应用模糊数学优化潭邵高速公路软基处理方案

地基方案的优化选择,就是使所选的地基方案满足多种要求。地基方案的优化受技术、经济、环保等多项指标的影响。聚类分析是模糊数学中研究物以类聚的一种多元分析方法。针对潭邵高速公路软土地基处理的特殊性,应用模糊数学理论对工程设计方案中的模糊性进行了分析,最终确定,潭邵高速公路软土地基,采用砂桩处理方案进行加固。     

����Ⱥ����·���ڵ���Ҫ�����������о�

城市群是区域城市大系统中具有较强活力的子系统,区域优势显著,在空间联系上具有网络性特点. 客观准确地评价各节点的重要度,是城市群区域公路网布局规划中的一个重要环节. 针对城市群的特点,在公路网节点重要度评估中增加了城市流强度评价指标,节点重要度计算中采用了因子分析法进行客观赋权以避免各指标主观赋权的随意性,为进一步更好区分各节点的重要度等级,采用K-Means聚类方法客观划分了城市节点重要度的类别. 最后以中原城市群为例进行了节点重要度的实例计算,结果表明本文方法具有较好的应用价值.     

有限深度均匀流中圆柱桩群绕流对流速的影响*

通过有限深度均匀流水槽试验研究不同圆柱桩群对均匀水流的绕流影响程度.通过多组试验研究数据对比分析得到圆柱桩群对均匀水流的影响规律;不同圆柱桩群对同样水流影响性质与程度的差别;不同圆柱桩群对水流流场影响范围的规律;圆柱桩群的阻流率、排架密度、桩群长度对水流流场的影响规律;不同流速的水流对圆柱桩群影响的反应规律;水流绕流圆柱桩群时,横向断面流速的变化规律等.研究分析还发现:桩群阻力是桩群影响水流的一个因素但非唯一因素,桩群的排列方式和桩群迎流角对水流流速的变化有不同程度的影响,且其影响程度大于阻力影响.     

基于典型非机动车事故的自动驾驶汽车测试场景构建方法研究

基于交通事故卷宗、交通事故视频信息数据,研究机非混行交通环境下典型交通事故形态,构建了面向机非混行交通环境下的自动驾驶汽车测试场景,旨在针对我国较为特殊的机非混行环境下的自动驾驶汽车的测试场景及测试评价方法提供参考。本文首先分析了自动驾驶测试场景的构建需求,建立交通事故数据筛选标准,得到133例可用于构建自动驾驶汽车测试场景的机动车与非机动车交通事故数据集;其次基于《中华人民共和国道路交通安全法》行驶要求,对133例交通事故的发生地点、车辆行为、道路类型、环境光线等方面进行解构分析;最后通过聚类分析,建立了5类典型的自动驾驶测试场景模型,并分析了不同场景模型的关键要素,为实际道路测试提供理论指导。     

行人过街风格多参数融合识别

行人作为城市慢行交通系统中的弱势群体,探究其过街行为特征,对减少人车冲突,提升行人过街的本质安全具有重要的现实意义。本文针对特定交叉口,设计自由过街及心理临界过街两种场景状态下的行人过街试验。分析过街速度、心理临界时窗及视觉特性等参数之间的耦合关系,选取风险感知因子与临界安全系数作为心理临界过街状态下行人风格的表征指标,构建过街风格识别模型,透析不同风格行人过街行为可靠性。结果表明：年龄对过街速度和心理临界时窗有显著影响;心理临界时窗与过街速度和信号灯区域注视概率呈负相关关系,与车行区域注视概率呈正相关;心理临界过街状态下,依据风格识别结果,男性谨慎程度高于女性,冒险型行人过街失败率为65.22%;冒险型老年人对自身运动机能的评估能力以及视觉搜索特征存在重要缺憾。研究结果可为交叉口基础设施设计与优化,行人过街教育与培训等提供参考。     

Improving transportation service quality based on information fusion

How to improve transportation service quality and thus attract more passengers to use public transportation systems is an important concern for city governments around the world. In this study, we propose a novel information fusion model that addresses the dependent relationships among the various criteria for a method of non-additive weighted gap analysis aimed at evaluating and improving the service quality of transport systems. The hybrid model remedies prior shortcomings and should be more applicable to real-world situations. The proposed model is applied to a real case study of Taipei city bus companies to demonstrate its usefulness. The resulting analysis and the managerial applications for improving the bus service quality are also discussed with regards to the current policies of Taipei city.     

