基于模块度的复杂网络社团结构聚类新算法

针对传统的分裂聚类算法删除连边会影响节点划分的问题,结合K-means算法的思想并加以改进,提出了一种以邻居节点为聚类备选集,所有社团中心节点同时聚类的同步聚类算法。该算法根据节点中心度和最短路变化率确定社团中心节点集,然后以中心节点为社团中心,以邻居节点为聚类备选集合进行聚类,完成社团的划分。将算法应用到Zachary网络中并与GN算法、Top Leader算法进行比较,仿真结果表明该算法对网络有较好的划分。     

遗传算法在机翼气动优化设计中的应用

把最优化方法与机翼的气动力求解相结合,进行跨音速机翼的气动优化设计。采用最优化方法为遗传算法,机翼的气动力由三维欧拉方程的数值解来提供。与基准机翼相比较,优化设计的机翼其气动性能有较大幅度的提高,表明遗传算法在机翼优化设计中的可行性和有效性     

Upgrading CCIR's foF2 maps using available ionosondes and genetic algorithms

We have developed a new approach towards a new database of the ionospheric parameter $f_{o}F2$. This parameter, being the frequency of the maximum of the ionospheric electronic density profile and its main modeller, is of great interest not only in atmospheric studies but also in the realm of radio propagation. The current databases, generated by CCIR (Committee Consultative for Ionospheric Radiowave propagation) and URSI (International Union of Radio Science), and used by the IRI (International Reference Ionosphere) model, are based on Fourier expansions and have been built in the 60s from the available ionosondes at that time. The main goal of this work is to upgrade the databases by using new available ionosonde data. To this end we used the IRI diurnal/spherical expansions to represent the $f_{o}F2$ variability, and computed its coefficients by means of a genetic algorithm (GA). In order to test the performance of the proposed methodology, we applied it to the South American region with data obtained by RAPEAS (Red Argentina para el Estudio de la Atmósfera Superior, i.e. Argentine Network for the Study of the Upper Atmosphere) during the years 1958–2009. The new GA coefficients provide a global better fit of the IRI model to the observed $f_{o}F2$ than the CCIR coefficients. Since the same formulae and the same number of coefficients were used, the overall integrity of IRI’s typical ionospheric feature representation was preserved. The best improvements with respect to CCIR are obtained at low solar activities, at large (in absolute value) modip latitudes, and at night-time. The new method is flexible in the sense that can be applied either globally or regionally. It is also very easy to recompute the coefficients when new data is available. The computation of a third set of coefficients corresponding to days of medium solar activity in order to avoid the interpolation between low and high activities is suggested. The same procedure as for $f_{o}F2$ can be perfomed to obtain the ionospheric parameter M(3000)F2.     

空间科学任务协同设计过程优化

为了对概念设计阶段空间科学任务协同设计过程进行合理规划, 减少设计反馈, 降低系统耦合度, 提出了设计结构矩阵(DSM)过程建模和遗传算法(GA) 过程优化算法. 该方法采用DSM对空间科学任务设计活动序列进行建模, 通过DSM描述设计活动间的信息依赖关系, DSM上三角之和代表该设计活动序列设计反馈次数; 将 DSM对应的设计活动序列视为染色体, 采用GA进行序列优化, 最小化设计反馈次数. 通过过程优化算法获取最佳设计活动序列, 优化设计过程, 降低系统耦合度. 空间科学任务实例分析结果表明, 该方法能够有效应用于空间科学任务协同设计的过程建模和过程优化, 指导设计过程的制定.      

