SOM神经网络在塔式起重机常见故障诊断中的应用  

Application of Self-organizing Feature Map Neural Network for Common Fault Diagnosis of Tower Crane

作　　者：原思聪[1] 李斌[1] 刘道华[1] 张满意[1] 季文祥 

机构地区：[1]西安建筑科技大学机电工程学院,西安710055 [2]陕西建设机械股份有限公司,西安710032

出　　处：《煤矿机械》2008年第12期208-210,共3页Coal Mine Machinery

基　　金：陕西省自然科学基金项目(2007E218)

摘　　要：塔式起重机的故障具有多样性,出现故障后,难于在较短时间内准确判别故障类型。通过SOM神经网络对输入样本进行"聚类",实现对故障模式的自动分类。据此对故障进行诊断,并在MATLAB环境下给出了塔式起重机故障诊断的具体实例。结果表明该方法可以对故障进行有效、准确地诊断,从而为塔式起重机的故障诊断提供了一种新的途径。As a result of the diversity of the tower crane faults, after the failures, it is difficulty to accurately discriminate the fault type immediately. The SOM neural network＇s ＂clustering＂ effect on the importation of samples can be used to automatically realize the classification of the failure modes, and diagnose the faults, the specific example of the tower crane fault diagnosis in the MATLAB environment is given. The results show that the method can effectively and accurately diagnoses the faults. Therefore, a new way is provided for the tower crane common fault diagnosis.

关 键 词：塔式起重机 SOM神经网络 自动分类 故障诊断 

分 类 号：TH218[机械工程—机械制造及自动化] TP39[自动化与计算机技术—计算机应用技术]