摘要
通过对矿井通风系统可靠性运行状态的跟踪统计和分析,首先建立了一套适合于矿井通风系统可靠性评价指标体系;然后利用人工神经网络与粗糙集理论的优势互补,以粗糙集作为前置处理系统优化指标结构,构建了基于粗糙集神经网络的通风系统可靠性评价仿真模型,并依此模型进行了实例验证.结果表明,该模型的仿真结论与基于ANN的结论完全吻合,训练效率提高了数百倍.
Follow the tracks of statistic and analyses to ventilation system reliability running state, set up a suit of reliability evaluating indexes＇ system adapted to it firstly; and then making use of the complementary superiority of ANN and RS each other, taking RS for former disposal system to optimize the indexes structure, constructed ventilation system evaluating simulation model basing on RS and ANN, and put it up to the example validation. The results indicate that the simulating conclusion of this model corresponding to the ANN＇ s entirely, and the training efficiency of it increasing hundreds of times.
出处
《系统工程理论与实践》
EI
CSCD
北大核心
2005年第7期81-86,共6页
Systems Engineering-Theory & Practice
基金
辽宁省教育厅A类计划(202183383)
作者简介
王洪德（1963-），男，辽宁阜新人，博士学位，副教授，硕士生导师．主要从事系统可靠性工程理论、安全评价理论及其应用方面的科研和教学工作．
