We lately propose PE-GPT (2024 Sep), the first multimodal large language model (LLM) specifically designed for power electronics (PE) design. What impedes the adoption of LLMs in the industry of power electronics mainly includes their limited technical expertise and incapability to process PE-specific data. To address these challenges, PE-GPT is enhanced by retrieval augmented generation (RAG) with customized PE knowledge base, which empowers PE-GPT with PE-specific knowledge and reasoning, while model zoo and simulation repository equip it with the capability of processing PE-specific multi-modal data. Moreover, we propose a hybrid framework that integrates an LLM agent, metaheuristic algorithms, a Model Zoo, and a Simulation Repository.
Keeping "AI-for-Good" as our mission, PE-GPT strives to revolutionize the paradigm for diverse power electronics design tasks.
Fig. 1. The hybrid framework of PE-GPT.
Demo videos of using PE-GPT for the power electronics design tasks.
Demo Case 1:
- Modulation Optimization for DAB Converters - 1
Design.case.-.DAB1.mp4
Demo Case 2:
- Modulation Optimization for DAB Converters - 2
Design.case.-.DAB2.mp4
Demo Case 3:
- Circuit Parameter Design for Buck Converters
Design.case.-.buck.mp4
- To deploy PE-GPT on your PC, the first step is to setup your API call to OpenAI models, please see core/llm/llm.py for more details.
- If you want to interact with Plecs software to simulate the designed modulation for DAB, you need to enable the xml-rpc interface in Plecs settings,
and to add the directory "core/simulation/devices" in the device library searching path in plecs.
# clone the github repository
git clone https://github.com/XinzeLee/PE-GPT
# change the current working directory
cd PE-GPT
# install all required dependencies
pip install -r requirements_specific.txt
# run the GUI and chat with PE-GPT
streamlit run main.py
@reference: Fanfan Lin, Xinze Li, Weihao Lei, Juan J. Rodriguez-Andina, Josep M. Guerrero, Changyun Wen, Xin Zhang, and Hao Ma, "PE-GPT: a New Paradigm for Power Electronics Design", IEEE Transactions on Industrial Electronics.
@code-author: Xinze Li (email: [email protected]), Fanfan Lin (email: [email protected])
- This repository provides a simplified version of the PE-GPT methodology presented in our journal paper. Despite the simplifications, the released code preserves the overall core architecture of the proposed PE-GPT.
- This repository currently includes the following functions/blocks: Retrieval augmented generation, LLM agents, Model Zoo (with a physics-in-architecture neural network, PANN, for modeling DAB converters), metaheuristic algorithm for optimization, simulation verification, graphical user interface, and knowledge base. Please note that the current knowledge base is a simplified version for illustration.
This code is licensed under the Apache License Version 2.0.