Tehnički vjesnik, Vol. 32 No. 1, 2025.
Izvorni znanstveni članak
https://doi.org/10.17559/TV-20240108001253
Self-adaptive Teaching-Learning-Based Optimization with Reusing Successful Learning Experience for Parameter Extraction in Photovoltaic Models
Yang Du
; School of Automotive and Traffic Engineering, Hubei Key Laboratory of Power System Design and Test for Electrical Vehicle, Hubei University of Arts and Science, No. 296, Longzhong Road, Xiangyang, Hubei, 441053, China
Bin Ning
orcid.org/0000-0003-4822-716X
; School of Computer Engineering, Hubei Key Laboratory of Power System Design and Test for Electrical Vehicle, Hubei University of Arts and Science, No. 296, Longzhong Road, Xiangyang, Hubei, 441053, China
*
Xiaowang Hu
; School of Automotive and Traffic Engineering, Hubei Key Laboratory of Power System Design and Test for Electrical Vehicle, Hubei University of Arts and Science, No. 296, Longzhong Road, Xiangyang, Hubei, 441053, China
Bojun Cai
; School of Computer Engineering, Hubei University of Arts and Science, No. 296, Longzhong Road, Xiangyang, Hubei, 441053, China
* Dopisni autor.
Sažetak
This paper proposes a self-adaptive teaching-learning-based optimization with reusing successful learning experience (RSTLBO) to accurately and reliably extract parameters of different photovoltaic (PV) models. The key novelties of RSTLBO are: 1) Learners adaptively choose teacher or learner phase based on a selection probability according to their performance, balancing exploration and exploitation; 2) Successful learner experiences are reused to enhance search capability. Experiments on single diode, double diode and PV panel models demonstrate that RSTLBO achieves higher accuracy and faster convergence than state-of-the-art methods like P-DE, TLBO, GOTLBO, etc. Specifically, RSTLBO obtains the minimum RMSE across all models, outperforms compared methods in statistical results, and exhibits fastest convergence in almost all cases. The self-adaptive probability selection and experience reuse make RSTLBO effective for PV parameter extraction.
Ključne riječi
learning experience; optimization; parameter extraction; photovoltaic model; teaching-learning-based optimization
Hrčak ID:
325995
URI
Datum izdavanja:
31.12.2024.
Posjeta: 11 *