摘要
介绍了永磁球形步进电动机模型。在转子球面上进行随机编码;光电传感器的输出用来表示转子的位置。针对96个传感器的输出所构成的位置空间进行搜索。然而搜索空间较大,寻找一种高效的智能算法是解决位置检测实时性的关键。采用改进的混沌优化方法进行位置检测。根据优化的数学模型,将优化变量的定义域等分并重新组合成各个搜索子空间,并采用混沌变量的动力学方程映射到各子空间中进行并行搜索。仿真表明这种多搜索空间并行搜索方法能够快速地收敛到目标位置,大大提高了搜索效率;搜索精度取决于传感器的分辨率与搜索终止条件。
The permanent magnetic spherical step motor model was introduced in this paper, random encoder was painted on the surface of the sphere, and the outputs of optoelectronic sensors were used to identify the rotor position and search the orientation space consisted of the output combination of 96 photoelectric sensors. But the search space was comparatively large, searching an efficient and intelligent algorithm was a key to detect the sphere orientation in real time. The paper used an improved chaos optimization method to detect the orientation, subdivided the intervals of optimized variables and recombined them into many search subspaces based on mathematic optimization model, also mappings the dynamics equation of chaos variable to the subspaces and starts parallel search. Simulation showed that the parallel search method of more-subdivides had ability of fast convergence to the goal orientation, thus the search efficiency was greatly improved and the precision lied on the sensor resolution and searched stop condition.
出处
《微电机》
北大核心
2009年第2期32-35,53,共5页
Micromotors
基金
国家自然科学基金项目(50377010)
安徽省自然科学基金项目(03044103)
关键词
球形电机
步进电动机
光电传感器
并行
混沌搜索
优化
位置检测
Spherical motor
Step motor
Optoelectronic sensor
Parallel
Chaos search
Optimization
Position detect
