Research On Intelligent Traffic Light Unordered Control Algorithm Based On Vehicle Flow Detection

In recent years,the problem of urban traffic congestion has become more and more serious.The traffic light control system with a fixed change cycle cannot divert traffic based on real-time traffic volume,which causes the traffic system to face the problem of low utilization of road resources.Therefore,how to identify the real-time traffic flow on the road and realize the intelligent control of traffic lights according to the real-time traffic flow in all directions is a challenging research topic.Based on this background,this paper studies the problem of intelligent traffic light control,and conducts research work from two aspects: vehicle flow detection and intelligent control of traffic lights.The main contributions of this thesis are as follows:1.Propose algorithm VFDV(Vehicle Flow Detection algorithm based on Video).Different from the traditional method that uses the original sensor to detect vehicle flow,the algorithm uses road video monitoring data to detect vehicle flow in real time,which has the advantages of low cost,convenience and high accuracy.The algorithm first uses road video surveillance data to extract image frames,and uses image preprocessing algorithms and lane line detection algorithms to correct the images to improve image readability.Then use the background subtraction method to detect foreground vehicle targets.Afterwards,use the projection features to perform feature extraction on the single lane image to obtain the feature vector of the single lane.Finally,the K-nearest neighbor algorithm is used to obtain the traffic flow detection result of the current lane.Compared with the traditional traffic detection algorithm,the algorithm proposed in this paper uses a classification algorithm to detect vehicle flow,which avoids the false recognition caused by the traditional algorithm that needs to identify individual vehicles one by one.In the verification phase,the video taken at the real intersection is used as the data source,and the valid images in the video are used as the label samples.Designrelated experiments verify the effectiveness and superiority of the proposed algorithm.2.Propose algorithm NFTP(Not Fixed Time Plan of Intelligent traffic light control).In order to avoid the problem of road resource waste caused by the large difference in traffic flow of the same phase in the fixed timing scheme of traffic lights,an intelligent sequential control algorithm for intelligent traffic lights is proposed.The algorithm first uses a video-based vehicle flow detection algorithm to obtain the current traffic conditions in each lane in real time.The corresponding initial green light duration is assigned to each lane according to the traffic flow detection result.Then,based on the actual congestion at the current intersection,the initial green light duration of each lane is optimized.Afterwards,based on the principle of compatible lanes,a combined method is used to calculate the optimal traffic sequence for the current cycle.Corresponding priorities have been set for each lane to give priority to lanes with high traffic volume.Compared with the traditional traffic light fixed timing scheme,the algorithm proposed in this paper effectively solves the problem of adapting the green light time to the traffic volume of each lane,reduces the average waiting time of vehicles at intersections,and improves road resource utilization.Experiments are performed using simulation data.The experimental results show that the design scheme in this paper increases the number of vehicles passing in a unit time and shortens the average waiting time of the vehicles.