A model predictive control approach to the periodic implementation of the solutions of the optimal dynamic resource allocation problem

This paper proposes a model predictive control (MPC) approach to the periodic implementation of the optimal solutions of a class of resource allocation problems in which the allocation requirements and conditions repeat periodically over time. This special class of resource allocation problems includes many practical energy optimization problems such as load scheduling and generation dispatch. The convergence and robustness of the MPC algorithm is proved by invoking results from convex optimization. To illustrate the practical applications of the MPC algorithm, the energy optimization of a water pumping system is studied.     

Optimal feedback scheduling of model predictive controllers

Model predictive control （MPC） could not be reliably applied to real-time control systems because its computation time is not well defined. Implemented as anytime algorithm, MPC task allows computation time to be traded for control performance, thus obtaining the predictability in time. Optimal feedback scheduling （FS-CBS） of a set of MPC tasks is presented to maximize the global control performance subject to limited processor time. Each MPC task is assigned with a constant bandwidth server （CBS）, whose reserved processor time is adjusted dynamically. The constraints in the FS- CBS guarantee scheduler of the total task set and stability of each component. The FS-CBS is shown robust against the variation of execution time of MPC tasks at runtime. Simulation results illustrate its effectiveness.     

树型网格计算环境下的独立任务调度

任务调度是实现高性能网格计算的一个基本问题,然而,设计和实现高效的调度算法是非常具有挑战性的.讨论了在网格资源计算能力和网络通信速度异构的树型计算网格环境下,独立任务的调度问题.与实现最小化任务总的执行时间不同(该问题已被证明是NP难题),为该任务调度问题建立了整数线性规划模型,并从该线性规划模型中得到最优任务分配方案??各计算节点最优任务分配数.然后,基于最优任务分配方案,构造了两种动态的需求驱动的任务分配启发式算法:OPCHATA(optimization-based priority-computation heuristic algorithm for task allocation)和OPBHATA(optimization-basedpriority-bandwidth heuristic algorithm for task allocation).实验结果表明:在异构的树型计算网格环境下实现大量独立任务调度时,该算法的性能明显优于其他算法.     

一种基于QoS的自适应网格失效检测器

失效检测器是构建可靠的网格计算环境所必需的基础组件之一.由于网格中存在大量对失效检测有着不同QoS需求的分布式应用,对于一个网格失效检测器来说,为保持其有效性和可扩展性,应该既能够准确提供应用程序所需的失效检测QoS,又能够避免为满足不同QoS而设计多套失效检测器所产生的多余负载.基于QoS基本评价指标,采用PULL模式主动检测策略实现了一种新的失效检测器--GA-FD(adaptive failure detector for grid),可以同时支持多个应用程序定量描述的QoS需求,不需要关于消息行为和时钟同步的任何假设.同时,证明了GA-FD在部分同步模型下可实现一个◇P类的失效检测器,并给出了相应的实验及数据.     

Ant colony algorithm for automotive safety integrity level allocation

ISO 26262, the new automotive functional safety standard, aims to foster the design and development of safe products by ensuring that the risks posed by hazardous components are reduced to a residual level. Therefore, the standard defines and uses the concept of Automotive Safety Integrity Levels (ASILs) that classify the strictness of safety requirements to be assigned to the failure modes of the system based on the hazard they may cause. ASIL allocation can be described as a hard optimization problem focused on finding the optimal ASIL allocation that maximizes the safety requirements and minimizes cost. However, finding this optimal allocation among a set of possible allocations can represent a difficult task in large systems that contain a large number of components, which subsequently increases the search space. In this paper, we introduce a novel approach that uses the nature-inspired meta-heuristic Ant Colony Optimization (ACO) algorithm to solve the ASIL allocation problem and makes use of strategies that reduce the solution space. The problem was formulated as a construction graph, which the ants use to construct possible ASIL allocations. The search space reduction is accelerated considerably by both the effective performance of the ACO and the convergence of the algorithm on the optimal solution. This approach has been evaluated by applying it to a hybrid braking system and a steer-by-wire system. The results show a significant improvement over genetic-based, penguins search-based and tabu search-based approaches.     

