一种低开销非阻塞的协同式检查点算法

协同式检查点设置及卷回恢复技术是一种简单有效的容错手段，被广泛地运用于并行/分布式系统中。为进一步降低协同式检查点算法的开销，该文给出了一个基于可重建检查点的非阻塞协同式检查点算法。并行程序出错导致卷回恢复发生的概率远小于检查点设置概率，该算法利用这一特性，将检查点设置的部分开销转至卷回恢复阶段，降低了容错的开销，提高了系统的可扩展性。     

Linux下系统V共享内存的保存与恢复

设置检查点是保存和恢复进程运行状态的一种重要技术，是实现容错、卷回调试和进程迁移的重要手段。本文研究了全透明检查点系统Epckpt在系统Ⅴ共享内存方面的实现方法和不足，给出了自己的改进，从而更好地实现了系统Ⅴ共享内存的保存与恢复。     

双机容错系统中最佳检查点间隔的分析

设置检查点是容错计算机系统进行故障恢复的重要手段。因为检查点间隔选择过大或过小都将使系统性能受到影响，所以检查点间隔的适当选定是系统性能优化的一个重要指标。该文针对双机容错系统，采用检查点设置与回卷恢复的方法提出了一种系统模型，利用马尔科夫链得到了最佳检查点间隔的求解等式，通过实验证实了求解等式的正确性。     

一种改进的同步检查点设置算法

检查点设置与卷回恢复是集群系统中容错计算的重要手段.同步检查点方法在集群系统中得到了广泛应用.为了提高集群计算系统的工作效率,降低系统的容错开销,根据基于消息驱赶的同步检查点设置算法的性质和在实际应用中并行应用程序的通信特征,通过减小协同过程中的阻塞时间,降低系统中控制消息的数量,对基于消息驱赶的Syncand-Stop算法进行优化.改进的算法有效降低检查点设置的时间和空间开销,减小在系统应用中检查点设置的代价,进一步提高系统可扩展性和应用可靠性.     

工作流系统适应性检查点机制的研究

容错技术已经成为工作流的研究热点,设置检查点是一种常用的容错方法。针对工作流系统提出一种适应性检查点机制,该机制通过最优化检查点数量和动态设置检查点间隔,大大提高了错误发生情况下任务按时完成的比率,并通过实验验证了该机制优于传统的检查点机制。     

减少检查点开销的一种方法

设置检查点（ｃｈｅｃｋｐｏｉｎｔｉｎｇ）是容错计算机系统进行故障恢复的重要手段。设置检查点的开销则是影响其性能的一个主要因素。文章提出了一种预先保存部分检查点数据的新方法。该方法不仅能够有效地减少检查点开销,而且具有比较短的检查点延迟。     

复杂失效分布下的动态检查点设置

检查点技术是保障计算机系统可靠性的一种常用方法.通常假定系统失效的发生服从泊松分布,因此检查点是等间隔设置的.但近几年公布的现场数据表明以上假定不合实际.因此,首先利用现场失效数据对固定检查点间距方法的适应性进行分析,并提出两种检查点动态设置方法,它们根据系统前个阶段的失效信息动态地设置下一个检查点.模拟实验结果表明,所提出的方法在复杂失效分布下所获得的效果达到或者优于最优固定检查点的水平.     

ATC中FDP集群进程检查点的研究与实现

针对空中交通管制系统(ATC)中对飞行数据集群处理的可靠性要求,提出了一种基于Linux的用户级进程检查点设置与恢复方案.对基于该Linux用户级的进程检查点的飞行数据集群处理的各个主要模块进行了介绍,在此基础上给出了系统设计框架.从进程的初始化数据段、堆、栈和打开的文件的保存与恢复,给出了该方案的详细实现方法.该进程检查点设置与恢复方案不但可以在主机崩溃重启后恢复进程在重启前的运行状态,更重要的是可以在分布式系统通过进程迁移将保存的进程检查点迁移到其它主机运行,从而有效的提高系统的可靠性,减少运算损失.     

基于PVM的准同步检查点设置方法

检查点是并行系统中实现容错的重要手段，同步检查点方法已广泛应用在工作站机群系统中。PVM所提供的消息传递机制支持高效的异构网络计算，但不支持客错功能。为了降低同步检查点设置的时间开销，提出了一种基于PVM的准同步检查点设置方法，它吸取了同步检查点方法的优点，又通过消息记录方式实现各节点间独立进行状态保存，大大降低了检查点的同步开销，提高了检查点操作效率，该方法在PVM环境下得以实现，实验结果表明所提出的方法具有较好的客错性能。     

一种基于检查点的卷回恢复与进程迁移系统*

ChaRM是一种并行程序后向故障恢复与进程迁移系统.它不仅实现了对工作站机群系统瞬时故障的恢复,而且通过检查点设置时的Mirror存储技术和进程迁移技术,实现了对机群系统结点永久故障的恢复,并支持系统软硬件的在线维护、处理机资源的排他/限时使用和动态负载平衡等功能.文章主要介绍ChaRM系统的检查点设置与回卷恢复、进程迁移等实现技术,并给出了部分性能评测结果.     

