A method of the neural identification of the moisture content in brick walls of historic buildings on the basis of non-destructive tests

The article proposes a method of neural identification of the moisture content in saline brick walls in historic buildings, which is based on non-destructive testing. The method involves the use of artificial neural networks (ANNs), which are trained, tested and experimentally verified on a set of data constructed for this purpose. The set consists of the results of tests that were obtained using non-destructive methods on a selected representative group of masonry historic buildings. Based on numerical analyses, an appropriate type and structure of ANN, as well as a learning algorithm, were selected. Positive results were obtained, indicating the possibility of using the proposed method in practice. According to the authors, a wider use of the proposed method requires verification on other historic buildings. In order for other researchers to be able to verify the presented approach, a full set of data used for training and testing the ANN was provided.