论文部分内容阅读
分析了目前常用的故障预测方法种类及其特点,提出了一种最小二乘法与BP神经网络相结合的预测盾构故障的方法。先构建二者结合的预测模型,进而比较单一BP神经网络的预测结果与最小二乘法和BP神经网络结合的预测结果,结果发现二者结合的预测效果比单一的效果要好。
The types and characteristics of the commonly used fault prediction methods are analyzed. A method to predict the shield fault based on the least square method and BP neural network is proposed. First, a prediction model combining the two is constructed, and then the prediction results of the single BP neural network are compared with those of the least square method and the BP neural network. The result shows that the combined effect of the two is better than the single one.