基于均值属性测度聚类分析的推土机故障诊断

Fault Diagnosis of Bulldozer Based on Cluster Analysis of Mean Attribute Measurement

提出了一种基于均值属性测度聚类分析的推土机柴油发动机故障诊断模型。模型以训练样本中各分类样本平均值表示其分类中心,建立各判别因子的未确知测度函数,然后计算单指标未确知测度和样本的均值属性测度,以样本均值属性测度进行等级判别;模型回判的正确率为100%。研究表明,该方法是种用于推土机发动机工作状态分类和故障诊断识别的有效方法。

A model for fault diagnosis of diesel engine of bulldozer is put forward based on cluster analysis of mean attribute measurement. The average of classification indicates the classification center of training samples, and the undetermined measurement functions of discrimination factors of these training samples are rigorously established to calculate the attribute measurements of means of single index and samples. Finally the mean attribute measurement of samples is applied to classify the bulldozer conditions. According to back-discrimination method, the correctness rate is 100%. The study shows this method is effective in working condition classification and fault diagnosis of engine on bulldozer.