A Novel ANN-Based GMPPT Method for PV Systems Under Complex Partial Shading Conditions

In this article, a new two-stage global maximum power point tracking (GMPPT) algorithm based on artificial neural network (ANN) is proposed. The novel ANN architecture is presented first, which requires fewer sampling points than other ANN-based GMPPT approaches, thereby reducing both the tracking t...

Full description

Saved in:
Bibliographic Details
Published in:IEEE transactions on sustainable energy 2024-01, Vol.15 (1), p.328-338
Main Authors: Ye, Song-Pei, Liu, Yi-Hua, Pai, Hung-Yu, Sangwongwanich, Ariya, Blaabjerg, Frede
Format: Article
Language:English
Subjects:
Accuracy
Algorithms
Artificial neural network (ANN)
Artificial neural networks
Computer simulation
global maximum power point tracking (GMPPT)
Illuminance
Maximum power point trackers
Maximum power tracking
Neural networks
Performance evaluation
Performance indices
photovoltaic (PV)
Photovoltaic systems
Shading
Solar panels
Temperature requirements
Temperature sensors
Voltage control
Voltage measurement
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:In this article, a new two-stage global maximum power point tracking (GMPPT) algorithm based on artificial neural network (ANN) is proposed. The novel ANN architecture is presented first, which requires fewer sampling points than other ANN-based GMPPT approaches, thereby reducing both the tracking time and power loss involved in the tracking. In addition, it does not require costly illuminance or temperature sensors and can be realized using a low-cost digital signal controller. According to the simulation results, the proposed method has the best performance among all methods in terms of tracking speed, tracking accuracy, and tracking loss for all the three tested shading patterns (SPs). The simulated results of 252 SPs show that the performance indexes (PIs) of the proposed method are the best among all the compared methods, which are: the average tracking time 0.18 seconds, average power loss 0.01 W. In addition, the proposed method can correctly predict the GMPP positions of all 252 SPs. Furthermore, the PIs of the proposed method are also the best among all the compared methods according to the experimental results, which are: the tracking speed 0.21 seconds, tracking accuracy 99.66%, and tracking loss 12.58 W, all the above are average values.
ISSN:1949-3029
1949-3037
DOI:10.1109/TSTE.2023.3284866