveness.