• CSCD核心库收录期刊
  • 中文核心期刊
  • 中国科技核心期刊

ELECTRIC POWER CONSTRUCTION ›› 2021, Vol. 42 ›› Issue (8): 18-28.doi: 10.12204/j.issn.1000-7229.2021.08.003

• Original article • Previous Articles     Next Articles

Static Voltage Safety Analysis Based on ELM for Active Distribution Power Grid

HAN Chenyu1, XU Peng1, ZHU Hong2, LIU Shaojun2, WANG Beibei1   

  1. 1. Scool of Electrical Engineering,Southeast University, Nanjing 210096, China
    2. Nanjing Power Supply Branch of State Grid Jiangsu Electric Power Co., Ltd., Nanjing 210000, China
  • Received:2020-11-27 Online:2021-08-01 Published:2021-07-30
  • Supported by:
    research program “Research on Deep Reinforcement Learning Strategy of Active Distribution Network Voltage Control under Power Internet of Things”of State Grid Corporation of China

Abstract:

With the increase of uncertainty caused by the single phase access of distributed renewable generation (DRG) in the distribution power grid, the static safety analysis for overvoltage and imbalance of the distribution power grid with DRG faces new challenges. This paper proposes to use the extreme learning machine to analyze the complex relationship between the input and output of the three-phase power flow calculations in distribution network. The trained network can greatly improve the efficiency of power flow calculation in different scenarios caused by DRG fluctuations with this topology. On this basis, a voltage safety analysis method considering multiple scenarios of distribution power grid is proposed, which can quickly analyze and judge the safety of grid nodes. Finally, this paper uses IEEE 13-node,IEEE 33-node and IEEE 118-node systems with DRG for simulation calculation. The experimental results show that the method proposed in this paper is more effective than the traditional three-phase power flow calculation method without any convergence problem, and has higher efficiency and accuracy than BP neural network. The effectiveness of the proposed method is verified.

Key words: extreme learning machine, overvoltage analysis, safety analysis, neural network

CLC Number: