面向汽车的CPS中混合关键级调度模型及算法研究

Several functions with different criticalities will be integrated into the same platform in mixed-criticality cyber-physical system, and they have different requirements on real-time, safety and reliability. While the current design methodology of embedded system assumed that all functions are with the same level of criticality, thus mixed-criticality cyber-physical system put forward great challenges for existing computation theory and technology of embedded system. To tackle these challenges, this project research on function-level mixed-criticality scheduling model, which can formally model and describe the relation between criticality and real-time, system's resources and cost; research on mixed-criticality scheduling algorithm that combines the time-triggered and event-triggered scheduling approaches, so as to realize the integration of system resource's static configuration and dynamic management, and the united optimization of several non-funtional properties; research on safe and accurate end-to-end real-time analysis algorithm, and to solve the uncertainty problem in real-time brought by networking and integration of functions. Finally, a physical prototype system is constructed to analyze and verify the effectiveness of the proposed models and algorithms. Through this project, it can realize the safety guarantee of mixed-criticality cyber-physical system, and also improve the resource efficiency and service provision. The research achievements of this project will promote the application of mixed-criticality cyber-physical system in key industries such as automotive and aviation, and also be used as the theory basis for the next stage of the development of itself.

混合关键级CPS的同一硬件平台上将同时集成具有不同关键级的多个功能，它们在实时性、安全性和可靠性等方面具有不同要求，但是现有嵌入式系统设计假设系统中所有功能的关键级相等，因此混合关键级CPS对已有嵌入式计算理论和方法提出了严峻挑战。为此，本项目研究功能级的混合关键级调度模型，对关键级与实时性、系统资源和成本之间的关系进行形式化的建模和描述；研究融合时间触发和事件触发的混合关键级调度算法，实现系统资源的静态配置和动态管理之间的有效融合和多个非功能属性的联合优化；研究安全、准确的端到端实时性分析算法，解决网络化和集成化带来的实时性不确定性问题。最后，通过搭建物理原型平台，对提出的模型和算法的有效性进行分析和验证。通过本项目的研究，可在保障混合关键级CPS安全、可靠运行的同时，促进其高效实现和提高其服务质量。研究成果可为混合关键级CPS在汽车、航空等关键工业领域的应用和其下一步发展提供理论依据。

汽车CPS的同一硬件平台上将同时集成具有不同关键级的多个功能，它们在实时性、安全性和可靠性等方面具有不同要求，但是现有嵌入式系统设计理论和方法不支持上述混合关键级系统的设计和实现，因此汽车CPS对已有嵌入式计算理论和方法提出了严峻挑战。为此，本项目拟从调度模型、调度算法、实时性分析和优化算法和汽车CPS原型系统构建技术等方面展开研究，以实现系统资源的静态配置和动态管理之间的有效融合和多个非功能属性的联合优化，解决网络化和集成化带来的实时性不确定性问题和信息安全问题，可在保障汽车CPS 安全、可靠运行的同时，促进其高效实现和提高其服务质量。本项目提出了面向车内CAN网络的混合安全调度模型、CAN多维非功能属性集成优化模型、可实现带宽利用率优化的CAN/CAN FD设计算法、汽车CPS实时性分析和优化算法，并采用英飞凌Aurix多核处理器平台等设计和实现了汽车CPS原型系统。基于上述研究成果，共发表了13篇学术论文（SCI4篇、国内权威期刊3篇）、授权了2项发明专利和2项软件著作权。相关研究成果可为CPS在汽车、航空等关键工业领域的应用和其下一步发展提供理论依据。