基于BP神经网络的安全生产事故统计情况预测分析

通过对1990—2008年安全生产事故统计情况进行分析,梳理了社会经济发展的各项指标,选择其中的人均GDP、第三产业比重、非农就业人口比重、城镇居民可支配收入、万人大学生数、万人医生数等反映经济发展的主要指标作为神经网络的输入端,以安全生产事故起数和死亡人数为输出端,以2007年全国各地区经济发展指标和事故统计情况为样本训练BP神经网络,进行不同省份安全生产情况的预测。其结果与实际情况符合性较好,为目前安全生产形势分析提供了一种新方法,该法对预测安全生产监督管理工作提供参考。

Statistical Prediction and Analysis of Work Safety Accidents Based on BP Neural Network

By analyzing the statistics of work safety accidents in 1990-2008,and summarizing the various indexes of social economic development,the main indexes were selected as the input of neural network,such as per capita GDP,the proportion of tertiary industry,the employment proportion of non-agricultural population,urban per capita disposable income,the number of college students per ten thousand people,the number of doctors per ten thousand people,and so on. The death toll and the number of accidents were selected as the output. The economic development indexes and statistic of work safety accidents across China in 2007 were taken as the training samples of BP neural network,and then the predictions for work safety situation in different provinces were conducted. The results show that the predictions are in good accord with the fact. This study provides a new method for analyzing work safety situation.