Research On Attitude Detection Of Spherical Motor Based On Multi-target Tracking Method Of Machine Vision

Permanent magnet spherical motor(PMSM)is a kind of motor with new structure.Due to the characteristics of rotor's structure,PMSM can complete three degrees of freedom space movement.The detection of rotor's attitude is an important part in the closed loop control of PMSM.Due to its three-dimensional characteristics,the traditional motor position detection method is not suitable for application to spherical motor(SM).Therefore,the research on the spherical motor rotor attitude detection is of great significance to the closed-loop control of the spherical motor.This paper reviews and analyzes the development status of SM and its attitude detection technology,and proposes a multi-target tracking algorithm based on IFDSST to obtain the attitude information of rotor from images.The main research work of this paper is as follows:1.The stator coordinate system is established based on the structure characteristics and motion form of the PMSM.In this coordinate system.Three attitude angles describing the motion of a spherical rotor.2.Firstly,the application of machine vision in industry and the basic principle of correlation filter tracking algorithm belong to machine vision are introduced.Then the principle of fast discriminant scale-space tracking(FDSST)algorithm,a target tracking algorithm in correlation filtering algorithm,is introduced.Then on the basis of pixel and image coordinate systems,the calculation method of rotor's attitude angle is introduced.3.The experimental platform of PMSM and the experimental device and process of rotor's attitude detection are introduced.The principles of experimental contrast methods of attitude detection are analyzed.The comparative test results of rotor's attitude detection are given to verify the practicability and accuracy of the proposed method for rotor's attitude detection of SM.Finally,the error analysis is made to the experimental results.