Skip to main content

Development of a spontaneous combustion TARPs system based on BP neural network

Journal Article


Abstract


  • Spontaneous combustion of coal is a major cause of coal mine fires. It not only poses a severe hazard to the safe extraction of coal resources, but also jeopardizes the safety of mine workers. The development of a scientific management system of coal spontaneous combustion is of vital importance to the safe production of coal mine. This paper provides a comparative analysis of a range of worldwide prediction techniques and methods for coal spontaneous combustion, and systematically introduces the trigger action response plans (TARPs) system used in Australian coal mines for managing the spontaneous heating of coal. An artificial neural network model has been established on the basis of real coal mine operational conditions. Through studying and training the neural network model, prediction errors can be controlled within the allowable range. The trained model is then applied to the conditions of Nos. 1 and 3 coal seams located in Weijiadi Coal Mine to demonstrate its feasibility for spontaneous combustion assessment. Based upon the TARPs system which is commonly used in Australian longwall mines, a TARPs system has been developed for Weijiadi Coal Mine to assist the management of spontaneous combustion hazard and ensure the safe operation of its mining activities.

UOW Authors


  •   Wang, Longkang (external author)
  •   Ren, Ting
  •   Nie, Baisheng (external author)
  •   Chen, Yang (external author)
  •   Lv, Changqing (external author)
  •   Tang, Hua (external author)
  •   Zhang, Jufeng (external author)

Publication Date


  • 2015

Citation


  • Wang, L., Ren, T., Nie, B., Chen, Y., Lv, C., Tang, H. & Zhang, J. (2015). Development of a spontaneous combustion TARPs system based on BP neural network. International Journal of Mining Science and Technology, 25 (5), 803-810.

Scopus Eid


  • 2-s2.0-84942549494

Ro Metadata Url


  • http://ro.uow.edu.au/eispapers/4645

Number Of Pages


  • 7

Start Page


  • 803

End Page


  • 810

Volume


  • 25

Issue


  • 5

Abstract


  • Spontaneous combustion of coal is a major cause of coal mine fires. It not only poses a severe hazard to the safe extraction of coal resources, but also jeopardizes the safety of mine workers. The development of a scientific management system of coal spontaneous combustion is of vital importance to the safe production of coal mine. This paper provides a comparative analysis of a range of worldwide prediction techniques and methods for coal spontaneous combustion, and systematically introduces the trigger action response plans (TARPs) system used in Australian coal mines for managing the spontaneous heating of coal. An artificial neural network model has been established on the basis of real coal mine operational conditions. Through studying and training the neural network model, prediction errors can be controlled within the allowable range. The trained model is then applied to the conditions of Nos. 1 and 3 coal seams located in Weijiadi Coal Mine to demonstrate its feasibility for spontaneous combustion assessment. Based upon the TARPs system which is commonly used in Australian longwall mines, a TARPs system has been developed for Weijiadi Coal Mine to assist the management of spontaneous combustion hazard and ensure the safe operation of its mining activities.

UOW Authors


  •   Wang, Longkang (external author)
  •   Ren, Ting
  •   Nie, Baisheng (external author)
  •   Chen, Yang (external author)
  •   Lv, Changqing (external author)
  •   Tang, Hua (external author)
  •   Zhang, Jufeng (external author)

Publication Date


  • 2015

Citation


  • Wang, L., Ren, T., Nie, B., Chen, Y., Lv, C., Tang, H. & Zhang, J. (2015). Development of a spontaneous combustion TARPs system based on BP neural network. International Journal of Mining Science and Technology, 25 (5), 803-810.

Scopus Eid


  • 2-s2.0-84942549494

Ro Metadata Url


  • http://ro.uow.edu.au/eispapers/4645

Number Of Pages


  • 7

Start Page


  • 803

End Page


  • 810

Volume


  • 25

Issue


  • 5