Fundamentals of Artificial Intelligence in Power Transformers Smart Design

Abstract

Fundamentals of artificial intelligence in power transformers smart design such as a generative design methods and models, methods of creating digital twins of power transformers, parametric and structural optimization by means of genetic algorithm are described. Ramified equivalent circuit models are used for modeling steady state and transient physical processes in power transformers such as electromagnetic, heat transfer, short-circuit withstanding and light impulse impact. Subsystems based on open mathematical tools are developed. Electromagnetic field models for exacting calculations are used. Combinations of ramified equivalent circuit models and field models lead to decreasing of estimated time of calculations, analysis parameters of processes and synthesis of optimal power transformer construction.