Gas Time Series Optimization Prediction Based On IABC-RBF Algorithm And Wavelet Analysis

Mainly on underground mining of coal mine in our country, along with our country coal mining are reaching the depth of mining, gas accident is easy to happen. it is an effective way to prevent gas accident to accurately predict the gas emission.In view of the traditional mine gas emission prediction accuracy is not high and only applies to certain mining area this problem, this paper puts forward a new method of prediction, time series analysis is based on historical gas for the optimization of gas time series prediction. First of all, after the validation of the gas with chaotic time series, using chaos theory in the two classical linear prediction method for gas time series prediction, Short-term prediction, simulation with maximum Lyapunov index prediction accuracy is higher; Secondly, according to the characteristics of gas dynamic nonlinear time series based on IABC-RBF coupling algorithm of nonlinear time series prediction method of gas,Simulation results show that the coupling algorithm has both IABC global optimization and fast convergence of the algorithm, and inherited the nonlinear mapping ability of the RBF neural network. Model output value rapidly approaching actual value, minimum prediction error is 0.0373. Finally, the noise of nonlinear time series of gas prediction, By identifying three optimal parameters of the wavelet analysis, Constructing wavelet-IABC-RBF algorithm gas time series prediction, By comparing the simulation to verify the prediction accuracy of the algorithm is better than the IABC-RBF algorithm.