Traffic intersection vehicle flow single detector detection method based on decision theory

Abstract

The invention relates to a traffic intersection vehicle flow single detector detection method based on decision theory. The method comprises the following steps of: a first step, scanning and recording whether a vehicle exists above the detector in real time; a second step, when starting detection, zeroing the vehicle flow N; a third step, calculating the passing time delta t of the vehicles on each lane; a fourth step, converting the delta t into the vehicle flow increment delta N of each lane; a fifth step, accumulating each delta N and adding the accumulated delta N into the vehicle flow N; a sixth step, judging whether to continue to execute detection, if so, returning back to the third step, otherwise, ending the detection, and taking the vehicle flow N accumulated right now as the traffic intersection vehicle flow within a predetermined detection time range. The method provided by the invention is simple and easy to carry out, high in accuracy and higher in cost performance; the data volume is smaller; the method can directly obtain the vehicle flow indicated by a standard vehicle equivalent; and the method can use the single detectors, such as a single coil detector or a single magnetism sensitive detector and on the like, and is suitable for a wide range.

Description

technical field [0001] The invention relates to a traffic intersection vehicle flow detection method, in particular to a traffic intersection vehicle flow single detector detection method based on decision theory, which belongs to the technical field of urban intelligent transportation. Background technique [0002] As far as the applicant knows, the traffic flow at traffic intersections is generally represented by the equivalent of a standard car. Single-coil detection is usually used to detect traffic flow, and double-coil detection is often used to improve accuracy. [0003] When using single-coil detection, the traditional method is: when the vehicle enters the coil, the inductance of the coil will change, and the vehicle detector detects this change and immediately outputs a logically true signal, indicating that the vehicle has just entered the coil. The term is called Vehicle occupancy; when the vehicle leaves the coil, the inductance of the coil will return to the va...

AI Technical Summary

Problems solved by technology

[0005] However, the existing improved detection methods using single coils have problems in the stages of data acquisition, preprocessing and feature extraction: (1) in the acquisition stage, first of all, it is impossible to recognize that there is a vehicle passing above the detector when the detection is started, This also makes it impossible to detect the accurate traffic increment of passing vehicles; secondly, the main factors affecting the frequency change of the coil detector oscillation circuit are the vehicle chassis and vehicle speed. When the vehicle chassis is uneven and the vehicle speed is uneven, the frequency waveform will be affected. Third, due to factors such as measurement devices, measurement environment, measurement methods, and human factors, there will inevitably be errors in the original measurement data. Although some software algorithms have a certain ability to deal with errors, for system errors and negligent errors, existing algorithms are difficult to overcome

(2) In the data preprocessing stage, due to factors such as unevenness of the vehicle chassis, the frequency waveform has no obvious boundary, and multiple processing (such as secondary differentiation, dimensionality reduction, etc.) is required, which will inevitably lead to misjudgment of the vehicle type

(3) In the identification stage, the first problem encountered is that when the vehicle stops or changes speed on the upper part of the coil, it will greatly affect the accuracy of identification, and this phenomenon inevitably occurs frequently; secondly, the acquisition The frequency change waveform also has the problem of too high dimensionality. When the vehicle length changes or the vehicle speed is slow, the curve needs to be re-analyzed to extract eigenvalues; finally, there are many types of vehicles in my country, and the frequency waveforms of various vehicles are relatively close to each other. , and affected by the detector itself, it is difficult to accurately obtain typical waveforms of all types of vehicles

[0007] However, the existing detection method using double coils has the following problems: (1) the installation cost of the double coil is high, and the installation and maintenance are difficult; (2) in order to accurately detect the driving time and model of the vehicle, it must be ensured that There are no other vehicles within a certain distance before and after the tested vehicle. At present, the double-coil detection method of known car models that has been used in the world for decades points out that this distance must be greater than the detection range of 6.5 meters, otherwise the detection results will have serious deviations, which is difficult Ensure the accuracy of measured traffic flow

Images