Improved Control of DC-DC Converter for PV Array and Battery Hybrid System Based on Sine Cosine Algorithm and RL-TD3 Agent

Based on the relatively widespread use of photovoltaic (PV) arrays as an energy production system, this article presents the main components of an energy production and storage system, where generation is done by a PV array and energy storage is done by a battery, based on a benchmark. The general o...

Full description

Saved in:
Bibliographic Details
Published in2023 3rd International Conference on Electrical, Computer, Communications and Mechatronics Engineering (ICECCME) pp. 1 - 6
Main Authors Nicola, Marcel, Nicola, Claudiu-Ionel
Format Conference Proceeding
LanguageEnglish
Published IEEE 19.07.2023
Subjects
Benchmark testing
DC-DC Converter
Machine learning algorithms
Mechatronics
Photovoltaic systems
PI control
Production systems
PV Array
Reinforcement learning
SCA Algorithm
Online AccessGet full text
DOI10.1109/ICECCME57830.2023.10253414

Cover

More Information
Summary:Based on the relatively widespread use of photovoltaic (PV) arrays as an energy production system, this article presents the main components of an energy production and storage system, where generation is done by a PV array and energy storage is done by a battery, based on a benchmark. The general objective is to maintain V DC voltage from DC circuit at a constant value, and the main performance indicators are response time, V DC voltage error ripple, and fractal dimension (DF) of the V DC voltage signal. The types of controllers used are classical Proportional Integral (PI), PI tuned using the Sine Cosine Algorithm (SCA), and a classical PI controller combined with a special machine learning like Reinforcement Learning namely Twin Delayed Deep Deterministic Policy Gradient (RL-TD3) type agent. The validation of the performance of the presented control systems are achieved in the Matlab/Simulink environment.
DOI:10.1109/ICECCME57830.2023.10253414