Please use this identifier to cite or link to this item: http://hdl.handle.net/10266/6255
Title: Modeling and Parameter Estimation of Solar Photovoltaic Array
Authors: Chauhan, Abhishek
Supervisor: Prakash, Surya
Keywords: Modeling of solar photovoltaic;Environmental conditions;Optimisation;Parameter Estimation;Renewable energy;Solar Photovoltaic
Issue Date: 9-Aug-2022
Abstract: Various advantages allied with the solar PV, also faces numerous challenges that comprise the designing of the best possible configurations for solar PV arrays, optimised power converters with converter protocols, optimised maximum power point tracking (MPPT) techniques, and prediction of generated power under real-time environmental conditions. Hence, various simulations are to be executed at specific software platforms through various solar PV electrical models, where modeling of these quantified electrical equivalent models should be reliable, robust, and accurate. In view of this it is important to comprehend that the precision of a commercially available solar PV simulation software resides on the electrical PV model selected, the model parameter extraction technique incorporated, and the preciseness in estimated model parameters through various techniques. This research aims the assessment of precise solar PV modeling parameters, validated through experimental current-voltage (I-V) data, and to achieve this goal, the research is divided into three key objectives. The first major contribution of the thesis includes the physical interpretation and behavior of the solar PV, various equivalent models are studied and characterized, where single diode model (SDM), double-diode model (DDM), and three-diode model (TDM) are selected and considered for the implementation of various parameter estimation approaches. Whereas, the rest two objectives comprise the implementation of five new methodologies for parameter estimation under different environmental conditions. As the parameter estimation problem is a non-linear engineering problem, where the literature recommends precise results through metaheuristic and hybrid techniques, hence three novel metaheuristic methodologies and two new hybrid approaches are instigated in this research. Initially, a new Emperor Penguin Optimisation (EPO) based parameter estimation approach for an SDM is presented, that is cohesively analysed under different sets of temperature (T) and irradiance (G). Validation of the proposed technique is perceived on the analysis of performance characteristics i.e. I-V and power-voltage (P-V) of KC200GT, PWP201, and STP6 120-36 PV modules under various simulations conditions. Moreover, estimated results are compared with the experimental data and several established parameter estimation techniques in the literature for validation and demonstrate the proposed technique with highly precise outcomes having reduced computational cost for the PV model parameter extraction problem. The variations of estimated parameters with variables ‘T’ and ‘G’ are also been presented for all three modules considered. Furthermore, six case studies are considered to estimate the parameters of SDM, DDM, and TDM through two metaheuristic techniques i.e. Pheromone Value Black Widow Optimisation (pv-BWO) and Cannibalism Black Widow Optimisation (cn-BWO). The results are compared on the basis of the cost associated with the objective function considered i.e. minimum of root mean square error (RMSE), and the convergence speed to the optimal solution, that leads towards the extraction of modeling parameters. Where, it is asserted that pv- BWO method gives a more reliable, stable, and precise solution for unknown parameters under investigation, on the other hand, cn-BWO exhibits high convergence ability for the parameter estimation problem. An experimental test is also performed in outdoor environmental conditions, where a 149 point I-V data of EIL 75 W and SFTI 60P modules are recorded through MECO 9018BT solar PV analyzer, where ‘G’ and ‘T’ sensor is used to record the irradiance and temperature of the module under test. This experimental data is also incorporated for the validation of proposed techniques. In literature, hybrid techniques are considered the most recent techniques for parameter estimation problems. Hence, two-hybrid approaches using two new metaheuristic algorithms, i.e. the sailfish optimisation (SFO) and spotted hyena optimisation (SHO) framework along with the Newton-Raphson method (NRM) are proposed and compared, to estimate the parameters for SDM, DDM, and TDM of solar PV. Where five case studies are presented for the investigation of the proposed methods, validated and compared through the convergence speed, accuracy of estimated parameters, analysis of I-V, and P-V performance characteristics. All the proposed methodologies are state-of-the-art and certainly not implemented for the parameter estimation problem yet in literature, where the comparison with some existing techniques and experimental data exhibits the competence of proposed methodologies. It is important to pronounce that the preciseness of the estimated parameters depends on the closeness of the objective function (OF) to zero. As the information of exacted parameter values is not accessible, hence the degree of preciseness depends on the experimental data only. Any diminution in the OF i.e. RMSE is observed as a significant improvement towards the preciseness of real unknown parameter values, where the same is well observed with the proposed methodologies that exhibits more promising results when compared with the techniques in the literature.
Description: Ph.D thesis of Abhishek Chauhan (951504002)
URI: http://hdl.handle.net/10266/6255
Appears in Collections:Doctoral Theses@EIED

Files in This Item:
File Description SizeFormat 
951504002_Abhishek Chauhan_PhD Thesis.pdf19.31 MBAdobe PDFView/Open    Request a copy


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.