基于HERA-JANUS模型的空管人误认知分析

空管人误分类分析是空管人误研究的基础。为了对管制员人误进行系统的分类研究,结合空管业务知识和认知心理学理论,对欧洲航空安全局和美国联邦航空局合作开发的HERA-JANUS模型的工作原理和流程进行较详细地分析。运用该方法模型,对我国一起空管不安全事件案例进行分析后得到3个由管制员所产生的人误差错,并对这3个人误差错分别从人误类型、人误认知、相关因素3方面进行详尽的分析研究,最后得出该不安全事件的21项人误结果。结果表明,HERA-JANUS模型能较全面地从深层次分析管制员的人误,其分类形式也便于开展空管人误统计。     

基于事故深度调查的城市交通事故特征分析

为探索适合我国的事故数据深度采集标准,并分析城市道路交通事故特征及致因,基于《道路交通事故深度调查信息采集表》(简称采集表),调查人员随交警赴事故现场随机详细调查87起城市道路交通事故。借鉴"Haddon Matrix"思想建立致因分析矩阵系统,分析事故的致因。发现采集表对事故地点、事故形态及原因项分类更加具体、明确,女性驾驶员的事故发生率略低于男性驾驶员,驾驶员年龄超过60岁后,发生事故的危险性显著提高,"交叉口影响区"事故50%由变更车道引起,非机动车驾驶员未戴安全头盔是造成严重伤害的重要原因。     

Characteristics of animal-related motor vehicle crashes in select National Park Service units—United States, 1990–2013

Objectives: Nationally, animal–motor vehicle crashes (AVCs) account for 4.4% of all types of motor vehicle crashes (MVCs). AVCs are a safety risk for drivers and animals and many National Park Service (NPS) units (e.g., national park, national monument, or national parkway) have known AVC risk factors, including rural locations and substantial animal densities. We sought to describe conditions and circumstances involving AVCs to guide traffic and wildlife management for prevention of AVCs in select NPS units.

Methods: We conducted an analysis using NPS law enforcement MVC data. An MVC is a collision involving an in-transit motor vehicle that occurred or began on a public roadway. An AVC is characterized as a collision between a motor vehicle and an animal. A non-AVC is a crash between a motor vehicle and any object other than an animal or noncollision event (e.g., rollover crash). The final data for analysis included 54,068 records from 51 NPS units during 1990–2013. Counts and proportions were calculated for categorical variables and medians and ranges were calculated for continuous variables. We used Pearson’s chi-square to compare circumstances of AVCs and non-AVCs. Data were compiled at the park regional level; NPS parks are assigned to 1 of 7 regions based on the park’s location.

Results: AVCs accounted for 10.4% (5,643 of 54,068) of all MVCs from 51 NPS units. The Northeast (2,021 of 5,643; 35.8%) and Intermountain (1,180 of 5,643; 20.9%) regions had the largest percentage of the total AVC burden. November was the peak month for AVCs across all regions (881 of 5,643; 15.6%); however, seasonality varied by park geographic regions. The highest counts of AVCs were reported during fall for the National Capital, Northeast/Southeast, and Northeast regions; winter for the Southeast region; and summer for Intermountain and Pacific West regions.

Conclusions: AVCs represent a public health and wildlife safety concern for NPS units. AVCs in select NPS units were approximately 2-fold higher than the national percentage for AVCs. The peak season for AVCs varied by NPS region. Knowledge of region-specific seasonality patterns for AVCs can help NPS staff develop mitigation strategies for use primarily during peak AVC months. Improving AVC data collection might provide NPS with a more complete understanding of risk factors and seasonal trends for specific NPS units. By collecting information concerning the animal species hit, park managers can better understand the impacts of AVC to wildlife population health.     

Epidemiology of alcohol and drug screening among pedestrian fatalities in the United States, 2014–2016

Objective: U.S. pedestrian fatalities increased by 25% between 2010 and 2015. Risk factors include distractions, the built environment, urbanization, economic variables, and weather conditions. Of interest is the role of alcohol and drugs in premature death among pedestrians. This study sought to explore the prevalence of substance use screenings among pedestrian fatalities in the United States between 2014 and 2016.

Methods: Data were collected from the Fatality Analysis Reporting System provided by the NHTSA. Pedestrian crash variables included demographics as well as information regarding alcohol or drug testing status. Frequency and cross-tabulation tables were constructed to assess the prevalence of screening by person, place, and time. Log-linear analyses were completed to explore age, race, and sex differences. A 3-year examination period was used to control for yearly fluctuations and to incorporate an increasing trend in cases.

