Developing an interactive graphical user interface for exploring near-optimal solutions in energy system models
Loading...
URL
Journal Title
Journal ISSN
Volume Title
School of Electrical Engineering |
Master's thesis
Authors
Date
Department
Major/Subject
Mcode
Language
en
Pages
71
Series
Abstract
The following thesis contributes to the development of an interactive dashboard used to explore near-optimal solutions in energy system models. The aim of the work lies in the improvement of decision-making processes through interaction-based exploration of the trade-offs between different sources of energy in a fixed feasible region. The dashboard uses the data produced by the Compact Energy System Model (CESM), which yields a large and diverse set of precomputed feasible system configurations meeting strict cost and emission constraints. The methodological basis of the dashboard lies in the Modeling to Generate Alternatives (MGA) method. This method is applied in the backend logic of the dashboard using linear programming to find the feasible endpoints within the convex hull of CESM outputs. Users define priority orders and directional constraints for energy sources, after which the dashboard solves for new feasible configurations and provides interpolation tools to navigate between them. This allows continuous and flexible exploration of the feasible energy pathways while maintaining transparency and feasibility guarantees at each step. This enables exploration of the feasible energy pathways while ensuring transparency and feasibility guarantees at each step. The implementation of the dashboard is available in Python, and the graphical interface is based on Streamlit. The Pandas and NumPy libraries are used to work with the data, Gurobi as an optimizer, and the Plotly library as a dynamic visualizer. It is helpful for energy analysts and stakeholders to have a better understanding of the flexibility and robustness of the design of future energy systems by providing them with intuitive feedback in real time according to the effects of the user-specified decisions. This study fills the gap between static MGA modeling exercises and interactive, user-driven tools, hence providing a more practical and useful decision-support system to discuss complex energy systems under uncertainty. As a demonstration, the dashboard is used on a German case study focusing on expanding PV and reducing coal, showing how the tool fits existing policy objectives.Description
Supervisor
Lehtonen, MattiThesis advisor
Steinke, Prof. Dr. FlorianHajikazemi, Sina