论文部分内容阅读
针对液压系统的泄漏问题，提出了基于ＢＰ神经网络，以液压系统压力动态过渡过程为分析对象的故障诊断方法。该方法在通常ＢＰ神经网络的基础上，采用对学习样本加噪声的方法，提高了ＢＰ网络对噪声的抑制能力。它比传统方法，具有可靠性高，适用性广，而且成本低廉的特点。
Aiming at the problem of hydraulic system leakage, a fault diagnosis method based on BP neural network and taking the pressure transient process of hydraulic system as the analysis object is proposed. Based on the usual BP neural network, this method adopts the method of adding noise to learning samples to improve the ability of BP network to restrain the noise. Compared with the traditional method, it has the characteristics of high reliability, wide applicability and low cost.