Missing Filling Method of Traffic Flow Parameters Based on Multi-source Detector Data

Abstract

The invention discloses a traffic flow parameter deletion filling method based on multi-source detector data, and relates to the intelligent traffic technical field; the traffic flow parameter deletion filling method comprises the following steps: S01, loading history traffic flow parameters of the road network into a multi-source detector, and calculating to determine an effective road segment set; S02, aiming at the effective road segment set and building a multi-source detector traffic parameter characteristic set, wherein the characteristic set is a traffic parameter characteristic matrixFeature formed by n-row-z-column; S03, using the traffic parameter characteristic matrix Feature to train a filling model used for representing traffic flow parameters; S04, using the loaded filling model to calculate the traffic flow parameters corresponding to deletions on the source detector, thus obtaining filling estimated values of the deleted traffic flow parameter. The method is suitable for multi-source various traffic parameter deletion estimations, thus filling the deletion data in real time.

Description

technical field [0001] The invention relates to the technical field of intelligent traffic, in particular to a method for filling missing traffic flow parameters based on multi-source detector data. Background technique [0002] In the intelligent transportation system, real-time traffic flow parameters are the basis of traffic control, traffic guidance and traffic information services. At present, the fixed detection equipment mainly used to detect real-time traffic flow parameters are: coil detector, electric police bayonet and video detector. The types of traffic parameters that can be obtained by different detection sources are different. For example, the coil detector can obtain the lane occupancy rate, but the electric police bayonet cannot. At the same time, the same traffic parameters obtained by different detectors have different definitions, such as The traffic flow speed obtained by the coil detector is the instantaneous vehicle speed of the section, and the traf...

AI Technical Summary

Problems solved by technology

[0004] In order to fill in the missing traffic flow parameter data, technicians have developed a variety of estimation methods to fill in the missing data, such as the historical average method and interpolation method. Although the above two methods are simple to operate, they cannot adapt to real-time traffic changes.

For other methods such as machine learning methods, the above methods need to ensure the integrity of the input fields when performing model training or real-time application of the model, so it is difficult to carry out engineering implementation

Moreover, the existing methods are mainly based on single-source detection data to fill and restore the missing data of a single traffic parameter, without considering the connection between different sources of detection data and different missing data. suitability

Images