基于改进的新型神经网络在煤矿企业安全管理中的应用  

Application of Improved New Neural Network Based in Safety Management of Coal Enterprised

作　　者：赵飞[1] 王则胜[2] 

机构地区：[1]江苏大学安全工程系,江苏镇江212003 [2]镇江船艇学院,江苏镇江212003

出　　处：《煤炭技术》2011年第4期248-251,共4页Coal Technology

摘　　要：针对煤矿企业生产过程中安全生产状况的模糊性、多因素性等特点,利用Levenberg-Mrquardt优化算法改进BP神经网络并对其进行训练和仿真。并与标准BP算法和动量BP算法进行比较,发现经过改进的网络比标准BP算法和动量BP算法具有更好的效果。According to Fuzzy and multifactorial characteristics of safety production status in chemical enterprises,the auther improved BP neural network and trained and simulated it by using Levenberg-Marquardt optimization algorithm.The auther found the improved BP neural network was better effect the standard BP algorithm and the momentum BP algorithm by comparing the standard BP algorithm and the momentum BP algorithm with the improved BP neural network.The improved BP neural network has better effectiveness for prediction of collision risk index,which can be used conveniently in ship automatic collision avoidance systems.

关 键 词：Levenberg-Mrquardt算法 神经网络 安全管理 煤矿企业 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]