基于神经网络的液压缸微小内泄漏数据分析及预测的研究

Research on Data Analysis and Prediction of Small Leakage in Hydraulic Cylinder Based on Neural Network

液压缸的内泄漏是工程机械故障中难以避免的,该故障会降低液压系统的工作效率,严重的内泄漏还会引发安全事故。构建一种实时测量液压缸内泄漏的系统,提出一种模拟液压缸微小内泄漏的方法,采用压力应变片将流量变化转换为应变信号的实验模型,对实验数据进行分析并建立应变-流量的数学模型,利用神经网络的学习与训练,对内泄漏量进行预测。最后将实际微小内泄漏量与神经网络的预测值相比较。实验结果表明,神经网络具有高精度和高效率的预测能力,为液压系统的微小泄漏监测奠定了基础。

The internal leakage of the hydraulic cylinder is unavoidable in the failure of construction machinery.This failure will reduce the working efficiency of the hydraulic system,and serious internal leakage can also cause safety accidents.A real-time measurement system for hydraulic cylinder leakage was constructed,a method of simulating the small internal leakage of hydraulic cylinders was proposed,and an experimental model using pressure strain gauge to convert the flow change into a strain signal was used to analyze the experimental data.A strain-flow mathematical model was established,and neural network learning and training was used to predict internal leakage.Finally,the actual small internal leakage was compared with the predicted value of the neural network.The experimental results show that the neural network has high-precision and high-efficiency predictive capabilities,which provides reference for the small leakage monitoring of hydraulic systems.