Optimal identification of model parameters for PVs using equilibrium, coot bird and artificial ecosystem optimisation algorithms

Abstract

Realising an accurate estimation of model parameters for solar cells and the Photovoltaic modules has serious importance for enhancing the performance of their control systems. Three neoteric metaheuristic methods of Artificial Ecosystem-based optimisation, Coot Bird-based optimisation, and Equilibrium optimiser have been applied and evaluated concerning the accurate estimation of various Photovoltaic models. The validation of the applied methods has occurred for valuing the model parameters of R.T.C. France solar cell, and Thin-film ST40 Photovoltaic module. The objective function has been formulated as the Root Mean Square Error between the actual and estimated data. Matlab/Simulink has been used for the verification of the optimisation methods. The outcomes demonstrate that: (1) The three optimisation algorithms can resolve the problem of the Photovoltaic parameter estimation; (2) There are small distinctions between the three algorithms concerning their best value of the impartial function; this distinction between the best and worst algorithm is 10-9 for R.T.C. France solar cell for SDM; (3) The best algorithm considering the best value of the objective function is Artificial Ecosystem-based optimisation for R.T.C. France solar cell; (4) Statistical results prove that the three algorithms have tracking efficiencies of 100%, 99.999%, and 98.285% for Artificial Ecosystem-based optimisation, Coot Bird-based optimisation, and Equilibrium optimiser, respectively, based on 10 individual runs for R.T.C. France solar cell for SDM. Moreover, the simulation results show that the I/V curves obtained employing Artificial Ecosystem-based optimisation, Coot Bird-based optimisation, and Equilibrium optimiser techniques were also matched with the corresponding datasheet curves with the Artificial Ecosystem-based optimisation and Coot Bird-based optimisation's predominance in standings the convergence speed, tracking efficiency, statistical indices, and solution accuracy.

Network performance oriented study of sensor nodes combining improved COOT and MOAHA algorithms

“…Lé vy flight is a random wandering search that obeys the Lé vy distribution, and perturbing COOT through the Lé vy flight strategy helps to increase the population diversity [8]. The inverse learning strategy, on the other hand, guides the algorithm to explore different search paths from non-optimal solutions, thus increasing the possibility of finding the global optimal solution [9]. The improved position update formula for the Levy flight strategy is shown in equation (1).…”

A New Hybrid Multi-Population GTO-BWO Approach for Parameter Estimation of Photovoltaic Cells and Modules

“…Another model described in this work is the TDM; this model includes a current source, two resistors, and triple diodes, as shown in Figure 2. Dual diodes are considered in the model and are similar to those of the DDM, due to the reassembly and connection losses, while the third diode is due to the losses of the reassembly flow zones and boundaries [65,67].…”

“…where N s denotes the number of solar cells combined in a series, while N p denotes the number of cells connected in parallel [67,68]. Seven parameters should be calculated in the PV panel circuit; these are A 1 , A 2 , R s , R sh , I d1 , I d2 , and I ph .…”

Photovoltaic model parameters identification using an innovative optimization algorithm

“…Throughout the past few years, researchers have used a variety of meta‐heuristic optimization approaches for the proposed problem, such as the Real Coded Genetic algorithm (RCGA) [24], Salp Swarm Algorithm [25], Crow search algorithm (CSA) [26], Particle swarm optimization [27], harmony search‐based algorithms [28], Firefly algorithm [29], Artificial bee colony [30], Cuckoo algorithm [31], Crow Whale optimization algorithm [32], A Genetic Algorithm Based on The Non‐Uniform Mutation [33], Directional Permutation Differential, Evolution Algorithm [34], Hybrid Grey Wolf Optimization and Cuckoo Search Algorithm [35], Biogeography Based Optimization [36], Enhanced JAYA [37], Brain Storming Optimization algorithm [38], Transient Search Optimization [39], Hybridized interior search algorithm [40], hybrid differential evolution with whale optimization algorithm [41]. Electromagnetic‐like Algorithm [42], Moth Search Algorithm [43], trust‐region‐reflective technique [44], shuffled frog leaping algorithm [45], Gradient‐based optimizer [46], Simplex simplified swarm optimization [47], Improved gradient‐based optimizer [48], Artificial ecosystem‐based optimization (AEO) [49, 50], Simplified swarm optimization [51], hybrid African vultures–grey wolf optimizer [52], modified social network search algorithm combined with the Secant method [53], improved stochastic fractal search [54], Random learning gradient‐based optimizer [55], comprehensive learning Rao‐1 [56], differential evolution [57‐59], arithmetic optimization algorithm [60], Fractional Chaotic Ensemble Particle Swarm Optimizer [61], supply–demand optimizer [62], Runge Kutta based optimization (RUN) [63]. Table 1 summarizes the main findings through the last 2 years.…”