六盘水师范学院学报2025，Vol.37Issue(3)：65-72,8.DOI:10.16595/j.1671-055X.2025.03.007
基于优化GBDT模型的工程机械液压系统故障诊断
Fault Diagnosis of Hydraulic Systems in Construction Machinery Based on Optimized GBDT Model
摘要
Abstract
To achieve accurate diagnosis of hydraulic system faults in construction machinery,a fault diagnosis algorithm based on gradient boosting decision tree was analyzed and tested using an excavator as an example.Through experimental test-ing,the average accuracy and recall of the gradient boosting decision tree unequal cost logistic regression model are 99.7%and 91.8%respectively,with an average F1 measure of 0.956.Meanwhile,the average non-uniform cost of the fault diagnosis model proposed in the study is only 16.2,far lower than other algorithms.When facing combination faults,the classification error rate of this model does not exceed 15%.It can be seen that the gradient boosting decision tree non-uniform,which costs logistic regres-sion model,has high recognition accuracy for both single and combined faults in excavator hydraulic systems.关键词
挖掘机/液压系统/故障诊断/梯度提升决策树/逻辑回归/非均等代价Key words
Excavator/Hydraulic system/Fault diagnosis/Gradient boosting decision tree/Logistic regression/Unequal cost分类
机械工程引用本文复制引用
李可,吴亚兰,王建军..基于优化GBDT模型的工程机械液压系统故障诊断[J].六盘水师范学院学报,2025,37(3):65-72,8.基金项目
安徽省教育厅提质培优项目"职业教育精品在线开放课程'数控车床技能综合实训'"(2020tzpy50) （2020tzpy50）
安徽省高校自然科学重点项目"电动汽车空调气动噪声与出风量协同优化设计及测试设备开发研究"(2022AH052369). （2022AH052369）
