摘要
研究煤矿安全风险准确评估问题,煤矿生产的复杂性导致煤矿事故的动态性、模糊性和随机性,且影响煤矿安全风险等级指标多,指标与风险等级之间呈复杂的非线性关系,导致传统评估方法的准确率低。为了提高煤矿安全风险评估的准确率,提出一种组合的煤矿安全风险评估方法。首先构建出煤矿安全风险评估指标体系,然后采用层次分析法计算各评估指标权重,且采用模糊方法建立判断矩阵,最后将其输入到BP神经网络学习建立煤矿安全风险评估模型。利用具体数据对模型性能进行了验证性测试。实验结果表明,相比较于其它评价方法,组合评价方法提高了煤矿安全风险评估的准确率,是一种有效的煤矿安全风险评估方法。
As the complexity of system, the diversity of safety factors, and nonlinear, the assessing accuracy rate of traditional methods is relatively low. In order to improve risk assessment accuracy of coal mine safety, the paper proposed a combined evaluation method （AHP -Fuzzy） based on BP neural network. First, build a mine safety risk assessment index system, then use the AHP - Fuzzy pre - assessment indicators. Assessment criteria has layers of data, and the expert＇s experience, knowledge, and learning ability were used as BP neural network inputs. And finally, using a mine safety risk data for a confirmatory test. The results show that, compared with the traditional BP neural network model, the algorithm can make full use of expert knowledge and experience and simplify the model structure to improve coal mine safety risk assessment accuracy.
出处
《计算机仿真》
CSCD
北大核心
2012年第2期194-197,共4页
Computer Simulation
关键词
煤矿安全
风险评估
神经网络
Mine safety
Risk assessment
Neural network
