面向服务智能协同的农业物联网动态自治与资源优化配置

The Internet of Things (IoT) is characterized by the mixture and heterogeneity of information from extremely different sources. In order to make a clear understand of the new developing network, collaborative intelligent service and dynamic autonomy are the key issues and also one of the most important problems of science. The project is going to study the basic theories and key technologies of collaborative intelligent service oriented dynamic autonomy and optimized resource allocation with the help of the theory and technology of intelligent network inspired by bio-network intelligence. The method of efficient dynamic information exchange and the self-organization algorithm of net elements in large-scale autonomous network are proposed to form the theory of dynamic autonomy in the IoT. Collaborative management and optimized allocation of network recourses are established to implement collaborative intelligent service innovatively in the autonomy of the IoT. The project will put forward intelligent discovery and monitoring method to construct the dynamic topology and implement self-evolutionary adaption of the service provided by the autonomy of the IoT. The new functions will be added to the information service platform for the IoT in agriculture of Shanghai. Then the prototype of collaborative intelligent service is developed to verify the feasibility of dynamic autonomy, optimized resource allocation and collaborative intelligent service for IoT in agriculture. The results can offer the solutions of various key issues for large-scale heterogeneous net elements in the IoT, such as dynamic autonomy, optimization with incomplete information, and collaborative intelligent service.

物联网存在高度的混杂性、异构性，其动态自治机理和服务智能协同是亟待解决的重要科学问题之一。本项目基于生物网络启发的网络智能理论与技术，研究面向物联网服务智能协同的动态自治与资源优化配置等基础理论和关键技术。研究物联网网元的高效信息动态交互机制和大规模自治网元的动态自组织算法，形成体系化的物联网动态自治理论和方法；建立面向服务智能协同的物联网资源协同管理机制和优化配置方法，实现物联网自治系统的资源配置管理和协同优化策略；创新物联网自治域的服务智能协同机制，提出物联网动态服务的智能发现与监测方法，实现物联网自治域服务的动态拓扑构建及自演化适配；在现有上海市农业物联网信息服务平台的基础上，建立面向农业物联网服务智能协同的原型系统，验证物联网的动态自治、资源优化配置与服务智能协同策略的正确性。以上成果将为物联网大规模异质网元的动态自治、不确定信息的优化利用、服务智能协同等关键问题的解决奠定基础。

本项目针对实际工程应用中存在的重要共性问题，将多路径路由算法与动态资源配置优化紧密结合，在基于物联网的分布式动态协同自治模型、物联网的动态资源配置优化方法、物联网自治域的服务智能协同机制和农业物联网服务智能协同应用的原型系统方面取得了重要进展。主要算法涉及三个方面，首先，提出了无线传感器网络的多路径路由算法与容错策略、农业物联网的动态资源配置优化方法和农业物联网的服务智能协同机制建模。其次，设计了无线传感器网络的多路径路由算法与容错策略包括异构无线传感器网络的智能路由容错策略、规则驱动无线传感器网络的动态免疫分簇多路径路由算法、基于能量均衡的WSN分簇多路径路由算法和事件和规则驱动的动态免疫分簇路由Sink移动策略。第三，提出的农业物联网的动态资源配置优化方法包括基于内分泌调控机理的农业物联网动态监测方法、面向农业物联网的可充电无线传感器网络、基于农业物联网的大田集成监测与智能灌溉系统。最后，农业物联网的服务智能协同机制包括面向农业物联网的智能协同预测、基于农业物联网的资源优化配置、面向工作流任务的云计算资源弹性容错调度、面向工作流任务的农业云计算资源重调度策略。以上成果为农业物联网的研究提供了创新理论和应用方法，并搭建了智能协同实验验证平台。.在国内外期刊与会议上发表论文35篇，其中SCI检索论文17篇；申请国家发明专项10项，登记软件著作权1项；培养博士生7名，硕士生10名。.