R-WELM model for predicting water inrush from coal seam floor
by Jiancong Fan; Qiang Wang; Yang Li
International Journal of Bio-Inspired Computation (IJBIC), Vol. 24, No. 1, 2024

Abstract: Water level variation in explorational boreholes serves as a direct indicator of water inrush risk in mining sites. Despite recent efforts to predict mine water inrush accidents using machine learning models based on diverse observational data, including borehole water levels, the accuracy, and timeliness of forecasting major accidents still need improvement. By integrating rough set theory to enhance the weighted extreme learning machine (WELM) model, a novel prediction model for coal seam floor water inrush is proposed. Firstly, the rough set theory calculates the roughness of positive and negative samples of bottom water inrush data. Then, the importance of the bottom water inundation data samples is calculated based on the roughness and approximation accuracy. Subsequently, the weight matrix is constructed using the sample importance and the number of samples, which is used to refine the WELM to improve classification prediction accuracy further. The experimental results show that the accuracy of the rough set weighted extreme learning machine (R-WELM) surpasses the traditional model, achieving a result of 91.66%. The method holds significant potential for diverse applications and can be a crucial reference for predicting water inrush in coal bed floors.

Online publication date: Wed, 24-Jul-2024

The full text of this article is only available to individual subscribers or to users at subscribing institutions.

 
Existing subscribers:
Go to Inderscience Online Journals to access the Full Text of this article.

Pay per view:
If you are not a subscriber and you just want to read the full contents of this article, buy online access here.

Complimentary Subscribers, Editors or Members of the Editorial Board of the International Journal of Bio-Inspired Computation (IJBIC):
Login with your Inderscience username and password:

Username:        Password:          Forgotten your password?

Want to subscribe?
A subscription gives you complete access to all articles in the current issue, as well as to all articles in the previous three years (where applicable). See our Orders page to subscribe.

If you still need assistance, please email subs@inderscience.com