Performance Evaluation of List Sphere Detector

IntroductionThis paper aims to address two problems ithe design of“list sphere detector”(LSD)[1]insystem with no system-level iteration between thMIMO detector and the channel decoder. The firsproblem is how to reduce its complexity. Here it ishown that a much smaller list size can be used ithe soft output is approximated appropriately. Alist size of 40 is adequate for a 4× 4 16QAMVBLAST MIMO system. Compared to a 1 02long list used for the same system in Ref.[1], thilist size redu…     

A Novel Algorithm Based on 3D-MUSIC Algorithm for Localizing Near-Field Source

A novel 3-D MUSIC algorithm based on the classical 3D-MUSIC algorithm for the location of near-field source was presented. Under the far-field assumption of actual near-field, two algebraic relations of the location parameters between the actual near-field sources and the far-field ones were derived. With Fourier transformation and polynomial-root methods, the elevation and the azimuth of the far-field were obtained, the tracking paths can be developed, and the location parameters of the near-field source can be determined, then the more accurate resuits can be estimated using an optimization method. The computer simulation results prove that the algorithm for the location of the near-fields is more accurate, effective and suitable for real-time applications.     

Two Models of DMFC under Effects of Cathode Humidification Temperature and Anode Flow Rate

This paper introduced a novel self-adjustment of parameters of fuzzy neural networks. Then, the effects of cathode humidification temperature and anode flow rate on the performance of direct methanol fuel celI（DMFC） were described respectively. Two dynamic performance models of DMFC under the influences of cathode humidification temperature and anode flow rate were established separately based on fuzzy neural networks. The simulation results show the accuracy of the models established is satisfactory.     

Synthesis and Characterization of Polyurethane and Epoxy Resin Interpenetrating Polymer Network

A series of Polyurethane (PU)/bisphenol A based Epoxy Resin (EP) Interpenetrating Polymer Networks(IPN) were synthesized and characterized by SEM, DSC, TGA and DMTA. It was found that IPN shows the best compatibility and damping properties when the ratio of PU/EP is 80 to 20. The results show that chain-extender and higher molecular weight of PPG are able to improve the properties of compatibility, damping and thermal properties.     

Splitting Balanced Incomplete Block Designs with Block Size 2×k

This paper investigated the existence of splitting balanced incomplete block designs with blck size 2 ×k.The necessary conditions for such a design are λ(v-1)≡0(mod k),and λv(v-1 )≡0 (mod 2k 2). It will show that the above necessary conditions are also sufficient for k = 3 with the definite exception(v,λ)= (10,1) and with several possible exceptions.     

新造船能效设计指数及其对船舶设计的影响

根据国际海事组织(IMO)海上环境保护委员会(MEPC)关于船舶能效设计指数(EEDI)的议题和结论,对目前市场上主流的三型散货船和三型油船进行了计算分析,得出的结论有一定的实用性,为船舶开发设计人员提供了一定的参考.     

叠船运输方式下的半潜船耐波性分析

以叠船运输方式下的半潜船为研究对象,采用CATIA软件进行三维建模。基于AQWA软件计算在规则波中不同浪向、不同航速和不同吃水下的响应幅值算子(Response Amplitude Operator, RAO),分析在不规则波中不同海况条件下货物绑扎设计关键点的加速度变化,发现浪向、航速和吃水等因素对RAO的影响规律。研究结果可为叠船运输方式下的半潜船安全运输提供运动预报,为货物绑扎设计提供依据。     

新的轮式驱动电动车电子差速控制算法的研究

提出了一种用于轮式驱动电动车的电子差速控制算法,将转弯时转矩分配计算和基于车轮滑移率的开关控制相结合,对车辆左右驱动轮输入不同的转矩,同时根据轮胎偏转角的变化率确定目标滑移率。仿真研究证明,与采用机械差速器相比,新的电子差速控制系统鲁棒性好,车辆的驾驶更安全平稳,并能获得更优异的转向性能和更快的响应特性。     

神经网络在客车车身造型美学评价中的应用研究

提出基于人工神经网络（ANN）对客车车身造型进行美学评价的计算机辅助方法,阐述了用于“准三维”造型评价的BP网络结构以及造型特征提取技术。初步创建了智能化客车造型美学评价软件,并将其应用到实际客车造型评价中,使用效果良好。     

船舶AIS大数据资源管理及分析应用架构设计

为充分探索船舶自动识别系统大数据在统计决策和安全监管方面的应用价值,系统性地提出了船舶AIS大数据资源管理与分析应用架构。首先,根据船舶AIS大数据特点,设计了船舶AIS大数据处理流程和存储策略,为后续高效计算提供支撑;然后,提出并实现了基于船舶AIS大数据的船舶轨迹重建算法、断面船舶流量统计算法、船舶进出港区识别算法、船舶航行状态分析算法四种基础算法;最后,基于上述计算方法,将船舶AIS大数据应用于断面流量统计、船舶规范使用船载AIS设备行为监控以及船舶规范执行进出港报告情况监控等场景,结果表明船舶AIS数据在统计决策和安全监管方面具有一定的应用价值。