Research on Substation Fault Handling Method Suitable for Digital Handover Scenarios

With the engineering application technology fully entering the digital era, it is particularly significant to achieve the digital handover of data. This paper develops a substation fault handling method based on deep reinforcement learning (DRL) suitable for digital handover scenarios. To accurately...

Full description

Saved in:
Bibliographic Details
Published in2022 7th Asia Conference on Power and Electrical Engineering (ACPEE) pp. 1431 - 1435
Main Authors Chen, Ran, Zhou, Li, Ke, Fangchao, Fang, Zhao, Li, Zhiwei, Liu, Wanfang, Liu, Wenbin
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.04.2022
Subjects
analytical model
Analytical models
Circuit breakers
deep reinforcement learning
digital handover
Fault diagnosis
fault handling
Neural networks
Reinforcement learning
substation
Substations
Training
Online AccessGet full text

Cover

More Information
Summary:With the engineering application technology fully entering the digital era, it is particularly significant to achieve the digital handover of data. This paper develops a substation fault handling method based on deep reinforcement learning (DRL) suitable for digital handover scenarios. To accurately reflect the faulty situation in the substation, the proposed method considers the real-time state influence of each protection and circuit breaker in the substation. Meanwhile, the proposed method uses the composite weighting method to design the contribution factor to improve the analytical model. Besides, the double deep Q network (DQN) algorithm employed in the proposed method can evaluate and select actions through different neural networks to reduce the estimation error. The effectiveness of the proposed method is verified by performance tests and analysis. The test results show that the proposed method has excellent fault handling accuracy and speed.
DOI:10.1109/ACPEE53904.2022.9783837