最小条件下一般二次曲面轮廓度误差的评定  

Evaluating general quadric profile error based on least condition principle

作　　者：王宇春[1,2] 孙和义[1] 唐文彦[1] 田思庆[2] 武俊丽[2] 

机构地区：[1]哈尔滨工业大学电气工程与自动化学院,哈尔滨150001 [2]佳木斯大学信息电子技术学院,佳木斯154007

出　　处：《仪器仪表学报》2014年第8期1803-1809,共7页Chinese Journal of Scientific Instrument

基　　金：国家自然科学基金(61203052)资助项目;黑龙江教育厅课题(12511556)资助项目

摘　　要：为准确、快速地评定任意位姿的一般二次曲面轮廓度误差,提出以改进粒子群算法(IPSO)和角度分割逼近法为基础的最小区域评定方法。首先,提出角度分割逼近法计算的所有测量点与一般二次曲面的距离,并以其中最大值作为任意位姿二次曲面方程拟合算法的目标函数。其次,结合模型中目标函数的特点,提出以最优化解析法思想、自然选择和杂交结合PSO算法的新IPSO算法解决误差评定模型的优化问题。实验采用上述方法对某抛物面天线进行评定,与最小二乘算法(LSM)、标准PSO算法的结果和进化曲线等进行比较,同时比较商用软件SA的评定结果。结果表明结合PSO、最优化分析法思想、角度分割逼近法的新方法可更快速、准确地解决任意位姿的二次曲面轮廓度误差评定问题。In order to evaluate profile error of general quadric in any position and orientation accurately and rapidly,an evaluating method based on Improved Particle Swarm Optimization （IPSO）and angle subdivision approximating algorithm with the requirements of the mini-mal zone is proposed.First,angle subdivision approximating algorithm is proposed to calculate the distances between measured points and general quadric.The minimal maximum in the distances is taken as the object function of the quadric fitting optimization.Second, Particle Swarm Optimization （PSO）improved by the idea of optimized analytical method,natural selection and crossover is proposed to solve the optimization in the error evaluation model according to the characteristics of the objective function.In experiment,a parabolic antenna was evaluated with the above method,compared to the results and the evolution curve of least squares method （LSM）,PSO and the commercial software SA.The results show that the method combined with IPSO and angle subdivision approximating algorithm can solve the profile error evaluation of general quadric more rapidly and accurately.

关 键 词：改进粒子群算法(IPSO) 角度分割逼近法 二次曲面轮廓度误差 最小区域法 

分 类 号：TB92[一般工业技术—计量学] TH161[机械工程—测试计量技术及仪器]