.png){kind=icon}
.png){kind=icon}
Return
A Data-Driven and Deep Learning-Based Method for Power System Operating Mode Identification
.png){kind=icon}
OA
AI
C
H
N
J
T
B
J
X
DOI：10.35833/mpce.2025.000634.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Efficient and accurate analysis for power system operating mode identification is crucial for handling the operation, planning, and stability analysis of future power systems. This paper proposes a data-driven and deep learning-based method for power system operating mode identification, while also addressing the needs for identifying future operating modes. Firstly, an operating mode analysis method is developed based on adaptive threshold affinity propagation (AP) clustering with dynamic time warping (DTW), which incorporates historical load sequences and employs a modified distance function and adaptive thresholds for clustering. Secondly, a parallel temporal fusion network (PTFN) model with temporal feature proj ection is proposed to address load uncertainty. Finally, based on the results of historical operating mode analysis and future load forecasting, a Shapley additive explanation with parallel temporal convolution network and the squeeze-excitation mechanism (SHAP-PTCN-SE) is proposed for identifying future operating modes. Numerical examples demonstrate the efficiency and accuracy of the proposed data-driven and deep learning-based method in identifying future operating modes, providing guidance for system monitoring and protection.
Keywords:
Power system
operating mode identification
parallel temporal fusion network (PTFN)
parallel temporal convolution network (PTCN)
affinity propagation (AP)
data-driven
deep learning
load forecasting
Journal
IF:
6.1
Papers:
1.6K
Citations:
6.0K
Organization
Cited Papers