Year: 2025
Type: conference-paper
Abstract: The increasing adoption of distributed Solar Photovoltaic (PV) systems presents a significant opportunity to decarbonize residential buildings and enhance energy resilience.However, determining the optimal size of a PV system, when coupled with battery storage, remains a complex challenge influenced by various factors including energy consumption, solar irradiance and the desired level of selfsufficiency.While current approaches rely on detailed simulations to estimate optimal system sizes, these methods are computationally intensive and may require several hours to complete.Instead, we propose a neural network-based approach as a fast and accurate alternative, capable of generating sizing recommendations within milliseconds, approximately 40,000 times faster than the baseline.By significantly reducing computation time, this approach aims to support consumers, planners, and policymakers in maximizing the economic and environmental benefits of PV deployment.
Authors: Julia Olivia Gschwind, Srinivasan Keshav
Institutions: ETH Zurich, University of Cambridge
Language: English
Topic: Smart Grid Energy Management
Subfield: Electrical and Electronic Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: gold
OpenAlex ID: W4411346325