基于蝙蝠算法优化的BP神经网络煤层冲击危险性智能综合评价研究

Research on the Intelligent Comprehensive Evaluation of Coal Seam Impact Risk Based on the BP Neural Network Optimized by the Bat Algorithm

通过深入分析开采深度、煤层顶底板性质、冲击倾向性、地质构造、开采技术等因素对煤层冲击危险性的影响,建立基于蝙蝠算法优化的BP神经网络煤层冲击危险性智能综合评价模型;将所确定的影响冲击危险性因素进行数量级划分,利用蝙蝠算法对BP神经网络的权值与阈值最优值选择进行优化,对煤层冲击危险性进行无冲击危险性、弱冲击危险性、中等冲击危险性、强冲击危险性的等级评价。应用蝙蝠算法优化BP神经网络煤层冲击危险性智能综合评价模型对内蒙古某矿的煤层及江苏某煤矿工作面进行实例验证,评价结果与综合指数法计算所得结果一致,表明该模型可以用于煤层冲击危险性评价。应用该模型对煤层进行冲击危险性评价时,改善了BP网络结构在权值和阈值确定上的随机缺陷,提高了算法稳定性,因而得到的评价结果更加合理。

An intelligent comprehensive evaluation model of coal seam impact risk based on the BP neural network optimized by the Bat algorithm was established through in-depth analysis of the in-fluence of mining depth,proper⁃ty of top and bottom floors of coal seams,impact tendency,geological structure,mining technology and other factors on coal seam impact risks.The determined factors in-fluencing impact risks are classified by order of magnitude,and the Bat algorithm is employed to optimize the selection of the optimal weight and threshold value of the BP neu⁃ral network,and the rating evaluation is performed for coal seam impact risks in terms of no-impact risk,weak im⁃pact risk,medium impact risk and strong impact risk.The intelligent comprehensive evaluation model of coal seam impact risk based on the BP neural network optimized by the Bat algorithm was used for in-stance validation of coal seams in a mine in Inner Mongolia and a face in Jiangsu Province,and the evaluation results were consistent with the results calculated by the synthetic index method,showing that this model could be used to evaluate coal seam impact risks.When this model is employed to evaluate coal seam impact risks,the random defects in the determina⁃tion of weight and threshold value of the BP network structure can be overcome and the algorithm stability can be im⁃proved,so the evaluation results obtained are more reasonable.