Forecasting Method of Traffic Flow Distribution in Expressway Network

Abstract

The invention discloses a prediction method for traffic flow distribution of an expressway network. According to the prediction method, it is assumed that future generation quantity of all entrance stations and future attraction quantity of all exit stations of the expressway network at a certain time scale are predicted already, current situation flow distribution and current situation utilization coefficients are calculated through historical flow distribution of the stations, and then prediction flow distribution is calculated to obtain a prediction result. The prediction method has the advantages that the prediction method is high in prediction accuracy in the aspect of expressway traffic flow distribution, algorithm convergence speed is high, because a sliding window technology is adopted, historical flow data can be used sufficiently to accurately and effectively correct parameters in a distribution predicting model and improve prediction accuracy, and the method is good in general applicability and not limited in a certain specific expressway section.

Description

technical field

[0001] The invention relates to the technical field of traffic flow prediction, in particular to a method for predicting traffic flow distribution of expressway networks. Background technique

[0002] Traffic distribution forecast is to use the current traffic generation and attraction to calculate the distribution of future traffic generation and attraction. It is one of the core contents of the feasibility study of highway construction projects. It is to evaluate the traffic status and comprehensively analyze the construction projects. The basis and premise of necessity and feasibility. The traffic distribution forecasting methods commonly used in engineering using the "four-stage method" for traffic planning include the growth coefficient method, the gravity model method, the intervention opportunity model method, and the system balance model method. However, the traditional methods have slow convergence speed and poor prediction accuracy when predicting ...

Images