Dynamic Configuration Method Of Intelligent Traffic Front-end System Resources For Crowding Control

At present and for a long period of time in the future,due to the continuous updating and improvement of information technology,traffic information collection,transmission and distribution equipment and related technologies will become more and more advanced,and the traffic management department will no longer be how to obtain real-time traffic.Data problems,but how to extract information from a large number of static and dynamic traffic data that can more intuitively,accurately and effectively reflect traffic conditions,or solve traffic problems to meet decision-making needs.Due to the one-sided,incomplete,high error rate,high uncertainty and small amount of information of single monitoring data,the traffic control department generally adopts the means of processing data information of multiple front-end systems in parallel.However,in the practical application,a prominent problem is reflected.In the background management configuration process of the front-end system resources,the theoretical technology in the field of traffic engineering is still in the subordinate position.Most of the regional data collection points and front-end cameras are still mainly used in decision-making.Relying on historical data and artificial experience,the actual response is difficult to cope with the complex and changing traffic environment.After the congestion of the road network,the effective and precise selection of the data source of the front-end system is the main research object.It is defined as the management configuration of the front-end system resources in the background.The spatial area is used as the basic unit of the configuration,and the segment based on the front-end data is studied.Spatial correlation feature,spatial data mining for road segments,assuming that there are different degrees of congestion in the road network,through the spatial traffic correlation analysis and spatial clustering of the road segments,the resource allocation range of multiple crowded areas in the background is determined.And give a ranking based on the contribution of the data source to the road network management control.The research aims to help determine the source of scientific data information and exclude invalid data for management decision makers to refer to.It mainly includes the following research contents:(1)Based on the detection principle and detection parameters,the data acquisition characteristics of different detectors are compared and analyzed.The data collection process of fixed detectors and the integrated calculation method of data in the system are studied.The Hurst index is used to judge whether the traffic parameters are in time series.Trends;research on platform management and control methods for front-end system resources;based on spatial information of detectors and spatial benefits of road network data,a resource management configuration method based on spatial regions is defined,which is quoted by Moran index and local Moran index.Spatial correlation verification of all regions and regional units.(2)The decision mechanism of the front-end system state recognition module is studied,and the traffic discriminant algorithm based on historical data fuzzy clustering is established.The type of congestion detected by the front-end system is defined,and the congestion warning is redefined according to the specific requirements of the management background resource configuration.Standard;based on density clustering,a road network congestion area partitioning algorithm is constructed.(3)The K-means algorithm is improved and the natural neighbor search rule is introduced to define the clustering algorithm of the road segment.Combined with the front-end system data,a method for calculating the congestion index of the road segment based on speed and occupancy is studied.Based on the Euclidean distance and congestion index,a new method is established.The congestion point distance calculation model is analyzed.The cluster-based resource point location configuration method is analyzed.By establishing the matching attribute table of resources and road segments,the configuration unit of the corresponding resources can be determined according to the road segment clustering;a resource configuration is established according to the traffic situation environment.The method of determining the priority;verifying the feasibility of the proposed model algorithm by establishing a simulated road network in the downtown area of Boise,Idaho.The resource configuration method studied in this paper can be used to determine the corresponding points of data detection and video monitoring,and the management configuration order,which are based on the area as the configuration unit,and the data is accurately captured in the background.