RECOGNITION OF STEEL REINFORCEMENT IN TUNNEL LINING USING GROUND PENETRATING RADAR AND ARTIFICIAL NEURAL NETWORKS

Based on the fact that the explanation of image acquired from Ground Penetrating Radar(GPR)is far from an easy job,a model test was carried out.First a physical model of mountain highway tunnel lining was built.Then a test was carried out using ground penetrating radar to simulate detecting steel reinforcement that embedded in final lining along the tunnel longitudinal direction.After data acquisition stage,a preprocessing procedure including Removing DC Component,Dewowing,Removing Global Background and Trimming Data Matrix was implemented on the raw data,in order to improve signal-to-noise ratio.Data that after preprocessing was divided into block matrices with an overlapping rate of 50 percent to make data proper for Artificial Neural Networks.Then they were used for training and testing the already established BP network.The test indicated that the first layer of steel reinforcement in the second tunnel lining can be effectively recognized with this method.