A hierarchical power system transient stability assessment method considering sample imbalance

Du, Yixing; Hu, Zhijian; Wang, Fangzhou

doi:10.1016/j.egyr.2021.08.052

Cited by 6 publications

(5 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In practical power systems, there are far fewer transient instability samples compared to stable samples, which leads to a model that is more biased towards predicting stable samples and increases the misjudgment of unstable samples. To address this class imbalance issue, there are currently two main methods [23]: at the data level, techniques such as oversampling, undersampling, and data augmentation are employed; at the algorithm level, weighted loss functions are used. In this paper, an improvement is made by introducing the focal loss function, which is formulated as follows:…”

Section: Improvement Of Loss Function Based On Unbalanced Samplesmentioning

confidence: 99%

Migratable Power System Transient Stability Assessment Method Based on Improved XGBoost

Qu,

Wang,

Cheng

et al. 2024

ENERGY

View full text Add to dashboard Cite

The data-driven transient stability assessment (TSA) of power systems can predict online real-time prediction by learning the temporal features before and after faults. However, the accuracy of the assessment is limited by the quality of the data and has weak transferability. Based on this, this paper proposes a method for TSA of power systems based on an improved extreme gradient boosting (XGBoost) model. Firstly, the gradient detection method is employed to remove noise interference while maintaining the original time series trend. On this basis, a focal loss function is introduced to guide the training of the XGBoost model, enhancing the deep exploration of minority class samples to improve the accuracy of the model evaluation. Furthermore, to improve the generalization ability of the evaluation model, a transfer learning method based on model parameters and sample augmentation is proposed. The simulation analysis on the IEEE 39-bus system demonstrates that the proposed method, compared to the traditional machine learning-based transient stability assessment approach, achieves an average improvement of 2.16% in evaluation accuracy. Specifically, under scenarios involving changes in topology structure and operating conditions, the accuracy is enhanced by 3.65% and 3.11%, respectively. Moreover, the model updating efficiency is enhanced by 14-15 times, indicating the model's transferable and adaptive capabilities across multiple scenarios.

show abstract

Section: Improvement Of Loss Function Based On Unbalanced Samplesmentioning

confidence: 99%

Migratable Power System Transient Stability Assessment Method Based on Improved XGBoost

Qu,

Wang,

Cheng

et al. 2024

ENERGY

View full text Add to dashboard Cite

show abstract

“…Particularly noteworthy are [129][130][131], in which the authors focused on ensuring system security through reinforcement learning. The assessment of the transient stability of energy systems using ML techniques was also discussed in articles [132][133][134][135][136]. For this purpose, CNN was used in [132], RNN was used in [133] and LSTM was used in [134,135].…”

Section: Power System Securitymentioning

confidence: 99%

“…The assessment of the transient stability of energy systems using ML techniques was also discussed in articles [132][133][134][135][136]. For this purpose, CNN was used in [132], RNN was used in [133] and LSTM was used in [134,135].…”

Section: Power System Securitymentioning

confidence: 99%

Application of Methods Based on Artificial Intelligence and Optimisation in Power Engineering—Introduction to the Special Issue

Pijarski,

Belowski

2024

Energies

View full text Add to dashboard Cite

The challenges currently faced by network operators are difficult and complex. Presently, various types of energy sources with random generation, energy storage units operating in charging or discharging mode and consumers with different operating characteristics are connected to the power grid. The network is being expanded and modernised. This contributes to the occurrence of various types of network operating states in practice. The appearance of a significant number of objects with random generation in the power system complicates the process of planning and controlling the operation of the power system. It is therefore necessary to constantly search for new methods and algorithms that allow operators to adapt to the changing operating conditions of the power grid. There are many different types of method in the literature, with varying effectiveness, that have been or are used in practice. So far, however, no one ideal, universal method or methodology has been invented that would enable (with equal effectiveness) all problems faced by the power system to be solved. This article presents an overview and a short description of research works available in the literature in which the authors have used modern methods to solve various problems in the field of power engineering. The article is an introduction to the special issue entitled Advances in the Application of Methods Based on Artificial Intelligence and Optimisation in Power Engineering. It is an overview of various current problems and the various methods used to solve them, which are used to cope with difficult situations. The authors also pointed out potential research gaps that can be treated as areas for further research.

show abstract

“…Sequence models such as the LSTM model [46] and GRU model [47] are also commonly applied to realize trajectory prediction, and the accuracy is improved. The BiLSTM model proposed by Yang et al [48,49] analyzes the interactive relationship characteristics among trajectories, and the long-term dependencies from the multivariate time series are studied, and the accuracy and robustness in the classification of the time samples are validated. Moreover, various Recurrent Neural Networks (RNN) and sequence models are combined in realizing traffic prediction [50].…”

Section: Trajectory Prediction Model Based On Deep Learningmentioning

confidence: 99%

KGCN‐LSTM: A graph convolutional network considering knowledge fusion of point of interest for vehicle trajectory prediction

Chen

Fan

Qian

et al. 2023

IET Intelligent Trans Sys

View full text Add to dashboard Cite

Urban vehicle trajectory prediction positively alleviates traffic congestion, avoids traffic accidents, and optimizes the urban transportation system. Since taxi trajectories are influenced by the driving intention, it is significant to consider the Points of Interest (POI) as the spatial features for trajectory prediction. A Knowledge Graph Convolutional Network Long Short-Term Memory (KGCN-LSTM) model is proposed here to improve the accuracy and robustness of trajectory prediction. POI information is considered as the prior-knowledge of the trajectory by the Graph Convolutional Network (GCN). Under multiple comparison experiments, Shopping POI gains the highest positive effect weight of 15% in holidays, and Hospital POI gains the highest weight of 16% in working days. In holidays, higher accuracy and robustness are achieved compared with benchmarks when performing the KGCN-LSTM model with POI of shopping, food, life service, scenic spots, and entertainment classes, while the performance is not improved with the rest of the POI classes. In working days, higher accuracy and stronger robustness are achieved compared with benchmarks when performing the KGCN-LSTM model with POI of hospital, life service, and exercise. While the performance is not improved with the rest of the POI classes.

show abstract

A hierarchical power system transient stability assessment method considering sample imbalance

Cited by 6 publications

References 7 publications

Migratable Power System Transient Stability Assessment Method Based on Improved XGBoost

Migratable Power System Transient Stability Assessment Method Based on Improved XGBoost

Application of Methods Based on Artificial Intelligence and Optimisation in Power Engineering—Introduction to the Special Issue

KGCN‐LSTM: A graph convolutional network considering knowledge fusion of point of interest for vehicle trajectory prediction

Contact Info

Product

Resources

About