Results: Pedestrian fatalities accounted for 84% of all deaths among vulnerable road users during the examination period. Those most at risk were white males between the ages of 45 and 64. Over all states, 74.7% of fatalities were tested for alcohol and 67.1% were tested for drugs; further, 66.5% of cases were tested for both alcohol and drugs and 24.8% were tested for neither substance. Cases screened for both alcohol and drugs ranged from 2.9% in North Carolina to 95.7% in Nevada and those testing for neither substance ranged from a high of 68.9% in Indiana to a low of 1.1% in Maryland. Log-linear regression revealed significant differences in alcohol screening by age and race but not by sex. Differences in drug screening were not identified for any demographic variable. Fatalities tested for alcohol were significantly more likely to be tested for drugs; only 8.2% were screened solely for alcohol and 0.05% were screened for drugs alone.

Conclusions: Preventive strategies become more important as pedestrian crashes and fatalities increase. Risk reduction in the form of policy change, alterations to the built environment, or interdisciplinary approaches to injury prevention is dependent upon best evidence supported in part by more deliberate and consistent screening.     

Investigating the effect of geometric dimensions of median traffic barriers on crashes: Crash analysis of interstate roads in Wyoming using actual crash datasets

IntroductionDespite the numerous safety studies done on traffic barriers’ performance assessment, the effect of variables such as traffic barrier’s height has not been identified considering a comprehensive actual crash data analysis. This study seeks to identify the impact of geometric variables (i.e., height, post-spacing, sideslope ratio, and lateral offset) on median traffic barriers’ performance in crashes on interstate roads.MethodGeometric dimensions of over 110 miles median traffic barriers on interstate Wyoming roads were inventoried in a field survey between 2016 and 2018. Then, the traffic barrier data collected was combined with historical crash records, traffic volume data, road geometric characteristics, and weather condition data to provide a comprehensive dataset for the analysis. Finally, an ordered logit model with random-parameters was developed for the severity of traffic barrier crashes. Based on the results, traffic barrier’s height was found to impact crash severity.ResultsCrashes involving cable barriers with a height between 30″ and 42″ were less severe than other traffic barrier types, while concrete barriers with a height shorter than 32″ were more likely involved with severe injury crashes. As another important finding, the post-spacing of 6.1–6.3 ft. was identified as the least severe range in W-beam barriers.Practical applicationsThe results show that using flare barriers should reduce the number of crashes compared to parallel barriers.     

E-bikers’ braking behavior: Results from a naturalistic cycling study

Abstract

Objective: The number of e-bike users has increased significantly over the past few years and with it the associated safety concerns. Because e-bikes are faster than conventional bicycles and more prone to be in conflict with road users, e-bikers may need to perform avoidance maneuvers more frequently. Braking is the most common avoidance maneuver but is also a complex and critical task in emergency situations, because cyclists must reduce speed quickly without losing balance. The aim of this study is to understand the braking strategies of e-bikers in real-world traffic environments and to assess their road safety implications. This article investigates (1) how cyclists on e-bikes use front and rear brakes during routine cycling and (2) whether this behavior changes during unexpected conflicts with other road users.

Methods: Naturalistic data were collected from 6 regular bicycle riders who each rode e-bikes during a period of 2 weeks, for a total of 32.5?h of data. Braking events were identified and characterized through a combined analysis of brake pressure at each wheel, velocity, and longitudinal acceleration. Furthermore, the braking patterns obtained during unexpected events were compared with braking patterns during routine cycling.

Results: In the majority of braking events during routine cycling, cyclists used only one brake at a time, favoring one of the 2 brakes according to a personal pre-established pattern. However, the favored brake varied among cyclists: 66% favored the rear brake and 16% the front brake. Only 16% of the cyclists showed no clear preference, variously using rear brake, front brake, or combined braking (both brakes at the same time), suggesting that the selection of which brake to use depended on the characteristics of the specific scenario experienced by the cyclist rather than on a personal preference. In unexpected conflicts, generally requiring a larger deceleration, combined braking became more prevalent for most of the cyclists; still, when combined braking was not applied, cyclists continued to use the favored brake of routine cycling. Kinematic analysis revealed that, when larger decelerations were required, cyclists more frequently used combined braking instead of single braking.

Conclusions: The results provide new insights into the behavior of cyclists on e-bikes and may provide support in the development of safety measures including guidelines and best practices for optimal brake use. The results may also inform the design of braking systems intended to reduce the complexity of the braking operation.     

