基于语音特征迁移学习的驾驶疲劳检测

针对现有方法在实际应用时的标记样本稀缺与测试样本数据分布偏移等问题,提出一种基于语音特征迁移学习的驾驶疲劳检测方法。通过基于迁移学习的特征空间变换,对源领域有标记样本与目标域无标记样本数据间的边缘分布、条件分布、流形结构进行联合适配及降维处理,以解决样本数据分布偏移和特征维度过高的问题。以半监督学习的方式来迭代优化目标域样本的伪标记,并据此不断更新特征变换方式和迁移分类器,进而提高疲劳检测模型的精度和泛化能力。通过实验将本文方法与现有常用的监督学习、半监督学习和迁移学习等方法进行对比。结果表明,在测试时间、应用场景和被试个体均发生变化的情况下,本文所提方法的驾驶疲劳检测效果显著优于现有方法,正确率最高达到86.7%,具有实际应用价值。     

Exploring the Knowledge Dynamics Associated with Coastal Adaptation Planning

Scientific projections for climate change induced sea-level-rise highlight current and potential future consequences for low lying coastal areas. In response considerable attention has been directed toward the task of coastal adaptation planning. Experience to date indicates that adaptation planning is more complex and contested than anticipated. We argue that this is partly due to the constrained way adaptation planning is conceptualized, whereby limited attention is directed toward understanding coastal adaptation planning as a site of complex and contested knowledge dynamics. Consequently, we use a knowledge systems perspective to explore coastal adaptation planning in order to highlight some of the knowledge exchange dynamics involved. In doing so we draw on views expressed in semi-structured interviews with a diverse range of stakeholders with an interest in coastal management in Victoria, Australia. While the focus is on coastal adaptation planning in Victoria, the insights generated are intended to contribute to broadening the way in which adaptation is conceptualized.     

高速公路隧道交通安全问题及对策分析

高速公路隧道交通事故原因复杂且具有多样性。结合隧道特点和交通事故特征,以人-车-路-环境体系为基础,多角度、多方面分析了隧道交通事故发生原因,并提出了合理的解决方案。     

Network effects of intelligent speed adaptation systems

Intelligent Speed Adaptation (ISA) systems use in-vehicle electronic devices to enable the speed of vehicles to be regulated automatically. They are increasingly appreciated as a flexible method for speed management and control particularly in urban areas. On-road trials using a small numbers of ISA equipped vehicles have been carried out in Sweden, the Netherlands, Spain and the UK. This paper describes the developments made to enhance a traffic microsimulation model in order to represent ISA implemented across a network and the impact of this on the networks. The simulation modelling of the control system is carried out on a real-world urban network, and the impacts on traffic congestion, speed distribution and the environment assessed. The results show that ISA systems are more effective in less congested traffic conditions. Momentary high speeds in traffic are effectively suppressed, resulting in a reduction in speed variation which is likely to have a beneficial impact on safety. Whilst ISA reduces excessive traffic speeds in the network, it does not affect average journey times. In particular, the total vehicle-hours travelling at speeds below 10 km/hr have not changed, indicating that the speed control had not induced more slow-moving queues to the network. A statistically significant, eight percent, reduction in fuel consumption was found with full ISA penetration. These results are in accordance with those from field trials and they provide the basis for cost-benefit analyses on introducing ISA into the vehicle fleet. However, contrary to earlier findings from the Swedish ISA road trials, this study suggested that ISA is likely to have no significant effect on emission of gaseous pollutants CO, NO_x and HC.     

Energy and emissions impacts of a freeway-based dynamic eco-driving system

Surface transportation consumes a vast quantity of fuel and accounts for about a third of the US CO₂ emissions. In addition to the use of more fuel-efficient vehicles and carbon-neutral alternative fuels, fuel consumption and CO₂ emissions can be lowered through a variety of strategies that reduce congestion, smooth traffic flow, and reduce excessive vehicle speeds. Eco-driving is one such strategy. It typically consists of changing a person’s driving behavior by providing general static advice to the driver (e.g. do not accelerate too quickly, reduce speeds, etc.). In this study, we investigate the concept of dynamic eco-driving, where advice is given in real-time to drivers changing traffic conditions in the vehicle’s vicinity. This dynamic strategy takes advantage of real-time traffic sensing and telematics, allowing for a traffic management system to monitor traffic speed, density, and flow, and then communicates advice in real-time back to the vehicles. By providing dynamic advice to drivers, approximately 10–20% in fuel savings and lower CO₂ emissions are possible without a significant increase in travel time. Based on simulations, it was found that in general, higher percentage reductions in fuel consumption and CO₂ emission occur during severe compared to less congested scenarios. Real-world experiments have also been carried out, showing similar reductions but to a slightly smaller degree.     