Performance analysis of a dependable scheduling strategy based on a fault-tolerant grid model

The grid provides an integrated computer platform composed of differentiated and distributed systems. These resources are dynamic and heterogeneous. In this paper, a novel fault-tolerant grid-scheduling model is presented based on Stochastic Petri Nets (SPN) to assure the heterogeneity and dynamism of the grid system. Also, a new grid-scheduling strategy, the dependable strategy for the shortest expected accomplishing time (DSEAT), is put forward, in which the dependability factor is introduced in the task-dispatching strategy. In the end, the performance of the scheduling strategy based on the fault-tolerant grid-scheduling model is analyzed by an software package, named SPNP. The numerical results show that dynamic resources will increase the response time for all classes of tasks in differing degrees. Compared with shortest expected accomplishing time (SEAT) strategy, the DSEAT strategy can reduce the negative effects of dynamic and autonomic resources to some extent so as to guarantee a high quality of service (QoS).     

Modeling and designing aircraft taxiing patterns for a large airport

At large airports, aircraft takeoff and land simultaneously on multiple runways. As a result, taxiing is a complex and, at times, lengthy process. Planning for aircraft taxiing is an essential design element for the expansion of runways, taxiways, and terminals. Unnecessary taxiing should be reduced. In this study, we develop a taxiing model and calculate the taxiing time. We observed aircraft traffic patterns and listened to the communication of air traffic controllers (ATCs) in order to develop a taxiing model. The developed model was applied to Narita International Airport. We begin by estimating the time required for taxiing at Narita International for 2014. Next, we propose two strategies to reduce the taxiing time. The first strategy involves changes in departure times with controlled intervals between them. In this strategy, the average taxiing time is reduced by 16.5%. The other strategy requires adjustments in the time from the aircraft departure from the terminal to the start of taxiing. In this strategy, the maximum taxiing time is reduced by 11.5%, and the intervals between the terminal departures and taxiing were more uniform.     

Computationally efficient nonlinear Min–Max Model Predictive Control based on Volterra series models—Application to a pilot plant

The mathematical model used in Min–Max MPC (MMMPC) to predict the future trajectory of the system explicitly considers disturbances and uncertainties. Based on the future trajectory, the control sequence is computed minimizing the worst case cost with respect to all possible trajectories of the disturbances and uncertainties. This approach leads to a more robust control performance but also complicates the practical implementation of MMMPC due to the high computational burden required to solve the optimization problem. This computational burden is even worse if a nonlinear prediction model is used. In fact, to the best of the authors’ knowledge, there have not yet been reported any applications of nonlinear MMMPC to real processes. In this paper a nonlinear MMMPC strategy based on a second order Volterra series model is presented. The particular structure of the used prediction model allows to obtain an explicit formulation of the worst case cost and its computation in polynomial time. Real time applications with typical prediction and control horizons are possible because of the reduced complexity of the proposed control strategy. Furthermore, input-to-state practical stability for the proposed control strategy is guaranteed under certain conditions. The MMMPC strategy is implemented and validated in experiments with a continuous stirred tank reactor whose temperature dynamics are approximated by a second order Volterra series model. The control performance of the proposed MMMPC strategy is illustrated by the obtained experimental results.     

Posture Stabilization Strategy for a Trotting Point-foot Quadruped Robot

This paper proposes a posture stabilization strategy for achieving the stable trot gait of a point-foot quadruped robot. Specifically, a stepping strategy (foot placement strategy) has been developed to achieve a stable trot gait. Because in the trot gait of a quadruped robot the diagonal legs can be considered to contact and leave the ground at the same time, the trot gait can be considered as a virtual biped gait. Based on the dynamic model of a virtual biped gait, the stepping point (or the foot placement) that achieves the stabilization of the robot is determined. Finally, the effectiveness of the proposed posture stabilization strategy is validated experimentally.     

低碳环境下双渠道供应链线上线下广告策略的微分博弈分析

以一个由绿色制造商与零售商组成的双渠道供应链为研究对象,基于微分博弈分别比较分析集中式决策、采用竞争型广告策略的分散式决策与采用支持型广告策略的分散式决策模型下,双渠道供应链的最优均衡决策与最优利润.研究发现:集中式决策下的最优低碳水平、最优线上与线下广告水平均高于分散式决策,而产品最优销售价格可能高于、等于或低于分散式决策;分散式决策下,采用支持型广告策略的双渠道供应链的产品最优低碳水平、最优销售价格和最优线上广告投入水平与采用竞争型广告策略的双渠道供应链均衡策略一致;一定条件下,支持型广告策略使得分散式双渠道供应链得到Pareto改进,此时的最优线下广告投入水平优于竞争型广告策略下的最优线下广告投入水平.     

