基于LSTM网络的驾驶意图识别及车辆轨迹预测

自动驾驶汽车需具备预测周围车辆轨迹的能力,以便做出合理的决策规划,提高行驶安全性和乘坐舒适性。运用深度学习方法,设计了一种基于长短时记忆（LSTM）网络的驾驶意图识别及车辆轨迹预测模型,该模型由意图识别模块和轨迹输出模块组成。意图识别模块负责识别驾驶意图,其利用Softmax函数计算出驾驶意图分别为向左换道、直线行驶、向右换道的概率;轨迹输出模块由编码器-解码器结构和混合密度网络（MDN）层组成,其中的编码器将历史轨迹信息编码为上下文向量,解码器结合上下文向量和已识别的驾驶意图信息预测未来轨迹;引入MDN层的目的是利用概率分布来表示车辆未来位置,而非仅仅预测一条确定的轨迹,以提高预测结果的可靠性和模型的鲁棒性。此外,将被预测车辆及其周围车辆组成的整体视为研究对象,使模型能够理解车-车间的交互式行为,响应交通环境的变化,动态地预测车辆位置。使用基于真实路况信息的NGSIM（Next Generation SIMulation）数据集对模型进行训练、验证与测试。研究结果表明：与传统的基于模型的方法相比,基于LSTM网络的轨迹预测方法在预测长时域轨迹上具有明显的优势,考虑交互式信息的意图识别模块具备更高的预判性和准确率,且基于意图识别的轨迹预测能降低预测轨迹与真实轨迹间的均方根误差,显著提高轨迹预测精度。     

Shared autonomous electric vehicle (SAEV) operations across the Austin,Texas network with charging infrastructure decisions

Shared autonomous vehicles, or SAVs, have attracted significant public and private interest because of their opportunity to simplify vehicle access, avoid parking costs, reduce fleet size, and, ultimately, save many travelers time and money. One way to extend these benefits is through an electric vehicle (EV) fleet. EVs are especially suited for this heavy usage due to their lower energy costs and reduced maintenance needs. As the price of EV batteries continues to fall, charging facilities become more convenient, and renewable energy sources grow in market share, EVs will become more economically and environmentally competitive with conventionally fueled vehicles. EVs are limited by their distance range and charge times, so these are important factors when considering operations of a large, electric SAV (SAEV) fleet.This study simulated performance characteristics of SAEV fleets serving travelers across the Austin, Texas 6-county region. The simulation works in sync with the agent-based simulator MATSim, with SAEV modeling as a new mode. Charging stations are placed, as needed, to serve all trips requested (under 75 km or 47 miles in length) over 30 days of initial model runs. Simulation of distinctive fleet sizes requiring different charge times and exhibiting different ranges, suggests that the number of station locations depends almost wholly on vehicle range. Reducing charge times does lower fleet response times (to trip requests), but increasing fleet size improves response times the most. Increasing range above 175 km (109 miles) does not appear to improve response times for this region and trips originating in the urban core are served the quickest. Unoccupied travel accounted for 19.6% of SAEV mileage on average, with driving to charging stations accounting for 31.5% of this empty-vehicle mileage. This study found that there appears to be a limit on how much response time can be improved through decreasing charge times or increasing vehicle range.     

Incorporating the extended theory of planned behavior in a school travel mode choice model: a case study of Shaoxing,China

Travel to and from school can have social, economic, and environmental implications for students and their parents. Therefore, understanding school travel mode choice behavior is essential to find policy-oriented approaches to optimizing school travel mode share. Recent research suggests that psychological factors of parents play a significant role in school travel mode choice behavior and the Multiple Indicators and Multiple Causes (MIMIC) model has been used to test the effect of psychological constructs on mode choice behavior. However, little research has used a systematic framework of behavioral theory to organize these psychological factors and investigate their internal relationships. This paper proposes an extended theory of planned behavior (ETPB) to delve into the psychological factors caused by the effects of adults’ cognition and behavioral habits and explores the factors’ relationship paradigm. A theoretical framework of travel mode choice behavior for students in China is constructed. We established the MIMIC model that accommodates latent variables from ETPB. We found that not all the psychological latent variables have significant effects on school travel mode choice behavior, but habit can play an essential role. The results provide theoretical support for demand policies for school travel.     

Daily activity and travel sequences of students,faculty and staff at a large Canadian university

This paper examines the activity engagement, sequencing and timing of activities for student, faculty and staff commuter groups at the largest university in the Maritime Provinces of Canada. The daily activity patterns of all university community groups are modeled using the classification and regression tree classifier algorithm. The data used for this study are derived from the Environmentally Aware Travel Diary Survey (EnACT) conducted in spring 2016 at Dalhousie University, Nova Scotia. Results show that there are significant differences in activity and travel behavior between university population segments and the general population in the region, and between campus groups. For example, students participate in more recreation activities compared to faculty and staff. They also take more trips to and from campus, and are more flexible in their scheduling of trips. The insights gained from this study will provide helpful information for promoting sustainability across university campuses, and for development of campus-based travel demand management strategies.     

Aggregating mobile object trajectories: cumulative time geographic density estimation for GPS data

We present new approaches that expand upon the time geographic density estimation (TGDE) framework previously employed to estimate potential path trees. In the past, TGDE metrics have identified possible locations an individual moving object may have passed between, given known origin and destination points. This paper utilizes a new form of TGDE to investigate taxicab GPS traces over a specified time horizon with position ‘gaps’. To this end, we propose a new extension to the TGDE framework, TGDE-C, which is used to determine the cumulative TGDE values for a group of GPS traces, at a given location. These metrics are applied to multiple taxis and allow for time of day analysis. Additionally, we combine these new extensions with existing TGDE metrics that allow us to determine how accessible individual or groups of vehicles are to urban opportunities.     

Impact of regional population density on walking behavior

Land use can influence walking (measured by the number of steps) and so the health of people. This paper presents the result of empirical research on the impact of regional population densities (inhabitants per inhabitable area) on the number of steps (all steps, both outdoors and indoors). With data collected from almost 11,000 respondents in 148 Japanese regions, we estimate polynomial regression models, the total number of steps being the dependent variable and densities being the main independent variable. Regional population density significantly affects the number of steps after controlling for individual and household attributes. The estimated population density that maximizes the number of steps is around 11,000?persons/km². Increasing densities, up to levels of around 11,000?inhabitants/km², could increase walking and consequently the health of inhabitants. The population density elasticity of the number of steps is 0.046–0.049 in a simple log linear regression model without a peak.     

波浪中破损船舶的运动响应研究

波浪中破损船舶的运动会同时受到波浪激励和进出水的影响,而船体运动也会影响进出水过程,二者的相互影响机理十分复杂.本文重点研究波浪中破舱进出水对船舶运动响应的影响,文中首先基于势流理论建立了考虑破舱进出水的4DOF(横荡-垂荡-横摇-纵摇)相互耦合时域预报方法,在计算中假设舱内的液面水平,利用修正的伯努利方程模拟破舱进/出水,利用Ikeda's经验公式修正阻尼系数.然后以一艘ITTC破损稳性标模为例,研究了波浪中考虑破舱进出水的数学模型以及破舱进出水对运动响应的影响,并研究了不同自由度、破舱口位置对运动响应的影响.研究表明,本文基于势流理论建立的时域预报方法可以定量的预报破损船舶的运动响应.     

新能源汽车充电用户行为大数据分析及应用

随着新能源汽车行业快速发展,电动汽车充电功能及安全问题受到广泛关注。结合新能源汽车充电大数据可视化系统,从用户类型、车辆类型、充电方式、地域分布维度进行交叉分析,洞察用户充电行为统计分布规律;通过数据分析驱动充电安全设计,并结合数据资源和大数据分析技术,挖掘数据背后蕴藏的丰富价值,以改善当前新能源汽车充电问题和促进新能源汽车产业发展。     

城市快速路匝道最小间距模型

匝道间距是路线设计中的重要内容,对交通流有决定性的影响。根据城市快速匝道的特点,应用驾驶员行为理论,模拟了驾驶员城市快速匝道上的驾驶行为。认为匝道间距是影响城市快速路主线运行状况的关键因素。为了合理确定匝道最小间距,必须确定匝道组合模式和计算匝道加减速车道长度,并计算出车流从匝道汇入主线后,由于车流变道而形成交织车流长度。由此建立了不同匝道组合模式下的匝道最小间距模型。应用实例表明。当匝道间距不能满足最小间距时,车速降低,服务水平下降。     

钢绞线网-复合砂浆加固钢筋混凝土梁的受弯性能

分析了钢绞线网-复合砂浆加固钢筋混凝土梁的荷载-挠度曲线,研究了加固层与本体梁界面的粘结机理。将钢筋混凝土梁的受力性能分为未裂阶段、裂缝阶段和破坏阶段,在平截面假定的基础上,建立了高强钢绞线网-复合砂浆加固梁的截面弯矩-跨中挠度分析模型。采用换算截面法对加固梁在集中荷载作用下的抗弯性能进行全过程受力分析,并通过10根加...