Abstract
This paper investigates the problem of multi-objective optimization of a distribution network using BESS. For a test network model, the placement and capacity of BESS were optimized based on criteria such as energy losses, as well as the reliability metrics SAIDI, SAIFI and eENS. Calculations were carried out using quasi-dynamic simulation and Monte Carlo simulation, whilst the NSGA-II, NSGA-III, CMOPSO and SMSEMOA algorithms were used as optimiza-tion methods. The paper presents a comparative analysis of the convergence and performance of meta-heuristic algo-rithms. It is shown that genetic algorithms ensure a more thorough exploration of the design space and generate a broader set of non-dominant solutions, but require greater computational resources. A statistically significant influence of the distance to network tie open points, the distance to power supply centers, and the topological centrality of nodes on the effectiveness of BESS deployment has been established. It is shown that the greatest positive effect from the use of BESS is achieved in remote and topologically peripheral network parts with limited restoration capabilities.
The results obtained allow for the development of an approach to the preliminary selection of candidate nodes prior to running computationally complex meta-heuristic algorithms, which makes it possible to significantly reduce the design space and accelerate network optimization through the deployment of BESSs. Ref. 8, fig. 4, tables 2.