Research On Adaptive Optimal Control System For Urban Traffic Signal

Traffic congestion has a great negative impact on people's work and life.There are many methods to alleviate traffic congestion,among them,the most effective way to improve the utilization of the intersection is to establish a traffic signal timing plan according to the actual accident situa-tion.Besides,it is helpful to reduce the waiting time,as well as energy-saving emission reduction.In this thesis,with the urban intersection as the research object,an innovative signal timing optimization method was researched as the result of the phenomenon that the fixed timing plan are still widely used,and the existing intersection approach could not estimate the actual intersection traffic condition precisely.This approach is based on multiple performance evaluation index,then a multiple-objective signal timing optimization model is established.Besides,the model is solved under the guidance of the intelligent algorithm theory,so the improved signal timing scheme is obtained,and achieve the adaptive optimal control.The key techniques including the acquisition of traffic flow parameters,system modeling,weight distribution,signal timing optimization and so on.This paper focuses on the key technologies above,the main contents and results include:1.For traffic information collection problem,two different videovehicle detection method,which are used in day and night,are proposed after analyzing the characteristics of traffic information and the essential functions of information collection system.The results show that: during the day,the traffic detection algorithm,which was established by virtual detection band,shadow detection algorithms,dynamic template updates etc,can effectively identify the shaded area,with good detection effect;during the night,the vehicle identification methods established by the headlights of cars,also achieved good results.What's more,the related research results can not only be used to optimize the subsequent timing,can also play a role in road design,maintenance and management.2.For the systems modeling of traffic signal timing,the paper proposes a multi-objective programming model for signal timing based on three performance evaluation indexes,this method overcomes the shortcomings of the existing typical traffic signal timing optimization methods in China.Three indicators,including delay time,parking and traffic capacity,evaluate the efficiency from different aspects of the intersection,it solved a deficiencies that a single indicator always over-emphasis on certain aspects of performance deficiencies.3.To solve the problem of weight distribution,different weights should be considered under different situations.A method is proposed which is adopt artificial determination the importance of index,combined with fuzzy analysis method.Not only simplifies the calculation of the weight of the index,reduce the artificial disturbance,but it also enables optimized timing plans can be adjusted according to actual intersection,more in line with actual requirements.A new environment mutation operator was introduced to normal immune clone algorithm,because of considering the the advantages of immune clone algorithm in solving multi-objective problems.4.According to the simulation results of a typical four-phase intersection,as well as the results of the timing optimal experiment taking by Tongzhou-Jinqiao road reveal that: Compared with thetraditional timing method and inertia weight particle swarm optimization,the optimal green time obtained by present method could not only efficiently figure out high-quality solutions,but also in accordance with the actual operating conditions.It has a positive meaning to alleviate congestion,and improve travel efficiency.