Physics-Informed Neural Networks for Non-linear System Identification applied to Power System Dynamics

Varying power-infeed from converter-based generation units introduces great uncertainty on system parameters such as inertia and damping. As a consequence, system operators face increasing challenges in performing dynamic security assessment and taking real-time control actions. Exploiting the widespread deployment of phasor measurement units (PMUs) and aiming at developing a fast dynamic state and parameter estimation tool, this paper investigates the performance of Physics-Informed Neural Networks (PINN) for discovering the frequency dynamics of future power systems and monitoring the system inertia in real-time. PINNs have the potential to address challenges such as the stronger non-linearities of low-inertia systems, increased measurement noise, and limited availability of data. The estimator is demonstrated in several test cases using a 4-bus system, and compared with state of the art algorithms, such as the Unscented Kalman Filter (UKF), to assess its performance.

[1]  Ali Abur,et al.  Power System Dynamic State Estimation: Motivations, Definitions, Methodologies, and Future Work , 2019, IEEE Transactions on Power Systems.

[2]  P. Kundur,et al.  Definition and classification of power system stability IEEE/CIGRE joint task force on stability terms and definitions , 2004, IEEE Transactions on Power Systems.

[3]  Carson W. Taylor,et al.  Definition and Classification of Power System Stability , 2004 .

[4]  Gang Wang,et al.  Real-Time Power System State Estimation and Forecasting via Deep Unrolled Neural Networks , 2018, IEEE Transactions on Signal Processing.

[5]  F. J. Soares,et al.  State estimation in distribution smart grids using autoencoders , 2014, 2014 IEEE 8th International Power Engineering and Optimization Conference (PEOCO2014).

[6]  Spyros Chatzivasileiadis,et al.  Physics-Informed Neural Networks for Power Systems , 2020, 2020 IEEE Power & Energy Society General Meeting (PESGM).

[7]  Tilman Weckesser,et al.  Robust Frequency Control for Varying Inertia Power Systems , 2018, 2018 IEEE PES Innovative Smart Grid Technologies Conference Europe (ISGT-Europe).

[8]  Abhinav Singh,et al.  Estimating dynamic model parameters for adaptive protection and control in power system , 2015, 2015 IEEE Power & Energy Society General Meeting.

[9]  Thanh Long Vu,et al.  Remedial actions to enhance stability of low-inertia systems , 2016, 2016 IEEE Power and Energy Society General Meeting (PESGM).

[10]  Lang Tong,et al.  State Estimation for Unobservable Distribution Systems via Deep Neural Networks , 2018, 2018 IEEE Power & Energy Society General Meeting (PESGM).

[11]  Paris Perdikaris,et al.  Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations , 2019, J. Comput. Phys..

[12]  Jimmy Ba,et al.  Adam: A Method for Stochastic Optimization , 2014, ICLR.

[13]  Yoshua Bengio,et al.  Understanding the difficulty of training deep feedforward neural networks , 2010, AISTATS.

[14]  Gabriela Hug,et al.  Foundations and Challenges of Low-Inertia Systems (Invited Paper) , 2018, 2018 Power Systems Computation Conference (PSCC).

[15]  Alberto Del Angel,et al.  Estimation of rotor angles of synchronous machines using artificial neural networks and local PMU-based quantities , 2007, Neurocomputing.