Traffic flow sequence classification method based on density peak value clustering

Abstract

The invention provides a traffic flow sequence classification method based on density peak value clustering. The traffic flow sequence classification method utilizes distance variance for measuring similarity degrees of subsequences, and comprises the steps of: calculating local density of each subsequence, measuring mutual cluster degrees between sequences, and searching for a clustering center by combining the local density with sequence spacing, wherein the sequence spacing is used for measuring class separability; and classifying sequences of non-clustering center by utilizing a density value, so as to obtain reasonable groups of the traffic flow sequences, and finally output a clustering result. According to the traffic flow sequence classification method, the subsequences of the same class can adopt the same time period dividing scheme and signal control scheme, and the workload of time period division and signal optimization under a timed control strategy is reduced on the premise of guaranteeing operation efficiency of the traffic flow.

Description

technical field [0001] The invention relates to a method for dividing traffic flow sequences, in particular to a method for dividing traffic flow sequences based on density peak clustering, and belongs to the field of traffic control research. Background technique [0002] Most of the existing signal control systems have self-adaptive functions, which mainly rely on the detection information of the coil detection equipment to optimize the signal control scheme in real time. However, in practical applications, coil detectors have a high incidence of damage and failure, and other types of detector data, including video, microwave, and geomagnetism, are difficult to directly access to existing signal control systems, resulting in many signal control systems And signal controllers can only passively adopt a fixed timing scheme. In order to improve the traffic flow operation efficiency of the intersection as much as possible, the signal timing scheme under the timing control str...

AI Technical Summary

Problems solved by technology

[0003] At present, the traditional clustering method is mostly used in the time division of signal control, that is, the flow values ​​of all time periods in a day are regarded as samples, and according to the attributes of the samples themselves, mathematical methods are used to quantify the data according to certain similarity or difference indicators There are three problems in determining the closeness relationship between samples and clustering the samples according to the degree of closeness relationship: 1. Most methods cannot automatically optimize the reasonable number of clusters, and need to compare the experimental data multiple times to get the best results ; 2. Most of the methods with automatic output of cluster numbers and results are enumerated, and the calculation time complexity is relatively large; 3. All methods are limited to determining the time division scheme for the traffic data of a specific day, without considering There are also similarities in the flow change law between different days, and the same period division scheme can be used for several days with similar change rules

Images