用Boltzmann网络解任务安排问题

本文提出了一种解任务安排问题的Ｂｏｌｔｚｍａｎｎ网络算法，已在Ｔｒａｎｓｐｕｔｅｒ多处理机系统上并行模拟实现，该算法具有优化程度高，运算速度快等特点．     

微波站路径电测标准值估算——利用60Hz测试仪表电测各工作频段的微波路径

在微波通信工程中，判别路径的通断是一项重要的勘察设计工作，常用的方法是利用1:5万的地形图，绘出两站的断面判断出路径是否满足要求。在障碍物阻挡明显时，该方法行之有效；但是，如果障碍物处于临界状态时，图上作业则无法精确判断出路径的通断，需要进行实地电测。     

求解二维装箱问题的强化学习启发式算法

二维带形装箱问题是一个经典的NP-hard的组合优化问题,该问题在实际的生活和工业生产中有着广泛的应用.研究该问题,对企业节约成本、节约资源以及提高生产效率有着重要的意义.提出了一个强化学习求解算法.新颖地使用强化学习为启发式算法提供一个初始的装箱序列,有效地改善启发式冷启动的问题.该强化学习模型能进行自我驱动学习,仅使用启发式计算的解决方案的目标值作为奖励信号来优化网络,使网络能学习到更好的装箱序列.使用简化版的指针网络来解码输出装箱序列,该模型由嵌入层、解码器和注意力机制组成.使用Actor-Critic算法对模型进行训练,提高了模型的效率.在714个标准问题实例和随机生成的400个问题实例上测试提出的算法,实验结果显示：提出的算法能有效地改善启发式冷启动的问题,性能超过当前最优秀的启发式求解算法.     

用于二维不规则排样的离散临界多边形模型

提出了一个用于求解二维不规则排样问题的离散临界多边形模型.Burke等人的BLF算法是求解排样问题的一种有效算法,但其算法对一些特殊实例会产生非法的解.为了解决这个问题,提出了一种基于离散临界多边形模型,并对其正确性作了严格证明.新模型是只含有点和区间的简单模型,在大大降低原问题几何复杂性的同时,也使许多启发式策略可以更容易地求解该问题.计算结果表明,基于离散临界多边型模型的排样算法是很有效的.     

网格监控系统研究

The monitoring system is very critical for complex and large high-performance distributed system such as a Grid. With the help of a monitoring system the users and developers of a Grid can perform tasks such as fault-detec-tion and diagnosis, performance analysis and tuning and performance prediction. And a good monitoring system can give useful data to the scheduling module of the Grid. This paper gives a detailed presentation and analysis of the key techniques of a grid monitoring system, and points out the research directions of the grid monitoring system based on current research around the world.     

求解圆形packing问题的拟人退火算法

Circles packing problem is an NP-hard problem and is difficult to solve. In this paper, a hybrid search strategy for circles packing problem is discussed. A way of generating new configuration is presented by simulating the moving of elastic objects, which can avoid the blindness of simulated annealing search and make iteration process converge fast. Inspired by the life experiences of people, an effective personified strategy to jump out of local minima is given. Based on the simulated annealing idea and personification strategy, an effective personified annealing algorithm for circles packing problem is developed. Numerical experiments on benchmark problem instances show that the proposed algorithm outperforms the best algorithm in the literature.     

环形互连并行系统上的任务分配

This paper is about distributed processing of a divisible task in a ring of communicating processors,all works are based on ring topology.Out distribution scheme aims to guarantee minimal finish time of the whole task,which is divided into parts (subtasks)and scheduled to the individual processors.In our model the communication startup time and delay time are both taken into consideration;we also discuss the cases of returning results or not as well as sending data in prallel or not,and returning results through the ring are our focus for discussion.The volume of subtask assigned to each processor is gotten after creating and solving the linear equations.At last the empirical results are presented.     

Distributing and Scheduling Divisible Task on Parallel Communicating Processors

In this paper we propose a novel scheme for scheduling divisible task on parallel processors connected by system interconnection network with arbitrary topology,The divisible task is a computation that can be divided into arbitrary independent subtasks solved in parallel.Our model takes into consideration communication initial time and communication delays between processors.Moreover,by constructing the corresponding Network Spanning present the concept of Balanced Task Distribution Tree and use it to design the Equation Set Creation Algorithm in which the set of linear equations is created by traversing the NST in Post-order.After solving the created equations,we get the optimal task assignment scheme.Experiments confirm the applicability of our scheme in real-life situations.     

求解矩形装箱问题的一种近似算法

提出了利用近似算法求解二维矩形装箱问题的最小高度的一种方法。该方法基于启发式递归策略和遗传算法。利用启发式递归策略把所有大小各异的矩形都装入宽度固定的矩形容器中，并计算装完后所需容器的高度，用遗传算法的进化能力优化高度，使得所需容器的高度尽可能小。计算数据证明这种方法能够得到很好的结果，特别是对数据量大的测试问题，效果更好。     

构造宏流水线并行算法的一种有效方法

本文基于任务相关、作业排队等理论,提出构造宏流水线并行算法的一种有效方法,并且利用这种方法构成了一组用于数字信号处理的线性和非线性宏流水线并行算法.其中部分算法已应用于分布式自动控制与测试系统FZC-1中,具有并行计算效率高和性能价格比好等优点.