Adaptive quality of service‐based routing approaches: development of neuro‐dynamic state‐dependent reinforcement learning algorithms

In this paper, we propose two adaptive routing algorithms based on reinforcement learning. In the first algorithm, we have used a neural network to approximate the reinforcement signal, allowing the learner to take into account various parameters such as local queue size, for distance estimation. Moreover, each router uses an online learning module to optimize the path in terms of average packet delivery time, by taking into account the waiting queue states of neighbouring routers. In the second algorithm, the exploration of paths is limited to N‐best non‐loop paths in terms of hops number (number of routers in a path), leading to a substantial reduction of convergence time. The performances of the proposed algorithms are evaluated experimentally with OPNET simulator for different levels of traffic's load and compared with standard shortest‐path and Q‐routing algorithms. Our approach proves superior to classical algorithms and is able to route efficiently even when the network load varies in an irregular manner. We also tested our approach on a large network topology to proof its scalability and adaptability. Copyright © 2006 John Wiley & Sons, Ltd.     

多域光网络中一种支持区分可靠性的通路保护算法

区分可靠性的通路保护方案既能为光网络中具有不同可靠性需求的业务提供有效的故障保护,又能优化网络资源的利用,是一种高效的生存性机制。但目前此类问题的研究主要是集中于传统的单域光网络中,所得的研究成果无法直接运用于具有多域特点的下一代光网络中。针对这一问题,在对以往通路保护算法进行改进的基础上提出了一种适用于多域光网络中的支持区分可靠性的通路保护算法,并对其性能进行了计算机仿真研究。仿真结果表明,该算法不仅能满足多域光网络中不同业务的可靠性需求,而且能提高网络资源的利用率、降低业务连接的阻塞率。     

Dilatation Transformation on Curvilinear Reaction Path

We present a dilatation transformation anlysis on curvilinear reaction path, andanalyze the curvature effect of complex scaling on the prediction of molecular reaction thresholdresonance. An example of numerical calculation was carried out for the prediction of thresholdresonance of H + H₂ molecular scattering with and without consideration of the reactionpath curvature. We found, in the studied system, that the curvature provides a correction tothreshold resonance lifetime as large as 25%.     

PATH INTEGRAL SOLUTION OF NONLINEAR DYNAMIC BEHAVIOR OF STRUCTURE UNDER WIND EXCITATION

IntroductionThe dynamic behavior of the nonlinear structure under wind excitation has beenobserved very complicated.Taking guyed masts as an example,only a few collapsingaccidents occurred under extreme atmospheric conditions[1],many took place under mild…     

共边排样件激光切割路径的规划

排样软件的应用使材料利用率得到了很大提高，然而后续切割软件若不能保证有效地切割零件、保证零件质量及提高生产率，则排样软件在材料利用率上获得的收益将丧失。基于图论学理论，建立了规则与非规则零件共边排样时激光切割路径规划的数学模型，给出了在充分考虑加工质量、加工效率、制造成本情况下的激光切割路径优化目标：打孔点最少以及切割中割嘴空行程最短。提出了满足激光切割工艺要求的三个切割路径优化算法：用于求解理想情况下共边切割路径优化问题的一个新的欧拉回路算法；基于奇度顶点完全图最小权最大匹配算法来求解一般情况下共边切割路径优化问题的算法；利用废料区域进一步减少打孔点的处理策略与求解算法。给出了各种算法的运行实例，验证了所提出的算法的有效性。     

几类凝聚图的轮廓

设Ｇ是个图，｜Ｖ（Ｇ）＝ｎ｜对Ｇ上的任一个标号ｆ：Ｖ（Ｇ）→｛１，…，ｎ｝记，且当ｊ≠ｉ时，Ｇ中有边以ｆ（－１）（ｊ）及ｆ（－１）（ｉ）为两端点｝）．称Ｐ（Ｇ）＝ｍｉｎ｛Ｐ（ｆ）：ｆ是Ｇ上的标号｝为图Ｇ的轮廓．对以Ｗ表示Ｇ中Ｗ的边界．本文证明：ｉ）若Ｇ是凝聚图，ｆ及ｆ是Ｇ上一对互逆标号，则Ｐ（Ｇ）＝Ｐ（ｆ）的充要条件是ｆ为凝聚标号，且此时若Ｇ，Ｈ均是凝聚图，则存在阶梯标号。使得路、回、完全留之间的下列乘积图也是凝聚图，且其轮廓为     

一种改进RRT的多机器人编队路径规划算法

多机器人路径规划是机器人领域的一个热点问题,相比于单机器人路径规划,其算法难度和复杂度都有所增加,在规划时需要兼顾多机避障、相互协作等难点问题.本文提出一种改进快速扩展随机树的多机器人编队路径规划算法,用于解决多机器人在复杂环境下的编队路径规划问题.针对多机器人在编队规划中的位置约束问题,定义机器人之间的领航-跟随结构,并对机器人队形建模.针对规划过程中编队朝向变化问题,建立搜索树扩展方向与队形方向之间的联系,通过调整队形方向改变规划时的编队朝向.针对具有质点模型和非完整约束动力学模型两种不同模型的多机器人系统,分别进行了仿真实验.仿真结果表明该算法在处理多机器人编队路径规划问题时可以取得良好的效果.     

动态环境下基于多人工鱼群算法和避碰规则库的机器人路径规划

为了提高机器人路径规划的速度、环境适应能力和高效动态避碰问题,提出了一种基于多人工鱼群的机器人路径规划算法和基于避碰规则库的动态避障算法.该算法中,人工鱼以其与目标点的距离为食物浓度,两个邻近栅格的距离为步长,其觅食行作为默认行为,在一定条件下执行聚群或追尾动作,并采用两鱼群双向搜索机制在静态环境下规划出较优路径.在此基础上,机器人查询动态避障规则库获得避碰方法,从而实现与动态障碍的避碰.大量仿真实验结果表明,该方法具有较高的收敛速度和较强的搜索能力,能在非常复杂的动静态障碍环境中,迅速规划出一条安全避碰的优化路径.     

Capacity Efficiency and Restorability of Path Protection and Rerouting in WDM Networks Subject to Dual Failures

Resilient optical networks are predominately designed to protect against single failures of fiber links. But in larger networks, operators also see dual failures. As the capacity was planned for single failures, disconnections can occur by dual failures even if enough topological connectivity is provided. In our approach the design of the network minimizes the average loss caused by dual failures, while single failures are still fully survived. High dual failure restorability is the primary aim, capacity is optimized in a second step. For WDM networks with full wavelength conversion, we formulate mixed integer linear programming models for dedicated path protection, shared (backup) path protection, and path rerouting with and without stub-release. For larger problem instances in path rerouting, we propose two heuristics. Computational results indicate that the connectivity is of much more importance for high restorability values than the overall protection capacity. Shared protection has similar restorability levels as dedicated protection while the capacity is comparable to rerouting. Rerouting surpasses the protection mechanisms in restorability and comes close to 100% dual failure survivability. Compared to single failure planning, both shared path protection and rerouting need significantly more capacity in dual failure planning.     

Variational transition state theory without the minimum-energy path

In this paper we propose a method for carrying out variational transition state theory calculations without first obtaining a converged minimum-energy path (MEP). We illustrate the method in two ways, first of all by employing an unconverged MEP and secondly by using a dynamically optimized distinguished reaction path. Preliminary tests of the algorithm for the reactions OH+H₂→H₂O+H and C₂H₅→C₂H₄+H are very encouraging. Received: 22 January 1997 / Accepted: 11 March 1997