基于液压实验平台的故障预测技术研究  

Research on Fault Prediction Technology Based on Hydraulic Experimental Platform

作　　者：赵鹏[1] 田秋实 ZHAO Peng;TIAN Qiu-shi(Civil Aviation Flight University of China,Guanghan 618307,China)

机构地区：[1]中国民航飞行学院广汉分院,四川广汉618307

出　　处：《数学的实践与认识》2021年第16期256-261,共6页Mathematics in Practice and Theory

基　　金：中国民用航空飞行学院青年基金项目(Q2018-76)。

摘　　要：首先,通过AMESim软件模拟了液压泵内泄漏故障并搭建了液压系统内泄漏故障的物理实验平台,运行了物理实验平台并采集了内泄漏故障的实验数据.然后采取基于BP神经网络和基于小波神经网络的故障预测方法,在Matlab中建立了故障预测模型,通过建立的模型对采集的实验数据进行预测分析,将预测数据与实验数据进行对比,计算出预测的误差.讨论了基于不同预测方法建模对故障实验数据预测效果的影响,为液压实验平台选择合理的预测方法构建预测模型提供了良好的参考价值.结果表明:BP网络在前几组数据预测的精度较高,但随着样本的增大,预测的误差增大,小波神经网络则能较好地反应出数据变化的趋势;但BP网络往往需要较少的训练次数就可以逼近目标,而小波网络的次数则相对较多.Firstly,the experimental platform of hydraulic internal leakage is set up,and the experimental platform is built.The data of leakage in the experimental platform are obtained.Then,the fault prediction model is established in Matlab by using the fault data and the method of fault prediction based on BP neural network and wavelet neural network.The fault prediction model is established in Matlab.The established experimental data is used to predict and analyze the experimental data,and the predicted data is compared with the experimental data to calculate the prediction.Error.The influence of modeling based on different prediction methods on the prediction results of fault experimental data is discussed.It provides a good reference value for the hydraulic experimental platform to select a reasonable prediction method to construct the prediction model.The results show that the prediction accuracy of BP network is higher in the first few groups of data,but as the sample increases,the prediction error increases,and the wavelet neural network can better reflect the trend of data change;but BP network often needs Less training times can approach the target,while the number of wavelet networks is relatively high.

关 键 词：液压实验平台 故障预测 BP神经网络 小波神经网络 

分 类 号：TH137.51[机械工程—机械制造及自动化]