Assignment and scheduling communicating periodic tasks indistributed real-time systems

Presents an optimal solution to the problem of allocating communicating periodic tasks to heterogeneous processing nodes (PNs) in a distributed real-time system. The solution is optimal in the sense of minimizing the maximum normalized task response time, called the system hazard, subject to the precedence constraints resulting from intercommunication among the tasks to be allocated. Minimization of the system hazard ensures that the solution algorithm allocates tasks so as to meet all task deadlines under an optimal schedule, whenever such an allocation exists. The task system is modeled with a task graph (TG), in which computation and communication modules, communication delays and intertask precedence constraints are clearly described. Tasks described by this TG are assigned to PNs by using a branch-and-bound (B&B) search algorithm. The algorithm traverses a search tree whose leaves correspond to potential solutions to the task allocation problem. We use a bounding method that prunes, in polynomial time, nonleaf vertices that cannot lead to an optimal solution, while ensuring that the search path leading to an optimal solution will never be pruned. For each generated leaf vertex, we compute the exact cost using the algorithm developed by Peng and Shin (1993). The lowest-cost leaf vertex (one with the least system hazard) represents an optimal task allocation. Computational experiences and examples are provided to demonstrate the concept, utility and power of the proposed approach     

基于异步奖励深度确定性策略梯度的边缘计算多任务资源联合优化

移动边缘计算（MEC）系统中,因本地计算能力和电池能量不足,终端设备可以决定是否将延迟敏感性任务卸载到边缘节点中执行。针对卸载过程中用户任务随机产生且系统资源动态变化问题,提出了一种基于异步奖励的深度确定性策略梯度（asynchronous reward deep deterministic policy gradient,ARDDPG）算法。不同于传统独立任务资源分配采用顺序等待执行的策略,该算法在任务产生的时隙即可执行资源分配,不必等待上一个任务执行完毕,以异步模式获取任务计算奖励。ARDDPG算法在时延约束下联合优化了任务卸载决策、动态带宽分配和计算资源分配,并通过深度确定性策略梯度训练神经网络来探索最佳优化性能。仿真结果表明,与随机策略、基线策略和DQN算法相比,ARDDPG算法在不同时延约束和任务生成率下有效降低了任务丢弃率和系统的时延和能耗。     

性能受限下云中心异构服务器的能耗优化

针对保证云中心性能下最小化能耗的问题,提出云中心异构服务器之间优化能耗分配方法.首先,建立云中心能耗优化的数学模型;然后,通过拉格朗日乘子法获取该模型的最优解,得到计算最小能量的最小能耗(MPC)算法;最后,通过大量数值实验进行算法验证并与功耗相等分配(EP)基准方法进行了比较.实验结果表明:在相同负载、相同响应时间约束下,MPC算法比EP基准方法节省近30%的能耗,并随着负载增加节省能耗的比例更高.MPC算法可有效避免云中心能源配置过载,为云中心资源优化配置提供思路和参考数据.     

MSs-MEC中基于DRL的服务缓存和任务迁移联合优化算法

多服务移动边缘计算（multiple-services mobile edge computing,MSs-MEC）能根据需求自适应调整服务缓存决策,使得部署在用户侧的边缘服务器能够灵活处理不同服务类型的任务。但在实际应用中,特定类型任务的成功迁移依赖于服务环境的提前安装。此外,同时进行任务迁移和服务缓存可能会因时间冲突而导致计算延时。因此,针对上述相关问题,首先将任务迁移和服务缓存决策进行解耦,针对深度强化学习（deep reinforcement learning,DRL）在具有高维的混合决策空间的性能提升不明显的缺点（例如资源分配时利用率不高）,将DRL与Transformer结合,通过在历史数据中学习,输出当前时隙的任务迁移决策和下一时隙的任务决策,保证任务到达边缘服务器时能立即执行。其次,为了提高资源分配问题中的资源利用率,将问题分解为连续资源分配问题和离散的任务迁移与服务缓存问题,利用凸优化技术求解资源分配最优决策。广泛的数值结果表明,与其他基线算法相比,提出的算法能有效地减少任务的平均完成时延,同时在资源利用率和稳定性方面也有优异的表现。     

Output‐feedback model predictive control for stochastic systems with multiplicative and additive uncertainty

This paper studies the output‐feedback model predictive control (MPC) design problem for linear systems with multiplicative and additive random uncertainty. We first present an off‐line optimization algorithm to optimize feedback gains of the observer and the dual‐mode control policy. After that, by defining a cuboid tube whose center and boundary are both time‐varying variables, we develop a set sequence with increased freedom to contain stochastic system trajectories. A quadratic performance function with analytic upper and lower bounds is minimized such that it decreases exponentially to a finite range under the expectation. The resulting MPC algorithms are proved to guarantee practically stochastic input‐to‐state stability. A numerical example of the wind turbine model illustrates the properties of the MPC algorithms.     

