Research On Car Following Behavior Control Based On Fuzzy Inference

With the acceleration of urbanization,the number of cars has been increasing year by year,exceeding the capacity of the road network,which has led to many traffic problems increasingly prominent.Car following behavior modeling is the main research content of traffic flow theory and the core component of traffic simulation system.By analyzing the following behavior between the cars that are restricted to change lanes to overtake,and exploring its interaction,the establishment of a high-precision model of following vehicles will help to alleviate these problems,so as to improve the traffic road service level.The car following model is a typical microscopic traffic flow model,which can improve the traffic capacity of the road.Based on the following characteristics and driving safety,the following model is studied in this paper.Firstly,aiming at the extraction of fuzzy rules,an improved WM algorithm based on hierarchical clustering is studied,which realizes the adaptive extraction of fuzzy rules.Firstly,the hierarchical clustering algorithm is used to cluster the sample data.Secondly,the membership and trust degree of each sample are calculated by using the center points obtained from clustering.Finally,trust degree is added to extract fuzzy rules.In order to verify the validity of.the model,four kinds of sample data were designed and compared with three algorithms.Experiments show that the WM algorithm based on hierarchical clustering still has high completeness and robustness under the premise of adaptive extraction of fuzzy rules.Secondly,study the following characteristics,and use the driver’s reaction time to filter the car following data,so as to improve the reliability of the data.According to the analysis of the driver’s emergency braking safety interval,the expected workshop distance of the following driver can be obtained.It is found that the deviation and speed difference between the expected and the actual distance between the following car drivers will directly affect the control as the input of the car following model,and the following acceleration at the next moment is selected as the output.Thirdly,aiming at the problem of error in the de-fuzzification of the paste control theory and the lack of following data,an optimal control method for vehicle following is proposed based on safety considerations.Based on the improved WM algorithm based on hierarchical clustering,the vehicle following rules are obtained in this method.Considering the safety problem,the time headway is introduced to solve the optimal solution of acceleration by constructing the equation,so as to obtain the optimal control of vehicle following.Finally,on the basis of the above research,the vehicle following model proposed in this paper is verified by numerical simulation.Through the analysis of vehicle following characteristics,the vehicle following model proposed in this paper is verified to satisfy the following characteristics.In addition,under the conditions of starting,stopping and driving at a constant speed,the vehicle following simulation is carried out to verify the safety and reliability of the vehicle following model proposed in this paper.