Investigating heterogeneity in social influence by social distance in car-sharing decisions under uncertainty: A regret-minimizing hybrid choice model framework based on sequential stated adaptation experiments

The present study is designed to investigate social influence in car-sharing decisions under uncertainty. Social influence indicates that individuals’ decisions are influenced by the choices made by members of their social networks. An individual may experience different degrees of influence depending on social distance, i.e. the strength of the social relationship between individuals. Such heterogeneity in social influence has been largely ignored in the previous travel behavior research. The data used in this study stems from an egocentric social network survey, which measures the strength of the social relationships of each respondent. In addition, a sequential stated adaptation experiment was developed to capture more explicitly the effect of social network choices on the individual decision-making process. Social distance is regarded as a random latent variable. The estimated social distance and social network choices are incorporated into a social influence variable, which is treated as an explanatory variable in the car-sharing decision model. To simultaneously estimate latent social distance and the effects of social influence on the car-sharing decision, we expand the hybrid choice framework to incorporate the latent social distance model into discrete choice analysis. The estimation results show substantial social influence in car-sharing decisions. The magnitude of social influence varies according to the type of relationship, similarity of socio-demographics and the number of social interactions.     

Adaptation Pathways in Coastal Case Studies: Lessons Learned and Future Directions

Climate change adaptation presents a difficult challenge for coastal towns around the world, forcing local governments to plan for sea level rise in a contentious decision-making space. The concept of “adaptation pathways,” a diagnostic and analytical tool to assist in adaptive planning and decision-making, is gaining traction as a way of framing and informing climate adaptation. It provides decisionmakers a way to acknowledge the inter-temporal complexities and uncertainties associated with the novel dynamics of climate change and a mechanism to manage these challenges in the local context. In 2012, the Australian Government funded an 18-month program to provide decisionmakers in the coastal zone an opportunity to test the utility of the adaptation pathways concept for coastal climate adaptation. Using a selection of completed projects as case studies, we performed a document analysis to better understand the learnings from the projects. The main themes surrounded: (1) the utility of the adaptation pathway framework in developing options, (2) decision-making rationale and criteria, and (3) stakeholder participation in pathway development. A project participant survey was developed to further understand these themes. Our analysis reveals that “adaptation pathways” was generally framed narrowly and conservatively to emphasize extant economic, administrative and legal considerations over community, participatory, or exploratory ones. Although some case study projects were able to reach a point in the pathway discussion to actively involve stakeholders in their decision-making process, many case studies continued to build technical data as a method for defending policies and actions. These results indicate that coastal adaptation can take-up adaptation pathways as a useful concept for decision-making and planning; however, many councils may still require assistance in stakeholder communication processes in order to develop sociallyacceptable plans that take into account the full range of values affecting local coastal environments.     

气候条件对沥青路面路用性能的影响

结合我国沥青技术标准,参照美国SHRP研究成果,在调查陕西省和甘肃省部分地区气候特征的基础上,研究了气候条件对沥青路面路用性能的影响.     

美国凤凰城都会区新型轻轨车辆

为解决美国凤凰域都会区公交系统不够完备的问题,首次建设一条轻轨线路并研发适应凤凰城都会区特殊气候条件的新型轻轨车辆.在安全性、可视化、人性化、人机工程等方面进行创新设计,使这种车辆成为当今世界上最先进的轻轨车辆之一.     

Perspectives of the Expert and Experienced on Challenges to Regional Adaptation for Sea Level Rise: Implications for Multisectoral Readiness and Boundary Spanning

This study focuses on adaptation to sea level rise (SLR) in the specific context of the Hampton Roads region of southeastern coastal Virginia. It analyzes the perspectives of stakeholders who are experts in and have experience with SLR to develop an informed understanding of the region’s ability to address SLR and its readiness to pursue adaptation using a whole-of-community approach. A survey of the expert and experienced reveals (1) what they prioritized as the top three most challenging phases of regional adaptation and (2) what issues within those priority areas were the most pressing. Our findings point to the importance of meso- and micro-level boundary spanning across organizations and the whole of the community to ensure that multiple organizations and individuals in a region can be engaged in adaptation. Finally, this work briefly discusses how boundary-spanning approaches can provide a region a pathway toward overcoming barriers to effective climate-related adaptation.