Proven Forecasting Tools: from Project to Portfolio
This paper covers a range of forecast tools that represent a state-of-the-art system in operational wind power prediction. It is important to understand that a true state-of-the-art forecast should be made up of several different tools that are each tuned to perform specific tasks optimally. There are a number of ways to produce forecasts-ranging in complexity from the simplest forecast-assuming no change in the wind speed through to the most complex systems such as artificial neural networks, Numerical Weather Prediction (NWP) modelling augmented with probabilistic ensemble weighting algorithms and ramp identification software. Moreover, on top of all of the technical solutions-there is still sometimes value in the oversight of a human to adjust the forecasts to better-fit expected behaviour using the natural “neural network-the human brain. There are also a number of different forecasts that can be made. Some of these forecasts are primarily of use to wind project operators, such as forecasts that help with maintenance scheduling and seasonal forecast to help manage energy expectations. Some are primarily of use to wind project traders, such as hour ahead and day ahead forecasts as well as anything that might result in a better power purchase agreement (PPA). Some are primarily of use to power system operators, such as the region-wide forecasts. Some are even of primary important to construction firms such as the preoperational construction forecasts.This paper focuses on a case study of operational tools used to solve, but also highlights some of the alternative possibilities that could also be used in conjunction with the tools demonstrated or as benchmarks of the accuracy.
Forecasting Numerical Weather Prediction Modelling Scheduling Operation Integration
C.W.Potter K. Westrick B. Rogers
3TIER Pacific Rim Pty Ltd GPO Box 528, Hobart, Tasmania, Australia, 7001 3TIER Inc. 2001 6th Avenue, Suite 2100, Seattle, WA 98121 Development - Pacific Rim 3TIER Pacific Rim Pty Ltd
国际会议
北京
英文
14-19
2010-10-13（万方平台首次上网日期，不代表论文的发表时间）