Intelligent Assisted Travel Wheelchair Based on Image Recognition Technology

This paper introduces an intelligent image recognition system integrated into a wheelchair based on deep learning in cold environments,aiming to improve the convenience and safety of disabled individuals.The system adopts advanced image recognition technology to monitor road conditions in real-time through the camera and to detect and measure distance to foreign objects on the road.The system visualizes the detection results on the wheelchair screen to assist the user in avoiding and improving the safety of their daily travel.In addition,the system also includes crawler tracks,seat heating,snow and rain protection,and other functions.The wheelchair has a wide range of application prospects and development potential.It is expected to be widely used in the future,providing a strong guarantee for the safe travel of disabled individuals in China.

Travel time function for basic link considering signal control in network traffic model

In order to describe the travel time of signalcontrolled roads, a travel time model for urban basic roads based on the cumulative curve is proposed. First, the traffic wave method is used to analyze the formation and dispersion of the vehicle queue. Cumulative curves for road entrances and exits are established. Based on the cumulative curves, the travel time of the one-lane road under stable flow input is derived. And then, the multi-lane road is decomposed into a series of single-lane links based on its topological characteristics. Hence, the travel time function for the basic road is obtained. The travel time is a function of road length, flow and control parameters. Numerical analyses show that the travel time depends on the supply-demand condition, and it has high sensitivity during peak hours.

Travel time prediction with viscoelastic traffic model

Travel time through a ring road with a total length of 80 km has been predicted by a viscoelastic traffic model(VEM), which is developed in analogous to the non-Newtonian fluid flow. The VEM expresses a traffic pressure for the unfree flow case by space headway, ensuring that the pressure can be determined by the assumption that the relevant second critical sound speed is exactly equal to the disturbance propagation speed determined by the free flow speed and the braking distance measured by the average vehicular length. The VEM assumes that the sound speed for the free flow case depends on the traffic density in some specific aspects, which ensures that it is exactly identical to the free flow speed on an empty road. To make a comparison, the open Navier-Stokes type model developed by Zhang(ZHANG, H. M. Driver memory, traffic viscosity and a viscous vehicular traffic flow model. Transp. Res. Part B, 37, 27–41(2003)) is adopted to predict the travel time through the ring road for providing the counterpart results.When the traffic free flow speed is 80 km/h, the braking distance is supposed to be 45 m,with the jam density uniquely determined by the average length of vehicles l ≈ 5.8 m. To avoid possible singular points in travel time prediction, a distinguishing period for time averaging is pre-assigned to be 7.5 minutes. It is found that the travel time increases monotonically with the initial traffic density on the ring road. Without ramp effects, for the ring road with the initial density less than the second critical density, the travel time can be simply predicted by using the equilibrium speed. However, this simpler approach is unavailable for scenarios over the second critical.

Traffic Data： Bluetooth Sensors vs. Crowdsourcing——A Comparative Study to Calculate Travel Time Reliability in Calgary, Alberta, Canada

The City of Calgary did a comparative study between two techniques of traffic data collection： Bluetooth sensors and crowdsourcing, for measuring travel time reliability on two goods movement corridors in Calgary, Alberta. To estimate travel time and speed, we used the output of BIuFAX sensors, which were operated by monitoring Bluetooth signals at several points along a roadway. On the other hand, TomTom historical traffic data were extracted from the TomTom Traffic Stats portal. To calculate travel time reliability, we applied the buffer index, and the planning time index recommended by FHWA （Federal Highway Administration）. The Bluetooth traffic data were presumed as the benchmark in this study. Unlike the TomTom traffic data, the data provided by the Bluetooth technology met the minimum recommended sample size requirement, although data processing was time consuming and impractical for long study periods. Our study results showed that crowdsourcing technique can be a viable alternative and provide travel time reliability estimates with a reasonable accuracy, when there are adequate numbers of records registered. However, the TomTom sample sizes in Calgary were not large enough to provide a statistically reliable method of providing travel time indices. Further researches may verify the accuracy of crowdsourcing technologies for travel time studies.

