基于机器协同与理解的可重构智能工业网络理论及关键技术研究

With the application of Internet of Things (IoT) techniques and new computing paradigms in the industry, the smart flexible manufacturing featuring collaborative automation and intelligent optimization becomes a focal research topic. The related investigations focus on the information integration of industrial systems and knowledge-based decision making based on production data analytics. Compared to traditional IoT applications, the Smart Industrial Networks (SIN) have the specificities of tight-coupling between machines, intensive inter-machine communications and complexity of the functionalities. This results in the difficulties in accurate mapping between low-level system resources and physical operations and high-level information models, and technical gap in the reasoning and coordinating of machine operations. This project aims to investigate the theory and methods for the deeply integration of operation technologies and information system in SIN from the system and modeling perspectives considering the features of communication and computation of SIN, and explore the key techniques for software-defined Flexible Manufacturing System (FMS) based on inter-machine collaboration and understanding. The investigations are scheduled as follows: Firstly, building a software-defined industrial field machine collaborative operation mechanism using ubiquitous Machine-to-Machine (M2M) messaging based on non-centralized modularization design. Secondly, improving the interoperability of the system by integrating low-level machine functionality and relationship modeling with high-level industrial process modeling and describing it with universal expressions. Then, exploring new computing paradigm to provide an optimal distribution of computation resources to handle the system management tasks located at the center of the network and industrial fields computation at the edges of the network. This investigation is of great theoretical and practical significance to the Human-Machine-Integration (HMI) and and intelligent innovations of industrial networks and industrial applications.

随着物联网技术与先进计算模式在工业中的应用，以机器协同和智能优化为特征的智能化柔性制造成为研究热点。目前研究集中在系统信息化集成和基于数据分析的知识决策方法等，由于智能工业网络机器间紧耦合、交互密集通信和功能复杂性等特点，使底层资源和物理操作与上层信息模型很难完成准确映射并进行逻辑推理和协同操作。本项目旨在针对智能工业网络通信与计算的特征从系统和模型的角度来研究机器物理操作与信息系统空间融合的理论方法，探索基于工业现场机器智能协同与理解的软件定义可重构柔性制造的关键技术。首先，通过去中心模块化设计构建泛在M2M消息模型来实现软件定义机器协同操作机制；其次，通过对底层机器功能和相互关系与工业过程的统一建模与通用表达促进系统可互操作性；然后，尝试新型计算模式合理分配计算资源，优化位于网络中心和边缘工业现场的计算任务。该研究对于推动工业网络系统的人机融合与智能化升级具有重要理论意义和实际价值。

先进制造系统的信息化和智能化是目前学术研究热点领域，也是各国产业竞争的焦点。本项目针对先进制造智能工业网络系统信息化、智能化面临的难题，对底层资源与工业过程的信息感知与集成、信息计算服务框架以及系统可重构性关键技术进行了研究。首先，针对底层资源与操作过程信息感知难题，提出了集成无线供电、传感、通信的双模射频识别技术。设计并验证了超高频与高频无线供电和数据传输接口集成技术，建立了基于双模射频识别传感器的工业过程信息融合方法。其次，针对先进制造过程的复杂性，构建了面向智能工业网络感知和通信的参考计算框架。通过射频识别感知，将原材料信息、生产过程信息进行集成，并将物流过程产品参数变化获取与集成，供用户通过智能手机进行参数感知和历史数据查询。然后，针对先进制造网络对系统灵活性的要求，构建了智能工业网络的系统可重构技术方案。提出系统全局建模及其轻量化模型表达方法，设计了面向系统功能重构的网络连接模型和功能逻辑，以保证模块间紧耦合密集通信的效率及系统功能的灵活可重构性。.所提出基于双模射频识别传感技术的工业过程信息感知与集成方法，有效集成了超高频和高频无线能量与数据传输接口，为智能工业网络提供了更加灵活、无干扰的传感和通信新途径。基于双模射频识别传感器的参考计算框架将信息链路扩展到产品终端用户，为先进制造过程全生命周期信息感知、融合与共享提供了技术支撑。所构建的建模、表达及计算方法为智能工业网络系统重构及功能的灵活性提高提供了原理模型参考。项目研究对先进制造系统智能化研究及已有系统的升级改造具有重要理论意义和实用价值。.在项目资助下发表SCI检索学术论文5篇，授权实用新型专利1项，申请发明专利2项。培养博士研究生1名，硕士研究生4，其中2名已取得硕士学位。项目投入经费25.00万元，支出17.01万元，设备购置费剩余5.07万元，其余各项支出基本与预算相符。剩余经费7.99万元，计划用于本项目后续研究的支出。