Multimodal travel groups and attitudes: A latent class cluster analysis of Dutch travelers

For developing sustainable travel policies, it may be helpful to identify multimodal travelers, that is, travelers who make use of more than one mode of transport within a given period of time. Of special interest is identifying car drivers who also use public transport and/or bicycle, as this group is more likely to respond to policies that stimulate the use of those modes. It is suggested in the literature that this group may have less biased perceptions and different attitudes towards those modes. This supposition is examined in this paper by conducting a latent class cluster analysis, which identifies (multi)modal travel groups based on the self-reported frequency of mode use. Simultaneously, a membership function is estimated to predict the probability of belonging to each of the five identified (multi)modal travel groups, as a function of attitudinal variables in addition to structural variables. The results indicate that the (near) solo car drivers indeed have more negative attitudes towards public transport and bicycle, while frequent car drivers who also use public transport have less negative public transport attitudes. Although the results suggest that in four of the five identified travel groups, attitudes are congruent with travel mode use, this is not the case for the group who uses public transport most often. This group has relatively favorable car attitudes, and given that many young, low-income travelers belong to this group, it may be expected that at least part of this group will start using car more often once they can afford it. Based on the results, challenges for sustainable policies are formulated for each of the identified (multi)modal travel groups.     

Improving the estimation of total and direction-based heavy-duty vehicle annual average daily traffic

Abstract

The estimation of annual average daily traffic (AADT) is an important parameter collected and maintained by all US departments of transportation. There have been many past research studies that have focused on ways to improve the estimation of AADT. This paper builds upon previous research and compares eight methods, both traditional and cluster-based methodologies, for aggregating monthly adjustment factors for heavy-duty vehicles (US Department of Transportation Federal Highway Administration (FHWA) vehicle classes 4–13). In addition to the direct comparison between the methodologies, the results from the analysis of variance show at the 95% confidence level that the four cluster-based methods produce statistically lower variance and coefficient of variation over the more traditional approaches. In addition to these findings – which are consistent with previous total volume studies – further analysis is performed to compare total heavy-duty monthly adjustment factors, both directions of traffic, with direction-based monthly adjustment factors. The final results show that the variance as well as the coefficient of variation improve on average by 25% when directional aggregate monthly adjustment factors are used instead of total direction.     

基于驾驶人特性的双车道跟驰模型稳定性分析

为了研究交通拥堵问题，了解交通拥挤形成的过程及驾驶员自身特性对双车道交通流稳定性的影响.本文基于复杂网络聚类同步理论，对一类基于驾驶员特性的双车道跟驰模型的稳定性进行研究. 通过设计适当的控制器使得基于驾驶人特性的双车道车辆跟驰模型趋于稳定，并得到了模型稳定性的条件.此外，在双车道上的车辆受到随机外部扰动的情形下，利用具有外部扰动的复杂网络自适应H∞ 聚类同步理论，研究了外部扰动情形下基于驾驶人特性的双车道车辆跟驰模型的稳定性.最后，采用MATLAB仿真技术进行数值模拟，验证所设计控制器对双车道交通流稳定性的影响及不同的驾驶员性格特性对交通流运行的影响.     

用模糊逻辑实现动力定位船舶的艏向寻优

由于各种环境因素随时间不断变化，船舶作为动力定位作业时，最优艏向也是不断变化的，所以艏向寻优是动力定位的关键技术之一，本文用模糊逻辑实现艏向寻优，其优点是，不依赖于环境要素的测量值和数学模型，寻优与控制相结合，反应快，准确程度高。