一类CPS中基于模糊理论的系统建模方法研究

The mathematical theory of system model supporting performance analysis of a mixture of discrete and continuous behaviors and its joint dynamics is one of the crucial theory basis in Cyber-Physical Systems (CPS), which are necessary to ascertain and deepen research. This research proposal will use the interaction and mutual feedback control problem between Switched Reluctance Motor (SRM) for electric vehicle and embedded Electronic Control Unit (SRM-ECU) for SRM as the application research issue, and utilize T-S fuzzy system and adaptive fuzzy control theory to produce the system modeling approach of CPS, and then develop the research on the system modeling method based on fuzzy theory in a class of CPS. Firstly, we will propose a novel adaptive fuzzy system for discrete-continous time dynamic hybrid systems and develop its stability analysis, and then solve the compositionality problem that cuts across the heterogeneous cyber and physical aspects of CPS; Secondly, we will constrcut the intelligent control system architecture based on fuzzy theory, and then develop the intelligent control method for the above adaptive fuzzy system, while we also will establish a suitable adaptive hybrid control algorithm with output feedback and model predictive control techniques, therefore, we will solve the self-adaptability problem that can be able to change or be changed itself in order to deal successfully with new complex and dynamic situations; Thirdly, we will study the quantitative estimation of stability regions and algorithmic reachability analysis of adaptive fuzzy systems, which contributes to synthesis of safe initial states as well as switching conditions in order to satify safety system specification in the aforementioned CPS. The present proposal will reveal feedback mechanism of joint dynamics between cyber elements and physical processes along with vehicular real-time network, and then establish the theory foundation in the application domain of CPS. At last, some simulation and physical verification experiments will be given to illustrate the merit and effectivity of the present results, which are based on electric vehicle platform in the laboratory.

支持离散-连续动态性能分析及二者间交互的系统模型数学理论是目前信息-物理融合系统(CPS)的应用研究领域尚需探明与深入研究的关键理论基础之一。本项目以车用开关磁阻电机(SRM)驱动设备与其车载计算控制单元(SRM-ECU)二者间的相互影响与交互为应用对象，将模糊系统和模糊控制引入CPS建模，开展一类CPS中基于模糊理论的系统建模方法研究。提出离散-连续混合的新型自适应模糊系统模型，发展其稳定性理论，解决CPS的异构模型组合问题；构建基于模糊理论的分层智能控制系统结构，发展模糊系统的智能控制方法，建立预测式自适应输出反馈混合控制算法，解决CPS的自适应性问题；研究用于系统安全验证的系统状态稳定区域量化计算和可达性分析与综合的方法。揭示信息组件通过网络控制物理过程的动态融合反馈机制，为CPS的应用领域奠定理论基础。最后，以实验室电动汽车平台为基础展开CPS系统的仿真与物理实验，验证其理论结果。

支持离散-连续动态性能分析及二者间交互的系统模型数学理论是目前信息-物理系统(CPS)的应用研究领域尚需探明与深入研究的关键理论基础之一。本项目以车用信息组件-物理设备通过网络反馈的动态融合为应用对象，基于模糊理论和方法，开展了CPS应用系统建模方法研究。研究了基于CPS的车用开关磁阻电机(SRM)动力学分析与模糊控制方法；提出了离散-连续混合的新型自适应模糊系统模型，探讨了系统的稳定性分析，得到了一些系统与分析结果；构建了基于模糊理论的分层智能控制系统结构，发展了各种CPS应用模糊系统的H∞控制与智能优化方法，建立预测输出反馈的混合控制方法，得到CPS应用结果与控制算法；研究了用于CPS系统功能安全验证的系统状态稳定区域量化计算和可达性分析与综合的方法。同时，在实验室的电动汽车平台上展开了CPS系统的仿真与物理实验，验证了其理论结果。项目执行期间，发表了学术论文20多篇，其中SCI收录12篇，EI/ISTP收录8篇，中科院分区二区以上论文7篇，CCF-A类论文3篇。申请国家发明专利6项，其中1项已获授权。国内外学术会议交流8余次。培养硕士研究生14名（已毕业10名），独立指导博士生1名，联合指导博士生2名。