基于形态检测与深度学习的高空视频车辆识别

为了准确提取广域场景道路交通信息,本文融合形态检测与深度卷积网络,提出了无人机视频车辆定位及车型识别方法. 首先,基于形态检测建立候选目标提取算法,并构建了含244 520 个无人机视频车辆样本的深度学习图像基准库;然后,通过增加卷积层、池化层及调整网络参数等方法对AlexNet 进行重构,提出了改进模型AlexNet*;最后,建立了基于候选目标提取算法与AlexNet*的车辆识别方法. 验证分析显示：AlexNet*的图像分类F1 均值达 85.51% ,优于AlexNet(82.54% )、LeNet(63.88% )、CaffeNet(46.64% )、VGG16(16.67% ) 及 GoogLeNet(14.38%);本文车辆识别方法对小汽车及公交车的正检率、重检率和漏检率均值分别达94.63%、6.87%、4.40%,可有效识别无人机视频目标.     

Fuel economy comparison of conventional drive trains series and parallel hybrid electric step vans

A hybrid electric vehicle (HEV) is a vehicle that combines a conventional propulsion system with an on-board rechargeable energy storage system to achieve better fuel economy than a conventional vehicle HEVs do not have limited ranges like battery electric vehicles, which use batteries charged by an external source. The different propulsion power systems may have common subsystems or components. The objective of this study is to compare the fuel economies of a conventional step van, a series hybrid electric step van (HESV), and a parallel HESV by calculating the fuel consumption using the ADVISOR software by NREL. We also showed the results of the vehicles in different driving cycles including the Central Business District bus cycles, the New York City Cycle, and the US EPA City and Highway cycles.     

Development of an intelligent CAE system for auto-body concept design

The concept design phase is a critical step in auto-body design, as it has a great effect on later design work. This paper describes the implementation of an auto-body structure design in the early stages of a new auto-body developing program. In order to reduce the long design period and analysis error that plagues traditional auto-body concept design, an intelligent CAE system has been successfully developed and implemented based on the UGS NX/API opening platform. This system, the so-called ACD-ICAE (Auto-body concept design-intelligent computer aided engineering) System that means concept design-intelligent computer aided engineering system, employs a fully parametrized template method to build the conceptual auto-body geometry model and FEM model quickly and easily. It also integrates auto-body modeling, analysis and optimization on only one CAD platform via a parametric variables database. Moreover, all parametric variables are shared and updated in different phases of the ACD-ICAE System. A wizard User Interface (UI) based on knowledge of auto-body engineering was developed and used in this system. The procedures implementing the functional diagram of the ACD-ICAE system are also provided. A typical example of a car body concept design with four doors shows that the ACD-ICAE system is efficient and accurate.     

Belt routing and safety assessment through computer simulation

Most of the research on safety belt systems has involved crash simulation that only considered a dynamic human model. However, belt routing analysis, usually known as comfort level estimation, is also an important factor in safety belt design, considering that serious injuries of the abdominal region result from the infiltration of a belt into the neck or the chest. Thus, safety belt evaluations using kinematic human models are also needed. In this paper, a belt fit simulation method is suggested. Using the proposed process, both comfort and safety analyses can be performed under the same conditions continuously, and thus the safety belt design parameters, such as the location of anchor points, dummy posture and etc., can be evaluated. In conclusion, this computer process enables a belt system design to reduce injuries.     

Theoretical method for predicting the weight reduction rate of a box-type car body for rolling stock by material substitution design

In this paper, a theoretical approach is suggested for predicting the structural performances and weight reduction rate of a car body with a box-type section when its material is substituted with a lightweight material for weight saving. For the material substitution design of a car body for rolling stock, bending, axial, and twisting deformations should be considered at constant stiffness and strength conditions. To compare the weight reduction effects on different material applications, some new indices were derived from a structural performance point of view. The derived indices provide good measures to estimate weight reduction by material substitution design and can be effectively applied to the conceptual design of a car body.     

吊拉组合体系预应力连续加劲梁性能分析

按照６ 种不同分跨的设计方案, 每种方案又按照施工阶段采取几种不同预压荷载的计算比较来研究按照预压合龙施工工艺完成体系转换的预应力连续加劲梁吊拉组合体系桥的主梁结构性能, 并选择体系的最优分跨比．     

电流变流体的润滑性能研究

提出了电流变流体作为润滑剂使用的新设想,研究了电流变流体粘度的可控特性,通过试验获得了电流变流体摩擦力随电场强度、悬浮颗粒浓度、油膜厚度等因素的变化规律,为深入研究电流变流体的润滑特性奠定了基础。     

铁路货运站场拓展综合物流服务模式研究

结合铁路货运站场的实际，提出了拓展综合物流服务的基本模式，并从战略层和操作层分析了铁路货运站场拓展现代综合物流服务的运作重构。     

Dynamic modeling and active control of a strap-on launch vehicle

Dynamic modeling and active control of a strap-on launch vehicle are studied in this paper. In the dynamic modeling, the double-compatible free-interface modal synthesis method is used to establish dynamic model of the system, and its model precision is compared with those of finite element method(FEM), fixedinterface modal synthesis method and free-interface modal synthesis method. In the active control, the swing angle of rocket motor is used as design variable, and the control law design based on the model of mass center motion is adopted to validate the system. Simulation results indicate that the double-compatible model synthesis method can properly approximate the FEM which is used as the benchmark solution, and the model precision of the double-compatible modal synthesis method is obviously higher than those of the fixed-interface and freeinterface modal synthesis methods. Based on the control law design, the deflection of mass center of the launch vehicle is very small.     

Online fault diagnosis of modern process industry system based on color-spectrum

This paper proposes a new method to diagnose the system fault of the process industry based on the monitor data set of distributed control system (DCS). Firstly, it defines a homeomorphism product space named color phase space which is a Cartesian product of two-dimensional Euclidean space and three-dimensional color phase space. Secondly, it maps the DCS data to the color phase space in order to get a system color-spectrum which displays the inherent relationship of the whole system. Then, it diagnoses the system fault by observing the color change on the color-spectrum depending on the physiological characteristics that human’s eyes are more sensitive for the color change than data change.