基于混合π网的信息物理融合系统可信软件形式化建模与分析

项目名称： 基于混合π网的信息物理融合系统可信软件形式化建模与分析

项目编号： No.61202128

项目类型： 青年科学基金项目

立项/批准年度： 2013

项目学科： 计算机科学学科

项目作者： 于振华

作者单位： 中国人民解放军空军工程大学

项目金额： 24万元

中文摘要： 信息物理融合系统(Cyber-Physical Systems, CPS)集成了计算、通信与控制功能，是计算进程和物理进程深度协作和有机融合的下一代智能系统。CPS主要应用于安全关键系统中，对软件可信性要求高，因此提高软件可信性已成为CPS研究的热点和难点问题。形式化方法是可信软件研究中必不可少的手段，但单一形式化方法不能全面描述和分析系统的静态、动态结构及行为。本项目结合混合Petri网和π演算，提出一种集成的形式化方法-混合π网，从理论上突破单一形式化方法的不足；从多Agent系统的角度，以混合π网为语义基础，对CPS软件静态和动态结构进行建模，描述其非功能性需求，从而建立CPS可信软件形式化模型；针对上述模型，深入研究动态演化后系统的一致性问题，利用模型检测技术验证系统关键属性的正确性，为CPS软件设计提供可信保障。本项目对促进CPS可信软件理论与技术的发展具有重要的理论和实际意义。

中文关键词： 信息物理融合系统；Petri网；π演算；非线性动力学；可信性演化

英文摘要： Cyber-Physical Systems (CPS) are next-generation intelligent embedded systems that integrate abstract computation and physical processes, in which sensors, actuators, and embedded devices are networked to sense, monitor, and control the physical world. A CPS is coordinated, distributed, and interconnected, and depends on the synergy of computation, communication and control components. As CPS are typically used in safety-critical systems, CPS software must be dependable. Development of high-confidence software is critical to assure the safety and effectiveness of CPS, which has become an exciting and promising research area.Formal methods are indispensable approaches to developing high-confidence CPS software. However, using only one type of formal methods is typically not sufficient to completely specify and analyze the static and dynamic CPS architecture as well as CPS system behaviors. In this project, a new type of formal methods, called hybrid π-nets is presented, which integrates two complementary formalisms, namely hybrid Petri nets and π-calculus. From the perspective of multi-agent systems, we propse a CPS formal model (CPSFM) using the hybrid π-nets. CPSFM can not only support visualizing the static and dynamic CPS architecture, and specifying CPS system behaviors, but also support modeling CPS system

英文关键词： Cyber-physical systems；Petri net；π-calculus；nonlinear dynamics；trustworthiness evolution