A novel global MPPT technique using improved PS-FW algorithm for PV system under partial shading conditions

• Published in: Energy conversion and management 2021-10, Vol.246, p.114639, Article 114639
• Main Authors: Chai, Lucas Gao King, Gopal, Lenin, Juwono, Filbert H., Chiong, Choo W.R., Ling, Huo-Chong, Basuki, Thomas Anung
• Format: Article
• Language: English
• Subjects: Algorithms; Convergence; Exploitation; Fireworks; Global Maximum Power Point Tracking (GMPPT); Maximum power tracking; Mutation; Partial Shading Conditions (PSC); Particle swarm optimization; Particle-Swarm Fireworks Algorithm (PS-FW); Photovoltaic (PV); Shading; Simulation; Verification
• ISSN: 0196-8904; 1879-2227
• DOI: 10.1016/j.enconman.2021.114639

•Novel maximum power point tracking implementation using hybrid algorithm.•Convergence speed and tracking accuracy maintained from singular counterparts.•Succession of performance from singular counterpart algorithms.•Removal and supplementation of population for enriched solutions and faster tracking. Under partial shading conditions (PSC), conventional Maximum Power Point Tracking (MPPT) algorithms are unable to reach the Maximum Power Point (MPP) due to non-linear characteristics of the curve. In the literature, some algorithms, such as Fireworks algorithm (FWA) and Particle Swarm Optimization (PSO), have been used to obtain the MPP. However, they provide less than optimal convergence rate and tracking accuracy of FWA which impair the MPPT performance. Therefore, a novel algorithm for Global Maximum Power Point Tracking (GMPPT) under PSC is proposed in this paper to track the single global MPP (GMPP). In particular, an application of the hybrid method of PSO and FWA (PS-FW) algorithm for GMPPT is proposed. The difficulty of balancing exploration and exploitation is alleviated within the PS-FW through the PSO velocity operator and the FWA mutation and explosion sparks operator. The population health is maintained through abandonment and supplement strategy and an adaptive modification to the operators to enforce convergence is also described. The proposed GMPPT algorithm performance is first verified within a simulation environment under four partial shading patterns. PS-FW is further validated using experimental demonstration. The measurement results show that our proposed GMPPT algorithm can achieve better performance than their singular algorithm counterparts. The proposed algorithm successfully obtains a minimum of 18.51% better tracking speed of the GMPP under simulation verification at one initial population setting. The PS-FW under experimental verification is able to achieve at least a minimum of 23.45% better tracking speed of the GMPP in two initial population settings.