Technical gazette, Vol. 32 No. 5, 2025.
Original scientific paper
https://doi.org/10.17559/TV-20240909001979
A Novel Work-State Analysis Method for Modelling Photovoltaic Arrays Under Multi-Illumination Conditions
Jun-Hong Zhang
; School of Electrical and information Engineering, Beijing University of Civil Engineering and Architecture, Beijing, 100044, China
*
Ling-Yu Li
; School of Electrical and information Engineering, Beijing University of Civil Engineering and Architecture, Beijing, 100044, China
Chao-Yu Zhai
; College of Artificial Intelligence, China University of Petroleum (Beijing) Beijing, 102249, China
* Corresponding author.
Abstract
This paper presents a novel work-state analysis method for accurate modeling of photovoltaic (PV) arrays under complex illumination conditions. Based on a refined seven-parameter model, we develop a segmented function approach that effectively captures array behavior under various lighting environments. The proposed method divides the operating regions into multiple intervals based on illumination levels, enabling precise characterization of array output. Experimental validation using a KG200GT module demonstrates modeling accuracy within 1.5% error under diverse lighting conditions. Results show that the model successfully divides the working range of local maximum power points and predicts multiple maximum power points in series configurations and accurately captures the unified characteristics in parallel arrangements. Thus, the working characteristics of series parallel arrays were analysed. The model achieves 98.5% accuracy in predicting array performance under partial shading, offering significant improvements over traditional modeling approaches. This work provides a robust mathematical framework for optimizing PV array design and maximum power point tracking algorithms in real-world applications.
Keywords
multi-illumination environment; output characteristic; photovoltaic array; segmented function model; work-state analysis method
Hrčak ID:
335064
URI
Publication date:
30.8.2025.
Visits: 202 *