Optimal Implementation Method Of Multiple Traffic Sensing Equipment For Intelligent Highway

In recent years,with the rapid increase of per capita car ownership in China,problems such as highway traffic safety and congestion have become increasingly prominent.intelligent highway is an effective way to solve the above problems.It uses multi-sensory and modern communication technology to sense the traffic flow parameters and abnormal events of the expressway in real time and accurately,and feeds back the decision-making after big data calculation to the high-speed vehicles.Effectively reduce the probability of accidents and improve traffic efficiency.However,how to reasonably deploy multi-sensing devices on the highway,while satisfying the corresponding sensing performance indicators and at the same time obeying the constraints of deployment costs,is still a key problem to be solved urgently.To this end,this paper proposes a highway section division method based on DBSCAN clustering and an optimal layout method for roadside equipment based on genetic algorithms,and builds a intelligent highway perception equipment layout simulation platform,using Hangzhou-Ningbo Expressway G92(Hangshao section)and The historical traffic data of a section of Shanghai-Kunming Expressway(hereinafter collectively referred to as G92Expressway)was used as the research object to verify the method proposed in this paper.The main research content of this paper is described as follows:First of all,this paper summarizes and summarizes the research on the deployment of sensing devices from four aspects: travel time estimation,event detection,traffic state prediction,and combined deployment of traffic sensing devices.It also conducts a full investigation on the types and performance parameters of various common roadside sensing equipment.On this basis,this paper proposes a road section division method based on the DBSCAN ordered clustering model,which uses road section mileage and speed or traffic flow as clustering indicators,takes the sample size of all road section units as input,and then sets the initial Value scanning radius and the minimum number of included points,randomly select an unvisited sample for cluster division,and then traverse all samples,and divide them into different clusters according to the similarity of speed or flow.The road segment mileage ensures that the samples classified into one class are adjacent and ordered.Therefore,expressways with similar characteristics of traffic flow parameters are divided into different road sections,which will prepare for the subsequent optimal implementation of corresponding road sections.In addition,on the basis of studying the single-objective optimal layout model,this paper proposes a multi-objective optimization model considering the travel time and layout cost of road sections.When solving the model,the multi-objective function is transformed into a single objective function by the weighted summation method,and then the transformed new objective function is used as the optimization objective,and the genetic algorithm is used for iterative solution to obtain the optimal objective function value.In order to theoretically verify the feasibility and accuracy of the proposed method,we constructed a 3-kilometer simulation test road,and simulated simulation data with different traffic flow characteristics by designing different traffic events in each kilometer road section.And use this data to verify the method proposed in this paper.After a large number of experiments,it can be seen that with the increase of the number of sensing devices deployed,the travel time estimation error will not continue to decline after it drops to a low level,but is in a state of fluctuation.The multi-objective optimization model proposed in this paper selects the layout scheme with the lowest equipment cost when the estimated error of the road segment travel time tends to the ideal value.Through this scheme,the error between the travel time obtained by the sensing device and the true value is only 0.0255 s.Finally,we bidirectionally coupled the traffic flow simulator SUMO and the network simulator OMNe T++ to construct a simulation platform for optimal implementation of intelligent highway sensing devices.Taking the historical traffic data of G92 expressway as the research object,the optimal layout method proposed in this paper is applied to actual cases.Firstly,the 78.075-kilometer expressway is divided into 8 sections by using the DBSCAN ordered clustering method.Referring to the industry-recognized roadside perception equipment system construction guidelines for high-speed traffic conditions,the divided road sections are divided into different perception performance levels.In each level of road section,a multiobjective optimization model is constructed considering the estimated error of road section travel time and layout cost,and is solved by genetic algorithm.The results show that the travel time obtained by the perception system is very close to the real value due to the full-coverage layout scheme of the fusion of omnidirectional radar and video detector used in the L3 road section;the optimal implementation method proposed in this paper is applied to L1 and L2 road section.The L2-level road section adopts the Levision all-in-one machine as the sensing device for optimal implementation,and its detection accuracy can reach 98.5%;while the L1-level road section chooses the video detector for optimal implementation,and the detection accuracy can reach 84.2%;the reason is that the video detector The detection range is small and the sensing accuracy is low.If the whole line is laid out,the cost will be very high.After comprehensively considering the detection accuracy and equipment cost,the method in this paper compromises the detection accuracy to a certain extent.But overall,the travel time estimation accuracy of each road section can reach a high level,which verifies the effectiveness of the method proposed in this paper.