# A Novel Koopman-Inspired Method for the Secondary Control of Microgrids with Grid-Forming and Grid-Following Sources

**Authors:** Xun Gong, Xiaozhe Wang  
**arXiv:** [2301.01461](https://arxiv.org/abs/2301.01461)  
**Date:** 2023-01-04  
**Summary page:** https://www.summarizepaper.com/en/paper/2301.01461v1/

## Key points

- The paper presents a new online data-driven method for secondary voltage and frequency control in microgrids
- The method is based on Koopman-inspired modeling and does not require warm-up training
- The method guarantees bounded-input-bounded-output (BIBO) stability and asymptotic stability under mild conditions
- The method adapts to handle microgrid nonlinearity and uncertainty by estimating the Koopman state space model adaptively
- Case studies were conducted on 4-bus and 13-bus microgrid test systems subject to changes in operating conditions or large disturbances such as microgrid mode transitions, generation/load variations, measurement noises, and time delays.
- Results show that the proposed method is more accurate, has faster convergence speed, and is more computationally efficient than typical data-driven methods.
- This approach offers a practical solution for real-time secondary voltage and frequency control in microgrids without extensive offline modeling or tuning efforts which can lead to improved power quality, reliability, and resilience against disturbances.
- This paper contributes to the growing body of research on data-driven control methods for renewable energy systems.

## Summary

This paper presents a novel online data-driven Koopman-inspired method for secondary voltage and frequency control in microgrids with grid-forming and grid-following sources. The proposed method does not require any warm-up training and guarantees bounded-input-bounded-output (BIBO) stability as well as asymptotic stability under mild conditions. The method estimates the Koopman state space model adaptively to effectively handle microgrid nonlinearity and uncertainty. To demonstrate the effectiveness and robustness of their proposed identification and control method, the authors conducted case studies on 4-bus and 13-bus microgrid test systems subject to changes in operating conditions or large disturbances such as microgrid mode transitions, generation/load variations, measurement noises, and time delays. The results show that the proposed method is more accurate, has faster convergence speed, and is more computationally efficient than typical data-driven methods. This approach offers a practical solution for real-time secondary voltage and frequency control in microgrids without extensive offline modeling or tuning efforts which can lead to improved power quality, reliability, and resilience against disturbances. This paper contributes to the growing body of research on data-driven control methods for renewable energy systems.