中国安全科学学报 ›› 2018, Vol. 28 ›› Issue (3): 84-89.doi: 10.16265/j.cnki.issn1003-3033.2018.03.015

• 安全工程技术科学 • 上一篇    下一篇

基于双耦合算法的煤与瓦斯突出预测模型

付华1,2 教授, 丰胜成2, 高振彪1, 杨玉岗1 教授   

  1. 1 辽宁工程技术大学 电气与控制工程学院, 辽宁 葫芦岛 125105;
    2 辽宁工程技术大学 安全科学与工程学院, 辽宁 阜新 123000
  • 收稿日期:2017-12-03 修回日期:2018-02-04 出版日期:2018-03-28 发布日期:2020-11-09
  • 作者简介:付 华 (1962—),女,辽宁阜新人,博士,教授,博士生导师,主要从事现代传感技术及系统、煤矿瓦斯智能检测和控制工程方面的研究。E-mail: fxfuhua@163.com。
  • 基金资助:
    国家自然科学基金资助(51274118,71371091)。

Study on double coupling algorithm based model for coal and gas outburst prediction

FU Hua1,2, FENG Shengcheng 2, GAO Zhenbiao 1, YANG Yugang1   

  1. 1 College of Electrical & Control Engineering, Liaoning Technical University, Huludao Liaoning 125105, China;
    2 College of Safety Science and Engineering, Liaoning Technical University, Fuxin Liaoning 123000, China
  • Received:2017-12-03 Revised:2018-02-04 Online:2018-03-28 Published:2020-11-09

摘要: 为提高煤与瓦斯突出预测精度,有效预防瓦斯突出灾害,将等距映射(IsoMap)算法与优化加权向量机耦合算法 (DDICS-WLS-SVM)相结合,建立煤与瓦斯突出双耦合算法预测模型。首先利用非线性流形学习IsoMap算法对煤与瓦斯突出高维数据进行数据挖掘,提取其低维本质特征参量;然后通过逐维改进布谷鸟(DDICS)算法对加权最小二乘向量机(WLS-SVM)的正则化参数λ和高斯核参数σ进行寻优;最后对双耦合算法预测模型进行仿真试验,将IsoMap算法提取的低维本质特征作为该预测模型的输入,煤与瓦斯突出强度值作为模型的输出,并与PSO-SVM、LS-SVM方法的预测结果进行对比。结果表明:双耦合算法预测模型的平均相对误差为1.825%,最大相对误差为2.63%,该预测模型具有较高的预测精度。

关键词: 煤与瓦斯突出, 加权最小二乘向量机(WLS-SVM), 等距映射(IsoMap)算法, 耦合算法, 预测

Abstract: In order to improve the accuracy of coal and gas outburst prediction, a double coupling algorithm based model was built on the basis of combining the IsoMap algorithm with the coupling method (DDICS-WLS-SVM). The IsoMap algorithm is used to excavate the low-dimensional essence features of coal and gas outburst, followed by usage of DDICS algorithm to optimize the regularization parameter and the Gaussian kernel parameter of WLS-SVM. Finally, the low-dimensional essence features are extracted as the input to the prediction model which subsequently outputs the intensity of coal and gas. The results show that the relative errors in prediction by the model are not greater than 2.63%, and that the model is better than both the PSO-SVM and LS-SVM in prediction accuracy.

Key words: coal and gas outburst, weighted least squares support vector machine (WLS-SVM), isometric mapping (IsoMap) algorithm, coupling algorithm, prediction

中图分类号: