Mine working face gas prediction based on weighted LS-SVM

  • Authors:

  • Tiezhu Qiao;Meiying Qiao

  • Affiliations:

  • Institute of Measurement and Control Technology Tanyuan University of Technology Tanyuan, China;School of Electrical Engineering and Automation Henan Polytechnic, University Jiaozuo, China

  • Venue:
  • AICI'11 Proceedings of the Third international conference on Artificial intelligence and computational intelligence - Volume Part III
  • Year:
  • 2011

Quantified Score

Hi-index 0.00

Visualization

Abstract

Because coal and gas outburst prediction are very complex. In recent years, using least square support vector machine (LS-SVM) time series forecasting model to predict mine working gas is proposed. However in the search support vector solution process, inequality constraints become equality constraints in the LS-SVM, its advantage is to improve the algorithm speed, at the same time the sparse of support vectors and robustness to model are loss. In this paper, weighted LS-SVM is proposed to improve sparse and robustness and its time series prediction model is used to analysis short-time mine working face gas emission. Under MATLAB2009b environment, using LS-SVM1.7 toolbox, specific algorithm model is established, further model is verified by Hebi 10th 1113 mine and gas outburst working face time series data. The results showed that: weighted LS-SVM can achieve a better short-time gas prediction than standard LS-SVM; meanwhile its model has a better robustness.