A Wavelet Neural Network Based Non-linear Model Predictive Controller for a Multi-variable Coupled Tank System

In this paper, a novel real time non-linear model predictive controller(NMPC) for a multi-variable coupled tank system(CTS) is designed. CTSs are highly non-linear and can be found in many industrial process applications. The involvement of multi-input multi-output(MIMO) system makes the design of an effective controller a challenging task. MIMO systems have inherent couplings,interactions in-between the process input-output variables and generally have an complex internal structure. The aim of this paper is to design, simulate, and implement a novel real time constrained NMPC for a multi-variable CTS with the aid of intelligent system techniques. There are two major formidable challenges hindering the success of the implementation of a NMPC strategy in the MIMO case. The first is the difficulty of obtaining a good non-linear model by training a non-convex complex network to avoid being trapped in a local minimum solution. The second is the online real time optimisation(RTO) of the manipulated variable at every sampling time.A novel wavelet neural network(WNN) with high predicting precision and time-frequency localisation characteristic was selected for an MIMO model and a fast stochastic wavelet gradient algorithm was used for initial training of the network. Furthermore, a genetic algorithm was used to obtain the optimised parameters of the WNN as well as the RTO during the NMPC strategy. The proposed strategy performed well in both simulation and real time on an MIMO CTS. The results indicated that WNN provided better trajectory regulation with less mean-squared-error and average control energy compared to an artificial neural network. It is also shown that the WNN is more robust during abnormal operating conditions.     

Adaptive gain control for spike-based map communication in a neuromorphic vision system

To support large numbers of model neurons, neuromorphic vision systems are increasingly adopting a distributed architecture, where different arrays of neurons are located on different chips or processors. Spike-based protocols are used to communicate activity between processors. The spike activity in the arrays depends on the input statistics as well as internal parameters such as time constants and gains. In this paper, we investigate strategies for automatically adapting these parameters to maintain a constant firing rate in response to changes in the input statistics. We find that under the constraint of maintaining a fixed firing rate, a strategy based upon updating the gain alone performs as well as an optimal strategy where both the gain and the time constant are allowed to vary. We discuss how to choose the time constant and propose an adaptive gain control mechanism whose operation is robust to changes in the input statistics. Our experimental results on a mobile robotic platform validate the analysis and efficacy of the proposed strategy.     

不确定环境下三级应急系统部分转运策略

针对应急系统中的多点库存共享问题,研究了需求为随机模糊变量情形下的应急调货策略。考虑一个三级多品种的应急供应系统,当缺货发生时,各供应点之间可依据就近应急转运的原则共享部分库存,据此建立了有需求满足时间约束和各供应点库容空间限制的系统总费用随机模糊期望值模型,提出了一种粒子群优化算法和模拟退火算法相结合的先进计算方法(PSO-SA算法)对模型进行了求解,结合算例分析了转运点、就近转运时间、单位物品库容空间等因素变动对部分转运的影响,并验证了算法的有效性和模型的适用性。     

基于随机模型预测控制的插电式混合动力汽车 多目标能量管理策略

为了改善插电式混合动力汽车的燃油消耗和排放, 开展多目标随机模型预测控制策略的研究. 首先, 建立适用于模型预测的多元线性回归的发动机和电池模型, 建立融合燃油消耗和排放的多目标价值函数的模型预测控制, 随后, 基于随机驾驶员模型未来时刻的车速, 结合交通信息并利用动态规划(DP)算法进行参考电荷状态(SOC)优化, 进而建立多目标随机模型预测控制策略. 最后, 通过与DP, MPC等策略进行对比验证, 及给出两组不同权值进行多目标控制效果分析. 结果表明, 该策略的燃油消耗和排放最接近DP的控制效果, 且设置不同权重值可获得相应的控制目标, 说明该策略对提升燃油消耗和排放的多目标性能的有效性.     

Optimization of the separation of two species in a chemostat

In this work, we study a two species chemostat model with one limiting substrate, and our aim is to optimize the selection of the species of interest. More precisely, the objective is to find an optimal feeding strategy in order to reach in minimal time a target where the concentration of the first species is significantly larger than the concentration of the other one. Thanks to the Pontryagin Maximum Principle, we introduce a singular feeding strategy which allows to reach the target, and we prove that the feedback control provided by this strategy is optimal whenever initial conditions are chosen in the invariant attractive manifold of the system. The optimal synthesis of the problem in presence of more than one singular arc is also investigated.