Intelligent evaluation model of rock slope stability

The factor of safety is usually used to evaluate the stability of rock slope. Wavelet neural network combines the time-frequency domain localization properties of wavelet transform and self-learning ability of traditional feed-forward neural network. Wavelet neural network is used to approximate the non-linear relation between factor of safety and control factors of rock slope stability The stability of rock slope can be analyzed and forecasted with wavelet neural network trained by engineering case as learning samples. The numerical example shows that the precision of approximation and prediction of the proposed model can satisfy the quantitative evaluation for the stability of rock slopes. The model has the good capability of noise-resistance. Time-frequency analysis for training data is implemented to determine the original structure of wavelet network. A new method is proposed for iterative elimination of the redundant neurons. In this method, rough sets theory is used to calculate the dependency between the network output and the nodes in the hidden layer. Lastly, this paper builds the evaluation model of one rock slope in the Three Gorges Project permanent shiplock.