Bluetooth as a traffic sensor for stream travel time estimation under Bogazici Bosporus conditions in Turkey

Travel time estimation is an integral part of Intelligent Transportation Systems, and has been an important component in traffic management and operations for many years. Travel time, being spatial in nature, requires spatial sensors to measure it accurately. Bluetooth is emerging as a promising technology for the direct measurement of travel time, and is reported in a few studies from homogenous traffic conditions. At the same time, there have been no studies on the applicability of Bluetooth for travel time estimation in heterogeneous traffic seen in Istanbul and even that Turkey. Bluetooth data collected from a busy urban road in Istanbul city have been analyzed and the penetration rate was found to be about 5 %. Two wheelers and light motor vehicles have been detected using the Bluetooth sensor and the data have been extrapolated to estimate travel times of other classes of vehicles. The study developed linear relationships between speeds of different classes of vehicles through weighted linear regression methods and were used for the estimation of stream travel time. The results obtained were promising and show that Bluetooth is a cost-effective technology for estimation of travel time for heterogeneous traffic conditions.

Temporal Stability and Transferability of Non-Motorized and Total Trip Generation Models

Transportation systems provide a means for moving people and the goods from which they are spatially separated. Of the two means of surface transportation, the motorized mode is used extensively for utilitarian travel in developed countries. The increasing reliance on motorized travel has contributed to increased traffic congestion, air pollution, and greenhouse emissions. Non-motorized travel has recently received significant attention as a means to reduce congestion and environmental problems and improve human health. However, non-motorized modeling is generally underdeveloped. This study investigated some changes in non-motorized and total travel and the characteristics of the traveling public in 1990, 1995, 2001, and 2009 using a national travel survey. The study also investigated the temporal transferability of linear-regression trip generation models for non-motorized and total travel under such changes. High-income households made fewer non-motorized trips in 1990 and 1995 compared to 2001 and 2009. Persons aged 50 and over showed an increased demand for non-motorized travel, whereas children aged 0 - 15 showed a decreasing preference for non-motorized travel over time. Regarding temporal stability, only the coefficient for single-adult households with no children was stable across all of the analysis years. For both non-motorized and total travel, most model parameter estimates were stable short term but not long term. In general, the total travel models transferred better than non-motorized models, both short term and long term. Despite not finding universal stability in model parameter estimates, the models were marginally able to replicate travel in 2009 relative to the locally estimated 2009 model.

可达性对城市群多模式交通碳排放的空间异质性影响

当前,在交通行业碳减排存在巨大挑战以及未来中国城市群交通发展长远规划双重背景下,如何通过改善交通可达性提高居民出行效率和减少碳排放是亟待解决的关键问题之一。本文基于城际出行手机信令数据,从居民出行的角度提出城际多模式交通客运碳排放量方法,并采用梯度提升决策树(GBDT)模型及多尺度地理加权回归(MGWR)模型探讨可达性对区域碳排放量的空间异质性影响。以关中平原城市群为例进行验证,结果表明:城际公路客运碳排放量远大于铁路,呈现沿交通基础设施线路分布的特征;在整体区域范围内,可达性指标对碳排放水平具有一定的正向边际效应;MGWR能够刻画碳排放与可达性指标关系的空间异质性及尺度差异;经济潜能可达性、介数中心性及接近中心性对城际碳排放具有显著的正向空间异质性影响,但影响尺度不同;公路客运碳排放对介数中心性及接近中心性要素较为敏感,经济潜能对碳排放的影响较为平稳;铁路出行可达性的提升对中心城市的影响效应低于周边区县城市。

基于手机信令数据的城市居民动态OD矩阵提取方法

现有的城市居民出行调查周期较长,交通小区划分粒度粗糙,导致调查不能及时准确地获取居民出行信息。针对该问题,提出了一种基于手机信令数据的城市居民动态OD矩阵提取方法。首先,针对信令数据中的两种复杂噪声:乒乓切换和漂移数据,提出了基于窗口阈值的检测与等效位置替换方法,以及复杂漂移点的检测和标记处理方法;然后,提出一种改进的ST-DBSCAN聚类方法,引入一种等时化方法将时间信息与空间信息相结合,识别出行过程中的驻留点;最后,基于地理信息系统构建含有道路关键节点的路网,将居民出行OD与路网节点相匹配,有效推导出城市居民动态OD矩阵。实验结果表明:与ST-DBSCAN算法相比,所提改进的ST-DBSCAN算法在聚类效果和识别速度上分别提升了6.10%和5.26%;与统计方法和二阶统计量方法相比,基于改进的ST-DBSCAN算法的动态OD矩阵提取方法在均方误差(MSE)上分别降低了16.98%和21.55%。以北京市为例,运用提出的动态OD矩阵提取方法,能够及时有效地分析城市居民日常与高峰时段的出行特征。

