Research On Position Detection Of Permanent Magnet Spherical Stepper Motor Based On Machine Vision

Because the spherical structure could work at any orientation in three-dimensional (3D) space, such as motion control of the humanoid robots, 3D space measurement of intelligent instruments, and industrial control in multi-dimensional space where high precision is essential, the research of spherical motor has attracted wide attention in recent years. Position detection for rotor of spherical motor is the key technology in motor control system because which is indispensable in feedback controlling, and its differentiate rate could affect the control accuracy of system directly. So position detection is also an urgent problem to be solved in the research of spherical motor.The object studied in this paper is permanent magnet spherical stepper motor. Based on the research of position detection methods for spherical motor, a vision-based position detection method proposed by Kok-Meng Lee et al is used to obtain the position of rotor. In order to achieve the absolute orientation information of the rotor, a new grid pattern and its generation method are presented according to the requirement of rotor detection. A new image analysis method to get the characteristic points is presented based on the feature of grid image also.At the end, the simulated grid image according to the new grid pattern processed by the new image analysis method is implemented in Matlab by C. The experimental results indicated the feasibility of this method.