Electric Power ›› 2016, Vol. 49 ›› Issue (3): 148-153.DOI: 10.11930/j.issn.1004-9649.2016.03.148.06

• New Energy • Previous Articles     Next Articles

Wind Farm Power Curve Optimization Based on Actual Operating Data

RAO Risheng1, YE Lin1, REN Cheng1, SONG Xuri2, LANG Yansheng2, JIN Jingxin3   

  1. 1. College of Information and Electrical Engineering, China Agricultural University, Beijing 100083, China;
    2. Power Automation Department, China Electric Power Research Institute, Beijing 100192, China;
    3. Inner Mongolia Water Resources and Hydropower Survey and Design Institute, Hohhot 010021, China
  • Received:2015-10-08 Online:2016-03-20 Published:2016-04-08
  • Supported by:
    This work is supported by National Natural Science Foundation of China (No. 51477174, 51174290, 51077126) and Specialized Research Fund for the Doctoral Program of Higher Education(SRFDP) of China(No. 20110008110042); Science and Technology Project of SGCC (No. DZB51201503568).

Abstract: This paper analyzes the statistical relationship between wind farm’s wind speed and power output using actual operating data. It proposes a methodology to optimize the power output curve for a wind farm based on the traditional wind speed partitioning approach. Historical observation data is equally distributed into each wind speed interval. A piecewise least squares linear regression approach is used to determine the best fir between the wind speed and the power output, and to establish the optimal power output curve using actual operating data. The power output curve is further tested with observed data in real time. Case studies indicate that the power output curve optimization method has high accuracy and can better reflect the actual operating characteristic of a wind farm. The methodology contributes to the study of wind farm operational characteristics and wind power integration research.

Key words: power system, operating data, observed data, wind farm, wind speed vs. power characteristic, piecewise fitting, least square method, linear regression method, optimal power curve

CLC Number: