| In recent years, the freeway mileage shows an increasing trend year by year with the number of motor vehicles and driver boom, which still cannot meet the rapid growth of traffic demand. In peak period and holidays, traffic congestion on freeway has become the norm. In congested traffic condition, driving behavior is variable, time and space distribution of vehicle speed become more discrete, which will promote the probability of traffic crash. Crash on freeway is more dangerous and more serious than on the general road because of its high speed characteristic. In order to prevent crashes on freeway, active traffic management (ATM) strategy became an important means of freeway traffic management. Therefore, research on how active traffic management strategies effect the freeway potential crash rates (PCR) has important significance for freeway managers to formulate effective traffic management measures, improve the freeway traffic safety level and prevent the occurrence of traffic crashes.Firstly, on the basis of the existing research, the paper chose road conditions, traffic flow characteristics and weather environment as the influence factors of freeway PCR. A Logit prediction model of the freeway PCR is built with 1000 crash samples and non-crash samples selected from 140mile-190mile section on I-5 freeway in Washington State. And the prediction accuracy of the PCR model is verified by the receiver-operating characteristic curves (ROC). Then, the paper chose the section between 160mile and 166mile of 1-5 freeway where traffic congestion always appears to build a simulation scene in VISSIM. Also the car following and lane changing parameters are corrected considering average vehicle speed and traffic flow of three cross sections in the front, the middle and the rear of the chosen section. Finally, the paper generated variable speed strategy considering length of the speed restrictive block and the speed difference between adjacent block, generated ramp metering strategy considering ramp demand and the occupancy downstream, generated temporary shoulder use strategy considering traffic flow upstream. And the VisVAP module is used to realize the simulation of the three ATM strategies in VISSIM. And with its result the paper analyzed the influence of different ATM strategies on the PCR of the control section and the time sensitivity of ATM strategies.The study found that, in morning peak period, all the three ATM strategies can improve the PCR in the control section, and variable speed limit strategy have the optimal effect, which made the average potential crash rates of the control section decreased 29%. Ramp metering strategy and temporary shoulder use strategy made the PCR in the control section decreased by 5% and 9% separately. Time sensitivity of the three ATM strategies is obviously different. In flat peak period, the effect of the variable speed limit strategy on the PCR is obviously weakened compared with the peak period. The improvement effect of ramp control strategy in flat peak time is the same as pea k period. However, temporary shoulder use strategy not only failed to improve the safety, but increased the PCR in control section... |
PDF Full Text Request