Comprehensive Evaluation of Machine Learning MPPT Algorithms for a PV System Under Different Weather Conditions
- Authors
- Nkambule, M.S.; Hasan, A.N.; Ali, A.; Hong, J.; Geem, Z.W.
- Issue Date
- Jan-2021
- Publisher
- SPRINGER SINGAPORE PTE LTD
- Keywords
- DC–DC boost converter; Machine learning (ML); Maximum power point tracking (MPPT); Partial shading conditions (PSC)
- Citation
- JOURNAL OF ELECTRICAL ENGINEERING & TECHNOLOGY, v.16, no.1, pp.411 - 427
- Journal Title
- JOURNAL OF ELECTRICAL ENGINEERING & TECHNOLOGY
- Volume
- 16
- Number
- 1
- Start Page
- 411
- End Page
- 427
- URI
- https://scholarworks.bwise.kr/gachon/handle/2020.sw.gachon/79669
- DOI
- 10.1007/s42835-020-00598-0
- ISSN
- 1975-0102
- Abstract
- The rapid growth of demand for electrical energy and the depletion of fossil fuels opened the door for renewable energy; with solar energy being one of the most popular sources, as it is considered pollution free, freely available and requires minimal maintenance. This paper investigates the feasibility of using machine learning (ML) based MPPT techniques, to harness maximum power on a PV system under PSC. In this study, certain contributions to the field of PV systems and ML based systems were made by introducing nine (9) ML based MPPT techniques, by presenting three (3) experiments under different weather conditions. Decision Tree (DT), Multivariate Linear Regression (MLR), Gaussian Process Regression (GPR), Weighted K-Nearest Neighbors (WK-NN), Linear Discriminant Analysis (LDA), Bagged Tree (BT), Naïve Bayes classifier (NBC), Support Vector Machine (SVM) and Recurrent Neural Network (RNN) performances are validated and proved using MATLAB SIMULINK simulation software. The experimental results demonstrated that WK-NN performs significantly better when compared with other proposed ML based algorithms. © 2020, The Korean Institute of Electrical Engineers.
- Files in This Item
- There are no files associated with this item.
- Appears in
Collections - IT융합대학 > 에너지IT학과 > 1. Journal Articles
Items in ScholarWorks are protected by copyright, with all rights reserved, unless otherwise indicated.