Method for acquiring and predicting road traffic flow

Abstract

The invention discloses a method for acquiring and predicting road traffic flow. According to the method, an annular sensing coil, a vehicle detection module, a traffic flow acquisition module and traffic flow data pre-processing and prediction software are adopted. The road traffic flow data pre-processing and prediction software runs in an upper computer (a personal computer (PC)) and is used for reading traffic flow data in the traffic flow acquisition module (a security digital (SD) card) through a network interface. In order to improve the reliability of prediction, a method for pre-processing and predicting the road traffic flow data comprises the following steps of: firstly, de-noising the traffic flow data by using a wavelet analysis method and a least square method; and secondly, establishing a traffic flow prediction model by using an improved back propagation (BP) neural network, so that prediction of the traffic flow is realized, and reference is provided for a control timing scheme for road traffic optimization and road traffic planning. The method has the advantages that: road traffic parameters such as vehicle flow, average vehicle speed, occupancy and traffic density in a rated period can be obtained, so that the prediction of the road traffic flow is realized, and the accuracy of data acquisition and the accuracy of prediction of the road traffic flow are improved.

Description

technical field [0001] The invention belongs to the technical field of ITS (Intelligent Transportation System), and in particular relates to the design of a road traffic flow data acquisition device, and a method for performing denoising preprocessing on the collected data and predicting the road traffic flow. Background technique [0002] Traffic flow is an important parameter in traffic information. Traffic flow collection and prediction is an important means to control the flow of vehicles and ensure safe and smooth roads. It has an important role and significance in urban traffic control and management. [0003] Usually, the acquisition of traffic flow is accomplished by various traffic flow detection devices, which use motor vehicles as detection targets to detect the existence or passing status of vehicles. Existing traffic flow detection methods mainly include induction coil detection, microwave detection and video detection. Among them, induction coil vehicle detecti...

AI Technical Summary

Problems solved by technology

[0005] The defects of the traditional induction coil traffic flow detection system in detection: when detecting the induction coils of multiple lanes, the detection system adopts the method of setting the oscillation circuit to generate different oscillation frequencies to avoid the coupling between the induction coils of each lane effect

Therefore, in practical applications, it will cause problems such as low prediction accuracy and long calculation time, making it difficult to apply in road traffic signal control systems.

Images