煤矿复杂巷道多无人机群组协作监测与区域分析知识迁移

Coal mine wireless monitoring and communication is an important insurance method for automatic production and disaster rescue in coal mine complex environment. The complex environment has great influence on multiple unmanned aerial vehicles group monitoring and communication network structure, energy efficiency design, monitoring efficiency and reliability. Our research proposal is focus on the following. (1) The coal mine complex tunnel’s characteristic will be described by set theory and multi-component system. The area operation scheme for communication and data processing will be modeled. The flight path of unmanned aerial vehicles will be optimized with the influencing factors. (2) Time delay tolerant opportunistic communication and optimization method for sensing data based on cooperative multi unmanned aerial vehicles in divided area will be researched. (3) The area coverage optimization scheme for multi unmanned aerial vehicle group and multi area virtualization mechanism for the object monitoring and communication tasks will be explored. We will study the object monitoring data correlating method between unmanned aerial vehicles. (4) The transfer law between the unmanned aerial vehicle group for area network effective collaboration monitoring and data communication method will be researched. We will analyze the transfer method not only between multi unmanned aerial vehicles but also between multi area networks for collaboration monitoring. In this study, there will be a significant breakthrough in the energy consumption and monitoring efficiency for multi unmanned aerial vehicles, which can solve the contradiction between energy limitation and energy consumption for each task. The research results expect to have important practical application value and scientific significance on improving the automatic production level ,establishing the wireless monitoring and emergency rescue of communication network in coal mine complex environment.

煤矿无线监测与通信是矿井复杂环境中自动化生产和灾害救援的重要保障。巷道复杂环境对多无人机群组监测与通信的网络结构、能效设计、监测效率与可靠性有重要影响，本项目研究：①用集合与多元组方法描述煤矿复杂巷道特征，构建通信与数据处理的区域运算方法；结合无人机飞行影响因素，优化飞行路线。②研究基于协作多无人机群组分区域的感知数据时延忍受机会通信与优化方法。③研究多无人机群组的区域覆盖优化方案和面向目标监测与通信任务的多分区虚拟化机制，分析无人机之间目标监测数据关联方法。④研究多无人机群组区域网络之间的有效协作监测和数据通信方法的迁移规律，分析多无人机群组之间和协作感测多区域网络之间数据分析知识迁移方法。本研究在多无人机群组能效与监测效率方面将有显著突破，解决无人机能量受限与能耗的矛盾。研究成果对于提高煤矿复杂环境中自动化生产水平、快速建立无线监测和抢险救灾应急通信网络有重要的实际应用价值和科学意义。

煤炭是我国能源安全的重要战略保障，国家八部委联合发布《关于加快煤矿智能化发展的指导意见》要求加快智能化煤矿技术研究，着力解决智慧矿山开采的战略性前沿性关键技术难题。煤矿无线监控与通信是井下智能平台系统的重要技术基础，集成通信和图像感知技术的无人机系统是矿井复杂环境中自动化生产和灾害救援的重要保障。巷道复杂环境对多无人机群监测与通信的网络结构、能效设计、监测效率与可靠性有重要影响。针对煤矿井下复杂环境中的无人机飞行优化、群组协作通信、覆盖优化与监测目标数据关联、监测图像质量提升、目标识别与无人机群组分析知识迁移的研究内容，本项目开展2方面内容的研究：1、通信网络组网与控制优化，任务调度与数据管理；2、图像处理方法，目标检测识别方法。研究了基于边缘通信架构和神经网络增强的智能矿山平台，分析了无人机监控系统通信网络的可靠性与检测数据的可靠传输方案，智能数据采集的群智感知方法、信任模型与控制方法，矿山物联网系统的安全架构与无线充电方法；研究了机载网络的任务卸载方法，分析缓存优化架构与关键研究问题；研究了监控图像的传输、高分辨率补齐、增强与重建方法，研究基于智能工业物联网、深度学习和迁移学习的煤矿工人检测、语义分割、识别与人员重识别方法。发表期刊论文33篇，其中，SCI检索26篇，核心期刊论文5篇，SCI源刊在线出版1篇；发表学术会议论文4篇；获江苏省信息技术应用学会科学技术奖三等奖1项。获江苏省科学技术奖三等奖1次。项目负责人于2022年9月获批国家自然科学基金面上项目1项。培养研究生19人，其中1名博士研究生进行了论文答辩。完成了本项目的研究任务。本项目研究成果对于切实推进煤矿“无人”、“少人”值守的智能化生产方法研究、促进煤矿复杂巷道网络的快速智能化巡检方法研究、提升煤矿安全生产和应急救援保障能力有重要科学意义，对煤矿智能无人机、智能机器人及机群的研发和应用有重要的实际价值。