Babenko T. Methods and models for artificial intelligence in the AСSTP of ceramic industry

In the dissertation an actual scientific problem related to analysis and synthesis for an package of effective direct and inverse models of artificial intelligence of physicochemical processes in ceramic industry. The main idea concerning its resolving lies in establishing direct and inverse relationships between incoming and outcoming parameters of formation processes in the ceramic disperse system and half-finished goods that determined by the structure of model's package of artificial neural networks, multitude of weighting coefficients for synaptic connections, and selected type of activation function of neurons. The approach to development of support system for a decision-making process in the ceramic industry based on integration technologies of artificial intelligence with the techniques of imitation modelling for physicochemical processes in ceramic industry with various controlling strategies is proposed. The developed system ensures accelerated analysis of technological process condition as well as decision-making and has high quality and flexibility.