面向区域无线监控的多自主体移动刚性组网与优化控制

Area remote wireless monitoring，such as local catastrophe ，fire prevention of forest, has important practical value.The most methods for the problem are to use normally the random deployment of sensors to realize monitoring, which needs a lot of redundancy sensors and it is difficult to fulfill the complete coverage and reliable communication transmission. In the project, we'll research the problem with respect to the complete coverage of sensors placement and wireless network control for area remote monitoring via multiagent systems(MAS). At first, considering the efficiently and reliably complete coverage of sensors placement, we propose the topology generation approach of k repetitive coverage for sensors placement by using the circle topology extension scheme. Then considering the case that there exist some directed sensors, the time-space covering efficiency is analyzed and computed under the rotating sensors, which can be eneficial to the coverage. In order to improve the reliability of wireless communication transmission, the optimal scheme of rigid topology network for wireless communication is studied. In the study, the schemes of optimally rigid topology are presented via algebra approach. And then the corresponding algorithms of optimally persistent topology are presented via the centre-neighbored extension scheme. We apply these algorithms into wireless communication network, which can not only ensure communication but also can decrease the communication complexity. To realize the obtained topologies, we use multiagent systems to carry out mobile control. Finally, the platform for simulation and experiment is set up to provide technical support and applied protomodel.

局部灾变、森林防火等区域无线监控具有重要的现实意义。现有研究普遍利用无线静态传感器随机部署来实现，需要大量的冗余传感器，且很难实现完全覆盖监控和可靠通信传输。本项目针对区域监控高效可靠覆盖传感器部署和无线多自主体组网优化控制问题开展研究。首先针对区域高效可靠覆盖问题，提出基于圆形结构拓扑扩展策略的传感器节点部署k 重覆盖拓扑生成方法，进一步研究存在有向传感器时基于旋转监控的时空覆盖效率分析与计算，为传感器节点部署提供指导；为提高信息无线通信传输可靠性，研究无线通信刚性组网拓扑优化问题，提出基于代数方法的最优刚性拓扑生成算法，并提出基于中心-邻居拓扑扩展策略的最优持久拓扑生成方法，应用在无线通信组网算法中，保证通信连通鲁棒性的同时降低通信复杂度；提出采用多自主体分布式移动控制实现上述拓扑结构。最后,为实际应用需求，建立多自主机器人移动组网控制仿真实验平台，为实际应用提供技术支撑和应用原型。

本项目针对区域远程监控多自主体移动组网与优化控制问题，围绕传感器节点覆盖控制、组网传输通信拓扑优化，多自主体移动编队控制及实验验证等方面开展了研究。首先，针对无线传感器网络覆盖部署问题，提出了基于最优持久拓扑覆盖策略的无缝覆盖控制算法。其次，考虑区域远程监控大规模传感器通信组网传输问题，提出了一种基于最小刚性图通信拓扑三级分簇优化算法，进一步设计了对任意最优刚性拓扑图都适用的最优持久编队生成算法，较好的保证通信的鲁棒连通，同时有效地简化了通信传输拓扑结构。进一步，基于提出的二重优化无缝覆盖策略，通过设计分布式观测器估计自主体速度和加速度，提出了一种基于虚拟leader跟踪的多自主体编队控制策略，有效地实现了目标区域监控的二重优化无缝覆盖监控。最后，通过购置SRV-1机器人和相关实验耗材，对现有多机器人编队控制实验系统进行了改造，并通过自主研发设计了2个水下机器人，建立了水下机器人组网与控制实验系统，用于对上述研究成果的实验验证和应用拓展。研究成果发表在IEEE Transactions on Cybernetics、IEEE Transactions on Industry Electronics、IET Control Theory and Applications、Information Sciences、自动化学报等国内外著名期刊及会议上，发表和录用论文37篇，其中SCI检索14篇, EI检索18篇，获授权发明专利6项，申请公开专利3项。培养毕业博士研究生1名,2018拟毕业博士研究生1名，毕业硕士研究生8名，获中国海洋学会OI中国水下机器人大赛三等奖1项，优胜奖1项，获燕山大学优秀硕士学位论文1篇。