Multi-objective approach for protection of microgrids using surrogate assisted particle swarm optimization (SAPSO)

Autor: A. Senthilselvi, R. Hannah Lalitha, G. Emily Manoranjitham, D. Weslin
Rok vydání: 2021
Předmět:
Zdroj: Applied Nanoscience. 13:1401-1414
ISSN: 2190-5517
2190-5509
DOI: 10.1007/s13204-021-02044-7
Popis: Despite of several formulations discussed for operation of Directional Over current Relays (DOCRs) in micro grids the coordination of DOCR is yet a great tasked to be solved. A micro grid is a group of generators preferably a renewable energy source and loads interconnected. The modes of micro grid operation, grid connected mode and autonomous mode are to be considered simultaneously to design an optimal protection scheme for the micro grid. Several conflicting objective functions that must be optimized simultaneously during the relay setting process make the coordination problem to be treated as multi-objective. In this work a multi-objective approach yielding better coordination is presented. There are two major objectives as follows; former is the synchronization period between the main and secondary relays, later is the relay operation time. The conventional Multi-Objective Evolutionary Algorithms (MOEAs) may take thousands of function evaluations, which is not practical for real-world problems. Hence surrogates are introduced in those MOEAs to approximate the fitness function. The relay coordination is devised as a constrained multi-objective problem. The Surrogate Assisted Particle Swarm Optimization (SAPSO) Algorithm is applied to accomplish better coordination between the relays. The scheme is validated using standard IEEE systems and the results are compared with the conventional Multi-Objective Particle Swarm Optimization (MOPSO). The optimal values of pickup current and Time Dial Settings are computed for both modes of micro grid operation. The sum of operating time of all relays and the synchronization period between the main and secondary relays are minimized. The Inverted Generation Distance (IGD) is employed as the parameter for computing the performance of the algorithm. Also the convergence characteristic is analyzed. From the results it is proved that SAPSO have good stability than MOPSO and offers better Pareto fronts, convergence and distribution.
Databáze: OpenAIRE
Externí odkaz: