Exploitation of Artificial Intelligence Methods for Prediction of Atmospheric Corrosion

摘要

The contribution deals with the use of artificial neural networks for prediction of corrosion loss of structural carbon steel. Nowadays there is certain chance to predict a corrosion loss of materials by rtificial intelligence methods, especially by neural networks. A model of neural network for prediction of corrosion loss of structural carbon steel based on the input environmental parameters affecting the corrosion of metals in the atmospheric environment (temperature, relative humidity, air pollution by sulphur dioxide and the exposition time) was created. The model enables to predict corrosion loss of steel with a ufficiently small error.