Prediction of absolute humidity of concrete gravity dam based on artificial neural network

In artificial neural network(ANN)method, the information treatment of the network are finished through interaction of neurones of the network. There are a series of advantages in the methodology,such as high degree non-linear,self-adaptation,self-learning,etc. Therefore the ANN method is used widely in the fields of prediction of physical paramters. In most cases, seepage equations show strong non-linear characteristics. This paper presents and establishes an ANN model based on the training method of learning into groups. Combining the practice of Huanglongtan Power Station, application of the ANN model to prediction of absolute humidity of concrete dam which can be used to separate dry concrete space from moist space is studied. There are high degree accuracy in the prediction result through using the ANN method. The results demonstrate that this method is widely available for the fields of dam safety monitoring and operation.