Prediction of coal seam water conduction fault zone height based on Prediction of coal seam water conduction fault zone height based on

Abstract

Aiming at the problems of low prediction accuracy for the height of water-conducting fracture zone and difficult parameters optimization, we propose a guide height prediction method based on improved sparrow search algorithm SSA optimized BP (LC-SSA-BP) neural network model based on logistic chaotic map. This method overcomes the problems of slow convergence, poor stability and easy to fall into local optimization of traditional BP neural network model methods. By optimizing the weight and threshold of BP neural network, the search ability of the group is improved, so as to increase the optimization and optimize the prediction performance. The mining depth, mining thickness, overburden structure, working face slope length and coal seam inclination are selected as the main influencing factors of the height of water-conducting fracture zone. Using 39 groups of training samples and 4 groups of test samples, the LC-SSA-BP neural network prediction model is established and compared with BP neural network algorithm. The results show that the maximum relative errors of BP neural network and LC-SSA-BP neural network are 30.77% and 9.05% respectively. The prediction accuracy of LC-SSA-BP neural network is higher. Finally, using this model, the height of water-conducting fault zone in Shuguang Coal Mine 90301 working face is predicted to be 51.6 m, the error value compared with the engineering verification result is 5.1%.

Keywords

• mining water hazards
• height of water-conducting fracture zone
• lc-ssa optimization algorithm
• bp neural network
• prediction model