摘要
针对煤矿瓦斯监测系统的非线性、时变性和多参数强耦合性问题,提出了一种差分进化改进蝙蝠算法(DEBA)优化加权D-S证据融合的煤矿安全监测策略。通过核独立主元分析算法(KICA)对煤矿井下多种传感器采集的原始数据初步处理,进行特征提取,再应用加权D-S证据理论进行数据融合处理,同时采用DEBA算法对加权平均D-S理论的权重进行优化,建立煤矿瓦斯监测模型,可对矿井下的瓦斯状态作出及时判断,并相应地采取决策。MATLAB仿真结果表明:该监测模型能够显著提高煤矿瓦斯监测的精确度和泛化能力以及全局决策的快速性与合理性。
In order to solve the problems of non-linear,time-varying and multi-parameter strong coupling of coal mine gas monitoring system,a coal mine safety monitoring strategy based on differential evolution improved bat algorithm(DEBA)optimized weighted D-S evidence fusion was proposed.The primary data collected by various sensors in coal mines is processed through kernel independent principal component analysis(KICA)algorithm to extract features.Then weighted DS evidence theory is used for data fusion processing.At the same time,DEBA algorithm is used to optimize the weights of weighted D-S theory.The establishment of a coal mine gas monitoring model can make timely judgments on the state of gas under the mine and take corresponding decisions.The simulation results of Matlab show that the monitoring model can significantly improve the accuracy and generalization ability of coal mine gas monitoring and the rapidity and rationality of global decision making.
作者
付华
梁小飞
李涛
司南楠
FU Hua;LIANG Xiaofei;LI Tao;SI Nannan(College of Electrical and Control Engineering,Liaoning Technical University,Huludao 125105,China;Huludao Power Supply Company of Liaoning Electrical Power Company of State Grid,Huludao 125105,China)
出处
《传感器与微系统》
CSCD
2019年第8期143-146,共4页
Transducer and Microsystem Technologies
基金
国家自然科学基金资助项目(71371091)
辽宁省重点实验室项目(LJZS)
