多传感器数据融合在煤矿安全预警中的研究与应用

发布时间：2018-07-11 17:21

本文选题：多传感器 + 数据融合　；参考：《宁夏大学》2015年硕士论文

【摘要】：煤矿开采过程中,由于自然环境因素复杂多变,对井下灾害进行事前安全预警较为困难。针对该问题,本文研究利用多传感器数据融合技术对井下安全状态预警的方法。论文对现有多传感数据融合模型进行分类评估,提出井下自然灾害安全预警模型的一般设计原则。按照上述设计原则,提出一种井下多传感器数据融合预警模型的设计方案。该模型采用两级分层融合结构,通过特征层和决策层两层实现数据融合。特征层面向针对单一危险源指标,将主成分分析和神经网络算法相结合,实现危险源特征提取。选取合适的权重系数,利用神经网络对样本数据进行训练,输出预警指标的危险程度。在决策层利用基于D-S证据理论的改进算法,构造BPAsO函数,给出单一危险源的预警决策输出。最后,论文利用MATLAB对预警模型框架进行仿真,以刘庄煤业瓦斯数据为例,将预警结果与实测数据进行比对研究。仿真结果表明,瓦斯单一危险源经融合后,其各种状态的安全判断评估与实测数据相对具有较好的一致性,结果符合实际情况。预警模型的研究对实际井下安全预警具有一定的指导意义。本文研究工作获得国家自然科学基金项目《无线传感器网络数据汇聚传送关键技术研究及在半干旱设施农业中的应用》(基金编号：612610001)的支持。
[Abstract]:In the process of coal mining, due to the complex and changeable natural environment factors, it is difficult to pre-warn the downhole disaster. In order to solve this problem, this paper studies the method of using multi-sensor data fusion technology to predict the underground safety state. In this paper, the existing multi-sensor data fusion model is classified and evaluated, and the general design principles of downhole natural disaster safety early warning model are proposed. According to the above design principle, a design scheme of downhole multi-sensor data fusion early warning model is proposed. The model adopts two-level hierarchical fusion structure, and implements data fusion by feature layer and decision layer. In view of the single hazard index, the feature level is extracted by combining principal component analysis and neural network algorithm. Select the appropriate weight coefficient, use neural network to train the sample data, output the danger degree of early warning index. An improved algorithm based on D-S evidence theory is used to construct the BPAsO function at the decision level, and the decision output of the single hazard source is given. Finally, the paper uses MATLAB to simulate the early warning model frame, taking Liuzhuang coal industry gas data as an example, the early warning results are compared with the measured data. The simulation results show that after the gas single hazard source is fused, the safety judgment and evaluation of its various states are relatively consistent with the measured data, and the results are in line with the actual situation. The research of the early warning model has certain guiding significance to the actual underground safety early warning. The work of this paper is supported by the National Natural Science Foundation of China "Research on key Technology of data Convergence and Transmission in Wireless Sensor Networks and its Application in Semi-arid Facility Agriculture" (Fund No.: 612610001).
【学位授予单位】：宁夏大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TD76;TP212

【参考文献】