基于BP神经网络的煤矿安全形势预测分析

Coal Mine Security Situation Prediction Based on BP Neural Network

根据2000年至2014年中国煤矿事故统计,运用BP神经网络方法,以MATLAB7.5为操作平台,对煤矿百万吨死亡率进行归一化处理,预测2015年至2020年煤矿安全形势。研究表明,2000年至2014年百万吨归一化死亡率与BP神经网络预测百万吨死亡率吻合程度较高,可为煤矿安全生产提供参考。

According to the accident statistics from 2000 to 2014 in China,by means of BP neural network method and with MATLAB7.5 as the operation platform,unitization processing is conducted on onemillion-ton mortality to predict the may-be security situation from 2015 to 2020.Study shows that the mortality rate after unitization processing almost coincides with the predicted rate,which can serve as a reference for a coal mine production.