Lyapunov-Regularized Reinforcement Learning for Power System Transient Stability

Cui, Wenqi; Zhang, Baosen

doi:10.1109/lcsys.2021.3088068

Cited by 22 publications

(9 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For example, [95] proposed Lyapunov-regularized reinforcement learning for power system transient stability. In addition, physics-constrained and physics-informed deep learning [96] is also under development and can be integrated into MFRL to address security concerns.…”

Section: Online Implementationmentioning

confidence: 99%

Fusion of Model-free Reinforcement Learning with Microgrid Control: Review and Vision

She¹,

Li²,

Cui³

et al. 2022

Preprint

View full text Add to dashboard Cite

Challenges and opportunities coexist in microgrids as a result of emerging large-scale distributed energy resources (DERs) and advanced control techniques. In this paper, a comprehensive review of microgrid control is presented with its fusion of model-free reinforcement learning (MFRL). A highlevel research map of microgrid control is developed from six distinct perspectives, followed by bottom-level modularized control blocks illustrating the configurations of grid-following (GFL) and grid-forming (GFM) inverters. Then, mainstream MFRL algorithms are introduced with an explanation of how MFRL can be integrated into the existing control framework. Next, the application guideline of MFRL is summarized with a discussion of three fusing approaches, i.e., model identification and parameter tuning, supplementary signal generation, and controller substitution, with the existing control framework. Finally, the fundamental challenges associated with adopting MFRL in microgrid control and corresponding insights for addressing these concerns are fully discussed.

show abstract

Section: Online Implementationmentioning

confidence: 99%

Fusion of Model-free Reinforcement Learning with Microgrid Control: Review and Vision

She¹,

Li²,

Cui³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…This relaxation can result in very conservative assessments of stability and does not yet scale to moderate or large systems. More recently, attempts have been made to learn a Lyapunov function parameterized by neural networks [12], [14]. However, it is challenging to verify that the learned neural networks are actually Lyapunov functions.…”

Section: Introductionmentioning

confidence: 99%

“…Hence, it is essential to characterize stability for a set of possible equilibria. In addition, the power electronics on the DERs allow their damping coefficients to be adjusted [14], [16]. But optimizing these coefficients using existing approaches are nontrivial, since they involve solving complicated nonconvex problems.…”

Section: Introductionmentioning

confidence: 99%

Equilibrium-Independent Stability Analysis for Distribution Systems with Lossy Transmission Lines

Cui¹,

Zhang²

2022

Preprint

Self Cite

View full text Add to dashboard Cite

Power distribution systems are becoming much more active with increased penetration of distributed energy resources. Because of the intermittent nature of these resources, the stability of distribution systems under large disturbances and time-varying conditions is becoming a key issue in practical operations. Because the transmission lines in distribution systems are lossy, standard approaches in power system stability analysis do not readily apply and the understanding of transient stability remains open even for simplified models.This paper proposes a novel equilibrium-independent transient stability analysis of distribution systems with lossy lines. We certify network-level stability by breaking the network into subsystems, and by looking at the equilibrium-independent passivity of each subsystem, the network stability is certified through a diagonal stability property of the interconnection matrix. This allows the analysis scale to large networked systems with time-varying equilibria. The proposed method gracefully extrapolates between lossless and lossy systems, and provides a simple yet effective approach to optimize control efforts with guaranteed stability regions. Case studies verify that the proposed method is much less conservative than existing approaches and also scales to large systems.

show abstract

“…This risk measures the violations of the Lyapunov conditions. This NN-based approach can assess the transient stability of power systems in [23], [24], [25], however, this approach was never used to consider the system's transient response within the ACOPF.…”

Section: Introductionmentioning

confidence: 99%

Transient Stable Corrective Control Using Neural Lyapunov Learning

Bellizio

Cremer

Štrbac

2022

IEEE Trans. Power Syst.

View full text Add to dashboard Cite

This paper proposes a method to compute corrective control actions for dynamic security in real-time and quantifies the economic value of corrective control. Lowered inertia requires fast control methods in real-time to correct system operation and maintain system security when equipment fails. However, using corrective control beyond such emergency failure measures does not make fully use of them. The key contribution of this work is the optimal use of corrective control applications in combination with preventive strategies to enhance the network utilisation, reduce the normal operating costs while maintaining adequate security levels. The proposed approach learns a neural network for safety certificates and models the predicted safe dynamic post-fault state as algebraic constraints in an AC optimal power flow (OPF) deciding close to real-time on the optimal corrective control. Considering these safety constraints within the ACOPF can balance simultaneously the system transient stability with the costs for preventive and corrective control. This proposed approach outperforms sub-optimal approaches aiming at sequentially finding the balance. Case studies were based on the IEEE 9-bus system with integrated electrical vehicles and shares of wind power up-to 40% and on the IEEE 39-bus and 118-bus systems. The proposed approach outperforms baseline control approaches in stability, economics, and carbon emissions. One baseline approach was preventive wind curtailment, against which the proposed approach reduced operating costs by up-to 60%, decreased unstable operations by 50% and reduced carbon emissions by 60% in the IEEE 9-bus. In the IEEE 39-bus and 118-bus systems, the approach was promising for larger systems.

show abstract

Lyapunov-Regularized Reinforcement Learning for Power System Transient Stability

Cited by 22 publications

References 14 publications

Fusion of Model-free Reinforcement Learning with Microgrid Control: Review and Vision

Fusion of Model-free Reinforcement Learning with Microgrid Control: Review and Vision

Equilibrium-Independent Stability Analysis for Distribution Systems with Lossy Transmission Lines

Transient Stable Corrective Control Using Neural Lyapunov Learning

Contact Info

Product

Resources

About