A novel hybrid parameter estimation technique of solarPV

Abstract

Summary This paper develops a novel hybrid parameter estimation technique of the solar photovoltaic (PV) cell. First, a mathematical model of the solar PV cell, which is a function of three unknown parameters, namely, series resistance, shunt resistance, and thermal voltage, is developed. Then, trust‐region algorithms are used to evaluate these three unknown parameters. The accuracy and efficacy of the proposed technique are tested with five different PV modules of mono‐ and poly‐crystalline materials by estimating their various benchmarks, such as relative maximum power error (RMPE), root mean square error (RMSE) and mean absolute error (MAE) of the current. In addition, the results of the proposed technique are compared with the experimental as well as the results obtained from other established techniques. The %RMSE, as obtained with the proposed technique, is found in order of 10−4 while the other techniques provide the value in the range between 10−1 and 10−3. Similarly, %MAE and %RMSE, as obtained with the proposed technique, are found 0.064 and 0.17, respectively, which is smaller than the other existing techniques. Hence, the proposed hybrid technique provides characteristics that are almost similar to the real characteristics of solar PV. Highlights The characteristic equation is developed in terms of Rs, Rp, and Vth. The hybrid technique is used to evaluate five unknown parameters of solar PV. Various performance indicators have been estimated for different types of solar PV. Characteristics obtained with the proposed technique are similar to real characteristics. The proposed technique reduces the complexity of parameter estimations.

An efficient data sheet based parameter estimation technique of solar PV

“…Furthermore, due to the limited spaces requirement for their installation, such as the rooftops and side of the buildings, making it more reachable to both commercial as well as domestic users. Unfortunately, due to significant installation cost and suboptimal conversion efficiency, it faces the challenges in terms of competitiveness against the conventional electrical generation systems 1 .…”

“…Several mathematical structure of solar PV, including single exponential model (SEM), double exponential model (DEM), fuzzy logic models, and ANN based model have already been studied in the research for the exact analysis of solar PV 1 , 4 . However, each of these models has their own merits and limitations.…”

“…Though, these methods are able to get acculturate characteristics but on the other hands faces challenges of the requirement of extensive experimental data for the validation of effectiveness and accuracy. Furthermore, low convergence rate, proper selection of initial value, large numbers of unknown variables, and use of assumptions & approximations are the other few major limitation of these methods 1 , 5 , 8 . Usually, the knowledge of slope, i.e ., at the short circuit region and at the open circuit region , are very crucial for the analytical and numerical methods 8 – 10 .…”

“…Furthermore, low convergence rate, proper selection of initial value, large numbers of unknown variables, and use of assumptions & approximations are the other few major limitation of these methods 1 , 5 , 8 . Usually, the knowledge of slope, i.e ., at the short circuit region and at the open circuit region , are very crucial for the analytical and numerical methods 8 – 10 . Here, is the shunt resistance at short circuit region and is the series resistance at open circuit voltage.…”

“…Moreover, the use of assumptions and approximation, in the analytical methods, makes characteristics very far from the real. For instant, in most of the work, based on analytical methods, the value of ‘ A ’ is taken in the range of 1–1.5, the value of ‘ R sh ’ is considered as infinity, the value of ‘ R s ’ is considered as zero, and I pv is taken as I sc 1 , 8 . Though, these assumptions reduce the complexity but on the other side improves the inaccuracy in real characterization 1 .…”