讨论了转子运动和单截面信息融合的相关理论,提出了基于双截面信息融合的旋转机械故障诊断方法,建立了基于双截面融合能量谱的旋转机械常见故障诊断BP神经网络模型。模拟实验结果表明:与基于单截面数据的诊断结果对比,将双截面融合应用于旋转机械常见故障诊断,可有效提高故障诊断的准确率。
The rotor motion and the information fusion of single section were discussed; the fault diagnosis method for rotary machinery based on the information fusion of two sections was put forward, and the back propagation neural network model was established. Engineering practice indicated that the fault diagnosis accuracy based on the information fusion of two sections was higher than that based on the information fusion of single section.