• Home|About JOPS|Editorial Board|Contact us|Chinese
Design of Grid Fluctuation Warning System for Grid-connected Distributed Photovoltaic Generation
Received:January 24, 2019  Revised:April 09, 2019
  View/Add Comment  Download reader
Keywords:distributed  photovoltaics grid  irradiation fluctuations  early warning  inverters  bp neural network
Fund Project:
XIONG Ning State Grid Jiangxi Economic Research Institute xiongn135246@126.com
Wenting Cui Key Laboratory of Smart Grid, Ministry of Education xiongn135246@126.com
Hits: 29
Download times: 0
      When the current grid-connected distributed photovoltaic power generation predicts solar radiation fluctuations, the algorithm is too simple to accurately determine the range of radiation fluctuations and easily fall into the local optimal solution. The accuracy of the prediction results is low, and the solar radiation fluctuations cannot be accurately predicted. Design a new grid-connected distributed photovoltaic power generation radiation fluctuation warning system. The DC power emitted by the solar array is transmitted to the micro-inverter through the photoelectric sensor module, and the DC power is converted into AC power and transmitted to the single-machine radiation fluctuation warning subsystem. Predicting solar radiation fluctuations in the solar irradiance information acquisition equipment, if abnormal fluctuations occur, transmitting to the central master control for correlation processing; using the exponential smoothing method combined with the trend moving average method to obtain the irradiation fluctuation range of the prediction period, using The irradiance fluctuation prediction method based on ant colony BP neural network obtains the optimal solution of the system prediction results, and finally completes the design of the grid-connected distributed photovoltaic power generation radiation fluctuation warning system. The experimental results show that the designed system can effectively predict solar radiation fluctuations under different weather conditions, and the accuracy of early warning in many weathers is not less than 98%, and the average warning time is 24.08s, which has high early warning accuracy and efficiency.