A Data-Driven Approach to Improve Wind Dispatchability

作者：Qiu, Feng; Li, Zhigang^*; Wang, Jianhui

来源：IEEE Transactions on Power Systems, 2017, 32(1): 421-429.

DOI：10.1109/TPWRS.2016.2554629

• Chance constraint
• data driven
• mixed-integer linear program
• wind dispatchability

Wind power has been treated as a non-dispatchable resource until recent development of active wind dispatch strategies in several electricity markets. In markets such as ISO New England, a dispatch range for each wind farm is determined based on security analysis. Wind power will be fully absorbed unless it is out of the ranges. This approach, though aiming at improving wind utilization with system security considered, relies solely on wind power forecasting, which could be inaccurate by nature, and might result in unnecessary wind curtailment. In our work, we recognize the discrepancy between wind power forecast and the actual wind power dispatched and develop a data-driven approach to better capture the uncertainties in wind power dispatch. The computational experiments demonstrate that the dispatch ranges determined by our data-driven approach can dispatch more wind without endangering the system security and that solution is also efficient.

• 出版日期2017-1
• 单位华南理工大学