煤矿本质安全管理评价的GA-NN模型及应用

GA-NN Model and Application of Coal Mine Inherent Safety Management Evaluation

为评价煤矿本质安全管理水平,根据煤矿本质安全管理的内涵,构建了评价指标体系。针对以往评价方法主观设置指标权重的缺陷,采用神经网络对煤矿本质安全管理进行评价。鉴于遗传算法全局搜索最优解的特点,将遗传算法用于获取神经网络权重的最优值,建立了煤矿本质安全管理评价的遗传优化神经网络模型(GA-NN)。测试结果表明,基于GA-NN模型的评价方法具有较高的精度,且无需人为设置指标权重,避免了人的主观因素对评价结果的影响,能够更客观、准确地得出评价结果,有利于监管部门评价煤矿安全管理水平及企业内部的评比与管理,对建立煤矿安全管理机制具有重要作用。

In order to evaluate the level of the mine inherent safety management,an index system was designed base on the connotation principle of the mine inherent safety management.According to the disadvantage of index weight setting by subjective idea in the former method,neural networks were used to assess the level of coal mine in-herent safety management.Due to the advantages of global search optimal solution of genetic algorithm,neural net-works weights optimization method was proposed based on genetic algorithm,and genetic algorithm-neural network model of coal mine inherent safety management assessment was established.The test results show that the genetic al-gorithm-neural network model has higher evaluation accuracy,and the advantage of no need artificial setting the in-dex weight and absence of the subjective factors influence to evaluation results.It can more objectively and accurately obtain the evaluation results.Not only conducive to evaluate the level of mine safety management for coal mine safety supervision department and the competition and management of enterprise interior ,but also play important roles to es-tablish a mechanism for coal mine safety management.