不同警告方式下驾驶人应激反应能力模拟试验研究

为研究在道路突发危险场景下先进驾驶辅助系统的不同警告方式对驾驶人应激反应能力的影响，利用自主开发的驾驶人应激反应能力测试软件，以计算机模拟与驾驶模拟器为试验平台，以实际驾驶视频为试验场景，选取操作准确率和反应时间为测试指标，分析不同警告方式下驾驶人的应激反应能力。研究结果表明:视觉警告可有效缩短应激反应时间；视听觉组合警告中，视觉警告占主导作用，听觉警告起辅助作用；在真实场景视频试验环境下，驾驶模拟器模拟试验的操作效果优于计算机模拟试验。     

Improving pedestrian safety using combined HOG and Haar partial detection in mobile systems

Objective: The objective of this study is to develop a novel algorithm on a mobile system that can warn drivers about the possibility of a collision with a pedestrian. The constraints of the algorithm are near-real-time detection speed and a good detection rate.

Method: Histogram of gradients (HOG)-based detection is widely used in pedestrian safety applications; however, it has low detection speed for real-time systems. Hence, it has no direct usage for mobile systems. In order to achieve near-real-time detection speed, partial Haar transform predetections are applied to an image before HOG detection. The partial and HOG detections are merged and a score-based confidence level is defined for the final detection phase. In this way, the outcome is prioritized and different warning levels can be issued to warn the driver before a possible pedestrian collision.

Results: The proposed algorithm provides an increase in detection speed (from 46 to 76 fps) and detection rate (from 80 to 91%) with respect to HOG-based pedestrian detection. It also improves confidence of the results by multidetection merging and score assignment to detections.

Conclusions: Performance improvement of the algorithm is compared with respect to state-of-the-art detectors/algorithms. Based on the detection rate and detection speed performance, it can be concluded that the proposed algorithm is suitable to be used for mobile systems to warn drivers about the possibility of collision with a pedestrian.     

Pediatric electric bicycle injuries and comparison to other pediatric traffic injuries

Objective: The objective of this study was to conduct a comprehensive analysis of demographics, injury characteristics and hospital resource utilization of significant pediatric electric bicycle (e-bike) injuries leading to hospitalization following an emergency department visit in comparison to pediatric injuries caused by other traffic related mechanisms.

Methods: A retrospective review of all pediatric traffic injury hospitalizations following an emergency department visit to a level I trauma center between October 2014 and September 2016 was conducted. Data regarding age, sex, number of computed tomography (CT) scans obtained, number of major procedures, length of hospital stay (LOS), Injury Severity Score (ISS), and number of injuries per patient were collected and compared between e-bike injuries and other traffic injuries.

Results: Three hundred thirty-seven admissions were analyzed: 46 (14%) were due to e-bike injuries (29% of patients >12 years). Age, proportion of brain injuries, and use of CT were significantly increased compared to mechanical bicycle injuries (13.1?±?3.4 vs. 10.6?±?3.6, 13% vs. 3%, 1 [0–3] vs. 1 [0–1], P < .01, P = .03, P = .05). Age, LOS, and use of CT were significantly increased compared to injuries caused to automobile passengers (13.1?±?3.4 vs. 7.4?±?5.3, 1 [1–3] vs. 1 [1–2], 1 [0–3] vs. 0 [0–1], P < .01, P = .03, P = .01), as well as ISS and number of injuries per patient (P = .04, P < .01). Injuries caused by e-bikes were similar to injuries caused to pedestrians, except for age (13.1?±?3.4 vs. 8.5?±?3.7, P < .01). Multivariable analysis revealed a significant association between mechanism of injury and ISS, with increased ISS among e-bike injuries compared to mecahnical bike injuries (OR 2.56, CI 1.1–5.88, P = 0.03) and automobile injuries (OR 4.16, CI 1.49–12.5, (P < .01).

Conclusion: E-bikes are a significant cause of severe injury in children compared to most other traffic injuries, particularly in older children.     

未按规定导向车道行驶行为干预方法研究

为保障信号交叉口的正常交通秩序,充分遏制机动车未按规定导向车道行驶行为,亟需探究该行为的影响因素及干预方法。以北京市内4个信号交叉口处共35 h的1 666条监控视频数据为基础,对未按规定导向车道行驶行为进行定义并将其分为9类,分别对频率较高的5类未按规定导向车道行驶行为构建二元Logit模型,以确定其关键影响因素,并据此提出干预方法。结果表明,排队车辆数、大车比例、时段、车流量、照明条件等因素会不同程度地影响5类未按规定导向车道行驶行为的发生概率,其中排队车辆数及时间因素影响最为显著。在此基础上,从交通工程设施及驾驶人安全意识角度,提出优化交叉口渠化设计及信号配时、采用智能标线、强化监管力度及完善交通管控设施、加强驾驶人安全教育4种未按规定导向车道行驶行为干预方法。