Modelling and solving grid resource allocation problem with network resources for workflow applications

A problem of allocating resources of a grid to workflow applications is considered. The problem consists, generally, in allocating distributed grid resources to tasks of a workflow in such a way that the resource demands of each task are satisfied. Grid resources are divided into computational resources and network resources. Computational tasks and transmission tasks of a workflow are distinguished. We present a model of the problem, and an algorithm for finding feasible resource allocations. A numerical example is included, showing the importance of the resource allocation phase on a grid. Some conclusions and directions for future research are given.     

Task allocation for maximizing reliability of distributed computing systems using honeybee mating optimization

This paper deals with the problem of task allocation (i.e., to which processor should each task of an application be assigned) in heterogeneous distributed computing systems with the goal of maximizing the system reliability. The problem of finding an optimal task allocation is known to be NP-hard in the strong sense. We propose a new swarm intelligence technique based on the honeybee mating optimization (HBMO) algorithm for this problem. The HBMO based approach combines the power of simulated annealing, genetic algorithms with a fast problem specific local search heuristic to find the best possible solution within a reasonable computation time. We study the performance of the algorithm over a wide range of parameters such as the number of tasks, the number of processors, the ratio of average communication time to average computation time, and task interaction density of applications. The effectiveness and efficiency of our algorithm are demonstrated by comparing it with recently proposed task allocation algorithms for maximizing system reliability available in the literature.     

数据驱动的多星任务网络预测调度算法

多星任务调度是具有NP-hard特性的优化问题,随着卫星资源规模和任务需求规模的双重增长,传统调度方法求解效率不高.在轨卫星在常年运行过程中积累了丰富的调度数据.针对大规模多星任务调度场景,建立多星多波束任务调度模型,并提出数据驱动的多星任务网络预测调度算法对其求解.以分割的思想,实现多星场景下任务可调度性预测.从历史调度数据中,提取设定的3个静态特征和5个动态特征,构建并训练预测网络,预测任务被不同卫星完成的概率,并以冲突避免、负载均衡等为原则,得到初始任务和资源卫星的分配方案.进一步设计双链结构的进化算法,以双链编码形式表征上述关系,配合设计的交叉、修复等进化算子,优化初始方案中的任务序列与资源分配关系,输出最终任务调度方案.仿真结果表明,与改进蚁群算法、混合遗传算法和数据驱动并行调度算法相比,所提出算法在运行时间、方案收益和卫星负载均衡3方面均有较好的表现.     

Robust model predictive control under redundant channel transmission with applications in networked DC motor systems

In networked systems, intermittent failures in data transmission are usually inevitable due to the limited bandwidth of the communication channel, and an effective countermeasure is to add redundance so as to improve the reliability of the communication service. This paper is concerned with the model predictive control (MPC) problem by using static output feedback for a class of polytopic uncertain systems with redundant channels under both input and output constraints. By utilizing the min–max control approach combined with stochastic analysis, sufficient conditions are established to guarantee the feasibility of the designed MPC scheme that ensures the robust stability of the closed‐loop system. In terms of the solution to an auxiliary optimization problem, an easy‐to‐implement MPC algorithm is proposed to obtain the desired sub‐optimal control sequence as well as the upper bound of the quadratic cost function. Finally, to illustrate its effectiveness, the proposed design method is applied to control a networked direct current motor system. Copyright © 2016 John Wiley & Sons, Ltd.     

Stochastic model predictive control with joint chance constraints

ABSTRACT

This article investigates model predictive control (MPC) of linear systems subject to arbitrary (possibly unbounded) stochastic disturbances. An MPC approach is presented to account for hard input constraints and joint state chance constraints in the presence of unbounded additive disturbances. The Cantelli–Chebyshev inequality is used in combination with risk allocation to obtain computationally tractable but accurate surrogates for the joint state chance constraints when only the mean and variance of the arbitrary disturbance distributions are known. An algorithm is presented for determining the optimal feedback gain and optimal risk allocation by iteratively solving a series of convex programs. The proposed stochastic MPC approach is demonstrated on a continuous acetone–butanol–ethanol fermentation process, which is used in the production of biofuels.     

