基于BP神经网络的轴承温度预测

Bearing Temperature Prediction Based on BP Neural Network

针对影响轴承温度的因素如转速、轴向载荷、径向载荷和润滑脂含量等,建立了基于误差回传神经网络(BP网络)的数学模型,分析并采用高速轴承测试台实测了7006C的2套轴承在各种因素影响下的温度,构建了BP网络并预测了轴承的温度。结果表明:BP网络预测温度与实际测量温度吻合较好,用BP网络对轴承温度进行预测的方法可行可靠。

Aiming at the factors affecting the bearing temperature rise,such as speed,axial load,radial load and grease content,a mathematical model based on error feedback neural network(BP network)is established.The temperature of two sets of 7006c bearings under the influence of various factors is analyzed and measured by high-speed bearing test-bed,a better BP network is constructed and the bearing temperature is predicted.The results show that the predicted temperature of BP network is in good agreement with the actual measured temperature,and the prediction method of bearing temperature by BP network is feasible and reliable。