Large language models (LLMs) have proven the ability to assist diverse users in conducting a variety of individual tasks via intuitive and natural conversations. This thesis discusses a utilization of LLMs as a tool for informed decision-making in energy investments and operations. One major goal is to transform the way the consumers engage with home energy management systems and reduce expertise-related dependencies.
Thesis Type |
|
Status |
Open |
Supervisor(s) |
Stefan Decker |
Advisor(s) |
Beyza Cizmeci |
Contact |
beyza.cizmeci@fit.fraunhofer.de |
As Europe pushes towards significant CO2 emission reductions, the Prosumer GPT, a project funded by Fraunhofer FIT, project addresses the urgent need for consumers to become proactive energy prosumers. In the scope of this project, we plan to explore the development of an LLM that acts as a virtual consultant for prosumers, guiding them through the complexities of managing photovoltaic systems, heat pumps, and storage systems. The project emphasizes user-friendly interfaces, backed by a Python-based engine, to simplify interactions with home energy management systems (HEMS).
Objectives:
– Identifying the potential of LLMs to serve as an interactive user interface for energy management decision-making
– Identifying the potential of a combined tool based on ad hoc trained LLM combined with a Python based engine exploiting the available knowledge on building modelling
– Identifying the key factors for successful development and implementation of an LLM-powered chatbot integrated with a Python-based engine for enhanced home energy management decision support.
– Implementing and evaluating the developed platform
Tasks:
– Critical literature review on the use of LLMs btw end users and expert systems and combination of LLMs with rule-based engine
– Design and implement a user-friendly LLM interface combined with a Python-based engine for HEMS
– Identifying gaps and shortcomings of the first implementation
– Conduct testing and validation to ensure the tool’s effectiveness in predefined scenarios
– Critical assessment and evaluation of the results
Contact:
Interested to work on this topic for your thesis? Write an email to beyza.cizmeci@fit.fraunhofer.de or foroogh.sedighi@eonerc.rwth-aachen.de.
– Master student in computer science, electrical engineering, or related fields
– Interest in inter-disciplinary research topics
– Good knowledge in Python and common machine learning libraries (e.g., NumPy, pandas, scikit-learn, etc.)
– Practical experience in machine learning is a plus
– Basic knowledge of power systems is also a plus