Stochastic task selection and renewable resource allocation

A class of renewable-resource-allocation problems is studied for the processing of dynamically arriving tasks with deterministic deadlines. The model presented explicitly considers time available, time required, resources available, resources required, stochastic arrivals of multiple types of tasks, importance of tasks, timeliness of processing, and accuracy of resource allocation. After state augmentation, the problem becomes a Markovian decision problem, and can be solved, at least in principle, by using a stochastic dynamic programming (SDP) method. Effects of key system parameters on optimal decisions are investigated and analyzed through numerical examples     

An algorithm for cooperative task allocation in scalable,constrained multiple robot systems

The current trends in the robotics field have led to the development of large-scale multiple robot systems, and they are deployed for complex missions. The robots in the system can communicate and interact with each other for resource sharing and task processing. Many of such systems fail despite the availability of necessary resources. The major reason for this is their poor coordination mechanism. Task planning, which involves task decomposition and task allocation, is paramount in the design of coordination and cooperation strategies of multiple robot systems. Task allocation mechanism allocates the task in a mission to the robots by maximizing the overall expected performance, and thereby reducing the total allocation cost for the team. In this paper, we formulate a heuristic search-based task allocation algorithm for the task processing in heterogeneous multiple robot system, by maximizing the efficiency in terms of both communication and processing cost. We assume a set of decomposed tasks of a mission, which needs to be allocated to the robots. The near-optimal allocation schemes are found using the proposed peer structure algorithm for the given problem, where the number of the tasks is more than the robots present in the system. The cost function is the summation of static overhead cost of robots, assignment cost, and the communication cost between the dependent tasks, if they are assigned to different robots. Experiments are performed to verify the effectiveness of the algorithm by comparing it with the existing methods in terms of computational time and quality of solution. The experimental results show that the proposed algorithm performs the best under different problem scales. This proves that the algorithm can be scaled for larger system and it can work for dynamic multiple robot system.     

Utilization Bounds for EDF Scheduling on Real-Time Multiprocessor Systems

The utilization bound for earliest deadline first (EDF) scheduling is extended from uniprocessors to homogeneous multiprocessor systems with partitioning strategies. First results are provided for a basic task model, which includes periodic and independent tasks with deadlines equal to periods. Since the multiprocessor utilization bounds depend on the allocation algorithm, different allocation algorithms have been considered, ranging from simple heuristics to optimal allocation algorithms. As multiprocessor utilization bounds for EDF scheduling depend strongly on task sizes, all these bounds have been obtained as a function of a parameter which takes task sizes into account. Theoretically, the utilization bounds for multiprocessor EDF scheduling can be considered a partial solution to the bin-packing problem, which is known to be NP-complete. The basic task model is extended to include resource sharing, release jitter, deadlines less than periods, aperiodic tasks, non-preemptive sections, context switches, and mode changes.     

遥感影像并行处理中基于优先级的任务分配策略

对集群环境下大规模遥感影像并行计算中任务分配效率低、负载不均衡的问题进行分析讨论,在此基础上建立多机任务分配模型,提出一种基于计算节点优先级的任务分配算法。该算法综合考虑计算节点的负载和性能,在任务分配时实时地收集各个节点的信息,计算出各个计算节点的优先级,按照优先级的高低分配任务,保证在满足集群间负载均衡的前提下能合理地将任务分配到计算节点。实验结果表明,该算法能快速实时地进行任务分配,任务的分布更加合理和均匀,并且当任务个数增多时,算法的执行效率要比轮转调度算法高出约2倍。     

Bandwidth‐aware divisible task scheduling for cloud computing

Task scheduling is a fundamental issue in achieving high efficiency in cloud computing. However, it is a big challenge for efficient scheduling algorithm design and implementation (as general scheduling problem is NP‐complete). Most existing task‐scheduling methods of cloud computing only consider task resource requirements for CPU and memory, without considering bandwidth requirements. In order to obtain better performance, in this paper, we propose a bandwidth‐aware algorithm for divisible task scheduling in cloud‐computing environments. A nonlinear programming model for the divisible task‐scheduling problem under the bounded multi‐port model is presented. By solving this model, the optimized allocation scheme that determines proper number of tasks assigned to each virtual resource node is obtained. On the basis of the optimized allocation scheme, a heuristic algorithm for divisible load scheduling, called bandwidth‐aware task‐scheduling (BATS) algorithm, is proposed. The performance of algorithm is evaluated using CloudSim toolkit. Experimental result shows that, compared with the fair‐based task‐scheduling algorithm, the bandwidth‐only task‐scheduling algorithm, and the computation‐only task‐scheduling algorithm, the proposed algorithm (BATS) has better performance. Copyright © 2012 John Wiley & Sons, Ltd.