时间触发以太网确定性调度设计与研究

随着航空航天领域电子系统向分布式发展,对机载网络高带宽、低延迟、高可靠性的要求越来越严格。传统的CAN、ARINC429等总线正在逐渐被AFDX、TTP、TTE等新一代航空总线技术代替。针对TTE交换式网络,设计了一种新的兼容时间触发流量和标准的事件触发流量的TTE以太网交换机方案,保证了时间触发流量的延时是可控的,并且固定在一定的范围内。经wireshark、Iperf3等软件测试检验了交换机的交换速率、抖动和丢包率;经modelsim软件仿真,并在FPGA硬件实验板上通过嵌入式逻辑分析仪和示波器检验了时间触发流量传输的可靠性、延迟范围和时间同步精度,根据同步周期的不同使同步精度达到了us级、ns级。     

基于时间触发的高可靠性实时系统架构

时间触发架构（TTA）为设计和实现高可靠性分布式嵌入式系统提供了软件和硬件的模型。在TTA系统中，全局时基和各种接口的精确定义，简化了节点间通信，保障了实时应用时限和错误检测。该文描述了TTA模型和TTA设计原则，并提出了基于TTA简单的车载节点模型。     

光纤通道──连接未来网络和通道的桥梁

从结构、协议和应用方面对光纤通道作了较为全面的介绍。     

基于光纤通道的通信系统研究与设计

提出了一种嵌入式平台下基于光纤通道的通信系统设计方案,论述了它的系统结构及实现方法。光纤通道技术是一种高速率的数据通信技术,文章在介绍了光纤通道的技术背景后,分析了基于光纤通道的FCP协议层次,并针对当前嵌入式平台数据传输需求进行了高速通信协议的应用设计。以此为基础,设计并实现了基于光纤通道的通信软件系统的各个模块,并介绍了其中主要模块的实现以及模块间的关系。     

光纤通道网络设计

通过设计Ｇｉｇｗｏｒｋｓ１０６２ＰＣＩ Ａｄａｐｔｅｒ构造的光纤通道网络监控测试软件和基于ＴＣＰ、ＵＤＰ网络应用软件,实现了对光纤通道数传率、拓扑结构、主要类服务和协议层次结构的研究。     

基于DSP的时间触发总线设计与实现

网络通信系统是现代航空电子系统以及其他分布式实时容错系统的中枢,基于时间触发架构的网络通信系统是一个重要的发展方向。首先对于TTP/C协议的基本架构以及上电群启动过程,全局时钟同步,容错策略等关键技术进行介绍,然后提出一种基于TTP的时间触发总线协议,并基于DSP实现协议控制器进行时间触发通讯。结果表明时间触发总线具有良好的时间确定性,容错特性以及较高的总线利用率,适用于航空,铁路,汽车等安全关键系统的底层网络通信。     

光纤通道综述

随着互联网规模的持续扩大和应用上的革命,其性能成为用户至为关心的内容。网络测量作为了解互网性能的基本手段,成为关注热点。介绍了对网络性能进行测量的历史和意义,对网络测量内容和方法进行了分类,着重介绍了时延、带宽、流量等网络性能指标的测量方法,最后就网络测量中误差产生的原因及其处理方法进行了相关论述。     

基于FPGA的时间触发协议控制器实现

时间触发协议是TTA架构必需的通信协议,用于在要求高可靠性的分布式容错实时系统中电子模块之间的互连;目前作为时间触发通信系统重要组成部分的时间触发控制器主要是采用处理器来实现协议的处理,协议开销比较大;基于FPGA的时间触发协议控制器的设计,采用了具有较好同步能力的编码方式和合理的帧格式,在建立全局时间基准的基础上优化了协议处理状态机,利用FP-GA的并行处理能力,降低了协议开销,增加了总线的效率,同时也提高了时钟同步精度和容错能力;仿真结果表明,基于FPGA的时间触发协议控制器具有较好的性能.     

