基于BP神经网络的事故致因分析方法

Analyzing Methods of Contributing Causes for Accidents Based on BP Neural Network

根据目前所采用的事故致因因素分析方法,提出了一种动态的分析方法,即神经网络分析方法,利用其中的一种神经网络(即BP网络)的原理和方法,根据专家系统提供的较为可靠和代表性数据训练网络,利用训练后的网络对所研究对象之间的相互关系进行分析。分析结果表明,该方法更为直观、准确,尤其在一些其他方法使用困难的场合,更有优越性。

Acconting to analyzing methods of contributing causes for accidents, a dynamic method, namely BP network is provided which is based on the artificial neural network theory. With the reliable and typical data provided by experts, a trained BP network would be obtained which can analyze the contributing causes. The results show that this method is more convenient, prompt and precise.