Research And Implementation Of Route Planning Algorithm Based On Decision Tree Quantization

In recent years,the problem of traffic congestion in urban roads has become more and more serious.This has not only brought chaos and inefficient in transportation,but also brought inconvenience to city life.It has also caused problems in many other ways,such as environmental pollution and excessive logistics transportation costs.Traffic congestion management has become one of the hot topics nowadays.In this paper,the subject will be studied from two aspects as roads condition forecasting and research on route planning algorithm.In traditional path planning and navigation strategies,people tend to relieve traffic congestion by planning driving route to maximum the use of road segments.However,this topic will focus on the inherent attributes of the road,through the study of road attributes to define the internal factors which could cause congestion.This process could be complex and lack of order.While in the face of complex classification and attribute division,the decision tree algorithm can offer good results.This paper makes the use of association function to try to solve the problem that the traditional decision tree algorithm is more inclined to select attributes with more possible options.At the same time,this study uses formalized expression to describe the traffic flow and proposes an urban traffic flow model.This model describes the traffic patterns of ordinary road sections and intersections and defines the ability of road capacity and its impact factors accordingly.Through the analysis of the influence factors of road capacity,this paper presents some main road attributes which could help analyze and improve decision tree algorithm.After collecting and screening the data sample sets,two decision trees based on different levels of congestion were created.Then the paper compares the results of road prediction between the decision tree algorithm and traditional methods to illustrate the applicability and effectiveness of the decision tree algorithm under this issue.Besides,this article developed a quantitative study of road connectivity and applied the results to the collected data.The discussion and analysis of road connectivity makes it possible to use this property to develop a route planning algorithm.First,the actual road is abstracted into a finite graph as a simulation experiment environment.For the edges and nodes in the figure,the corresponding attributes are defined for further preparation.After analyzing the characteristics of common path-finding algorithms,this study designed an algorithm based on road connectivity.The algorithm is on the base of the attributes of road connectivity,and its derivative property—the road connectivity could be used as the value of edge weight.Finally,this study compares the route planning algorithm based on road connectivity with the reference algorithm under simulation conditions and analyzes the advantages and disadvantages of the algorithm.