A Neural Network Approach for Existing Bridge Evaluation Based on Grid

Existing bridge state is often evaluated periodically so that the bridges with high risks can be maintained. This study presents a method for estimating the status of RC beam bridges using an artificial neural network based on grid. The inputs of the network for training and testing are corresponding to the criteria of bridge evaluation and inspection result. A computer program written in MS C++.Net was used for the implemented of grid schedule and ANN toolbox of MATLAB is used for predictions. As a conclusion, when the calculated bridge rating and evaluation time compared with the ANN method, it is proven that the proposed algorithm provided results similar to those obtained by experts, but can improve efficiency of bridge state assessment. We expect that this algorithm can be used as an effective assessment method for existing bridge structures in regular inspection stages.