In the upcoming five-year period, the planned research activities primarily involve the development of advanced models for numerical modeling of electromagnetic phenomena in the electric power system. The research activities of the group can be divided into following more focused segments:

1. The development of a new electromagnetic model of the grounding system based on principles previously developed for harmonic and transient analysis of grounding systems in multilayered media is planned. An analysis of the feasibility of using various modern regression models of machine learning to enhance the method of representing multilayered soil in the aforementioned model is also planned. Furthermore, an analysis of the justification for introducing frequency-dependent soil parameters into the grounding model is planned. In addition to this, an optimization procedure for minimizing the number of frequency samples while maintaining a high level of accuracy is also planned. Finally, the optimal way to introduce the nonlinear effect of soil ionization into the grounding system model based in the frequency domain will be investigated.
2. Further development and improvement of the originally developed quasi-static electromagnetic model for computing low-frequency electric and magnetic fields of power lines and facilities are planned.
3. Further development and improvement of the originally developed numerical models for calculating electromagnetic transients in power networks, with an emphasis on transmission lines in the time domain using a combination of the finite element method and transmission line models, are planned.
4. Further development and improvement of numerical algorithms for analyzing angular, voltage, and frequency stability in modern power systems with a high share of renewable energy sources using the finite element method are planned.
5. The improvement of existing numerical models of electric arc for the purpose of analyzing switching overvoltages in the power system and their inclusion in the existing originally developed numerical algorithms based on the finite element method is planned.

Throughout all the above, the possibilities of applying regression models of machine learning and classification models of machine learning to a range of practical scenarios that will be simulated using some of the previously mentioned models will be simultaneously explored. These scenarios include for example the analysis of the effective length or effective area of the grounding electrode/system and the analysis of stability in the electric power system.