基于RF-SVM的发动机缸体孔组加工质量分级监控  

Machining quality classification monitoring for engine cylinder hole group based on RF-SVM

作　　者：杨泽青[1] 范敏 彭凯[1] 刘丽冰[1] 黄凤荣 YANG Ze-qing;FAN Min;PENG Kai;LIU Li-bing;HUANG Feng-rong(School of Mechanical Engineering,Hebei University of Technology,Tianjin 300131,China)

机构地区：[1]河北工业大学机械工程学院,天津300131

出　　处：《计算机工程与设计》2020年第5期1411-1419,共9页Computer Engineering and Design

基　　金：国家自然科学基金项目(51305124);天津市自然科学基金项目(16JCYBJC19100);河北省自然科学基金项目(E2017202294);河北省重点研发计划基金项目(16211803D)。

摘　　要：为实时动态监控发动机缸体顶面孔组的加工质量,提出基于随机森林(random forest,RF)和支持向量机(support vector machine,SVM)相结合的工序节点处加工质量分级监控模型。设计在工序间快速获取发动机缸体孔组图像的机器视觉系统,提取单缸孔7个特征参数及3个相邻孔间距;用主成分分析法对特征参数进行降维处理,建立样本集合训练孔组整体加工质量RF分级监控模型及单孔加工质量SVM分级监控模型。应用该模型对某发动机缸体顶面孔组加工质量进行在线监控,结果表明,与决策树模型、RF模型和SVM模型相比,所提模型对孔组整体加工质量分级精度可达97.778%,单孔分级精度可达99.167%,能快速响应发动机缸体制造过程质量反馈控制,可有效解决相关工程实际问题。To dynamically monitor the machining quality of hole group of the engine cylinder top face,a machining quality classi-fication monitoring model at process nodes based on the combination of random forest(RF)and support vector machine(SVM)was proposed.The machine vision system was designed for quickly obtaining the hole group image of the engine cylinder between the processes,and seven characteristic parameters of the single cylinder hole and three characteristic parameters of distance of adjacent holes were extracted.The principal component analysis(PCA)method was used to reduce the dimensionality of the characteristic parameters.The sample sets were established to train the RF classification monitoring model of the hole group machining quality and the SVM classification monitoring model of the single hole machining quality.The model was used to online monitor the machining quality of hole group of the engine cylinder top face.The results show that the proposed model’s machining quality classification accuracy can reach 97.778%for the whole and 99.167%for single hole,compared with the decision tree model,RF model and SVM model.It shows that the proposed model can quickly respond to the quality feedback control in the manufacturing process of engine cylinder,which can effectively solve the relevant engineering practical problems.

关 键 词：发动机缸体 加工质量 随机森林 支持向量机 分级监控 

分 类 号：TP278[自动化与计算机技术—检测技术与自动化装置]