TY - GEN
T1 - Operation Reliability Assessment and Diagnosis Method Based on Random Matrix Distribution Network
AU - Le, Hongxi
AU - Ni, Chunhua
AU - Lu, Yu
AU - Zhu, Hongcheng
AU - Xie, Jing
AU - Sun, Zhihao
AU - Gao, Congzhe
N1 - Publisher Copyright:
© 2019 IEEE.
PY - 2019/8
Y1 - 2019/8
N2 - The reliability and safety of distribution network operation have great significance for daily power supply security. In order to make full use of multi-source distribution network data, diagnose and evaluate the evaluation index of distribution network operation reliability, this paper proposes a long-short-Term memory neural network and random matrix operation diagnosis method for distribution network based on the combination of physics and data model. The long-shortterm memory neural network and random matrix theory are analyzed first and the long-short-Term memory neural network is used to train the node load data to obtain the prediction model. According to the definition of the multilayer physical index system, the system physical index is calculated. The description is established combined with the random matrix. Data indicators for overall operational posture. Principal component analysis is used to extract the main physical indicators and integrate them with the data indicators to obtain the diagnostic indicators for the distribution network operation assessment. The actual operation data proves that the proposed evaluation and diagnosis method can effectively reflect the running state of the urban distribution network, and has good multi-source data application capability and practicability.
AB - The reliability and safety of distribution network operation have great significance for daily power supply security. In order to make full use of multi-source distribution network data, diagnose and evaluate the evaluation index of distribution network operation reliability, this paper proposes a long-short-Term memory neural network and random matrix operation diagnosis method for distribution network based on the combination of physics and data model. The long-shortterm memory neural network and random matrix theory are analyzed first and the long-short-Term memory neural network is used to train the node load data to obtain the prediction model. According to the definition of the multilayer physical index system, the system physical index is calculated. The description is established combined with the random matrix. Data indicators for overall operational posture. Principal component analysis is used to extract the main physical indicators and integrate them with the data indicators to obtain the diagnostic indicators for the distribution network operation assessment. The actual operation data proves that the proposed evaluation and diagnosis method can effectively reflect the running state of the urban distribution network, and has good multi-source data application capability and practicability.
KW - Distribution Network
KW - Evaluation and Diagnosis
KW - Long and Short Time Memory Method
KW - Multi-level Evaluation System
KW - Principal Component Analysis
KW - Random Matrix
UR - http://www.scopus.com/inward/record.url?scp=85075927809&partnerID=8YFLogxK
U2 - 10.1109/ICSGEA.2019.00132
DO - 10.1109/ICSGEA.2019.00132
M3 - Conference contribution
AN - SCOPUS:85075927809
T3 - Proceedings - 2019 International Conference on Smart Grid and Electrical Automation, ICSGEA 2019
SP - 562
EP - 570
BT - Proceedings - 2019 International Conference on Smart Grid and Electrical Automation, ICSGEA 2019
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 2019 International Conference on Smart Grid and Electrical Automation, ICSGEA 2019
Y2 - 10 August 2019 through 11 August 2019
ER -