Management of Renewable Energy Integration into Local Energy Systems Under Unstable Uncertain Grid Support Conditions

Abstract

The study is based on the use of methods of mathematical modeling of power system regimes, time series forecasting based on neural network models LSTM/Transformer, predictive control, as well as methods of optimizing the flow distribution of electric energy taking into account network constraints and uncertainty of renewable energy generation. To ensure timely adoption of control decisions, a system of continuous monitoring of the regime parameters of the power system was used, which collects, processes and analyzes information about the network status in real time. The algorithm adaptation is implemented by dynamically changing the weight coefficients of the objective function in accordance with the current regime state of the power network. A predictive -adaptive algorithm for optimizing the operating modes of a local power system has been developed, which combines neural network forecasting of future operating parameters with active control based on predictive regulation and a system for continuous monitoring of the power system. The proposed algorithm forms optimal control influences for generation, energy storage systems and controlled load, takes into account the uncertainty of forecasting renewable generation and automatically adjusts the optimization parameters depending on the current state of the network. The use of a monitoring system ensures prompt detection of deviations in operating parameters, increases the accuracy of forecasting and the efficiency of making control decisions. This allows reducing frequency deviations, reducing the risk of overloading network elements, using energy storage more efficiently and increasing the share of integrated renewable generation without violating technical restrictions.