Abstract: The theory and the algorithm of the artificial neural network are applied in the research of the double glow plasma surface alloying processing. On the basis of the training of network, the mathematics model on the relationship between the double glow plasma surface alloying technique and the composition, the gross mass fraction of element, the thickness of surface alloying layer as well as the absorption rate is built. The calculation results are in good agreement with the experimental results.

Key words: Double glow Artificial neural network Prediction model