Modeling and Predictive Analysis of Small Internal Leakage of Hydraulic Cylinder Based on Neural Network

The internal leakage of a hydraulic cylinder is an inevitable system failure that seriously affects the working efficiency system. Therefore, it very important to accurately identify and predict data in cylinder. In this paper, model proposed simulate small cylinders, convert amount oil into strain signals through high-precision gauges train collected using various neural networks form computational obtain prediction results from model. applied paper are convolutional networks, BP T-S Elman networks. predicted network compared with actual amount. show accuracy above four kinds all 90%, which most accurate. This research provides scientific technical support for measuring predicting leaks.