Research Of Automobile Classifying Method Based On Inductive Loop

In intelligent transportation systems, traffic control system, by means of the timely data collecting and processing on the traffic condition, traffic parameters, weather conditions and equipment operation, can determine the corresponding traffic control strategies so as to ensure the road running in the best condition with its capacities to a maximum degree. Among them, such capabilities as the calculation, statistics of traffic parameters and vehicle classification that follows rapid, accurate, and automatic data collection and effective data processing are basic to achieve dynamic traffic control. With sound adaptability, stability and efficiency, loop vehicle detector has been widely applied in vehicle monitoring.The feature that the same types of vehicles have similar induction waveform can be used to cluster vehicles. The rapid clustering could be achieved for complex waveforms by the method of Adaptive Resonance Neutral Network(ART2), pre-process the collected induction waveform by means of taking the sampling quantitative values as input feature vectors for the ART2 Neutral Network, and thus the quick and automatic clustering vehicles can be achieved via ART2 Neutral Network.The paper attempts to select the appropriate ART2 Neutral Network parameters from the input training of random samples and put them into the ART2 Neutral Network in accordance with the predetermined order so as to correspond the ART2output logos with the provisioned vehicle types, finally form the mature ART2 Neutral Network suitable for vehicle classification.In addition, the paper applied the system software VC + + to develop transport parameters with the ART2 Neutral Network embedded into the system and achieved the accurate classification of vehicle types in the engineering application. Meanwhile, according to collected information the paper can make calculation and statistic on the traffic parameters. The system establishes connection with the database while instantly displaying the traffic parameters which can be used for future inquiry.