Triboelectric Nanogenerator Device Development, Analysis and Prediction for Energy Harvesting

Abstract

As the global need for renewable energy sources grows, research into clean and cost effective harvesting of energy devices has become vital. The purpose of this thesis is to explore the design, production, and optimizing of Triboelectric Nanogenerator (TENG) technologies using reusable waste materials. The goal is to limit waste accumulation and boost sustainable energy generation by reusing materials. The current thesis focuses on the use of reusable material aluminium foil as an electrode in a device. As a dielectric material, the gadget used a laboratory film known as Parafilm, which the researchers compared to other low-cost materials such as OHP, PET, and PMMA. created several Triboelectric Nanogenerator (TENG) architectures and observed that the output response improved significantly from 4 V to 200 V. This enabled the TENG to power tiny devices such as digital watches, calculators, thermometers, and even LEDs containing up to 85 LEDs. To validate the practical results, with the help ofsimulation platform called COM SOL tool, by considering various parameters for simulation, such as the hardware design materials, their thickness, and properties, keeping them consistent with the practical setup. The simulation results closely matched the practical findings, reinforcing the accu racy of their experiments. To further improve the prediction of output power for unknown load conditions, incorporated algorithms for to reduce the loss in the accuracy prediction used deep learning and machine learning algorithms. In Machine learning random forest regression and in deep learning Adamdelta is giving less loss compared to existing algo rithms. The outcome of this research lead to reduce the waste materials and helpful to the society in terms of air, water, and soil pollution. The proposed research offers the design of TENG device with less cost, validate the practical results with simulation results with scaling factor and for the complex designs to predict the output power for unknown load conditions using AI