基于蚁群算法的机器人足球通信机制

给出了基于蚁群算法实现机器人足球比赛中球员之间协作和通信的机制.选用两只球队分别采用了传统的通信机制和"黑板"通信机制进行实验比较,发现虽然两个球队的整体结构和高层策略相差无几,但是无论从速度、力量还是传接球成功率等方面,Novauro都比ujdb强.该通信机制具有灵活,鲁棒性强的特点,不会由于局部的失误而影响全局的性能.     

Within‐nest abundance of a tropical cockroach Pseudoanaplectinia yumotoi associated with Crematogaster ants inhabiting epiphytic fern domatia in a Bornean dipterocarp forest

Many epiphytic ferns in tropical forest canopies have domatia that provide habitat for arboreal arthropods such as ants. Two species of fern that were predominantly occupied by Crematogaster difformis ants were collected from the canopy of a forest in Borneo. In the fern domatia, two cockroach species, Pseudoanaplectinia yumotoi and Blatta sp., were found to live with C. difformis. Pseudoanaplectinia yumotoi were found in large numbers and accounted for approximately 20% of the total arthropod population, while Blatta sp. comprised only 1% of the arthropods living in the domatia. Behavioural experiments were conducted to evaluate the response of C. difformis workers to newly introduced cockroaches of each species and to allospecific ant workers. Crematogaster difformis workers were highly aggressive and quickly attacked allospecific workers. Both cockroach species could fend off ant attacks; however, dead Blatta sp. were often attacked by the ants while dead P. yumotoi were infrequently bothered.     

Algorithms for degree-constrained Euclidean Steiner minimal tree

A new problem of degree-constrained Euclidean Steiner minimal tree is discussed, which is quite useful in several fields. Although it is slightly different from the traditional degree-constrained minimal spanning tree, it is also NP-hard. Two intelligent algorithms are proposed in an attempt to solve this difficult problem. Series of numerical examples are tested, which demonstrate that the algorithms also work well in practice.     

预警机空天协同作战典型样式效能仿真分析及对策

随着信息技术和空天技术装备的融合发展,预警机空天协同作战已是大势所趋,将大幅提升整体作战效能。以空天一体化作战为核心,以预警机为对象,围绕空天战场环境下预警机协同进攻作战典型样式开展研究;以电子侦察卫星信息支援下预警机协同作战飞机开展远程打击为背景,引入蚁群优化算法对空天协同作战任务方案和作战效能进行仿真分析,从作战理念、力量建设、作战训练和人才培养等方面提出了对策建议。     

基于s-模的反模糊M-半群的性质

本文给出基于s-模的反模糊M-半群的概念,并且探讨基于s-模的反模糊M-半群的性质.     

基于蚁群算法的多目标跟踪方法

提出了一种新的基于蚁群算法的多目标跟踪方法.方法采用蚁群算法实现多目标跟踪中的数据关联,首先将多目标跟踪中的数据关联问题表示为具有约束条件的优化问题.用蚁群算法对该优化问题求解,得到的解即为最优关联.为验证该算法的有效性,在两种状态估计方法EKF(extended Kalman filter)和S1S(sequential importance sampling)的基础上进行了多目标跟踪实验,并且与传统的NN(nearest neighbor)方法进行了比较.在与SIS框架结合时,算法中采样粒子包括状态矢量和关联矢量,状态矢量通过序贯重要性重采样获得,关联矢量通过蚁群算法求得.实验结果表明,将蚁群算法融合进SIS算法进行多目标跟踪是有效的.     

基于MMAS的MIMO-OFDM系统上行多用户检测

在多输入多输出-正交频分复用(multiple input multiple output-orthogonal frequency division multiplexing,MIMO-OFDM)系统上行多用户检测(multi-user detection,MUD)中,针对基本蚁群算法(ant colony optimization,ACO)在搜索过程中易出现过早停滞及收敛于局部最优解等问题,提出一种基于最大最小蚁群系统(max-min ant system,MMAS)的MUD新算法.该算法在蚁群每次循环结束后,仅处于最优路径上的单只蚂蚁释放信息素;同时,通过限制每条路径上信息素的取值范围,避免路径间信息素的差值过大,从而使蚂蚁在每次循环时尽可能地选择不同的路径,提高算法的搜索能力.仿真结果表明,MMAS算法能够有效降低蚁群陷入局部最优解的概率,进而提高了检测性能;同时,随用户数的增加,该算法的计算复杂度却远低于最大似然(maximum likelihood,ML)检测算法,因此,该算法能够在检测性能与计算复杂度之间取得较好的折中.     

基于蚁群搜索理论的优化卡尔曼步频预测

为了使智能下肢能够实时地调节膝关节阻尼,使其固有频率与实际步频相符,达到更加接近正常步态的效果,提出了一种基于蚁群搜索理论的卡尔曼步频预测算法.利用蚁群算法在规定的连续域中寻找卡尔曼预测方程中系统误差与测量误差两个参数的最优组合,实现对卡尔曼预测方程的优化.通过获取六组人行走的步频数据进行实验,结果表明:经蚁群算法优化后,卡尔曼预测算法预测得到的步频与后验值之间的误差比跟随方式的误差分别减少了44.10%,43.42%和36.17%,证明了基于蚁群搜索理论的优化卡尔曼步频预测方法在智能下肢控制上具有良好的应用前景.     

基于蚁群几何优化算法的全局路径规划

将改进的蚁群算法与路径几何优化相结合,用于解决移动机器人的全局路径规划问题.算法结合机器人的越障性能对移动机器人的环境空间进行建模.通过设置初始信息素加快蚂蚁的搜索速度,同时设置自适应信息素挥发机制,解决特定地图中初始信息素的干扰问题;设置自适应路径长度,筛选规划路径的优劣;提出由路径优劣程度决定的信息素散播策略,并从几何原理出发,对规划路径进行优化处理,加快最优解的收敛速度.仿真结果验证了该算法的有效性和普遍应用性,在随机给定的环境地图中,该算法能够迅速规划出最优路径.     

蚁群算法求解离散最小约束去除问题

摘要： 引入蚁群算法解决最小约束去除运动规划问题,在求解过程中对蚁群算法的启发函数以及信息素更新策略进行改进,使其不再易于陷入局部极值并适合求解该问题.仿真实验结果表明,该算法在解的质量和收敛速度上优于精确搜索与贪心算法.