【作者】 马国胜；

【导师】 李彬；

【作者基本信息】 武汉理工大学 ， 控制科学与工程， 2010， 硕士

【摘要】 在我国煤炭行业中,瓦斯事故是当前煤矿生产的最大灾害,目前已成为制约煤矿安全生产的主要因素。因此,为了加强对煤矿生产的安全监控,本文提出了一种基于多传感器融合技术的瓦斯监控系统,可以有效的预防和遏制煤矿事故的发生。本文首先对瓦斯监控系统的发展现状进行了简单的介绍,并针对其研究现状及面临的挑战提出一种新的基于多传感器数据融合方法的瓦斯监控系统。由于煤矿井下环境极其复杂,如果用单一传感器只能获得井下环境的部分信息,不利于对瓦斯的准确检测,经常需要使用多个同类或异类传感器的组合来获得互补的信息或增加抗干扰的能力。因此,本文运用多传感器信息融合的方法对煤矿井下环境的多个数据进行融合处理,以提高整个监控系统的可靠性。本文首次提出了运用局部融合和全局融合两级融合的方法对煤矿井下多传感器数据进行融合。采用多传感器算数平均值与分批估计相结合的算法作为局部融合算法,采用D-S证据理论方法作为全局融合算法。本文对这两种方法进行了研究和理论分析,并结合试验数据验证了这两种融合方法的有效性。根据相关国家标准和行业标准设计了监控系统的总体方案,阐述了监控系统工作原理和要实现的功能。系统能够同时对井下气体中甲烷、一氧化碳等多种危险气体的浓度和井下的温度、湿度等进行检测。当有某气体浓度或现场温度等超标时,井下分站能够在很短时间内将警报信息通过CAN总线传至上位机,上位机能够迅速给出指令对下位机正在进行的工作状态进行调整和报警。根据系统功能进行了瓦斯监控系统硬件电路和软件的设计,硬件部分选用ARM LPC2290作为系统分站的核心处理器,设计了数据采集电路、CAN总线接口电路、后备电源及电源管理电路等。软件部分设计了各个模块的工作流程框图,并分别对其进行了论述。重点分析了CAN总线传输报文的特点,设计了适用于井下分站的传输协议。结果表明整个系统各方面的综合功能和性能,已经初步达到了预期的目的和要求,能够实现对煤矿矿井数据有效的采集,为下一步的改进和完善也提供了重要的数据资料。更多还原

【Abstract】 In China coal industry, coal mine gas accident is currently the biggest disaster, which has become a major constraining factor in production of coal mine safety. Therefore, in order to enhance safety monitoring for coal mine production, a multi-sensor fusion technology based on gas monitoring system is present in this paper, which can effectively prevent and contain coal mine accidents.This paper first simply introduces the development status on gas safety inspection system. According to the modern research status and challenges, this paper puts forward a new gas monitoring system based on multi-sensor data fusion method. As the in-well coal mine environment is complex, if using a single sensor only part information under well can be obtained, which is not easy to do the accurate gas detection. So multiple combinations of the same or different sensors usually need obtain complementary information or increase the capacity of interference. Therefore, in order to improve the reliability of the monitoring system, this paper uses multi-sensor information fusion method to fusion multi-sensor data underground coal mine.The paper first proposes two fusion methods of local integration and global integration to fusion multi-sensor data underground coal mine. Arithmetic average of multi-sensor approach combined with the batch estimation algorithm is used as the local integration algorithm, and the evidence theory is used as a global fusion algorithm. This paper theoretically studies and analysis these two methods, and demonstrates the effectiveness of the two fusion methods by combining the test data.According to the relevant national standards and industry standards the overall program on monitoring system is designed, which explains its working principle and function. The system can simultaneously monitor the concentration of methane, carbon monoxide and other dangerous gases and temperature and humidity in the coal mine. When a certain gas concentration or field temperature exceeds the limits, sub-station in coal mine can quickly transmit alert information to host computer in a very short period of time via CAN bus, then host computer can quickly give instruction to the sub-station in coal mine to adjust the ongoing work status and operation alarm. According to its function of gas monitoring system, respective hardware and software are designed. As for hardware, ARM LPC2290 is used as the core processor of the system, and a data acquisition circuit, CAN bus interface circuit, backup power and power management circuits are designed. As for software, work flow diagram of each module is designed, and the features of each module and work flow chart are discussed. The CAN bus transmission characteristics of packet are mainly analyzed, and transport protocol for the sub-station in the coal mine is designed.The result shows that the system’s comprehensive functionality and performance have basically achieved the desired objectives and requirements, the system can realize effective gathering of coal mine pit data,and the important data information for the next step improvement and perfection can be provided.更多还原