时间触发总线验证技术研究

TTP协议定义了一种高确定性,无冲突,高安全的通信总线,能够满足包括飞行控制等的安全关键实时控制系统的应用要求。时间触发总线验证技术根据TTP协议规范要求,针对研制的节点进行测试,包括基本通信测试、时钟同步、故障注入等不同的测试场景,充分验证被测节点的各项功能、性能。通过这些测试,表明被测节点各项指标都满足研制需求,可用于安全关键实时控制系统。     

时间敏感网络中确定性传输技术综述

时间敏感网络是工业互联网基础设施转型升级的重要研究方向,确定性传输技术是时间敏感网络面向工业控制支撑确定性实时传输需求的核心关键技术,主要包括控制面的时间触发调度技术、转发面的混合传输技术以及确定性时延分析技术.调研了近年来时间敏感网络中确定性传输技术的研究现状并进行了系统地梳理和总结,首先介绍了时间敏感网络不同种类业务流的流量模型;然后基于该流量模型,介绍了控制面的时间触发调度模型、研究现状及其挑战,介绍了转发面时间敏感网络交换机的体系结构、多业务流混合传输策略以及当前方案的不足与改进方向;接着,对整个时间敏感网络的时延模型进行建模分析,介绍基于网络微积分的确定性时延分析技术、研究现状以及可能的改进方向;最后,总结时间敏感网络确定性传输技术所面临的挑战以及未来的研究展望.     

事件/时间触发嵌入式操作系统内核的设计

针对分布式实时系统对操作系统内核的新需要,为避免因时间触发任务相互影响而造成的抖动,研究了时间触发任务的调度器设计,提出了一种改进的调度策略;在开放源代码的uCOSII嵌入式操作系统内核的基础上扩展了时间触发功能,设计了支持事件/时间混合触发的嵌入式实时操作内核TTuCOSII（Time-Triggered Micro Operating System II）。仿真实验表明TTuCOSII具有较高的时间触发精度,良好的调度性能,可以满足事件/时间混合触发的要求。     

Event‐triggered finite‐time reliable control for nonhomogeneous Markovian jump systems

This article investigates the event‐triggered finite‐time reliable control problem for a class of Markovian jump systems with time‐varying transition probabilities, time‐varying actuator faults, and time‐varying delays. First, a Luenberger observer is constructed to estimate the unmeasured system state. Second, by applying an event‐triggered strategy from observer to controller, the frequency of transmission is reduced. Third, based on linear matrix inequality technique and stochastic finite‐time analysis, event‐triggered observer‐based controllers are designed and sufficient conditions are given, which ensure the finite‐time boundedness of the closed‐loop system in an H_∞ sense. Finally, an example is utilized to show the effectiveness of the proposed controller design approach.     

Event‐triggered approach for finite‐time state estimation of delayed complex dynamical networks with random parameters

This article emphasizes the finite‐time state estimation problem for delayed complex dynamical networks with random parameters. In order to reduce the amount of transmission process, an aperiodic sampled‐data event‐triggered mechanism is introduced to determine whether the measurement output should be released at certain time points which incorporate an appropriate triggering condition and sampling moments. Furthermore, a concept of finite‐time boundedness in the pth moment is proposed to access the performance of state estimator. The objective of this article is to design an event‐triggered state estimator to estimate the states of nodes such that, in the presence of time delays, uncertainties, and randomly changing coupling weights, the estimation error system is finite‐time bounded in the pth moment related to a given constant. Some sufficient conditions in form of linear matrix inequalities and algebraic inequalities are established to guarantee finite‐time boundedness. Finally, a numerical example is presented to show the effectiveness of the theoretical results.     

Adaptive event‐triggered global fast finite‐time control for a class of uncertain nonlinear systems

This article designs an adaptive event‐triggered controller to solve the problem of global finite‐time stabilization for a class of uncertain nonlinear systems. By using the symbol function technique, the event‐triggered error is completely compensated, the adaptive technique and the back‐stepping method are simultaneously applied to the controller design, and the new way of designing controller is completed on the basis of fast finite‐time stability theory. Subsequently, taking Lyapunov stability theorem into account, the system stability is proved, and the system is demonstrated by contradiction to be non‐zeno. Finally, giving a simulation example to display the feasibility of this method.     

Event‐triggered practical finite‐time output feedback stabilization of a class of uncertain nonlinear systems

This paper studies the event‐triggered practical finite‐time output feedback stabilization problem for a class of uncertain nonlinear systems with unknown control gains. First, a reduced‐dimensional observer is employed to implement the reconstruction of the unavailable states. Furthermore, a novel event‐triggered output feedback control strategy is proposed based on the idea of backstepping design and sign function techniques. It is shown that the practical finite‐time stability of the closed‐loop systems is ensured by Lyapunov analysis and related stability criterion. Compared with the existing methods, the main advantage of this strategy is that the observer errors and event‐trigger errors can be processed simultaneously to achieve the practical finite‐time stability. Finally, an example is adopted to demonstrate the validity of the proposed scheme.     

