神经网络动态预测软土路基沉降的研究

Some improved steps for the BP neural networks are introduced in this paper. The improved steps can increase the convergence speed of BP neural networks, and can improve the performance of BP neural networks. In this paper, on the basis of the improvement of BP neural network and field measuring information, the dynamic analysis method for predicting the highway settlement which varies with time is established. Since the model of this method is directly based on real samples, it can avoid the mistakes due to factitiousness in traditional methods. It is proved that the prediction model is accurate and the settlement has least error.