During my internship at the Tsinghua Sichuan Energy Internet Research Institute from August to December 2023, I had the opportunity to contribute to a highly technical project focused on power system analysis, specifically examining the impact of large-scale renewable energy integration on system inertia and frequency. This experience greatly enhanced my algorithm development, simulation, and technical communication skills. It provided a deeper understanding of the intersection between AI and energy systems.
Key Achievements
-
Algorithm Development:
- I developed and optimized a SFR power system frequency response model, using a parameter identification method based on Graph Neural Networks (GNNs). The model was designed to analyze the impact of large-scale renewable energy access on power system inertia and frequency, which are critical for maintaining grid stability. This required integrating advanced concepts from power systems and machine learning, providing a unique interdisciplinary challenge.
-
Algorithm Implementation:
- The algorithms were implemented using PyTorch, and the results were verified through PSD-BPA simulation. The models we established were directly applied to real-world industrial power grid scenarios, allowing me to see the direct impact of our work in a practical context. This step significantly improved my ability to develop technically sound and practically relevant algorithms.
-
Collaboration and Communication:
- I actively participated in milestone reporting for our partner companies throughout my internship. This involved preparing technical reports, feasibility study reports, technical declaration forms, and milestone progress presentations. I also contributed to writing and visualizing project-related papers, which allowed me to refine my technical writing and communication skills in a highly professional and rigorous environment.
What I Learned
-
Interdisciplinary Application of Machine Learning:
- One of the most valuable lessons I learned during this internship was how to apply machine learning techniques, specifically GNNs, to solve real-world engineering problems in power systems. This experience reinforced my understanding of how AI methods can be used to address the unique challenges of renewable energy integration and grid stability.
-
Advanced Simulation and Verification Techniques:
- Working with PSD-BPA simulations to verify our models exposed me to advanced simulation tools and techniques commonly used in power system analysis. This experience significantly broadened my technical skillset, particularly in the context of energy systems.
-
Technical Writing and Reporting:
- Writing technical reports and contributing to project papers helped me develop a more structured and concise writing style, essential for communicating complex technical concepts. It also taught me the importance of clarity and precision when presenting research findings to stakeholders, a critical skill when working with industry partners.
Conclusion
This internship experience allowed me to develop and apply advanced algorithms to a real-world problem in the energy sector while enhancing my ability to communicate technical results effectively. I gained a deeper understanding of how machine learning can be applied to power systems. I also learned to navigate the complexities of working in a highly technical and interdisciplinary environment.
As I continue my academic journey, I am excited to apply the skills and knowledge I gained during this internship to future AI and energy systems research. This experience has further solidified my passion for exploring the intersection of these fields and finding innovative solutions to the challenges posed by the transition to renewable energy.