Characteristics And Modeling Methods For Driving Behavior In Mix-traffic Environment

Mix-traffic is the main characteristic of large cities in China for a long time,which makes large cities exist higher potential safety problems and lowers the capacity of road.Mix-traffic consisting of automobiles and bicycles will certainly exist for a long time in our country,from the development of bicycles and the fastest-growing car ownership trend.The development of Intelligent Transportation Systems(ITS)is an important measure for relieving above transportation problems.As a complicate individual with ability of thinking,summarizing the experience and refining subjectivity,driver is the core of ITS.In mix-traffic environment,for mix-traffic information of low regularity,driver can percept,understand,judge,coordinate and control four elements including human,vehicle,road and environment through exerting the subjective dynamics.With the development of traffic technology,driving behavior has been becoming the key of ITS,and the theory basis of microscopic traffic flow simulation.So the research of the mix-traffic system should be focused on the driving behavior,and human-vehicle-road-environment also should be considered as a whole system.According to the demands of microscopic traffic flow simulation modeling and ITS,and with focus on driving behavior characteristics,the paper has studied the driving behavior in mix-traffic environment.The main contents are as follows:bicycle traffic characteristics in mix-traffic environment;disturbance state recognition based on genetic neural network;the car-following model based on neural network;the sphere of bicycle disturbance influence based on immune neural network.These studies can completely hold driving behavior characteristics in mix-traffic environment,which provide the theoretical foundation to the simulating realization of ITS and have an important theoretical value and practical significance for improving traffic safety and capacity.