基于云模型和危险理论的瓦斯监测系统研究

发布时间：2018-04-01 18:34

本文选题：瓦斯监测　切入点：多传感器信息融合　出处：《江西理工大学》2013年硕士论文

【摘要】：瓦斯是引起煤矿安全隐患的重大因素，一旦发生瓦斯爆炸，不仅对国家财产造成损失，还威胁着工人的生命。目前，瓦斯灾害的煤矿监测系统数据处理方式大都比较单一，整体智能化程度不高，对瓦斯、温度等监测参数采用各自传感器独立处理，缺少多传感器信息的多层次融合。本文将多传感器信息融合技术、云模型和免疫危险理论相结合，提出了具有多层次信息融合和智能化决策的煤矿瓦斯监测系统。采用分批估计理论进行数据层融合，降低井下冲击信号所引起的误差，提高传感器检测数据的准确性；运用遗传算法产生检测器；利用安全状态下运行数据进行训练和优化，以生成异常检测器；运用云模型作为危险感知器，提取危险程度信号。免疫危险理论作为多传感器信息融合的决策层融合，运用免疫危险理论的协同刺激特性对检测到的异常信号和危险信号进行综合处理，，使用加权平均法进行全局态势综合决策。仿真结果表明：分批估计方法可有效降低部分同质传感器受冲击信号的影响；使用免疫异常检测，以及基于云模型的危险信号协同实现瓦斯监测综合决策方法，可有效的提高综合决策的准确性。
[Abstract]:Gas is a major factor that causes the hidden danger of coal mine safety. Once gas explosion occurs, it will not only cause loss to national property, but also threaten the lives of workers. At present, most of the data processing methods of coal mine monitoring system for gas disaster are relatively simple. The whole intelligence degree is not high, the gas, temperature and other monitoring parameters are independently processed by their sensors, and the multi-level fusion of multi-sensor information is lacking. In this paper, the multi-sensor information fusion technology, cloud model and immune hazard theory are combined, and a coal mine gas monitoring system with multi-level information fusion and intelligent decision is proposed. Reduce the error caused by the downhole shock signal, improve the accuracy of sensor detection data; use genetic algorithm to generate detector; use safe operation data to train and optimize to generate abnormal detector; The cloud model is used as the hazard sensor to extract the signal of the degree of danger, and the immune hazard theory is used as the decision level fusion of multi-sensor information fusion. Based on the co-stimulation characteristic of immune hazard theory, the detected abnormal signal and danger signal are processed synthetically, and the global situation synthesis decision is made by weighted average method. The simulation results show that the batch estimation method can effectively reduce the impact of shock signals on some homogeneous sensors, and use immune anomaly detection and dangerous signals based on cloud model to achieve a comprehensive decision method for gas monitoring. Can effectively improve the accuracy of comprehensive decision-making.
【学位授予单位】：江西理工大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：TP274;TD712

【参考文献】