瓦斯浓度预测的混沌时序RBF神经网络模型

Model of chaotic sequence and RBF neural network on gas concentration forecast

为对煤矿瓦斯质量浓度进行精确预测,针对瓦斯质量浓度的非线性特点,在验证其时间序列具有混沌特性的基础上,建立了基于混沌理论和径向基神经网络的预测模型。将实测瓦斯质量浓度时间序列进行相空间重构得到训练样本,并利用MATLAB仿真软件进行编程预测分析。结果表明,相对误差为0~3%,均方差为0.005 6,预测效果良好。实例验证该预测模型切实可行。

Directed at the accurate prediction of the coal gas concentration,this paper proposes a chaos theory and RBF neural network-forecasting model,established according to the nonlinear characteristics of gas concentration and the validated chaotic characteristics of time series for gas concentrations.The reconstruction of the gas concentration time series for the training samples and the use of MATLAB simulation to forecasting analysis show that the relative prediction error ranging from 0 to 3% and the mean square error of 0.005 6 justify the feasibility of prediction model.