Development of Stand-Alone Photovoltaic System Test-Bed using Neural Network based Solar PV Array Emulator

IF 0.4 4区计算机科学 Q4 AUTOMATION & CONTROL SYSTEMS

Control Engineering and Applied Informatics Pub Date : 2023-09-26 DOI:10.61416/ceai.v25i3.8106

Ulaganathan M, Devaraj D, Muniraj R

引用次数: 0

Abstract

Research on solar power generation is gaining momentum in recent decade, which requires a costly and complex experimental setup. The Photo-Voltaic (PV) source emulator is a low cost and necessary equipment to evaluate the solar PV array performance, Maximum Power Point Tracking (MPPT) algorithm, power converters, and corresponding control algorithm. This paper proposes a novel Neural Network (NN)-based Solar Array Emulator (SAE) to emulate PV array dynamic characteristics under varying environmental conditions. The proposed SAE reference model has been developed using NN, which can replicate a PV array characteristics with a programmable DC power source’s support. A 640 W stand-alone PV system has been designed and tested using the proposed SAE to validate the performance of the developed prototype under various environmental conditions. The results demonstrate that the developed SAE has good accuracy in replicating the PV array characteristics than the conventional diodebased SAE. DOI: 10.61416/ceai.v25i3.8106

查看原文本刊更多论文

基于神经网络的太阳能光伏阵列仿真器单机光伏系统试验台的研制

近十年来，对太阳能发电的研究正在蓬勃发展，这需要昂贵而复杂的实验装置。光伏(PV)源模拟器是一种低成本的必要设备，用于评估太阳能光伏阵列性能、最大功率点跟踪(MPPT)算法、功率转换器以及相应的控制算法。本文提出了一种基于神经网络(NN)的太阳能阵列仿真器(SAE)，用于模拟不同环境条件下光伏阵列的动态特性。所提出的SAE参考模型是使用神经网络开发的，该模型可以在可编程直流电源的支持下复制光伏阵列的特性。一个640w的独立光伏系统已经被设计和测试，使用所提出的SAE来验证开发的原型在各种环境条件下的性能。结果表明，与传统的基于二极管的SAE相比，所开发的SAE在复制光伏阵列特性方面具有良好的准确性。DOI: 10.61416 / ceai.v25i3.8106

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Control Engineering and Applied Informatics 工程技术-

CiteScore

1.50

自引率

22.20%

发文量

审稿时长

6 months

期刊介绍： The Journal is promoting theoretical and practical results in a large research field of Control Engineering and Technical Informatics. It has been published since 1999 under the Romanian Society of Control Engineering and Technical Informatics coordination, in its quality of IFAC Romanian National Member Organization and it appears quarterly. Each issue has up to 12 papers from various areas such as control theory, computer engineering, and applied informatics. Basic topics included in our Journal since 1999 have been time-invariant control systems, including robustness, stability, time delay aspects; advanced control strategies, including adaptive, predictive, nonlinear, intelligent, multi-model techniques; intelligent control techniques such as fuzzy, neural, genetic algorithms, and expert systems; and discrete event and hybrid systems, networks and embedded systems. Application areas covered have been environmental engineering, power systems, biomedical engineering, industrial and mobile robotics, and manufacturing.