Date of Award

7-20-2023

Document Type

Thesis

Degree Name

Master of Science (MS)

Department

Systems Engineering

First Advisor

Kamran Iqbal

Abstract

Research and development for alternative energy sources that are cleaner, renewable, and have little to no environmental impact have been pushed by the ongoing rise in energy demand, the possibility of a decline in the use of conventional petroleum fuels, and concerns about environmental degradation. Electricity from photovoltaic (PV) systems is significantly better regarded among these alternative sources as a renewable energy source such as wind power, bioenergy, tidal energy, and hydroelectric with a wide application range because it is clean, accessible, and abundant with little to no environmental effect. However, solar energy usage is significantly impacted by the landscape, location, seasonal changes, and other natural factors, such as the recurring issue of sunlight direction and irradiance changing over time. All these factors contribute to the PV system's low efficiency. Therefore, the solar panel must be operated at its maximum power point to boost the photovoltaic system's efficiency. However, the maximum power point is not stable due to the nonlinear nature of the PV cells and changes in the weather (temperature and irradiance). How do we optimize the solar panel system to work best always to produce the maximum power even when the temperature and irradiance fluctuate? The perturb and observe (P&O) algorithm is well-known and the most often used Maximum Power Point Tracking (MPPT) algorithm because of its simplicity and ease of implementation. However, in the P&O algorithm, the operational point oscillates near the maximum power point, resulting in an unending oscillation in output power. There is also a divergence from the maximum power point when weather conditions suddenly change, leading to energy loss and lower efficiency. In this research, the PV panel maximum power point was tracked with the aid of NASA weather data-based Neural Network, which is a type of computational model, and this would be based on actual weather conditions from NASA temperature and irradiance data of a specified location (Littlerock). The critical contribution of this research is to utilize NASA weather data-based Neural Network to forecast the optimized system reference voltage in all weather conditions and to use it as a reference for the PI controller to ensure that the DC/DC boost converter generates maximum power and a steady output voltage in all weather conditions to the inverter and the grid. The effectiveness of this MPPT was compared to the Perturb & Observe (P & O) algorithms for tracking the maximum power point of solar PV panels. Using MATLAB Simulink software, the simulation results verified the proposed work's performance under diverse weather conditions.

Download

Share

COinS