Development of an automated control system for a hybrid energy complex using artificial intelligence to ensure the energy security of Ukraine

R&D registration number in UkrISTEI: 0120U102032

What priority area of science and technology does it correspond to: energy and energy efficiency

Research: applied (scientific and technical development)

Funding source: state budget, general fund of the state budget

The amount of financing: is more than 1000 thousand UAH (UAH 1619.925 thousand)

The total duration of the study (according to the plan): 2020-2021.

Perspective for further implementation in 2022-2023: completed

Research level: no analogues in Ukraine

Patent: no

Additional actions requiring further investigation: none

Participation of the institution in the existing regional targeted and integrated programs: Peter Mohyla Black Sea National University didn’t participate.

Brief description, advantages, further prospects for application.

Developed: simulation models of power system control based on wind turbine and solar panels based on hierarchical colored Petri nets in the CPN Tools environment; architecture of the intelligent system of the control system for the components of an autonomous hybrid energy system. The architecture of a neural network for controlling an energy system based on solar batteries was researched and developed using the algorithm for finding the maximum power point. The architecture of a neural network of the LSTM type was researched and developed to control the components of a hybrid energy system (photovoltaic panels, wind generator). The operating modes of the energy module on solar batteries and the energy module based on a wind generator for training neural networks of the LSTM type are determined. The control system software was developed based on neural networks of power system control systems based on the technology of searching for the point of maximum power and neural networks of the LSTM type. The most efficient multicriteria genetic algorithms for distribution problems (SPEA, NSGA2, E-MOEA) are determined. On their basis, software components have been developed for managing the distribution of resources in an autonomous hybrid energy system.

More details…