Study On Lane-changing Behavior In Bad Weather Based On Data Of Ramp Section

The bad weather will not only greatly affects driver’s psychology and sight, but also makes the road adhesion decreased and reduces braking, operation stability of vehicles, which will cause the vehicle easy to slide or slip and increase the probability of accidents accordingly. Lane-changing behavior is a part of microscopic traffic simulation and it can disturb the traffic flow greatly.The behavior is an important cause of traffic congestion, but in bad weather, it will be more complex. Because of the waving section of ramp area, the traffic in ramp region is more complex than the link, which can lead to more lane-changing and more data of lane-changing can be acquired. Hence, based on the ramp area, this paper’s study about behavior of lane-changing under bad weather will be beneficial to the further development of microscopic traffic simulation, and it will provide reliable theoretical basis for making future transport policies.The method of videotaping is used to investigate the data in some ramp section in Chengdu this paper needs. Introducing the weather factor and based on random utility theory, the lane-changing behavior under bad weather has been studied. Details are as follows:The qualitative and quantitative analysis of bad weather’s effects on drivers and vehicles as well as on road traffic flow are done. Combined with the gap acceptance theory, introduced the weather factor and the utility maximization principle is established to describe the discrete choice model of driver’s lane changing behavior, including the target lane choice model and the gap acceptance model.The driving track in the ramp section is recorded by video. Data of lane-changing in the image is captured using artificial statistics method. According to the rules of the road and the research contents of this paper, data are sorted. After extracting from the dataset the characteristic variables that describing driver’s lane changing behavior have been preliminarily analyzed.On the basis of the acquired data, estimation and calibration for lane changing model parameters using maximum likelihood method under bad weather and normal weather are both completed. Results confirm that the model is of high accuracy. According to the calibration of the model parameters, a detailed analysis of the weather factors and the differences between the lane changing behavior under two different weather is implemented. So the length of the critical gap is.