基于独立元分析-最小二乘支持向量机的冷水机组故障诊断方法

发布时间：2018-03-24 20:59

本文选题：冷水机组　切入点：故障诊断　出处：《北京工业大学学报》2017年11期

【摘要】：冷水机组作为复杂系统,其变量间相关性严重,并且故障时的症状和原因具有多样性,导致了冷水机组的故障诊断较为困难.为了降低数据冗余性,提高故障诊断效率,提出一种基于独立元分析与最小二乘支持向量机相结合的冷水机组故障诊断方法.首先,运用独立元分析法提取冷水机组变量的独立元信息;然后,将提取的独立元信息作为最小二乘支持向量机的输入值进行故障类型的识别.利用北京某高校的地铁车站通风空调实训平台的实验数据验证该模型的故障诊断性能,并与传统的冷水机组故障诊断方法进行对比.比较结果证明基于独立元分析与最小二乘支持向量机相结合的冷水机组故障诊断方法优于传统方法.这表明该方法可以有效提取数据的高阶统计信息,提高故障诊断的效率.
[Abstract]:As a complex system, chillers have serious correlation among variables, and the symptoms and causes of the faults are various, which makes the fault diagnosis of chillers more difficult, in order to reduce the data redundancy and improve the efficiency of fault diagnosis. A fault diagnosis method for chillers based on independent element analysis (ICA) and least square support vector machine (LS-SVM) is proposed. Firstly, the independent element analysis method is used to extract the independent element information of chiller variables. The extracted independent meta-information is used as the input value of the least square support vector machine to identify the fault types. The fault diagnosis performance of the model is verified by the experimental data of the ventilation and air conditioning training platform of a certain university in Beijing. The comparison results show that the fault diagnosis method based on ICA and LS-SVM is superior to the traditional method. High order statistics of data can be extracted effectively, Improve the efficiency of fault diagnosis.
【作者单位】：北京工业大学信息学部;城市轨道交通北京实验室;数字社区教育部工程研究中心;计算智能与智能系统北京市重点实验室;
【基金】：国家自然科学基金资助项目(61174109,61364009)
【分类号】：TB657

【相似文献】