A road traffic flow prediction method, system, medium and equipment

Abstract

The invention relates to a road traffic flow prediction method, system, medium and equipment. The method includes: training the free-flow velocity and single-lane traffic capacity of the road section based on RFID and GPS data; according to the free-flow speed of the road section, the single-lane traffic capacity and the flow density curve of the road section, the parameter calibration of the segmental flow delay function in different flow speed states ;Adopt the system clustering analysis method to classify the segmental flow delay function of each road segment according to the physical characteristics of the road segment, and obtain various types of segmental flow delay functions of different road grades; apply the segmental flow delay function to the road traffic model for road traffic flow forecasting. The embodiment of the present invention calibrates segmental traffic delay functions based on RFID and GPS big data resources, provides accurate and sustainable localized road traffic delay functions, realizes accurate simulation and prediction of current and planned road traffic, and supports relevant department leaders and technical personnel analysis and decision-making.

Description

technical field [0001] The invention relates to the technical field of traffic models, and in particular, to a method, system, medium and device for predicting road traffic flow. Background technique [0002] Road traffic is an important part of the urban comprehensive traffic system. The planning, design and management of the road traffic system must be supported by a macro-quantitative road traffic model. The flow delay function is the basic function for the simulation and prediction of macroscopic road traffic models. It can support the analysis of the current situation and planned road function positioning, scheme comparison and route design, and its importance is self-evident. [0003] The Volume Delay Function (VDF) is the key basic function of route selection and flow allocation in the macroscopic road traffic allocation model. It mainly reflects the mathematical relationship between the travel time of the road and the flow of the road. And the way to divide the stag...

AI Technical Summary

Problems solved by technology

[0004] Traditional data sources for calibrating traffic delay functions, such as road traffic and vehicle speed, are generally obtained through time-consuming and laborious methods such as manual surveys, video observations, or car speed measurements. The survey costs are high, the coverage is small, and the calibration function accuracy is low, making it difficult to effectively apply

There are relatively many theoretical studies on the flow delay function in foreign countries, and a large number of functional forms such as BPR function, Conical function, Akcelik function and Logit function have been produced, but some functions have many parameters, and calibration is difficult. Difficult to adapt to the status quo of my country's urban roads with various levels, rich road types, and various types of vehicles, cannot realize localized application of parameters, and has poor function practicability

In addition, the function adopts a single type, and the sensitivity of road traffic distribution flow overflow is low, so it is difficult to effectively solve the problem of flow overflow when the road traffic distribution flow exceeds the road capacity, which affects the accuracy of road traffic flow distribution

Images