Vehicle identification method using neural network modeling technique

摘要

In the detection of the vehicle passing through the road, when the vehicle passes the induction loop coil buried in the road surface, it detects the change of the magnetic field generated in the loop coil so that the vehicle type of the passing vehicle can be recognized quickly The method comprising the steps of: collecting information on each vehicle type through an inductive loop detection means and then executing a predetermined process through data analysis means to set and store reference data for each vehicle type; A step of processing the information about the vehicle to be recognized into a predetermined state and comparing and analyzing the information with the set reference data to recognize the corresponding vehicle occupying the road and detecting occupancy time and non occupancy time and traffic volume, The setting of the data includes grouping the collected vehicle information by vehicle type and then normalizing each grouped information, Calculating a mean value of each vehicle type from the learned value, and setting the calculated average value as a reference value; and recognizing the vehicle type includes calculating the inductance of the vehicle based on the detected inductance Calculating a mean value after processing the vehicle information by a technique of the general learning vector quantization neural network after normalizing the vehicle information and comparing the average value of the set vehicle groups with the average value of the set vehicle types, And then recognizing the corresponding vehicle.