Open AccessData-driven management mechanism of demand response resources in active distribution networks
Author(s) -
Jianbing Yin,
Zhiwei Xu,
Cheng-Xian Lin,
Dairui Li
Publication year - 2021
Publication title -
iop conference series. earth and environmental science
Language(s) - English
Resource type - Journals
eISSN - 1755-1307
pISSN - 1755-1315
DOI - 10.1088/1755-1315/645/1/012022
Subject(s) - demand response , computer science , electricity , mechanism (biology) , distribution (mathematics) , deep learning , distributed computing , artificial intelligence , engineering , electrical engineering , mathematical analysis , philosophy , mathematics , epistemology
As power distribution networks have been transferring from passive to active, conventional physical-driven models are incompetent to deal with the challenges of flexible and fast-changing operating conditions. Alternatively, data-driven methods, especially deep learning methods, have unique advantages in tackling those challenges. To this end, the paper first introduces the opportunities brought by the massive amount of data in active distribution networks (ADNs). Then the paper employs neural networks to identify load categories and their potential capacity of demand response to interact with ADNs. Furthermore, a generalized dispatch framework is presented to coordinate flexible loads for peak-valley electricity alleviation, with long-short-term-memory networks as an example. Last, the paper sheds light on the limitations of deep learning applications such as data cleaning and cybersecurity attacks.