学生通学出行的家长接送方式的选择

研究家长接送方式选择的影响因素,对于早晚高峰交通出行政策的制定具有重要意义.考虑家长、学生的个人属性与家庭属性,以及后续行程目的、生理成本与延误时间等出行属性因素,基于多项logit(multinomial logit,MNL)模型分析对于家长接送学生交通方式选择的影响因素,并通过中国海宁市居民出行数据进行验证.研究结果表明,家长年龄越大,选择步行接送的可能性越大;学生性别对于家长接送方式的选择没有显著影响;家长性别和学生年龄对接送方式具有显著影响,其中,女性接送者更偏向于选择非机动车接送,学生年龄越小,家长更偏向于选择非机动车接送;后续行程目的为上班的家长更偏向于选择机动车接送;小汽车接送者的接送时间相对较长,但该接送方式承担的生理成本和延误时间相对较少;电动车接送者承担的生理成本和时间成本都较低.依据研究结果,从家庭、学校和交通管制等维度对学生通学出行提出相关改善建议,能够为相关部门从学生通学的角度制定交通规划管理政策、方案提供指导.

考虑交通信号和交通密度的城市路段个体行程时间建模

为探究城市路段中交通密度与交通信号对个体车辆行程时间的影响规律,提高城市道路行程时间的预测精度,基于车牌识别数据与信号机配时数据等多种融合数据,提出了由自由流行程时间、密度延误与随机误差项组成的行程时间关系函数,并在该函数基础上构建了考虑个体驾驶习惯的行程时间预测模型。首先,基于道路限速条件、车辆进入路段时的下游交叉口信号状态构建自由流行程时间函数;其次,提出了一种通过密度阈值判断车辆能否在当前信号周期通过下游交叉口的方法,并基于个体车辆进入路段时的路段密度计算出车辆所需等待信号周期个数及等待前方排队疏散所需时间,从而构建密度延误函数;然后,将个体实际行程时间与自由流行程时间、密度延误的差值作为随机误差项,通过高斯混合模型拟合随机误差项的概率分布;最后,选用安徽省宣城市多个路段作为案例,分析各路段行程时间函数各部分的具体表现,对该行程时间预测模型进行验证。结果表明:行程时间预测模型的MAPE,MAE,RMSE分别为6%~12%,5~15,13~35,在准确率方面优于其他算法,是一种有效的城市路段个体行程时间预测方法。

考虑机会充电与行程时间可靠性的区域多车型电动公交调度优化

为提高电动公交系统运营效率,降低运营成本,本文提出一种考虑机会充电和行程时间可靠性的电动公交调度优化方法。首先,基于区域调度场景,提出在线路始末站配备快速充电桩,利用车次接续时间进行机会充电的策略;然后,考虑行程时间的随机波动,以表征特定可靠性的预留行程时间作为模型输入生成调度方案,同时,将发车延误成本纳入目标函数中,综合考虑公交企业从规划到运营阶段的整体效益,构建以总成本最小为目标的区域多车型电动公交调度优化模型,针对模型特点,设计自适应大邻域搜索算法进行求解;最后,以北京市大兴区4条公交线路为例,验证模型和算法的有效性。结果表明:基于本文方法得到的最优调度方案相较于传统单线路单车型调度方案,能使企业日均总成本下降37.93%,平均每辆车的发车延误时长减少5.63 min,说明本文所提方法能有效降低企业成本,提升公交系统可靠性。相较于不考虑机会充电和行程时间可靠性的区域多车型运营模式,本文最优方案能使总成本下降28.67%。此外,通过灵敏度分析,建议公交企业以240 kW的充电功率进行快速充电资源的配置,以90%的行程时间可靠性进行电动公交调度方案的编制。

