Parallel implementation method of multi-source heterogeneous traffic data fusion

Abstract

The invention relates to a parallel implementation method of multi-source heterogeneous traffic data fusion. Based on the D-S (Dempster-Shafer) evidence fusion theory, parallel fusion is performed to traffic data acquired by a Sydney coordinated adaptive traffic system (SCATS) and a GPS (global position system) in a traffic system. The parallel implementation method includes firstly, collecting the data acquired by the SCATS and the GPS; secondly, diving a traffic area into subareas by a two-dimensional quadtree method; thirdly, performing parallelism to the SCATS data by the algorithm level to roughly convert the SCATS data into an average vehicle speed; fourthly, performing parallelism to the GPS data by the data level to convert the GPS data into an average vehicle speed; and fifthly, performing parallel evidence fusion to the average vehicle speeds from the SCATS and the GPS to evaluate the traffic state. The parallel implementation method has the advantages of short operation time, high accuracy, large coverage area and the like, real-time traffic state can be evaluated efficiently, and more accurate traffic information is provided for travellers and traffic managers.

Description

technical field [0001] The present invention belongs to the technical field of intelligent transportation, and in particular relates to a parallel implementation method for the fusion of multi-source and heterogeneous mass traffic data, in particular to a parallel transformation and processing of traffic data collected by ground sensing coils (SCATS) and vehicle navigation (GPS). Fusion method. Background technique [0002] The transportation system is the necessary foundation for the development of the country and society. However, the existing transportation infrastructure cannot meet the growing number of vehicles, which will bring serious hidden dangers to the national economy, people's health, and social security. There is an urgent need for a solution to traffic congestion with low energy consumption, less land occupation, and low cost. Intelligent Transportation System (ITS) came into being. Intelligent transportation system refers to the application of information...

AI Technical Summary

Problems solved by technology

However, the neural network requires a large amount of true data for training, which is difficult to achieve in practical applications.

Choi et al published the paper "A data fusion algorithm for estimating link travel time" ("A Fusion Algorithm for Evaluating the Travel Time of Road Sections") on "Intelligent Transportation System" ("Intelligent Transportation System") in 2002, using fuzzy reasoning algorithm To fuse the travel time data of road sections, but this algorithm is highly targeted and poor in portability

Automatic incident detection algorithms and advanced driver assistance systems have also been proposed, but none of these technologies can process massive traffic data in real time

Images