SFLA-Verhulst组合模型预测矿井瓦斯涌出量

针对矿井瓦斯涌出量的时变性、波动性、非线性以及不确定性等特征，提出了SFLA-Verhulst组合预测模型，用于对具有非线性动态特征的瓦斯涌出量进行预测。该模型通过蛙跳算法对Verhulst模型的背景值参数寻优，并引入一次指数平滑法对原始数据进行优化处理，建立了基于混合蛙跳算法的SFLA-Verhulst组合预测模型；结果使模型在原始数据不准确或存在误差干扰的情况下仍能进行精度较高的预测。将新模型应用于某矿瓦斯涌出量预测，并对模型的预测结果进行检验分析，结果表明:该模型在结合蛙跳算法的全局寻优特点后预测精度较传统的GM(1,1)模型有明显的提高，适用性更强。

Prediction of gas emission in mine by SFLA-Verhulst combined model

In view of the characteristics of gas emission in mine, such as the time-varying, volatility, non-linear and uncertainty, a SFLA-Verhulst combined prediction model was put forward to predict the gas emission with the characteristics of non-linear and dynamic. In this model, the background value parameters of Verhulst model were optimized through the frog leaping algorithm, and the optimizing processing of the raw data was carried out by introducing into the single exponential smoothing method, then the SFLA-Verhulst combined prediction model based on the shuffled frog leaping algorithm (SFLA) was established. It showed that the model could still obtain the prediction results with a higher accuracy when the raw data were not accurate or the error interference existed. The new model was applied to the prediction of gas emission in a certain mine, and the prediction results of the model were verified. The results showed that after combining with the characteristics of global optimization of the frog leaping algorithm, the prediction accuracy of this model was much higher than that of traditional GM (1,1) model, with a greater applicability.