The Research On Vehicle Dynamic Route Guidance Method In Intelligent Traiffc System

With the continuous development of intelligent transportation System and thecontinuous improvement of the traffic acquisition technology, using real-time trafficdata to predict the driving time and route has become the problem which requireimmediately solution in Dynamic Route Guidance System DRGS (Dynamic RouteGuidance System, DRGS).When using the traditional Dijkstra algorithm which based on physical distance tosearch the shortest route, the searching range is big and the efficiency is low. Thispaper takes the directionof the urban road, intersection delay into account.Moreover, by the influence of the traffic conditions, the vehicles will comeacross certain when going through the intersection. Relevant network model wasestablished the according to the actual traffic environment. This model utilizes thebasic elements such as the intersection, sections and important node of links todescribe road network. The specific work was as follows:Firstly,minimum travel time and the traffic load are loaded as the mainoptimization object and the status variable respectively in the research of dynamictraffic distribution. Queue length models at the crossing are established based ontraffic flow gather-disperse theory and dynamic travel time is described by the conceptof density; Regarding delay at the crossing, the delay time is calculated according todifferent forward direction, so travel time functions of different forward direction androute weight of vehicles in dynamic traffic flow are detemined. The feasibility of thealgorithm has proved by simulation, and the simulation results were analysised indetail.Then, route network model and the optimization route algorithm are studied. theDijkstra algorithm’s searching range and efficiency was optimizated according to thecharacteristics of road network in Lanzhou. The Dijkstra algorithm was optimizated byrectangular limited searching range, and then sections characteristics were describedby the relationship among three adjacent points. The urban route network model,which incarnates the direction, crossing delay and restraint of urban traffic isintroduced. The route network model is destribed by some basic components, such ascrossing and route. Route characters are described by using united structure ofnode-arc. Using the thought of directed graph in graph theory, route network is turned into the mathematical model.Finally, the resolution to optimization route of turning delay and restraint isintroduced on the basis of improved Dijkstra algorithm. Basing on the above theory,experiment uses NetworkX to build the Chengguan District and the Qilihe District roadnetwork model. Finally basing on this model, several optimal routes with differenttarget selection were found. The feasibility and practicability of the research wereconfirmed in this paper. Compared with the classical Dijkstra algorithm which onlycontains the length information of traditional network, the improved algorithm wasmore practical and efficiency.