The gas concentration forecast based on RBF neural network and chaotic sequence

For getting the accurately coal gas concentration, according to the its nonlinear characteristics and time series chaotic characteristics, established a forecasting model, using chaos theory and RBF neural network. To get the training samples, it reconstructed gas concentration time series. Using MATLAB simulation to forecast analysis, the result shows that the relative prediction error is from 0 to 3%, and the mean square error is 0.0056. The result is well, and the examples show prediction model is feasible.