Short-time traffic flow prediction method based on state pattern

Abstract

The invention provides a short-time traffic flow prediction method based on a state pattern. The method includes a) establishing a historical standard sample database; b) acquiring a traffic flow state vector M(t)=[q(t-l+1), q(t-l+2), ..., q(t)]; c) acquiring the traffic flow state pattern vector Md(t)=[r(t-l+1), r(t-l+2), ..., r(t-l)]; d) obtaining the similarity of the state patterns of the current points and points in the historical standard sample database by means of an Euler's distance calculating method; e) acquiring a filtered set A={q(t1), q(t2), ..., q(tn)}; f) further filtering the set A to obtain a set B of filtered points; and g) acquiring the prediction result of traffic flow. The traffic flow prediction method improves the accuracy and real-time property of short-time traffic flow prediction and is an effective short-time traffic flow prediction method, and the prediction results are the basis for the traffic guidance and control service of traffic management departments.

Description

technical field [0001] The present invention relates to a traffic flow forecasting method, more specifically, to a state model-based short-term traffic flow forecasting method. Background technique [0002] In recent years, with the accelerated process of urbanization and motorization, the number of cars, especially the vigorous development of private cars, has caused the traffic flow to continue to rise, and the contradiction between urban traffic demand and supply is getting bigger and bigger. Therefore, taking reasonable traffic demand control and road traffic management measures has become the key to solving urban traffic problems. How to accurately predict road traffic flow has become the bottleneck of reasonable induction, control and management of traffic travel. Traffic flow forecasting is to predict the traffic flow at time t+△t and beyond at time t based on the existing historical traffic flow data, and the forecast time is generally less than 15 minutes. Only wi...

AI Technical Summary

Problems solved by technology

The current single short-term traffic flow forecasting method requires unique information characteristics and specific applicable conditions, resulting in a single forecasting model with low accuracy for complex traffic flow forecasting, and often requires a lot of calculations to select the best traffic flow before forecasting. method, which is not conducive to real-time traffic flow prediction

Images