自适应增强优化的瓦斯涌出量预测模型

Prediction model of gas emission based on adaptive enhancement optimization

为提升回采工作面瓦斯涌出量的预测能力,提出一种基于门控循环单元(gate recurrent unit,GRU)预测模型,利用瓦斯涌出相关影响因素对瓦斯涌出量进行预测。对麻雀搜索算法的初始化过程加以改进,采用改进后的麻雀算法对影响GRU预测模型的超参数进行优化,提高瓦斯涌出量的预测精度;利用AdaBoost算法的自适应增强能力,构建自适应增强优化的瓦斯涌出量预测模型(ISSA-GRU-AdaBoost模型),并通过核主成分分析提取预测指标特征,提升预测的快速性。将所建模型与PSO-ELM模型、QPSO-LSTM模型、PSO-BP模型,以及SSA-SVM模型进行对比实验,结果表明ISSA-GRU-AdaBoost预测模型的预测精度最高。

In order to improve the prediction ability of gas emission in mining face,a prediction model based on gate recurrent unit(GRU)is proposed to predict the gas emission by using the relevant influencing factors of gas emission.The initialization process of the sparrow search algorithm is improved,and the improved sparrow algorithm is used to optimize the hyper parameters affecting the GRU prediction model,so as to improve the prediction accuracy of gas emission.Combined with the adaptive enhancement ability of AdaBoost algorithm,an adaptive enhanced optimization gas emission prediction model(ISSA-GRU-AdaBoost model)is constructed,and the prediction index features are extracted by using principal component analysis to improve the rapidity of prediction.The model is compared with PSO-ELM model,QPSO-LSTM model,PSO-BP model and SSA-SVM model.The results show that the prediction accuracy of ISSA-GRU-AdaBoost prediction model is higher.