职住学空间测度及儿童通学决策的影响机理——以昆明市双职工家庭为例

当代城市家庭选择居住地时往往将临近学校作为重要参考,在缩小学住距离的同时加剧了职住分离。选取昆明市561个有学龄儿童的双职工家庭为研究对象,分析家庭空间分异对儿童通学决策的影响。首先,基于时间地理学理论构建考虑职住学三个锚点的家庭活动空间及测度理论模型。其次,根据测度结果将家庭活动空间划分为职住学就近型和职住学分离型两类。最后,建立结构方程模型分别探究两种模式下父母工作时长以及父母儿童组合时间窗对儿童通学时空行为的影响。研究结果表明:相对于学住距离,父母的职住距离和工作时长对通学行为的影响更为显著,因此就近入学政策对于促进儿童步行上学的作用有限;对于职住学分离型家庭,父母下班与儿童放学组合时间窗的宽裕程度对儿童出行影响显著,组合时间窗越宽裕,儿童小汽车出行的可能越大,因此弹性放学时间的效果也有待商榷;儿童在通学行为中的性别差异主要在职住学分离型家庭中体现。

重大节假日出行方式选择模型研究

为优化节假日高速公路免费政策,选取出行者变量、出行方式相关变量和行程特性变量作为效用变量,构建了出行方式位于上层、出行距离位于下层的NL模型。利用弧弹性法分析不同收费标准和收费里程条件下出行者的行为特性,并利用场景仿真评价多种收费组合对出行方式的影响。研究结果表明:提高收费标准,汽车出行比例会显著降低,且高速铁路承担了大部分的转移交通量。随着收费里程的增加,出行者逐渐倾向于出行成本较小的普通铁路。对长距离出行收费可以提升公共交通分担率,缓解节假日高速公路的交通拥堵。

考虑空间异质性的短距离上学方式选择机理

短距离出行是学生日常出行的重要组成部分,合理引导短距离上学群体交通方式选择对于促进学生健康发展和缓解高峰期间道路交通拥堵具有重要意义.本研究以中国南京市为例,利用贝叶斯分类法识别出南京市学生的短距离出行阈值为2.7 km,在此基础上,分别构建多项logit(multinomial logit,MNL)模型与地理加权多项logit(geographically weighted multinominal logit,GWMNL)模型,探讨个人属性、出行特征、家庭特征及建成环境对短距离学生群体上学方式选择的影响及其空间异质性.结果表明,相比MNL模型,GWMNL模型具有更好的拟合度和解释能力,说明学生上学交通方式选择行为因居住位置不同存在明显差异.与居住在学校附近的低龄学生相比,核心区与近郊区较短出行距离的高龄学生更倾向于乘坐小汽车上学,而远郊区(六合区和江宁区)的高龄学生更倾向于选择积极出行(步行和自行车)和电动自行车上学.小汽车数量对学生选择积极出行、电动自行车及公共交通上学均有消极影响,且这种消极影响从市中心向外围区域逐渐增强.最近公交站距离对居住在核心区与近郊区的学生选择公共交通具有明显促进作用,而对部分远郊地区的学生选择公共交通具有抑制作用,且在南京市的北部和东部地区尤为明显.研究结果可为不同区域学生群体短距离积极出行引导策略的制定提供参考.

基于需求层次理论的城市公共交通发展战略研究

针对目前城市交通供需结构不平衡的问题,以马斯洛需求层次理论模型为参考,结合城市交通需求和供给特性,分析其供求关系及层次结构,提出城市交通供需的响应关系和理想化城市交通出行结构。考虑时间、经济、舒适度等交通需求因素和环境、运营、基础设施、能耗等交通供给资源,以现状需求、资源、环境污染、运营、发展规模、时间为约束,从宏观和微观两个角度构建双层规划模型。以大连市为研究对象,基于现有交通出行结构开展成本分析,计算现有出行需求情况下,大连市城市交通结构的适宜比例,并将计算结果与现有出行结构进行比较。研究表明,优化的出行结构可以更好地满足现有城市交通的出行需求,满足城市交通可持续发展的目标,研究结果可为今后大连市公共交通发展提供参考。

