Multicores and GPU utilization in parallel swarm algorithm for parameter estimation of photovoltaic cell model

Ting, T. O.; Ma, Jieming; Kim, Kyeong Soo; Huang, Kaizhu

doi:10.1016/j.asoc.2015.10.054

Cited by 23 publications

(8 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The first group makes the use of available data in the manufacturer's datasheets to derive consistent parameter identification [17][18][19]. The second group uses an experimental I-V curve to utilise non-linear regressions and then uses non-linear programming algorithms to solve the non-linear optimisation problem so as to extract the model parameters [6,[20][21][22][23][24]. Rahman et al proposed a modelling technique, which can determine all the PV panel parameters without any explicit repetitive iteration, utilising only the quantities provided in the manufacturer's datasheet [25].…”

Section: Model-based Methods Use the Equivalent Circuit Of Pv Cell Tomentioning

confidence: 99%

“…The influence of large measurement errors on the results of estimation would be decreased and better parameter estimation would be achieved. Assuming measured variables contain output current, output voltage, irradiance, and temperature of the PV array, the robust parameter estimation for PV array model can be formulated based on the correntropy estimator: (see (24) σ I , σ V , σ S , σ T are the kernel widths of the measured variables; N I is the number of measurement data points on the output current, and N V is the number of measurement data points on the output voltage. Note that when irradiance and temperature of the PV array are not measured, the corresponding variables can be considered as the model parameters to be estimated.…”

Section: Fig 4 I-v Curve Of the Model And The Experimental Data Undementioning

confidence: 99%

See 1 more Smart Citation

Correntropy‐based parameter estimation for photovoltaic array model considering partial shading condition

Zhang

Chen

et al. 2019

IET Renewable Power Generation

View full text Add to dashboard Cite

Analytical modelling of photovoltaic (PV) array is crucial for studying the current-voltage (I-V) characteristic of PV array and maximum power point tracking. A PV array model generally contains some undetermined parameters and the values of the parameters cannot be measured by sensors. It is difficult to correctly determine those model parameters. They should be estimated based on experimental data. Since the experimental data gathered from the solar panel equipment usually contain random and gross errors, a robust parameter estimation method, correntropy-based parameter estimation (C-PE) is proposed for PV array model considering partial shading condition here. First, the theoretical model of PV array considering partial shading condition is investigated. Second, compared with the most common estimator, weighted least squares (WLS), robustness of the proposed correntropy estimator is analysed by using influence function (IF), and then C-PE method is developed for the PV array model. The WLS-based parameter estimation (WLS-PE) and C-PE methods are used in the simulation example. The results show that the C-PE method is more robust than WLS-PE method. Finally, the experimental data of PV array under ideal condition and partial shading condition are also used to demonstrate the feasibility and effectiveness of C-PE method.

show abstract

Section: Model-based Methods Use the Equivalent Circuit Of Pv Cell Tomentioning

confidence: 99%

Section: Fig 4 I-v Curve Of the Model And The Experimental Data Undementioning

confidence: 99%

Correntropy‐based parameter estimation for photovoltaic array model considering partial shading condition

Zhang

Chen

et al. 2019

IET Renewable Power Generation

View full text Add to dashboard Cite

show abstract

“…To decrease software experiment execution time, one can use faster hardware or optimize underlining algorithms. Some hardware options to decrease execution time include FPGA [49,50], GPU [51], faster processors [52] or computer clusters [53,54]. Algorithm optimization examples are found in studies by Gou et al [55], Naderi et al [56] and Sánchez-Oro et al [57].…”

Section: Computer Clustermentioning

confidence: 99%

Parallelism Strategies for Big Data Delayed Transfer Entropy Evaluation

2019

View full text Add to dashboard Cite

Generated and collected data have been rising with the popularization of technologies such as Internet of Things, social media, and smartphone, leading big data term creation. One class of big data hidden information is causality. Among the tools to infer causal relationships, there is Delay Transfer Entropy (DTE); however, it has a high demanding processing power. Many approaches were proposed to overcome DTE performance issues such as GPU and FPGA implementations. Our study compared different parallel strategies to calculate DTE from big data series using a heterogeneous Beowulf cluster. Task Parallelism was significantly faster in comparison to Data Parallelism. With big data trend in sight, these results may enable bigger datasets analysis or better statistical evidence.

show abstract

“…Since the solutions of (19) vary in a very tiny set, choosing any point for R s and V t in this tiny range has very small effect on the final I − V curve. Therefore, it can be almost assumed that the solutions of dynamics (19) converge to unique point (R s * , V t * ) belongs to the small tiny set. In the other words, if one considers the small tiny ball of reduced parameters as Θ r (solutions to optimisation problem (17)), then for any points…”

Section: Parameter Estimationmentioning

confidence: 99%

“…Many researchers have applied datasheet-based algorithms to different equivalent electrical circuits for PV module (cell) identification. These algorithms include analytical approaches [10][11][12], iterative methods [13][14][15][16], soft computing and artificial intelligence techniques [2,[17][18][19][20][21].…”

Section: Introductionmentioning

confidence: 99%

Constrained optimisation approach for parameter estimation of PV modules with single‐diode equivalent model

Moshksar

Ghanbari

2018

IET Renewable Power Generation

View full text Add to dashboard Cite

This study deals with estimation of five unknown parameters in single-diode equivalent model of photovoltaic (PV) modules. First, to simplify the problem, the unknown parameters are reduced to series resistance and diode thermal voltage. These two parameters have significant role for PV model identification. On the other hand, PV model has the least sensitivity to the choice of parallel resistance. Hence, an approximation is utilised for parallel resistance and large value is assigned to this electrical parameter. Thanks to the proposed approximation, a novel cost function is designed for the reduced model such that all of its optimum solutions remain in a small interval of the reduced model. A set of inequality constraints are defined to generate an almost convex optimisation problem with all solutions located in a very small set. The gradient update laws are developed to find the solutions in the tiny set forced by the constructed optimisation problem. The proposed estimation technique generates an accurate model for PV modules, especially at voltage values lower and equal to maximum voltage value.

show abstract

Multicores and GPU utilization in parallel swarm algorithm for parameter estimation of photovoltaic cell model

Cited by 23 publications

References 22 publications

Correntropy‐based parameter estimation for photovoltaic array model considering partial shading condition

Correntropy‐based parameter estimation for photovoltaic array model considering partial shading condition

Parallelism Strategies for Big Data Delayed Transfer Entropy Evaluation

Constrained optimisation approach for parameter estimation of PV modules with single‐diode equivalent model

Contact Info

Product

Resources

About