决策树算法在声速探头故障诊断中的应用

Application of Decision-tree Algorithm in Fault Diagnosis of Sound Velocity Probe

为增强声速探头自动故障诊断能力,将决策树算法应用于实际声速探头故障诊断问题。简述了决策树算法原理与优势,分析了声速探头工作原理与常见故障,面向声速探头故障诊断问题及数据特点,编程实现了基于CART决策树算法的故障诊断模型。依据收集的真实探头故障数据,经预处理后输入给模型进行故障诊断。实验结果表明,模型所得决策树分类结构清晰,易于理解且分类精度高,实现了故障诊断自动化,提高了故障诊断效率。

In order to enhance the automatic fault diagnosis ability of sound velocity probe,the decision-tree algorithm is ap⁃plied to the actual problem in the fault diagnosis of sound velocity probe.The principle and advantages of decision tree algorithm are briefly described,and the working principle and common faults of sound velocity probe are analyzed.In view of the fault diagnosis problems and data characteristics of sound probe,the fault diagnosis model based on CART algorithm is programmed.Based on the collected real probe fault data,it is input to the model for fault diagnosis after preprocessing.The experimental results show that the decision-tree classification structure obtained by the model is clear,is easy to understand and has high classification accuracy,which realizes the automation of fault diagnosis and improves the efficiency of fault diagnosis.