基于遗传算法优化支持向量机的船用柴油机气门漏气故障智能诊断方法  

Intelligent Diagnosis Method Based on Genetic Algorithm Optimization and Support Vector Machine for Leakage Faults of Gas Valves of Marine Diesel Engines

作　　者：蔡一杰 陈俊杰 王君[1] 张云东 杨建国[3] CAI Yijie;CHEN Junjie;WANG Jun;ZHANG Yundong;YANG Jianguo(School of Mechanical Engineering,Hubei University of Technology,Wuhan 430068,China;Hubei Key Laboratory of Modern Manufacturing Quality Engineering,Hubei University of Technology,Wuhan 430068,China;School of Energy and Power Engineering,Wuhan University of Technology,Wuhan 430070,China;A Direct Branch of China Coast Guard,Sanya 572000,China)

机构地区：[1]湖北工业大学机械工程学院,武汉430068 [2]湖北工业大学现代制造质量工程湖北省重点实验室,武汉430068 [3]武汉理工大学能源与动力工程学院,武汉430070 [4]中国海警局某直属局,三亚572000

出　　处：《内燃机工程》2022年第2期71-76,84,共7页Chinese Internal Combustion Engine Engineering

基　　金：湖北工业大学高层次人才基金项目(BSQD2020010);智能中速柴油机关键技术研究项目(工信部装函[2019]360号);现代制造质量工程湖北省重点实验室开放基金项目(KFJJ-2021012)。

摘　　要：针对船用柴油机气阀漏气故障的问题,提出一种结合遗传算法(genetic algorithm,GA)与支持向量机(support vector machine,SVM)的船舶柴油机气阀漏气振动诊断方法,称之为遗传算法优化支持向量机(GA-SVM)。通过分析静态与动态工况下的缸盖振动信号,提取训练SVM特征参数,利用GA-SVM的惩罚因子与核函数参数对故障进行识别。试验结果表明,GA-SVM方法完善了SVM参数选取方法,可有效识别柴油机气门漏气故障。优化后的整体故障诊断准确率为99.333%,相比于未优化前的测试集,故障诊断正确率提高了约2%。To solve the problem of valve leakage on marine diesel engines, a diagnosis method combining genetic algorithm(GA)and support vector machine(SVM)was proposed and named as GA-SVM.Through the analysis of cylinder head vibration signal in stable states and working states,the characteristic parameters for SVM model training were extracted and the faults were identified by the penalty factors and kernel function parameters of GA-SVM. Results show that the GA-SVM method improves the selection of SVM parameters,and the recognition method for valve leakage fault is effective. The overall fault diagnosis accuracy rate after optimization is 99. 333%,which is about 2% higher than that before optimization.

关 键 词：柴油机 故障诊断 气门漏气 遗传算法 支持向量机 故障模式识别 

分 类 号：TK428[动力工程及工程热物理—动力机械及工程]