基于支持向量机和免疫算法的故障检测与诊断

发布时间：2018-01-12 16:40

本文关键词：基于支持向量机和免疫算法的故障检测与诊断　出处：《华东理工大学》2011年硕士论文　论文类型：学位论文

【摘要】：随着现代技术的不断发展,工业规模不断扩大,生产设备也越来越复杂,工业生产过程中的安全性越来越受到人们的重视,因此过程监控与故障诊断成为近年来的研究热点。近年来,支持向量机作为一种新型的机器学习方法得到广泛的应用,并且在小样本的情况下表现出其优势。本文考虑到工业过程系统复杂等特点,对支持向量机和免疫算法在工业过程的故障诊断中的应用进行了深入研究。本文对工业过程中的故障检测和诊断方法进行了详细的综述,比较了各种检测及诊断方法的性能,介绍了统计学习理论和用于分类的支持向量机的基本原理,研究了支持向量机核函数参数对故障检测及诊断效果的影响。为了改善支持向量机故障诊断的性能,提出了将免疫算法与支持向量机相结合的故障检测及诊断算法,分别对免疫遗传算法和克隆选择算法进行改进,提出了基于改进免疫遗传算法的支持向量机和基于改进克隆选择算法的支持向量机,并将其应用于故障检测和诊断。为了验证所提出的方法的有效性,以标准仿真模型TE模型为平台,将其用于TE过程的故障诊断。仿真结果表明,本文提出的基于改进免疫遗传算法的支持向量机和基于改进克隆选择算法的支持向量机具有较高的故障诊断效率。
[Abstract]:With the development of modern technology, the scale of industry is expanding, and the production equipment is becoming more and more complex. People pay more and more attention to the safety in the process of industrial production. Therefore, process monitoring and fault diagnosis have become the focus of research in recent years. In recent years, support vector machine as a new machine learning method has been widely used. Considering the complexity of industrial process system, the application of support vector machine and immune algorithm in fault diagnosis of industrial process is deeply studied. In this paper, the methods of fault detection and diagnosis in industrial process are reviewed in detail, the performance of various detection and diagnosis methods are compared, and the statistical learning theory and the basic principle of support vector machine for classification are introduced. In order to improve the performance of SVM fault diagnosis, the effect of kernel function parameters on fault detection and diagnosis is studied. A fault detection and diagnosis algorithm combining immune algorithm and support vector machine is proposed, which improves immune genetic algorithm and clonal selection algorithm respectively. Support vector machine based on improved immune genetic algorithm and support vector machine based on improved clonal selection algorithm are proposed and applied to fault detection and diagnosis. Based on the standard simulation model te model, it is used in the fault diagnosis of te process. The simulation results show that. The proposed support vector machine based on improved immune genetic algorithm and support vector machine based on improved clonal selection algorithm have high fault diagnosis efficiency.
【学位授予单位】：华东理工大学
【学位级别】：硕士
【学位授予年份】：2011
【分类号】：TH165.3;TP18

【相似文献】