基于支持向量机的混凝土泵车支腿故障诊断方法  

Fault Diagnosis for Concrete Pump Truck Outrigger Based on Support Vector Machine

作　　者：王涛[1] 李业学[1] 梁学战[1] 

机构地区：[1]湖北文理学院建筑工程学院,湖北襄阳441053

出　　处：《机械设计与研究》2015年第6期74-76,80,共4页Machine Design And Research

基　　金：湖北教育厅资助项目(B2015143)

摘　　要：针对混凝土泵车支腿频繁发生开裂的问题,提出了一种基于支持向量机的故障诊断方法。混凝土泵车在施工过程中,施工状态的不同,直接影响到支腿的承载情况。对支腿故障进行分析,找出支腿最易出现开裂的关键位置,研究出了影响支腿关键位置承载的主特征量,搭建关键位置受力的支持向量机预测模型。运用MATLAB编写支持向量机预测程序,对模型进行训练和验证,通过输出的应力曲线图对支腿故障进行预测。实例验证了支持向量机对支腿关键位置应力预测的可行性。该方法相对于BP神经网络在小样本上更加精确,并为泵车工况参数的选择提供了理论支持。According to the characteristics of outrigger fault occurs frequently, a fault diagnosis method based on support vector machine is proposed. In the construction process of concrete pump truck, different construction status directly affect the outrigger bearing force. Fault analysis was carried out on the outrigger, find out the key position of outrigger easy to cracking, determining the primary features vector influencing the key point stress of outrigger, the stress prediction model of outrigger key point was built based on SVM. Using MATLAB to write support vector machine prediction program, the model is validated and trained by using MATLAB, through MATLAB output forecast stress curve, making a prediction of outrigger fault. The example verifies the feasibility of the SVM prediction outrigger stress. Compared with the BP neural network is more accurate in small samples, and provides a basis support in selecting working narameters of numn truck.

关 键 词：支腿故障 支持向量机(SVM) 主特征量 应力预测 

分 类 号：TU646[建筑科学—建筑技术科学]