面向多目标站的灵活型公交路径优化调度模型

在已知乘客需求量、车辆载客容量和站点间行程时间的条件下,将车辆的运行时间和乘客出行时间最小化作为目标,构建面向多目标站的灵活型公交路径优化调度模型.该模型采用引力模型进行车辆路径初始化,采用启发式算法对车辆路径进行最优化求解.根据仿真案例结果发现,在乘客需求分布存在较大差异和不确定性时,模型仍能满足所有乘客需求,且车辆总行程耗时较为稳定,系统进行路径优化计算耗时较小,验证了模型及算法的实用性.研究结果表明,面向多目标站的灵活型公交路径优化调度模型能够最大程度满足乘客需求,并在企业成本、乘客时间成本与需求响应方面达到最大平衡,在实际交通中具有重要意义.

基于两阶段注意力层Transformer的弹道目标多站融合识别

多站弹道目标融合识别旨在利用多个雷达站点信息的互补性提升弹道目标识别性能,而传统多站下弹道目标识别方法未直接考虑多站数据间的关联特性,难以取得准确、稳健的识别性能。针对多站下基于雷达散射截面积(RCS)高速飞行目标的识别问题,提出了一种基于两阶段注意力的弹道目标多站融合识别方法。首先,在现有的Transformer模型上添加维度分段模块将多站雷达数据嵌入于二维向量中,保留站内数据时序及站间关联信息;然后,添加了两阶段注意力层,有效地捕获站内时序信息及跨站维度的依赖关系;最后,基于仿真动态RCS数据模拟多站场景开展了融合识别实验。实验结果表明该方法能够有效提升多站条件下的弹道目标识别性能。

基于角度引导Transformer融合网络的多站协同目标识别方法

多站协同雷达目标识别旨在利用多站信息的互补性提升识别性能。传统多站协同目标识别方法未直接考虑站间数据差异问题,且通常采用相对简单的融合策略,难以取得准确、稳健的识别性能。该文针对多站协同雷达高分辨距离像(HRRP)目标识别问题,提出了一种基于角度引导的Transformer融合网络。该网络以Transformer作为特征提取主体结构,提取单站HRRP的局部和全局特征。并在此基础上设计了3个新的辅助模块促进多站特征融合学习,角度引导模块、前级特征交互模块以及深层注意力特征融合模块。首先,角度引导模块使用目标方位角度对站间数据差异进行建模,强化了所提特征与多站视角的对应关系,提升了特征稳健性与一致性。其次,前级特征交互模块和深层注意力特征融合模块相结合的融合策略,实现了对各站特征的多阶段层次化融合。最后,基于实测数据模拟多站场景进行协同识别实验,结果表明所提方法能够有效地提升多站协同时的目标识别性能。

多站多目标雷达数据融合

多站多目标跟踪是雷达数据融合的主要问题。提出了一种对测量数据聚类,在类间进行数据融合,之后基于目标的融合状态,采用强跟踪滤波器(STF)对多目标进行跟踪。利用上述方法,进行了各种态势下的仿真实验。

载人月球科研试验站若干问题思考

为实现月球探测的可持续发展,在实施首次载人登月任务后,将探索建造载人月球科研试验站,实现人在月面长期生存和月球资源开发利用。针对此背景,调研了国外载人月球探测两次热潮提出的发展规划与现状,之后根据载人深空探测发展规律,分析并思考提出以验证人类长期地外生存技术与难点问题、开展月球深度科学研究、资源的探测与开发利用等任务为建站目标;在分析建站难点与发展趋势的基础上,探讨了未来可能分步实施、循序渐进的发展规划与路径;面向战略、科学、工程与经济4大要素,并重点考虑载人航天安全性要求,提出了建站选址的初步思路,最后分析了初步思考的6项关键技术及难点问题与发展规划。