Fixed‐time sliding mode cooperative control for multiagent networks via event‐triggered strategy

In this article, the problem of event‐triggered‐based fixed‐time sliding mode cooperative control is addressed for a class of leader‐follower multiagent networks with bounded perturbation. First, a terminal integral sliding mode manifold with fast convergent speed is designed. Then, a distributed consensus tracking control strategy based on event‐triggered and sliding mode control is developed that guarantees the multiagent networks achieve consensus within a fixed time which is independent of initial states of agents in comparison with the finite‐time convergence. Furthermore, the update frequency of control law can be considerably reduced and Zeno behavior can be removed by utilizing the proposed event‐triggered control algorithm. Simulation examples are used to show the effectiveness of the new control protocol.     

Robust event‐triggered control of second‐order disturbed leader‐follower MASs: A nonsingular finite‐time consensus approach

This paper addresses the finite‐time and the prescribed finite‐time event‐triggered consensus tracking problems for second‐order multi‐agent systems (MASs) with uncertain disturbances. The prescribed finite‐time event‐triggered consensus of the second‐order disturbed MASs was obtained for the first time and the controller is nonsingular. Furthermore, a new self‐triggered control scheme is presented for the finite‐time consensus tracking, and the continuous communication can be avoided in the triggering condition monitoring. Hence, the finite‐time consensus tracking can be achieved with intermittent communication. Moreover, Zeno behavior is excluded for each follower. The efficiency of the proposed algorithms is verified by numerical simulations.     

Dynamic self‐triggered output‐feedback control for nonlinear stochastic systems with time delays

In this paper, the dynamic self‐triggered output‐feedback control problem is investigated for a class of nonlinear stochastic systems with time delays. To reduce the network resource consumption, the dynamic event‐triggered mechanism is implemented in the sensor‐to‐controller channel. Criteria are first established for the closed‐loop system to be stochastically input‐to‐state stable under the event‐triggered mechanism. Furthermore, sufficient conditions are given under which the closed‐loop system with dynamic event‐triggered mechanism is almost surely stable, and the output‐feedback controller as well as the dynamic event‐triggered mechanism are co‐designed. Moreover, a dynamic self‐triggered mechanism is proposed such that the nonlinear stochastic system with the designed output‐feedback controller is stochastically input‐to‐state stable and the Zeno phenomenon is excluded. Finally, a numerical example is provided to illustrate the effectiveness of proposed dynamic self‐triggered output‐feedback control scheme.     

Event‐triggered Control for Cluster Consensus in Multi‐agent Networks

This paper considers the issue of cluster consensus for multiple agents in fixed and undirected networks. Agents in a network are supposed to split into several clusters, and a fraction of the agents in each cluster are pinned by virtual leaders. According to the Lyapunov stability theory and graph theory, some appropriate event‐triggered protocols are developed for consensus of the agents belonging to the same cluster, which can greatly reduce both the number of communication updates and that of control actuation updates. Finally, a numerical example is shown to demonstrate the effectiveness of the proposed theoretical results.     

Event‐triggered control via impulses for exponential stabilization of discrete‐time delayed systems and networks

This paper investigates the stabilization issue via event‐triggered controls (ETCs) for discrete‐time delayed systems (DDSs) and networks. Based on the recently proposed ETC scheme for discrete‐time systems without time delays, improved ETC (I‐ETC) and event‐triggered impulsive control (ETIC) are proposed for DDS. The algorithms for ETC, I‐ETC, and ETIC are given respectively to derive criteria of exponential stabilization of DDS. Moreover, the exponential stabilization and stabilization to ISS for discrete‐time delayed networks is achieved by employing the algorithms of ETC and ETIC. The issue of stabilization via ETCs for dynamical networks where different subsystems have different sequences of event instants is solved by introducing the check‐period into ETCs and establishing general ISS estimate of discrete‐time delayed inequality. In order to assess the performances of the control schemes, discussions on nontriviality are given by proposing the concept of rate of control and the function of control cost. Finally, two examples with numerical simulations are presented to demonstrate the effectiveness of theoretical results. From the obtained results on stabilization and the simulations, the ETIC is shown to have clear advantages and well performances than the classical state feedback control, the ETC recently proposed, I‐ETC, and the time‐based impulsive control on aspects of nontriviality, lower rate of control, lower cost of control, and robustness w.r.t. external disturbances.