Creating and sharing knowledge for telecommunications

Multiswarm spiral leader particle swarm optimisation algorithm for PV parameter identification

Nunes, H.G.G. ; Silva, P.N.C. ; Pombo, J. ; Mariano, S.J.P.S. ; Calado, M.R.A.

Energy Conversion and Management Vol. 225, Nº , pp. 113388 - , December, 2020.

ISSN (print): 0196-8904
ISSN (online):

Journal Impact Factor: 2,216 (in 2011)

Digital Object Identifier: 10.1016/j.enconman.2020.113388

Abstract
The ambition for more photovoltaic (PV) systems, the concern for optimal utilisation, and the uncertainty associated with its energy production have led to an accelerated research in the PV field. The precise modelling of PV systems under any operating condition is necessary to obtain accurate and reliable estimates of PV model parameters. In this paper, a novel multiswarm spiral leader particle swarm optimisation (M-SLPSO) algorithm is proposed to solve the PV parameter identification problem. The proposed M-SLPSO uses several swarms with different search mechanisms: each swarm is guided by a leader with a different spiral trajectory. In addition, the swarms can exchange search mechanisms and agents can migrate between swarms, thereby enabling a good balance between exploration and exploitation mechanisms. This algorithm maintains a diversity of exploratory trajectories throughout the entire search process when constructing new solutions, mitigating population stag- nation, and premature convergence. Furthermore, it adapts to the optimisation problem that is being solved and simultaneously explores different regions of the search space. The performance of the proposed M-SLPSO was evaluated and compared with other state-of-the-art metaheuristics by applying it to some benchmark functions and to the PV parameter identification problem, which considered two case studies: one using a standard dataset and the other using eight experimental datasets (under different operating conditions). Comparative and sta- tistical results comprehensively indicate that the proposed M-SLPSO has an extremely competitive performance and can determine highly accurate and reliable solutions.