考虑出行阻抗的城市群综合客运网络可达性

为阐明城市群综合客运交通网络的多层空间结构并丰富网络的功能属性,以耦合这一全新视角提出考虑城市交通换乘的城市群综合客运网络模型,并从出行费用与出行时间出发分析节点与网络的可达性。首先,基于复杂网络理论Space-P建模方法,将城市群中的客运站点抽象为网络中的节点并进行编号,以票价与出行时间标定边权,分别构建公路客运子网、铁路客运子网。其次,划设城市群中的各城市中心城区为独立的交通小区,将有城市公共交通、步行等换乘方式连接的节点间添加耦合边,构建耦合子网。再次,将两类客运子网通过耦合子网组合叠加为综合网络,改进传统网络效率指标,构建加权的阻抗效率指标评估节点与网络的阻抗可达性。最后,以成渝城市群为实例研究,构建成渝城市群综合客运网络并分析公路、铁路、综合网络的阻抗可达性。结果表明:多个节点的耦合使得这些节点的阻抗可达性较好,因此在研究城市群综合客运交通网络时,城市交通换乘是不可忽略的因素;综合网络的阻抗可达性相较于公路子网、铁路子网分别提升了83.4%,28.5%,公路与铁路客运子网的耦合提高了城市群客运交通网络的阻抗可达性;推进站点间、多种运输方式间的耦合协调,继而构建更加便捷顺畅的城市群综合客运交通网络,是降低旅客的出行成本的重要途径。

交通规划导向的科技赋能交通小区划分方法

文章以手机信令、路网大数据为基础,提出了针对交通规划的科技赋能交通小区划分方法,以路网数据划分聚类单元,利用手机信令数据准确识别聚类单元内居民出行时空分布特征,基于改进耦合空间聚类算法进行聚类,借助现状路网对聚类结果进行修正得到交通小区划分结果。宁波市实例研究表明,科技赋能交通小区划分方法能兼顾居民出行时空特征和城市路网匹配,并能较好保证交通小区的整体连续性,该方法适用性强,可为交通规划的科学制定提供借鉴。

不确定环境下收费与信息对路径选择行为影响的实验研究

在不确定环境下考虑信息发布与拥挤收费两种措施,基于交通网络均衡,推导了两路径简单路网的最优收费模型,并推广至一般路网。基于两路径路网设计逐日路径选择行为实验,分析了两种措施对路径选择行为的影响。结果表明:完美信息与收费的实施均能减少流量波动,单独实施收费的效果最佳,但只有两种措施相结合时,路网流量趋于用户均衡的趋势最稳定;完美信息会增加路径切换行为,而收费可有效抑制大幅的路径切换;无收费时路径切换往往会造成出行成本增加,而收费时则相反;收费与信息结合时的出行时间比两种措施单独实施时小。

基于机器学习的中型城市居民出行方式选择行为研究

为探索中型城市居民出行特征以及不同因素对出行方式选择行为的影响机制,本文以中国某中型城市居民出行数据为例,综合考虑传统离散选择模型和机器学习模型在预测精度和建模合理性上的优劣,以及机器学习模型超参数求解算法的特点和效率,引入变异程序,提出粒子群优化随机森林的中型城市居民出行方式选择预测模型,采用预测准确率、出行方式比例预测绝对误差和期望模拟误差这3项性能指标,量化对比粒子群优化随机森林模型与多种机器学习模型和多项Logit模型统计学上的预测性能差异,利用SHAP(SHapley Additive exPlanation)模型深入分析个人社会经济属性、出行属性及出行方式属性等相关因素与居民出行方式选择之间的非线性关系。结果表明:粒子群优化随机森林模型整体平均预测准确率最高,为0.856,出行方式比例预测平均绝对误差和期望模拟平均误差最低,分别为0.062和0.306,模型间指标差异在统计学检验下显著;距离对不同出行方式选择的影响最显著,步行和私家车出行对距离敏感性更高,不同距离下,两者选择概率变化超过35%;30岁以下群体不同出行方式选择概率差距大于其他年龄段;性别、是否拥有私家车或公交IC卡等因素显著改变公交车和私家车的选择概率。

考虑低碳出行意愿的地铁公交接驳满意度分析

现如今,城市地铁建设规模日益增大,公交作为最常见的接驳方式,接驳满意度直接影响着居民的出行体验。文章通过问卷调查分析城市居民低碳出行的意愿和对地铁与公交接驳的满意度,在处理问卷数据后,对量表进行信效度检验,并对变量进行相关性分析。结果表明:绝大部分居民倾向于使用低排放的交通方式出行,当小汽车拥有数量越多,居民对当前低碳出行的接受意愿和环境满意度、地铁公交接驳的满意度、接驳的需求程度和优化意愿均有所降低。