APPLICATION OF WAVELET NEURAL NETWORK BASED ON GEOLOGICAL STATISTICAL MODEL IN TO OVERBREAK PREDICTION OF UNDERGROUND OPENINGS
基于地质统计模型的小波神经网络在地下洞室超挖预测中的应用

The prediction of block overbreak in the openings is based on the wavelet neural network of the geological statistical model. Firstly,several basic theories,including the theory of geological statistical model,the theory of wavelet neural network,the theory of the opening overbreak,are introduced. Secondly,according to these theories,the process of programming is introduced and the calculation programs are designed. In geological statistical model,the basic theories of probability and statistics are adopted. In wavelet neural network,Sigmoid function is used by wavelet neural network. Opening overbreak prediction gives the predicted volume of block according to the field survey parameters. Lastly,according to the field survey data in a pump-storage hydroplant,several theories are used to predict overbeak of several openings. The prediction result provides a basis to avoid or decrease overbreak.