基于距离信息的多站多目标数据融合

在基于距离测量信息的多站系统中 ,将单传感器联合概率数据关联滤波 ( JPDA)推广到了多传感器场合。对 MSJPDA的两种常用实现结构 ,采用深度优先搜索算法 ( DFS)实现可行联合事件的快速生成 。

电子商务网站的规划与管理

本文论述了电子商务网站规则与管理的作用、意义和主要内容 ,以及网站管理的主要模式 ,提出了电子商务网站管理的层次 。

利用COPY-SITECOPYSIT功能快速创建新基站数据

针对移动通信网络维护和优化以及节假日、突发事件期间的移动通信保障工作中,经常遇到的增加新基站、调整基站归属、复制基站数据等工作,本文一改传统的通过命令行方式创建基站数据的复杂、烦琐且易于出错的流程,通过应用实例着重介绍了利用MOTOROLA OMC-R中的Copy-SITE功能直观、简单、快速增加新基站数据的方法,力求简化原有的操作步骤,达到数据准确、过程明晰的目的。

Urban Journalism： Street by Street

The website philadelphianeighborhoods.com provides news coverage about more than 30 underreported and underserved communities in Philadelphia. Affiliated with Temple University, the website and bimonthly printed editions focus on hyperlocal journalism, or journalism at the street-by-street level. The website has won numerous state, regional, and national awards for its coverage. The goals focus on many aspects of journalism, but this paper will address the successes and failures of this hyperlocal approach and the potential applicability for other organizations.

Tracking Target Identification Model Based on Multiple Algorithms

In view of current situation of bad data synchronization, image blurring and tracking station stability in tracking target identification, a kind of tracking target identification model based on multiple algorithms was put forward, firstly establishing the image degradation model, using the wavelet algorithm for image preprocessing, doing image edge segmentation by using Robert algorithm after pretreatment, then using the maximum variance threshold method for image threshold segmentation, then extracting target features from the segmented image, and finally using the ABS algorithm to finish target tracking. Experiments proved the proposed model practical and effective.

Multi-objective optimization of space station short-term mission planning

This paper studies the multi-objective optimization of space station short-term mission planning(STMP), which aims to obtain a mission-execution plan satisfying multiple planning demands. The planning needs to allocate the execution time effectively, schedule the on-board astronauts properly, and arrange the devices reasonably. The STMP concept models for problem definitions and descriptions are presented, and then an STMP multi-objective planning model is developed. To optimize the STMP problem, a Non-dominated Sorting Genetic Algorithm II(NSGA-II) is adopted and then improved by incorporating an iterative conflict-repair strategy based on domain knowledge. The proposed approach is demonstrated by using a test case with thirty-five missions, eighteen devices and three astronauts. The results show that the established STMP model is effective, and the improved NSGA-II can successfully obtain the multi-objective optimal plans satisfying all constraints considered. Moreover, through contrast tests on solving the STMP problem, the NSGA-II shows a very competitive performance with respect to the Strength Pareto Evolutionary Algorithm II(SPEA-II) and the Multi-objective Particle Swarm Optimization(MOPSO).

Application of Matter-element and Information Entropy in Distribution Analysis of Railway Stations Dealing with Refrigeration Containers

Distributing of the railway refrigeration container handling stations is influenced by multitudinous factors, such as technique and equipment conditions, intermodal traffic ability, foreground and social economy conditions of the district. In this paper a maiden adoption of the matter-element is applied to combine with information entropy theory to analyze distributing the railway refrigeration container handling stations based on concerned literature and historical data, and a multi-objective decision-making model on analyzing distributing the railway refrigeration container handling stations is set up based on matter-element and extension model. Based on establishing sutra field and controlled field, we apply relating function of matter- element and extension set to get relating function value of the railway refrigeration container handling stations. Finally a case study is carried out to verify the validity, objectivity and applicability of this model through calculated and compared with traditional analysis methods. Practicality application indicates that this method can avoid subjectivity of weighting, but enhance the extensibility, the model mentioned above has guiding significance and applied value for analyzing distributing the railway refrigeration container handling stations.