Method for detecting traffic flow parameters based on compound sensor

Abstract

The invention provides a method for detecting traffic flow parameters based on a compound sensor. The method includes: based on the time when a vehicle arrives and leaves, collecting statistics of the traffic flow, and at the same estimating the speed of the vehicle and the length of the vehicle; based on information of the axle, the length of the vehicle and the speed of the vehicle, conducting a I level vehicle type classification on the vehicle; based on the I level vehicle type classification result, extracting vehicle wave characteristics, and conducting II level vehicle type classification. The method, by utilizing an individual geomagnetic sensor, can conduct compound detection on the traffic parameters, especially the level identification of the vehicle type. The method herein reduces feature calculated amount, increases detection efficiency, and provides basic data for smart traffic systems, traffic planning, etc.

Description

technical field [0001] The invention relates to the technical field of acquiring traffic flow parameters, in particular to a method for detecting traffic flow parameters by using a single geomagnetic sensor. Background technique [0002] Acquisition of road traffic flow parameters is an important fundamental issue in traffic management, traffic safety assurance and traffic infrastructure monitoring and maintenance. Years of practice have shown that intelligent transportation systems have provided theoretical, technical support and practical directions for solving urban traffic problems. The subsystems of the intelligent transportation system, the advanced traffic information service system and the advanced traffic management system are all based on the perfect road network traffic information. [0003] In the existing methods of detecting traffic flow parameters using sensor groups, most of them use multiple geomagnetic sensors to detect vehicle information, and then obtain...

AI Technical Summary

Problems solved by technology

However, the structure of this sensor group is complex, the accuracy of detection results is not high, and the cost is high

For the vehicle type identification technology, most of the existing technologies only carry out a simple first-level classification, that is, they are only divided into small cars, medium-sized cars, large cars, etc., and there is no further subdivision of the specific types of vehicles. The provided basic data The types are limited, which limits the scope of research on intelligent transportation and transportation planning

Images