传感技术学报2025，Vol.38Issue(3)：477-486,10.DOI:10.3969/j.issn.1004-1699.2025.03.014
基于IDBO-HKELM-Adaboost的煤与瓦斯突出危险性预测方法
A Method for Predicting the Risk of Coal and Gas Outburst Based on IDBO-HKELM-Adaboost
摘要
Abstract
To achieve a more efficient and accurate completion of coal and gas prominence hazard prediction,a prediction model using the improved dung beetle optimizer(IDBO)enhanced by Adaboost algorithm to optimize the hybrid kernel extreme learning machine(HKELM)is proposed.First,kernel principal component analysis(KPCA)is used to process the influencing factors and extract effective feature quantities during data dimensionality reduction to obtain pre-processed sample data.The PWLCM chaotic mapping,nonlinear decreasing strategy,and neighborhood learning mechanism are incorporated into the dung beetle algorithm,and the pre-processed sample data are trained and tested for IDBO performance.The key parameters of HKELM are optimized by using IDBO,and the IDBO-HKELM coal and gas prominence hazard classification prediction model is constructed.The validation results show that the prediction method based on IDBO-HKELM-Adaboost has higher prediction accuracy than other models,and meets the requirements of accuracy and relia-bility of coal and gas prominence prediction with an accuracy rate of 97.44%while improving the computing efficiency.关键词
煤与瓦斯突出/突出预测/改进蜣螂算法/混合核极限学习机/核主成分分析/预测模型Key words
coal and gas outburst/highlight predictions/improved dung beetle optimizer/hybrid kernel extreme learning machine/Kernel principal component analysis/prediction model分类
计算机与自动化引用本文复制引用
李曼,徐耀松,王雨虹,王丹丹..基于IDBO-HKELM-Adaboost的煤与瓦斯突出危险性预测方法[J].传感技术学报,2025,38(3):477-486,10.基金项目
国家自然科学基金项目(51974151) （51974151）
辽宁省教育厅重点实验室项目(LJZS003) （LJZS003）
辽宁省教育厅辽宁省高等学校基本科研项目(LJ2017QL012) （LJ2017QL012）
辽宁省教育厅科技项目(LJ2019QL015) （LJ2019QL015）
