Data-Driven Power Flow Calculation Method: A Lifting Dimension Linear Regression Approach

Guo, Li; Zhang, Yuxuan; Li, Xialin; Wang, Zhongguan; Liu, Yixin; Bai, Linquan; Wang, Chengshan

doi:10.1109/tpwrs.2021.3112461

Cited by 29 publications

(16 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…• Constant Kernel Structure: The proposed VDK design exploits the graph structure and remains unchanged unless the graph-structure or admittance matrix itself changes 5 . This eliminates the need to solve any optimization problem for obtaining kernel structure.…”

Section: A Proposed Kernel Designmentioning

confidence: 99%

“…Although, this relative impact is not linear or directly quantifiable, it suggests that a network-swipe structure of blockdescend optimization can help in solving the high-dimensional AL problem Eq. ( 6) with the proposed VDK (5).The proposed network-swipe algorithm starts by solving Eq. ( 6) with respect to the x j , while keeping all other injections fixed, to learn V j (•).…”

Section: Active Learning Mechanismmentioning

confidence: 99%

“…The research work conducted at Los Alamos National Laboratory is done under the auspices of the National Nuclear Security Administration of the U.S. Department of Energy under Contract No. 89233218CNA000001 to capture the full non-linearity of power flow, and lifted approaches [5] tend to lose interpretability and require high sample complexity. Recently Deep Neural Networks (DNNs) have made significant advancements as universal function approximators with increased expressiveness, particularly via the idea of physics-informed machine learning [6].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

A Framework for Analytical Power Flow Solution Using Gaussian Process Learning

Pareek

Nguyen

2022

IEEE Trans. Sustain. Energy

View full text Add to dashboard Cite

This paper proposes a novel analytical solution framework for power flow (PF) solutions in active distribution networks under uncertainty. We use the Gaussian process (GP) regression to learn node voltage as a function of effective bus load or negative net-injection vector. The proposed approximation is valid over a subspace of load and provides an understanding of system behavior under uncertainty via GP interpretability. We interpret the relative variation extent of different node voltages using the quality ratio (QR) defined based on the hyper-parameters of GP. Further, the application of the proposed framework in calculation of voltage limit violation probability and dominant voltage influencer ranking has also been presented. Through test simulations for 33-bus and 56-bus systems, the proposed method achieves low mean absolute error (MAE) of order E-05 (pu) in voltage magnitude and E-04 (rad) in angle. The discussions on salient features of the proposed method and comparative analysis with large-scale Monte-Carlo simulations, and state-ofart methods is also presented for the proposed applications.

show abstract

Section: A Proposed Kernel Designmentioning

confidence: 99%

Section: Active Learning Mechanismmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A Framework for Analytical Power Flow Solution Using Gaussian Process Learning

Pareek

Nguyen

2022

IEEE Trans. Sustain. Energy

View full text Add to dashboard Cite

show abstract

“…Traditional power flow calculation methods [5][6][7][8] obtain voltage magnitude and phase angle by constructing differential equations [9,10] and solving them with iterative numerical methods; [11] considers DC, establishes a unified steady-state nonideal model, and proposes a power flow calculation model that retains the nonlinear algorithm; [12] used the decoupled linear power flow (DLPF) method to effectively calculate the voltage magnitude, but sacrificed the accuracy of the voltage phase angle calculation; [13] proposes a combination of Newton-Raphson-Seder and downslope algorithms with a combination of static and transient stability; [14] considers integrated energy sources and constructs a power flow calculation model based on a refined thermal network. With the development of measurement devices and storage equipment, the data-driven [15] power flow method was developed, and the nonlinear power flow model was linearized [16][17][18][19] to speed up the power flow calculation. [20] a linear regression model is established by constructing the mapping relationship between the active, reactive power and the voltage, phase angle of buses by partial least squares and Bayesian linear regression methods; [21] constructs constant impedance, current and power static load models and proposes a least-squares power flow linear regression method for distribution networks under three-phase imbalance.…”

Section: Introductionmentioning

confidence: 99%

Distribution network power flow calculation based on the BPNN optimized by GA‐ADAM

Liu,

Feng,

et al. 2023

The Journal of Engineering

View full text Add to dashboard Cite

Power system operation and control are based on power flow calculations. In order to solve the uncertainty of the increasing penetration of renewable energy, the voltage fluctuation at the load point increases in the distribution network, and the inaccuracy of the power flow calculation due to the insufficient power flow data collection capability of the traditional power system. In this paper, a data‐driven power flow analysis model is proposed, a back propagation neural network combined with genetic algorithm (GA) and adaptive moment estimation (ADAM) optimization algorithm model is constructed to analyze the power flow calculation method of distribution networks under stochasticity. Firstly, the power flow initial value information, topology characteristics, and power factor index are introduced to construct a training set, and the mapping relationship between bus voltage and power is fully explored by training the regression model. Second, the GA‐ADAM algorithm is used to optimize the initial values and weight parameters of the model. Finally, it is verified based on IEEE‐33 bus distribution model, and the model is used for power flow calculation, and compared with other methods through each relevant error evaluation indicators. The results show that the model constructed in this paper has small error indicators and high accuracy, which improves the efficiency and accuracy of power flow calculation.

show abstract

“…The use of the Koopman concept can boost the applicability of traditional linear analysis and control techniques, such as linear quadratic regulator and model predictive control (MPC), to act with an accuracy tantamount to non-linear control approaches with less computational burden while preserving optimization convexity. KO has been applied in various fields of power systems such as the identification of inter-area oscillations [24], power networks partitioning [25], state prediction [26], coherent identification [27], frequency regulation [28,29], damping control [30,31], power flow calculation [32], uncertainty quantification [33], and stability analysis [34].…”

Section: Introductionmentioning

confidence: 99%

Koopman model predictive control based load modulation for primary frequency regulation

Husham,

Kamwa,

Suprême

2023

IET Generation Trans & Dist

View full text Add to dashboard Cite

Conventional power systems function under the assumption that loads are uncontrollable, and that the generation control is the primary means of preserving system voltage, frequency, and stability. Thanks to recent advancements of power electronics technology and communication schemes, demand‐side resources are now capable of providing fast frequency regulation. Controllable loads can provide upward/downward reserve during frequency excursions. In this paper, the authors consider collective contribution of large clusters of controllable loads which modulate their aggregate demand power to regulate the primary frequency. Koopman model predictive control is designed to handle local frequency variations caused by various disturbances at each load bus, considering uncertain load models. The efficacy of the proposed method has been validated using the New‐England power system considering two scenarios, namely, load variation, and generation outage.

show abstract

Data-Driven Power Flow Calculation Method: A Lifting Dimension Linear Regression Approach

Cited by 29 publications

References 28 publications

A Framework for Analytical Power Flow Solution Using Gaussian Process Learning

A Framework for Analytical Power Flow Solution Using Gaussian Process Learning

Distribution network power flow calculation based on the BPNN optimized by GA‐ADAM

Koopman model predictive control based load modulation for primary frequency regulation

Contact Info

Product

Resources

About