基于关键路径和任务复制的多核调度算法

针对目前大多数多核处理器任务分配优化算法没有考虑关键路径上节点对任务完成时间的重要影响,导致任务完成总时间延迟的问题,提出了基于关键路径和任务复制(CPTD)的单任务调度算法。CPTD算法通过复制任务图中fork节点的方式将任务图转化为与之相对应的产品加工树;再在生成的产品加工树中找到关键路径,并采取使关键路径上节点的紧前节点尽早调度的方式,使关键路径上节点尽早开始执行,进而使产品加工树中节点完成时间得以提前,达到缩短任务执行总时间的目的。理论分析表明,CPTD算法能够实现应用程序在多核上充分并行处理,并能缩短任务完成时间。     

一类分段线性费用网络流问题

本文通过把一类分段线性费用网络流问题化成线性费用网络流问题,给出了求解这类分段线性费用网络流问题的算法。     

考虑弹体动态特性的机动效率约束导引律

针对考虑交会角和过载约束导引律在大机动时能量损失大的问题,提出一种考虑导弹机动效率的多约束制导律。应用最优二次型原理推导出考虑一阶弹体延迟的时变导引系数闭环次优制导形式,将导弹机动时刻阻力系数引入时变权系数,并通过迭代确定机动效率约束边界。将时变约束表示成剩余时间与弹体延迟时间的函数,代入制导指令,进行弹道仿真。结果表明,对于常值与机动目标,文中制导律与过载约束导引律同只考虑交会角约束的导引律相比,对目标均能实现末端弹道成型要求,而考虑机动效率的制导指令分配更为合理,在避免指令加速度饱和的同时有效降低了拦截末端速度损耗,提高制导精度与毁伤效果。且该制导律中时变权系数无须配平求解,在保证精度的同时极大地提高了迭代速度。     

高超音速化学平衡流粘性激波层数值计算

本文给出了层流化学平衡粘性激波层方程对于细长解析体的数值解。这些求解结果是通过空间步进方法得到的,并且与参考文献里的计算结果作了比较,以估价该方法的精确度。结果表明,这种控制方程完全耦合求解的方法能够给出相当精确和稳定的结果,这些结果与参考文献的计算结果相当吻合。     

基于邻域最优控制理论的中制导弹道簇生成

针对临近空间高超声速目标飞行速度快,跟踪预测难的特点,提出了在中制导阶段进行最优弹道设计与弹道簇生成用于对目标预测命中区域进行有效覆盖的方法。首先,通过分析临近空间高超声速目标对现有防空体系带来的挑战,阐明了在中制导段进行弹道簇设计与生成的必要性,其次,将中制导段的弹道规划问题视为求解满足多种约束条件下的最优控制问题,应用最优化理论方法得到了基准的最优弹道,再次,应用邻域最优控制理论,针对终端约束条件进行调整,设计了邻域最优弹道簇的生成算法。最后,通过仿真验证了所提方法的有效性。     

基于FPGA的某发控仪模拟器设计

针对某导弹发控仪模拟器集成度低,难以提高性能的问题,提出了基于FPGA的微型化设计方法,给出了具体的硬件电路和软件设计流程图,对典型电路进行了分析。仿真结果表明实现了原电路的功能,提高了集成度,实现了系统的功能优化。该设计为同类模拟器的研制提供了有效的方法和途径。     

一种控制学科在设计回路的多学科设计优化方法

提出了一种直接针对设计参数不确定性描述的鲁棒控制器设计方法,并将该方法的研究和多学科设计优化方法的研究结合起来,实现了控制学科在设计回路的多学科设计优化。以某无尾布局微型飞行器为例开展了气动、控制学科的并行设计优化研究,说明了该方法的可行性。     

An adaptive forecasting algorithm and inventory policy for products with short life cycles

Products with short life cycles are becoming increasingly common in many industries, such as the personal computer (PC) and mobile phone industries. Traditional forecasting methods and inventory policies can be inappropriate for forecasting demand and managing inventory for a product with a short life cycle because they usually do not take into account the characteristics of the product life cycle. This can result in inaccurate forecasts, high inventory cost, and low service levels. Besides, many forecasting methods require a significant demand history, which is available only after the product has been sold for some time. In this paper, we present an adaptive forecasting algorithm with two characteristics. First, it uses structural knowledge on the product life cycle to model the demand. Second, it combines knowledge on the demand that is available prior to the launch of the product with actual demand data that become available after the introduction of the product to generate and update demand forecasts. Based on the forecasting algorithm, we develop an optimal inventory policy. Since the optimal inventory policy is computationally expensive, we propose three heuristics and show in a numerical study that one of the heuristics generates near‐optimal solutions. The evaluation of our approach is based on demand data from a leading PC manufacturer in the United States, where the forecasting algorithm has been implemented. © 2004 Wiley Periodicals, Inc. Naval Research Logistics, 2004.     

Flow shop scheduling with earliness,tardiness, and intermediate inventory holding costs

We consider the problem of scheduling customer orders in a flow shop with the objective of minimizing the sum of tardiness, earliness (finished goods inventory holding), and intermediate (work‐in‐process) inventory holding costs. We formulate this problem as an integer program, and based on approximate solutions to two different, but closely related, Dantzig‐Wolfe reformulations, we develop heuristics to minimize the total cost. We exploit the duality between Dantzig‐Wolfe reformulation and Lagrangian relaxation to enhance our heuristics. This combined approach enables us to develop two different lower bounds on the optimal integer solution, together with intuitive approaches for obtaining near‐optimal feasible integer solutions. To the best of our knowledge, this is the first paper that applies column generation to a scheduling problem with different types of strongly ????‐hard pricing problems which are solved heuristically. The computational study demonstrates that our algorithms have a significant speed advantage over alternate methods, yield good lower bounds, and generate near‐optimal feasible integer solutions for problem instances with many machines and a realistically large number of jobs. © 2004 Wiley Periodicals, Inc. Naval Research Logistics, 2004.     

共面编队飞行卫星星座的控制

共面编队飞行的卫星星座的相对运动轨迹是一个椭圆,但卫星由于受摄动力的影响,其椭圆队形会发生变化。提出了一个简单而又节约燃料的队形保持控制方案,即在Hill轨道坐标系中,通过测量x轴和z轴,利用小推力喷气发动机进行控制,消除椭圆中心的相位漂移和各卫星轨道的不共面。