Abstract
This study proposes a method based on an artificial neural network for classifying truck systems for monitoring road structures. As a rule, in order to guarantee structural and operational safety and provide early warning of damage or wear before expensive repairs or even catastrophic collapse of road structures, it is necessary to provide a high-quality monitoring system based on innovative methods. Starting from the selection of suitable sensors and ending with the design of a structural condition assessment system, the development of an intelligent monitoring system for a large bridge that is really capable of providing information for assessing the integrity of the structure. This article explores the latest innovative achievements in the field of vehicle registration on road structures. Recognition and classification of each individual type of cargo vehicle is very difficult, because there are different types of them, there are different numbers of axles and their location, which cause ambiguity in the recognition and classification process. To cope with these classification ambiguities, the authors propose a method of recognition and classification based on an artificial neural network. Cargo vehicles are classified by applying an artificial neural network to the signals of truck systems received in the frequency domain. Once the neural network is trained on the received signals, it is used to determine the type of cargo vehicles using the received signals. The effectiveness of the classification was evaluated by experiments on the Freybrücke Bridge in Berlin, Germany. Model training was applied using the CNN method for vehicle classification, which resulted in an overall classification accuracy of about 95%.