Control sub-area dividing method considering traffic flow correlation of road sections

Abstract

The invention provides a control sub-area dividing method considering traffic flow correlation of road sections. According to the method, first, abstracting processing is performed on a road network and a road adjacency matrix is built, then traffic allocation is performed on the road network by using a user balance model so as to obtain traffic flow of each road section, the traffic flow of the road section is divided by an actual length of a road section so as to obtain quasi-traffic flow density of each road section, a similarity model associated with quasi-traffic flow density of the roadsection is built, a traditional Newman community fast partitioning algorithm is improved through introduction of an edge weight parameter with a road section as a node and similarity as an edge weightof nodes, so as to be used in a weighted traffic road network, and a finally divided traffic control sub-area further meets traffic control requirements. The dividing method is easy to use, optimizesdivision of control sub-areas of the road network in a macro-level, ensures that traffic flows in the same control sub-area have high homogeneity, facilitates coordinated control in the sub-area andbetween sub-areas, and increases maximum capacity of the road network theoretically.

Description

technical field [0001] The invention relates to the technical field of traffic road network management, in particular to a method for dividing control sub-areas considering the correlation of road section traffic flow. Background technique [0002] According to the topological structure of the traffic network and the network distribution characteristics of traffic flow, the huge and complex road network is divided into several independent sub-areas, and the effective control of the signals in each sub-area is achieved, which makes the entire road network system become The key to more efficiency, reliability and flexibility. [0003] When dividing the traffic control sub-areas, most of the intersection correlation degree models are used at present, and several intersections with similar traffic characteristics are divided into the same control sub-area by combining various partition algorithms. There are also some partitioning algorithms that take the road section as the clu...

AI Technical Summary

Problems solved by technology

[0004] However, there are two main disadvantages in the spectral clustering method: one is that the partition results are highly sensitive to parameter values; the other is that the eigenvalues ​​of the solution matrix are required, and for applications with large clustering scales, the pressure of calculation and storage is relatively high

In addition, the clustering-based traffic control area division algorithm, such as the k-means algorithm, also has the following disadvantages: the data to be classified—the number of clusters k, needs to be specified in advance, and in order to determine whether a value of k is appropriate, the user often needs to select several Experiment with different k values; due to the existence of singular value objects, the distribution of data may be distorted to a certain extent. The k-means algorithm is sensitive to noise and cannot solve the problem of data clustering of arbitrary shapes

[0005] To sum up, the existing division methods of traffic control sub-areas have certain limitations.

Images