摘要
目前高速公路施工机械设备故障检测主要采用聚类算法,但此方法易受到高频特征信号波动性的影响,导致检测准确性较低.为此,提出一种基于改进决策树的高速公路施工机械设备故障检测方法.首先,根据高速公路施工机械设备正常运行与异常运行的差异性,提取设备的故障信号特征;然后,基于信号特征的不同属性,引入增益率对高频信号进行分割,以平抑信号波动性,并以故障特征数据为输入,引入改进决策树算法,通过节点分裂计算与代价复杂度计算构建故障检测模型;最后,以此模型为依据,结合分类误差率对故障概率进行计算,实现施工设备故障检测.实验结果表明,所提方法能够有效且准确地检测出设备故障类型,检测效果较好.
The conventional clustering algorithm is mainly used for faults of highway construction machinery and equipment at present,but this method is susceptible to the fluctuation of high-frequency feature signals,resulting in low detection accuracy.Therefore,a study on fault detection of highway construction machinery equipment based on improved decision trees is proposed.Firstly,based on the differences between normal and abnormal operation of highway construction machinery and equipment,the vibration signal characteristics of equipment under abnormal operating conditions is collected,secondly,based on their different attribute values,gain rate is introduced to segment high-frequency signals to suppress signal volatility,fault feature data is used as input,and an improved decision tree algorithm is introduced to construct a fault detection model through node splitting calculation and cost complexity calculation.Finally,a fault detection model is constructed,the classification cumulative error function is used to calculate the abnormal probability of equipment,and then complete the fault identification.The experimental results show that the designed method has higher accuracy for fault identification,and the detection effect is good.
作者
葛俊海
GE Jun-hai(China Railway 18th Bureau Group Municipal Engineering Co.,Ltd.,Tianjin 300222,China)
出处
《兰州文理学院学报（自然科学版）》
2024年第3期74-78,共5页
Journal of Lanzhou University of Arts and Science(Natural Sciences)
关键词
改进决策树
高速公路
施工机械设备
故障检测
improve the decision tree
expressway
construction machinery and equipment
fault detection
