ویژگی های حوادث چرخ عقب پایان در تقاطع چراغدار با استفاده از مدل رگرسیون لجستیک چندگانه
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|24716||2005||13 صفحه PDF||سفارش دهید||7856 کلمه|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Accident Analysis & Prevention, Volume 37, Issue 6, November 2005, Pages 983–995
Multi-vehicle rear-end accidents constitute a substantial portion of the accidents occurring at signalized intersections. To examine the accident characteristics, this study utilized the 2001 Florida traffic accident data to investigate the accident propensity for different vehicle roles (striking or struck) that are involved in the accidents and identify the significant risk factors related to the traffic environment, the driver characteristics, and the vehicle types. The Quasi-induced exposure concept and the multiple logistic regression technique are used to perform this analysis. The results showed that seven road environment factors (number of lanes, divided/undivided highway, accident time, road surface condition, highway character, urban/rural, and speed limit), five factors related to striking role (vehicle type, driver age, alcohol/drug use, driver residence, and gender), and four factors related to struck role (vehicle type, driver age, driver residence, and gender) are significantly associated with the risk of rear-end accidents. Furthermore, the logistic regression technique confirmed several significant interaction effects between those risk factors.
Rear-end accidents are one of the frequently occurring types of accidents, accounting for almost one-third of all reported accidents in the US (1.848 million) and 11.8% of multi-vehicle fatal accidents (1923) (National Transportation Safety Board, 2001). Rear-end accidents are the most common accident type at signalized intersections since the diversity of actions taken increases due to signal change. Specific causes of rear-end accidents include the following drivers’ inattentive driving and following too closely. A proper space cushion is needed to provide a driver enough reaction time to recognize a hazardous situation and make a stop decision. Typically, driver, vehicle, and roadway/environment characteristics influence accident occurrence and injury severity. Moreover, since a rear-end accident is related to both driving behaviors and performances of the leading (struck) vehicle and the following (striking) vehicle, the accident risk is possibly associated with struck or striking role that a driver or vehicle would assume in a rear-end accident. The driver age and gender were considered as main driver characteristics that might be associated with a rear-end accident. There is general consensus among researchers that older drivers tend to process information and take a corresponding action more slowly than younger driver. Slower reaction times for older versus younger drivers contribute to a disproportionately heightened degree of risk especially when older drivers are faced with two or more choices of action (Staplin et al., 2001). However, the younger drivers, especially under 25 years, are more likely to be involved in aggressive attitude and inattentive driving. A previous study on rear-end accidents indicated that drivers younger than 18 years were most vulnerable to roadway accidents followed by 18–24-year-old drivers; the propensity of drivers involved in accidents showed a decreasing trend with increasing age until the age of 69, after which the drivers again showed a higher accident involvement propensity as compared to the drivers who were 25–69 year old (Santokh, 2003). Additionally, among 18–24-year-old drivers, male drivers were found to be more prone to accident involvement as compared to their female counterparts. It was also found that in rear-end accidents, drivers up to the age of 25 years are more likely to be in the striking role than in the struck role. As drivers get older, they tend to be in the striking role less often than in the struck role. For different type of vehicles, steering and braking performance are critical in the avoidance of accidents. Strandberg (1998) pointed out that except for the hazards due to unpredicted change in properties within one vehicle, differences between vehicles in braking performance are responsible for many rear-end accidents. Moreover, the size of the leading vehicle may influence the behavior of the following driver. Sayer et al. (2000) examined the effect that the leading vehicle size (specifically, height and width) has on a passenger car driver's gap maintenance under near optimal driving conditions (e.g. daytime, dry weather, free-flowing traffic). Results showed that passenger car drivers followed light trucks at shorter distances than they followed passenger cars, but at the same velocities. Specifically, it appears that when dimensions of lead vehicles permit following drivers to see through, over, or around them, drivers maintain significantly longer (i.e. safer) distances. Abdel-Aty and Abdelwahab (2004) examined the relationship between vehicle type (car or LTV including light truck, van, and SUV) and the role (striking or struck) played by each vehicle in the accident. Using a nested logit structure model, they analyzed the probabilities of the four rear-end accident configurations (car–car, car–LTV, LTV–car, and LTV–LTV) as a function of driver's age, gender, vehicle type, vehicle maneuver, light conditions, driver's visibility and speed. Results showed that driver's visibility and inattention in the following (striker) vehicle have the largest effect on being involved in a rear-end collision of configuration Car–truck (a regular passenger car striking an LTV). The critical road environment conditions could play a significant role in rear-end accidents and they may contain all kinds of non-driver related factors such as lighting conditions, the roadway surface conditions, highway characteristics, traffic volume, the weather conditions, and so on. Braking performance of vehicle is substantially reduced in icy and snowy road surface condition and deceleration capacity may decrease by more than 90% compared to dry condition (Strandberg, 1998). The heavy traffic volume results in the smaller headway between gaps between leading and following vehicles, which definitely increases the possibility of rear-end accidents. Khattak (2001) reported that a majority of the accidents (54.9%) occurred during the peak times 7:00–9:00 a.m. and 3:00–6:00 p.m. A small portion (10.8%) of the accidents occurred at night on unlit streets and a smaller portion (4.9%) occurred at night on lighted streets. However, of those previous research findings, relatively few studies used the accident database and related statistical model to explore the propensity of rear-end accidents that occurred at signalized intersections. This paper presents the results of a thorough investigation into the relationship between the rear-end accidents and a series of potential risk factors classified by driver characteristics, road environments, and vehicle type. The quasi-induced exposure concept is used to compare the relative accident involvements between different risk conditions based on the 2001 Florida accident database. For striking role and struck role in the rear-end accidents respectively, multiple logistic regression models are used for hypothesis testing to identify the significant factors that contribute to the rear-end accidents.
