Reinforcement Learning Based Real-Time Wide-Area Stabilizing Control Agents to Enhance Power System Stability

Hadidi, Ramtin; Jeyasurya, B.

doi:10.1109/tsg.2012.2235864

Cited by 63 publications

(36 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The use of RL to adjust the gains of adaptive decentralized backstepping controllers has been demonstrated in [11]. Wide-area stabilizing control, exploiting real-time measurements provided by WAMS, using RL has been introduced in [7].…”

Section: Model-based Vs Model-free Solution Methodsmentioning

confidence: 99%

“…While Q-learning based approaches have been proposed in previous works about oscillations damping [7]- [11], in the present work we propose to use a model-free tree-based batch mode RL algorithm to optimize supplementary inputs to existing damping controllers [14]. This choice is motivated by the following reasons [12]- [14]:…”

Section: Model-based Vs Model-free Solution Methodsmentioning

confidence: 99%

“…Different with other works using RL for oscillations damping [7]- [11] where Q-learning or some of its variants have been suggested, we use a model-free tree-based batch mode RL algorithm [12]- [14]. Furthermore, we propose to design a multi-agent system of heterogeneous non-communicating RLbased agents through separate sequential learning of individual agents [15], [16].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Trajectory-Based Supplementary Damping Control for Power System Electromechanical Oscillations

Wang

Glavić

Wehenkel

2014

IEEE Trans. Power Syst.

View full text Add to dashboard Cite

Abstract-This paper considers a trajectory-based approach to determine control signals superimposed to those of existing controllers so as to enhance the damping of electromechanical oscillations. This approach is framed as a discrete-time, multi-step optimization problem which can be solved by model-based and/or by learning-based methods. This paper proposes to apply a model-free tree-based batch mode Reinforcement Learning (RL) algorithm to perform such a supplementary damping control based only on information collected from observed trajectories of the power system. This RL-based supplementary damping control scheme is first implemented on a single generator and then several possibilities are investigated for extending it to multiple generators. Simulations are carried out on a 16-generators medium size power system model, where also possible benefits of combining this RL-based control with Model Predictive Control (MPC) are assessed.

show abstract

Section: Model-based Vs Model-free Solution Methodsmentioning

confidence: 99%

Section: Model-based Vs Model-free Solution Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Trajectory-Based Supplementary Damping Control for Power System Electromechanical Oscillations

Wang

Glavić

Wehenkel

2014

IEEE Trans. Power Syst.

View full text Add to dashboard Cite

show abstract

“…It shows the interaction of 1 or more agent with their environment to learn the goal‐oriented control laws. Optimal control policy means selecting the best action among all actions that exist for each state of the agent in the environment …”

Section: Mathematical Modelsmentioning

confidence: 99%

“…Stability of power systems is one of the most important aspects of the system operation, for example frequency and voltage should be maintained at their specific ranges in normal condition and during faults . Power systems can be considered as large and interconnected systems with high degree of non‐linearity . Interconnection of different parts of the power system together imposes various kind of oscillations to the power system .…”

Section: Introductionmentioning

confidence: 99%

Application of reinforcement learning for generating optimal control signal to the IPFC for damping of low-frequency oscillations

Younesi

Shayeghi

Moradzadeh

2017

Int Trans Electr Energ Syst

View full text Add to dashboard Cite

Summary In this paper, an adaptive design of interline power flow controller (IPFC) using a reinforcement learning (RL) approach is utilized for damping of the low‐frequency oscillations (LFOs) in power system. This Q‐learning‐based adaptive damping controller is applied to a single‐machine and multi‐machine power system. The main advantages of the Q‐learning based controller are its robustness and adaptive behavior to change in the operation condition; also, it does not need any knowledge about the control system, making the control strategy suitable for the realistic power systems with high nonlinearities. In order to demonstrate the performance of the proposed RL‐based damping controller in both single‐machine and multi‐machine power system, nonlinear simulations are carried out using MATLAB/Simulink. Three cases of simulations are performed; the system equipped with (1) only optimal classical power system stabilizer (PSS), (2) coordinated designed of PSS and IPFC, and (3) RL‐based optimized IPFC. Krill herd optimization algorithm is used for coordination design of PSS and IPFC in both single machine and multi‐machine power systems. Suitable time‐domain performance indices in multiple operating conditions and different fault types are calculated for all 3 cases of simulations and are compared with one another. Simulation results demonstrate the effectiveness of the proposed control strategy in damping of the LFOs in the power systems.

show abstract

Adaptive and Online Control of Microgrids Using Multi-agent Reinforcement Learning

Shayeghi

Younesi

2019

Power Systems

View full text Add to dashboard Cite

Reinforcement Learning Based Real-Time Wide-Area Stabilizing Control Agents to Enhance Power System Stability

Cited by 63 publications

References 26 publications

Trajectory-Based Supplementary Damping Control for Power System Electromechanical Oscillations

Trajectory-Based Supplementary Damping Control for Power System Electromechanical Oscillations

Application of reinforcement learning for generating optimal control signal to the IPFC for damping of low-frequency oscillations

Adaptive and Online Control of Microgrids Using Multi-agent Reinforcement Learning

Contact Info

Product

Resources

About