多目标融合的冗余空间机械臂碰前轨迹优化

面向在轨对接与装配任务,提出了多目标融合的冗余度空间机械臂碰前轨迹优化方法。通过分析任务特点,阐述了空间机械臂碰前的多重任务目标。针对由空间机械臂连续轨迹跟踪、碰撞姿态控制及碰撞冲量幅值优化三个目标组成的第一类任务,通过设置任务优先级融合多目标并采用基于主任务零空间的方法实现对碰前轨迹的优化;而对由空间机械臂点到点规划、碰撞姿态控制及碰撞冲量幅值优化三个目标组成的第二类任务,通过设置任务权重将多目标融合并采用遗传算法对碰前轨迹进行优化。以典型七自由度机械臂为例开展仿真实验,验证了提出的轨迹优化方法的有效性。     

Optimization of geosynchronous satellite constellation for independent regional navigation and positioning in Middle East region

Regional Navigation and positioning systems using geosynchronous satellites are emerging for different parts of the world. Optimal constellation geometry achievement for a set of geosynchronous satellites is a key sector in designing of such systems. Genetic algorithm is applied for optimization of constellations with 6 and 7 satellites which is to be used in Middle East region where these systems have been deprived of. Based on the fact that the coverage criterion is not sufficient enough to guarantee a good navigation service, positioning process by the satellite systems is discussed and the performance of the system on the basis of position dilution of precision is collected as the fitness criterion. Placing the satellites in groups to pass over the same ground track is applied as a constraint to the genetic algorithm. By simulation, performance of the best obtained results and GPS are compared in terms of position dilution of precision, visibility and the required epochs (time span) to resolve the ambiguities with a 99% success rate, in this region. Also, capability of the GA for such problems is demonstrated.     

基于伴飞模式的补给星轨道设计与优化

针对采用伴飞模式的星座燃料补给问题,本文对其补给星的轨道进行了设计与优化。在给出补给星的轨道设计之后,利用遗传算法对补给星轨道机动所消耗的燃料质量以及每次转移给工作星的燃料质量进行了优化求解,使得补给星燃料利用率达到最高。仿真结果同时给出了燃料利用率与工作星轨道高度以及星座中卫星数目之间的关系,该结论对采用伴飞模式的星座燃料补给研究提供了较好的理论参考。     

一种速度匹配动基座传递对准的方法

介绍了一种速度匹配的动基座传递对准的方法。主子惯导均采用平台系统的指北导航方式,推导了平台对准误差方程,利用主子惯导的速度差信息构建控制回路对子惯导平台的水平失准角进行反馈修正,同时用卡尔曼滤波器估计出方位失准角,从而实现子惯导的初始对准。建立了相应的模型,对该传递对准方法进行了仿真验证,在仿真时考虑了杆臂效应的影响,加入了杆臂效应的算法补偿机制。仿真结果数据跟理论推导基本吻合,验证了方法的可行性。     

钻取式自动采样机构螺旋钻杆结构参数的多目标优化

分析了螺旋钻杆的输出月壤原理和输出月壤阻力（阻力矩）以及综合考虑了螺旋槽内月壤对钻杆驱动力矩的影响、螺旋升角的取值范围和钻杆的轴向钻进功率后建立了钻杆驱动力矩、轴向加载力及功耗的模型。利用该模型分析了钻杆外径、螺旋槽深、槽宽比及螺旋升角等结构参数对钻杆驱动力矩和轴向加载力的影响,获得了钻杆结构参数对钻杆驱动力矩和轴向加载力的影响规律。以钻杆质量和功耗最小为双优化目标,以钻杆外径、螺旋槽深、槽宽比及螺旋升角等结构参数为设计变量,基于遗传算法对钻杆结构参数进行了优化,优化后的功耗较优化前降低了31.8%,钻杆质量减小了23.3%。总体减少了钻探风险,可为钻取式自动采样机构的设计提供理论依据。     

团队进步算法与遗传算法和粒子群算法进行天线阵综合的比较

团队进步算法（TPA）是近期由研究者提出的一种新型优化算法,具有与传统优化算法不同的寻优机制。函数测试结果表明,该算法是一种能够兼顾避免早熟收敛和计算速度的有效的优化算法。文章将团队进步算法、遗传算法和粒子群算法应用于阵列天线方向图综合,给定阵列天线合适的设计要求,用Matlab编制程序对阵列天线进行了优化计算。通过对三种优化算法的综合结果比较,表明新算法在应用于较复杂的阵列天线方面以及在优化性能方面的优越性,显示了新算法在天线设计中的广泛应用前景。