نتیجه گیری انگلیسی
Because signalized intersections are accident-prone areas for rear-end accidents, in order to develop effective rear-end accident countermeasures, it is important to understand the driver characteristics, vehicle types, and road environment features of rear-end accidents. Using the 2001 Florida traffic accident database, this study investigated the overall characteristics of rear-end accidents at signalized intersections based on the quasi-induced exposure concept and logistic regression method. The analysis showed that the risk of rear-end accidents for 6-lane highways is higher than 2-lane and 4-lane highways; the rear-end accidents are more likely to happen at divided highways than undivided highways. However, the crash trend for higher number of lanes and divided highways may be confounded by the traffic volume effect because the highways with high heavy traffic tend to have more lanes and be separated by medians; the results confirmed that the higher traffic volume contributes to the higher rear-end risk. It was found that the relative accident involvement ratio for night is apparently lower than daytime; compared to a dry road surface, the wet and slippery road surface could greatly contribute to rear-end accidents; furthermore, as the highway character is more complex, the rear-end accident is more likely to occur. When horizontal curve and grade are present at the same time, the rear-end risk could be twice as that for normal straight highways. Moreover, the analysis showed a clear trend that as the speed limit increases, the risk of the rear-end accidents increases, especially when the speed limit is higher than 40 mph. Corresponding to the adverse road environment factors, appropriate engineering countermeasures need to be considered to reduce the rear-end crash rate. From the perspective of the intersection design and operation, improvement of configuration conditions (geometrics) may contribute to reducing reaction and stopping times, eliminating motorist confusion, and improving visibility of traffic control devices. It is suggested to avoid designing intersections located on a horizontal curve and vertical curve if possible. For an existing intersection with a curve or up/down grade, sufficient sight distance not only to the signal head but also to the other approaches should be satisfied, in order that the drivers going-through the intersection can detect potential conflicting vehicles in time. In addition, motorist information countermeasures may provide advance information to the driver about the signal ahead, such as advanced warning signs installed upstream at the intersection. That may reduce sudden-stop behaviors because of insufficient reaction time due to a signal change (dilemma zone). In the area where drivers frequently drive in the rainy weather condition, the drainage design should ensure that the rainwater is removed from the intersection surface in time. If intersections with the 50 or 55 mph speed limit have been identified to have higher rear-end crash rates, reducing the speed limit to 40 or 45 mph may efficiently contribute to the lower accident rate. The modeling results indicated that the driver factors as the striking role have different effects on the rear-end risk compared to the struck role. The struck drivers are more likely to be middle-age and female drivers, while the striking drivers with relatively larger accident propensity tend to be male, younger (<26), or older drivers (>75). For striking drivers, the accident propensity appeared a decreasing trend with increasing age until the age of 56–65, and then increase to a higher accident involvement for the age group older than 75. In general, the middle age drivers are usually under-involved in accidents due to better physical conditions and more driver experiences compared to the younger and older drivers. However, as the struck role in rear-end accidents, the middle age drivers have no advantage to void the accidents and even their quick reaction to the hazard in front of them might contribute to the struck role in a rear-end accident. Based on vehicle type, the rear-end accident risk for the striking role is increasing with increment of vehicle size, but compared to the passenger car, the large trucks or buses are less likely to be in the struck role and light trucks are more likely to be struck. For the driver residence, the non-local drivers always have higher accident risk compared to the local drivers, whatever as the striking role or struck role in a rear-end accident. Corresponding to the higher risk driver populations, generally, younger drivers have less driving experience and tend to drive in situations conditions that increase their risk. An education program to emphasize the rear-end risk at signalized intersections is strongly suggested for the younger group. According to a precious study (Eby and Molnar, 1998), some of these young drivers may engage in risky driving behaviors because they are risk taking (i.e. they perceive the risk and do it anyway) and others engage in this behavior because they are risk ignorant (i.e. they do not perceive the risk in their behavior). Those who are risk ignorant may benefit from the education countermeasure and become safer drivers by having a better understanding of the risk inherent in their driving behaviors. For the older drivers, their higher rear-end risk may result from deteriorating physical conditions, decreasing judgment ability, and vision problem. It is necessary to make a further analysis of the criteria of driving license issuance for the older drivers related to driver age and health condition. In the other hand, properly designed and implemented, the avoidance warning systems (CAWS) would notify drivers about potential dangers from roadway departures and other automobiles, particularly in rear-end collisions (James and Sarah, 2003). As a sort of in-vehicle technologies, CAWS may help reduce the crash risk for those drivers with weak driving abilities; CAWS may also effectively help bus and truck drivers to perceive the brake behavior of the leading vehicle since they may have difficultly responding to brake light of the leading car with a small headway because of a higher eyesight position. Moreover, the analysis confirmed the substantial affect of alcohol/drug use on driver's safety. Even drivers who had been drinking under legal alcohol use level could be 9.58 times more likely involved in a rear-end accident than non-drinking drivers, which strongly supports that the threshold of legal alcohol use level on the road should be reduced. If the data of illegal blood alcohol concentration (BAC) are available in traffic accident database, the further quantitative analysis of relationship between BAC and accident risk (not limited in rear-end accident) is strongly suggested. Lastly, it is worth mentioning that the analyses in this paper pertain to identifying the significant factors and comparing relative accident involvements between different traffic conditions but not to predicting rear-end occurrence rate. The accident propensity analyses in this paper may provide a better understanding of the rear-end risks and more information to seek effective accident countermeasures.