基于改进粒子群算法的平面度误差评定研究  

Flatness Error Evaluation Based on Improved Particle Swarm Optimization Algorithm

作　　者：夏亚涛[1] Xia Yatao(Department of Mechanical and Electrical Engineering,Shangqiu Polytechnic,Shangqiu,Henan 476000,China)

机构地区：[1]商丘职业技术学院机电工程学院,河南省商丘市476000

出　　处：《工具技术》2022年第5期111-116,共6页Tool Engineering

基　　金：河南省科技厅项目(2014第1644号)。

摘　　要：采用改进粒子群算法提高平面度误差评定的精度,建立平面度误差评定模型,基于链球复合模型设计粒子空间结构模型。链为粒子信息传输和共享的通道,粒子分布空间结构类似圆球形,链球形可根据进化迭代进行动态调整,计算链球在搜索过程的贡献度,删除链球上的相似粒子,给出了算法流程。实例验证结果表明,改进的粒子群算法对平面度测量误差平均值为6.256μm,相比PSO、GA、FPCIG、IDE、IPSOFF、NN、GPSO算法,测量误差分别减少了25.59%,22.30%,20.79%,18.34%,16.43%,14.08%,7.40%,平面度误差评定数据波动小,测量精度高,明显优于其它算法。In order to improve the accuracy of flatness error evaluation,the improved particle swarm optimization algorithm is proposed.The flatness error evaluation model is established.The particle space structure model is designed based on the chain ball compound model,the chain is the channel of particle information transmission and sharing,spatial structure of particle distribution is spherical,and shape of chain ball is dynamically adjusted according to evolution iteration.Contribution of chain ball in the search process is calculated,and the similar particles of chain ball are deleted to improve the efficiency,and the process is given.The experimental results show that the average error of IPSO is 6.256μm,the measurement errors are reduced 25.59%,22.30%,20.79%,18.34%,16.43%,14.08%and 7.40%compared with PSO,GA,FPCIG,IDE,IPSOFF,NN and GPSO,the flatness error evaluation data has little fluctuation,measurement accuracy is high,which are obviously better than other algorithms.

关 键 词：链球 空间 平面度 误差 

分 类 号：TG115[金属学及工艺—物理冶金] TG61[金属学及工艺—金属学]