信息物理融合系统原型平台的系统建模与控制基础理论研究

Cyber-Physical Systems (CPSs) perfectly integrate control, communication and computing with dynamic physical workload, and the widespread applications of CPSs still face some enormous challenges. It's an effective approach to remove these obstacles by adopting event-based control method that includes workload model and interdisciplinary integration. However, the characteristics of CPSs such as dynamic workload, task complexity, resource-constraints, heterogeneous network and distributed systems, reveal that the traditional control methods can not fully meet the application requirements. In this proposal, in order to explore the theories and methods for improving the performance of CPSs, the workload modeling and control method for CPSs are regarded as the main research subjects, and the two means,i. e., the model simulation and prototype platform verification, are adopted. The concrete details include the following aspects: 1) the dynamic workload modeling for CPSs; 2) the event-based control algorithm for CPSs; and 3) the system modeling and implementation verification of CPSs prototype platform. Through this research, the influences of different factors, e. g., workload features, control methods, time parameters and resource allocation, on CPSs performance are revealed, and the CPSs control methods from the standpoint of multiple disciplines are established, which may provide the theoretical and technological foundation for CPSs design, and possess the important theory and practice significances.

信息物理融合系统（Cyber-Physical Systems, CPS）集成控制、通信、计算及动态变化的负载等，其广泛应用还面临许多问题，基于CPS负载模型并融合多学科的事件控制方法是解决该问题的有效措施。然而，CPS 负载随时空变化、任务特性复杂、资源受约束、网络异构、系统分布式等特点决定了传统的控制方法并不能很好的满足CPS应用需求。本项目以CPS的负载建模与控制方法为主要研究对象，采用系统性能分析模型仿真和原型平台验证两种手段，研究提高CPS性能的理论及方法。内容包括：CPS动态负载建模；融合多学科基于事件的CPS控制方法；CPS原型平台的系统建模与实施验证。本项目的实施，可揭示任务负载特性、控制方法、时间参数特性（如延迟、抖动）、资源分配策略等对CPS性能的影响规律，从多学科融合角度建立面向CPS的控制方法，为高性能CPS设计提供理论和技术基础，具有重要的理论意义和工程应用价值。

信息物理融合系统（Cyber-Physical Systems, CPS）集成控制、通信、计算及动态变化的负载等，其广泛应用还面临许多问题，基于CPS 负载模型并融合多学科的事件控制方法是解决该问题的有效措施。然而，CPS 负载随时空变化、任务特性复杂、资源受约束、网络异构、系统分布式等特点决定了传统的控制方法并不能很好的满足CPS 应用需求。本项目以CPS 的负载建模与控制方法为主要研究对象，采用系统性能分析模型仿真和原型平台验证两种手段，研究提高CPS 性能的理论及方法。内容包括：CPS 动态负载建模；融合多学科基于事件的CPS 控制方法；CPS 原型平台的系统建模与实施验证。经过项目的实施，提出CPS系统架构，围绕智能制造、车联网、智能家居、无线传感网络等领域展开理论研究与实验验证，累计发表36篇SCI期刊论文，包括8篇SCI高被引论文。项目负责人主持了6个SCI专刊。项目实施期间, 项目负责人作为General Chair主持了2016 International Conference on Industrial IoT Technologies and Applications (IndustrialIoT 2016)、7th EAI International Conference on Cloud Computing (CloudComp 2016)两个学术会议，培养硕士学位的青年教师1名，硕士研究生3 名。项目的实施，揭示任务负载特性、控制方法、资源分配策略等对CPS 性能的影响规律，从多学科融合角度建立面向CPS 的控制方法，为高性能CPS 设计提供理论和技术基础。