海面分布式传感器多信息融合技术

在海上军事指挥系统及其目标识别中,基于分布式的多传感器信息融合技术一直是海上作战体系中的重要研究课题。由于受海面环境噪声等因素影响,单个传感器对目标物数据的采集往往精度比较低甚至会出现误差。本文重点研究信息融合技术中的数据关联算法,提出一种联合概率数据关联算法,较好的克服了在海面噪声干扰及多径信道干扰条件下的多目标定位跟踪的精度。最后对算法进行实验仿真,通过分析得出此算法的有效性。     

集装箱班轮公司航次运力销售过程优化模型

基于种群演变和共生理论,采用Cobb-Douglas生产函数描述航运市场整体需求,从顾客的购买行为出发,以收益最大作为集装箱班轮公司的经营目标,以基于时间序列的运力与运价作为决策变量,构建了集装箱班轮公司航次运力销售过程优化模型。运用Taylor公式与最小二乘法等代数变换手段将非线性规划问题转化为线性规划问题,对关键参数进行了标定与敏感性分析,并利用MATLAB软件进行仿真验证。仿真结果表明:当单个集装箱班轮公司的运力为104 TEU时,采用常规的销售策略,集装箱班轮公司可售出的运力为7 534~9 966TEU,获得收益为1 233 158~12 915 936USD,采用提出的优化模型,可售出的运力为9 915TEU,获得收益为15 111 975USD,收益至少提高17%;当2个集装箱班轮公司的运力均为104 TEU时,采用提出的优化模型,2个集装箱班轮公司可售出的运力分别为9 920、9 947TEU,获得收益分别为14 241 771、9 737 528USD,达到纳什均衡;当3个集装箱班轮公司的运力均为104 TEU时,采用提出的优化模型,3个集装箱班轮公司可售出的运力分别为8 289、5 526、6 034TEU,获得收益分别为6 755 755、6 119 906、4 377 758USD,达到纳什均衡。可见提出的模型可描述多个集装箱班轮公司运力销售情况,且表现出显著的优化效果。     

极坐标系下的航迹灰色区间关联算法

研究了雷达组网中时变系统偏差存在时的航迹关联问题。在分析时变偏差影响的基础上,从航迹的不确定区域着手,提出一种基于极坐标测量参数的航迹灰区域表示方法,并以此建立了区间灰数关联模型,实现了无需对系统误差配准的航迹关联配对。仿真结果表明,算法是有效的,并且具有较好的抗差性能。     

Spatio-temporal variations in deep-sea demersal communities off the Balearic Islands (western Mediterranean)

The spatial and temporal variations of deep-sea megafaunal assemblages from the western Mediterranean are analysed in the present paper. The assemblages from two locations of the Balearic Islands situated 120 km apart were compared using data collected seasonally on a bathymetric stratum covering the 150–750 m depth range during six bottom-trawl surveys. The assemblage structure, in terms of species composition, species dominance and population sizes, was differentially affected by the spatio-temporal variables analysed (depth, location and fishing period). Although depth was the main factor determining the assemblage composition, the differences obtained between the two locations were also relevant. On the upper slope these between-location differences in the dynamics of megafaunal assemblages were found to be related to the effect of fishing exploitation. Population size-based metrics and biomass spectra were good predictors of meso-scale fishing effects, and were mainly reflected by elasmobranchs and demersal teleosts. Nevertheless, the effects of fishing depended on the species considered. Two dominant large-sized fish species found on the upper slope in both localities, Galeus melastomus and Phycis blennoides, had higher biomass values associated with lower fishing effort. Although the mean body weight (MBW) of both species and also the mean maximum body weight (MMBW) of G. melastomus agreed with this pattern, the P. blennoides MMBW did not. This last case could be indicative of natural size-trends such as the bigger–deeper phenomenon which refers to the displacement of large individuals towards the deeper limit of their bathymetric distribution, beyond the maximum depth sampled in this study for this species. By contrast, the target species of the upper slope fishery, the red shrimp Aristeus antennatus, was not negatively affected by the direct impact of fishing activity and other environmental factors, such as the presence of specific water masses could also be important.     

用改进的灰色关联度分析交通事故成因

改进的灰色关联模型与普通灰色关联模型相比,可以突出关联度的微小变化,且更加科学。通过实例,用改进的灰关联对道路交通事故成因进行了分析,并与普通灰色关联分析进行了比较,发现前者分析结果更具层次性、差异性,更容易识别事故主要、次要和潜在的成因。     

汽车雷达多目标跟踪的模糊数据关联

介绍了一种量测(新测目标)到目标(已知目标)之间的模糊数据关联方法，将模糊聚类均值方法和决策逻辑方法相结合，利用Mahalanobis距离代替聚类运算中的欧氏距离，使量测与目标预测值之间的相似性度量更准确，应用决策逻辑确定量测与目标预测值的关联关系。仿真表明这种方法简单，在杂波环境下，有较高的跟踪精度，这种方法可以用于道路条件下，基于汽车雷达的多目标的可靠跟踪。     

对Apriori算法的进一步改进

介绍了关联规则挖掘的情况，并在分析关联规则的数据挖掘算法的基础上，针对Apriori算法和Apriori-1算法进行深入研究，提出了Apriori-2算法，新算法在计算候选大项集支持度所涉及的记录数目将小于事务数据库中原始的记录数目，提高了原算法的效率，具有一定的实用性．     

编队协同作战中的数据关联技术研究

介绍了编队协同作战系统及其功能,分析了该系统的关键技术之一—数据关联,分析了当前主要的数据关联算法,并指出了其未来发展趋势。     

基于本体的信息融合方法

本体理论在信息融合领域受到越来越广泛的关注和应用。文章概述了当前国内外在本体信息融合理论中的研究概况及取得的部分研究成果,总结了目前本体理论在军事应用领域中较为成熟的一些方法。本体工程在信息融合的第二级别和第三级别的应用较为广泛,相应的本体开发工具包括OWL在文中进行了较为详细的论述。最后,截取C3I系统中一个较为简单的仿真场景来说明怎样进行基于本体信息融合的数据关联。     

Recognizing metro-bus transfers from smart card data

Transfer points between metro and bus services remain an elusive, yet critical junction for transportation practitioners. Based on massive Smart Card (SC) data, previous studies apply a one-size-fits-all criterion to discriminate between transfers. However, this is not sufficiently convincing for different transfer pairs. To counter this problem, this study applies an association rules algorithm and cluster analysis to recognize metro-to-bus transfers using SC data, and demonstrates transfer recognition in a case study based on SC data collected during a week in Nanjing, China. It is shown that 85% of the transfer-recognition results are quite stable through the whole week, and the median transfer time between metro and bus is below 20?min. The method proposed in this study can be used to identify the busiest transfer points and to obtain average transfer times, which facilitates a smarter and more efficient public transit network.