基于IFOA-RotGBM的矿用挖掘机发动机故障诊断

Fault Diagnosis of Mining Excavator Engine Based on IFOA-RotGBM

针对矿山挖掘机发动机工作机理复杂、故障诊断效率低且精度不高的问题,提出了一种基于IFOA优化RotGBM的矿用挖掘机发动机故障诊断方法。首先利用随机森林-递归特征消除法(RF-RFE)对采集的挖掘机发动机故障数据进行特征提取,剔除冗余不相关特征;其次提出了一种改进的果蝇优化算法(IFOA)对LightGBM进行超参数寻优;然后融合旋转森林和LightGBM生成RotGBM,构建了新的故障诊断模型;最后利用某矿山挖掘机发动机故障数据对模型进行了验证,并与其他常用方法进行了性能对比分析。仿真结果表明:所提方法的诊断性能优于其他诊断方法,能达到98.31%的诊断精度,0.22%的误报率和2.5%的漏检率,满足矿山挖掘机发动机的故障诊断要求。

Aiming at the complex working mechanism,low fault diagnosis efficiency and low accuracy in mining excavator engine,a fault diagnosis method of mining excavator engine based on RotGBM optimized by IFOA is proposed.Firstly,the ran-dom forest-recursive feature elimination method(RF-RFE)is utilized to extract the features from the fault data of excavator engine and eliminate redundant irrelevant features.Secondly,an improved fruit fly optimization algorithm(IFOA)is proposed to optimize the hyper-parameter of LightGBM.Then,RotGBM is generated by combining Rotation Forest and LightGBM,and a new fault diagnosis model is constructed.Finally,the fault data of a mine excavator engine is used to verify the model,and the performance is compared with other commonly used methods.The simulation results show that the proposed method has better diagnostic performance than other diagnosis methods,and can reach 98.31%diagnosis accuracy,0.22%false positive rate and 2.5%missed detection rate,which meets the requirements of fault diagnosis of mine excavator engine.