Optimization Of Solar Energy Efficiency Using Neural Network Controllers With Direct Current Converters

Author

Kasim Ali Mohammad, Prairie View A&M University

Date of Award

8-2024

Document Type

Thesis

Degree Name

Doctor of Philosophy (PhD)

Degree Discipline

Electrical Engineering

Abstract

Photovoltaic (PV) systems offer a renewable energy source by converting sunlight into electricity. This dissertation enhances PV system performance using advanced control mechanisms and DC-DC converters. PV systems, influenced by variables such as irradiance and temperature, require efficient Maximum Power Point Tracking (MPPT) for optimal energy conversion. Traditional MPPT methods like the Perturbation and Observation (P&O) algorithm often struggle with rapidly changing conditions. While Artificial Neural Networks (ANN) and Recurrent Neural Networks (RNN) have been explored for control purposes, this research applies Positive Output Super Lift Luo (P/O SLL), and Ultra Lift Luo (ULL) converters integrated with ANN and RNN controllers. These converters achieve higher output voltage increase compared to traditional converters like Boost, Cuk, and SEPIC by employing a super-lift technique. The study aims to design and implement these applications with AI-based controllers using MATLAB/Simulink. Research objectives include boosting DC voltage, minimizing power loss, optimizing voltage for varying conditions, and achieving optimized matching resistance. The optimization lies not only in integrating P/O SLL and ULL converters with ANN and RNN controllers but also in developing control algorithms that enhance adaptability and efficiency. Unlike existing approaches, the proposed system offers higher output voltage increase, reduced component stress, and improved efficiency. The AI controllers enable real-time adaptability to changing conditions, ensuring maximum energy extraction and enhanced MPPT performance. This research addresses static and dynamic learning rates in AI controllers, with static learning rates for steady-state conditions and dynamic learning rates for rapid adaptation. This dissertation provides a comprehensive solution for optimizing PV system performance, contributing to more efficient renewable energy systems and advancing PV system technology to support the global transition to sustainable energy.

Index Terms—Artificial neural network, dynamic learning rate, dc-dc converters, maximum power point tracking, photovoltaic system, recurrent neural network, static learning rate.

Committee Chair/Advisor

Sarhan M. Musa

Committee Member

Kelvin Kirby

Committee Member

Pamela H. Obiomon

Committee Member

Penrose Cofie

Publisher

Prairie View A&M University

Rights

Date of Digitization

9/13/2024

Contributing Institution

John B Coleman Library

City of Publication

Prairie View

MIME Type

Application/PDF

Recommended Citation

Mohammad, K. A. (2024). Optimization Of Solar Energy Efficiency Using Neural Network Controllers With Direct Current Converters. Retrieved from https://digitalcommons.pvamu.edu/pvamu-dissertations/53