Method and system for dividing intersection control period based on multidimensional time series segmentation

Abstract

The present invention provides a method and system for dividing intersection control periods based on multi-dimensional time series segmentation. By configuring intersection static information tables, parameters or indicators reflecting traffic demand characteristics are used as period division basis to obtain data time series; according to intersection static information The configuration table collects the traffic flow data at different spatial levels; determines the time series dimension d of traffic flow at the intersection according to the channelization of the intersection; collects the time dimension of the traffic flow detection data; constructs the T*d basic matrix C; The dynamic programming and correction algorithm of change point analysis determine the number of segments of the basic matrix C and the specific division scheme. This kind of intersection control period division method and system based on multi-dimensional time series segmentation, the automatic period division mechanism reduces manual intervention, and uses traffic demand characteristics as the basis for division, which reduces the technical requirements for signal configuration links in the application and improves the efficiency. The reliability and stability of time division.

Description

technical field [0001] The invention relates to a method and system for dividing intersection control periods based on multidimensional time series segmentation. Background technique [0002] Time-division and single-point traffic signal control is one of the most commonly used traffic signal control means. When configuring the scheme, consider the time-varying characteristics of traffic demand, and implement the corresponding signal control timing scheme in different time periods such as peak and off-peak. Therefore, the time division scheme directly affects the effect of traffic signal control. [0003] In current applications, the manual division of time slots is still the mainstream. This kind of division method is relatively rough, mainly based on the historical traffic flow rules, which will easily lead to waste of green light time, and even increase delays, affecting traffic flow efficiency. In order to improve the rationality of time division, certain research resu...

AI Technical Summary

Problems solved by technology

[0004] The purpose of the present invention is to provide a method and system for dividing intersection control periods based on multidimensional time series segmentation, using non-parametric multi-change point analysis dynamic programming and correction algorithms, based on E-Agglo layered aggregation and cycle fit Statistics, estimate the number of segments and the location of segment points of multidimensional time series, realize the automatic division of intersection signal control periods based on traffic demand characteristics, and solve the problems in the existing technology that cannot take into account the selection and clustering of the optimal number of clusters The optimization of the center and the results of division are easy to fall into local optimum, the stability of the algorithm is poor, and the effect is not good in practical applications.

Images