一种用于磨粒识别的基于改进PSO算法的支持向量机模型  

A SVM Model Based on Improved PSO Algorithm for the Recognition of Wear Debris

作　　者：吕纯[1] 张培林[1] 杨玉栋[2] 徐超[3] 张云强[1] 李一宁[1] 

机构地区：[1]军械工程学院七系,河北石家庄050003 [2]武汉军械士官学校四系,湖北武汉430075 [3]军械工程学院军械技术研究所,河北石家庄050003

出　　处：《润滑与密封》2016年第2期81-85,共5页Lubrication Engineering

基　　金：国家自然科学基金项目(51205405;51305454)

摘　　要：为提高磨粒智能识别的准确率,以传统支持向量机和粒子群优化(PSO)算法为基础,提出一种基于改进PSO算法的支持向量机(SVM)识别模型。该识别模型的惩罚参数和核函数参数可同时得到最佳优化,从而可建立模型参数最优的自适应SVM识别模型。采用该识别模型对油液中的磨粒进行智能识别,结果表明该模型识别准确率高达98%,明显优于BP神经网络模型。In order to improve the accuracy of wear debris intelligent recognition, a support vector machine（SVM） recognition model based on improved particle swarm optimization （PSO） algorithm is proposed on the basis of traditional SVM and PSO.The penalty parameter and kernel function parameters of the recognition model can be best optimized at the same time, so that self-adaptive SVM recognition model with optimal parameters was established.The wear debris in the lubricating oil was recognized is as high as 98%, whi by using the proposed recognition model.The resuhs show that the model recognition accuracy rate ch is significantly higher than the BP neural network model.

关 键 词：油液检测 磨粒识别 粒子群优化算法 支持向量机 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构] TH117.1[自动化与计算机技术—计算机科学与技术]