An Adaptive Wavelet Networks Algorithm for Prediction of Gas Delay Outburst

An adaptive wavelet networks algorithm for prediction of gas outburst is proposed in this paper. First, adaptive clustering algorithm is first used to determine initial parameters of wavelet network according to the results of the clustering. Then genetic algorithm and SVM-RFE is adopted to tune the structure of the wavelet network and adjust the network parameters to improve generalization performance. Finally, the simulation for prediction of gas outburst is discussed, and the result shows the validity of the proposed algorithm.