Dynamic Leader Multi-Verse Optimizer (DLMVO): A New Algorithm for Parameter Identification of Solar PV Models

Abstract

To efficiently extract the model parameters of photovoltaic (PV) modules, this paper proposed an identification method based on the Dynamic Elite-Leader Multi-Verse Optimizer (DLMVO) algorithm. An adaptive strategy was used to control parameters based on population evolution rate and aggregation rate to balance the exploitation and exploration to avoid the search falling into local optimization. In addition, this paper proposed a dynamic elite-leader-based variation strategy to enhance the probability of variation success and improve merit search speed. This proposed algorithm was applied to the parameter identification of two different PV modules and validated using six existing methods in the literature for comparison. The experimental results show that the DLMVO algorithm significantly reduced the standard deviation of the three models compared with the standard deviation of the MVO algorithm, the single diode decreased by nearly 40%, the single-component model decreased by about 28%, and the double diode exhibited the best effect, which decreased by 83%.

Comparative analysis of the hybrid gazelle‐Nelder–Mead algorithm for parameter extraction and optimization of solar photovoltaic systems

“…To provide a comprehensive comparison, Table 8 compares the estimated parameters and RMSE values for the SDM obtained using the GOANM algorithm with other more recent optimization approaches, including gradient based optimizer with criss‐cross and NM algorithms (CCNMGBO) [41], improved moth flame algorithm (IMFOL) [42], ranking teaching–learning‐based optimization (RTLBO) [43], dynamic leader multi‐verse optimizer (DLMVO) [45], reptile search algorithm with Cauchy mutation and opposition‐based learning strategies (OBL‐RSACM) [46], chaos game optimization‐least squares (CGO‐LS) [47], artificial humming bird optimization (AHO) [48], particle swarm optimization with random reselection (PSOCS) [49], elite learning adaptive differential evolution (ELADE) [50], coyote optimization algorithm (COA) [51], gradient based optimizer with differential evolution and sine cosine algorithm (SDGBO) [44], hunger games search algorithm with quantum Nelder–Mead (IHGS) [52], bald eagle search (BES) [53], differential evolution (DE) [56], improved tunicate swarm optimization (ITSA) [54], tree seed inspired algorithm (TSA) [55]. The results highlight the GOANM's superior performance, as it achieves the lowest RMSE value among all the compared methods.…”

“…Table 16 provides a comparison of the estimated parameters and RMSE values for the PV module obtained using the GOANM algorithm with other more recent optimization approaches, including CCNMGBO [41], IMFOL [42], RTLBO [43], DLMVO [45], OBL‐RSACM [46], CGO‐LS [47], AHO [48], PSOCS [49], ELADE [50], COA [51], SDGBO [44], IHGS [52], BES [53], DE [56], ITSA [54], TSA [55]. The results show that the GOANM algorithm achieves the lowest RMSE value among the compared methods, indicating its superior performance in accurately modelling the Photowatt‐PWP201 PV module.…”

“…The power–voltage ( P – V ) and current–voltage ( I – V ) curve characteristics of the optimized SDM using the GOANM algorithm closely matched the experimental data. Comparative analysis with other recent optimization approaches (Gradient‐based optimizer with Criss‐cross and Nelder–Mead algorithm [41], local escape operators based improved moth flame algorithm [42], ranking teaching–learning‐based algorithm [43], gradient based optimizer with differential evolution and sine cosine algorithm [44], dynamic leader multi‐verse optimizer [45], reptile search algorithm with Cauchy mutation and opposition‐based learning strategies [46], chaos game optimization‐least squares [47], artificial humming bird optimization [48], particle swarm optimization with random reselection mechanism [49], elite learning adaptive differential evolution [50], coyote optimization algorithm [51], hunger games search with quantum Nelder–Mead [52], bald eagle search [53], tunicate swarm optimization [54], improved tree seed inspired algorithm [55], differential evolution [56]) further highlighted the GOANM's superior performance, as it achieved the lowest RMSE value among all the methods considered. Similar observations were made for the DDM optimization using the GOANM approach, indicating its effectiveness in fine‐tuning the parameters of the DDM for the RTC France solar cell case study.…”