Learning to forecast wind at remote sites for wind energy applications. Final report

Abstract

Pacific Northwest Laboratory Operated for the u.s. Department of Energy by Battelle Memorial Institute C) Banelle PNL-4318 UC-60 D ISCLAIM ER This re po rt was pre pared as an acco u nt of wo rk sponso red by an agency of th e U nited States Governmen t. Nei t her the Un ited States Gove rn m ent no r any age ncy the reof, nor any of t heir employees, makes any wa r ra nty, ex press o r implied , o r ass um es any legal liabi li ty or responsi bi lity for th e accu racy, completeness , or usefu lness of any information , apparatus, prod uct , or p rocess d isclosed, o r repre sents that its use wou ld not in fringe private ly ow ned ri ghts. Re ference he re i n to any specifi c commerci al p roduct, process, or service by trade nam e, tra d em ark, man u fact urer, or othe rwise, does not necess aril y co nstitute or imply i ts endorsemen t, recommendat ion , o r favo ri ng by the U n ited States Gove r nmen t or any agency thereo f. The views and o p in ion s of au t ho rs expressed herein do not necessarily state or ref lect t hose of t he U nited States Government or any agency t hereof.

Test application of a semi-objective approach to wind forecasting for wind energy applications

“…A semi-objective (S-O) approach to wind forecasting for wind energy applications has been developed previously and documented in Learning to Forecast Wind at Remote Sites for Wind Energy Applications (Notis et al 1983). The S-O procedure has now been tested and refined by applying the procedure at three Department of Energy (DOE) candidate wind turbine sites.…”

“…In Section 5 conclusions are drawn regarding the usefulness of the S-O wind forecasting approach and regarding the need for future efforts in wind forecasting for wind energy applications. The three sites selected for verification of the S-O wind forecasting method, developed by Notis et al (1983), were chosen to represent different geographical and topographical regimes. Block Island is a coastal site, Finley is located in the northern Great Plains and the San Gorgonio site lies in a mountain pass.…”

Short‐Term Wind Power Prediction for Wind Turbine via Kalman Filter Based on JIT Modeling

“…Among them, a method based on the ARMA model known as the Box-Jenkins [3] and a method based on neural networks are being actively studied [4,5]. The first one is the methods based on a physical model known as numerical weather prediction (NWP).…”

“…These are methods that predict future data based on past measurement data and present data. Among them, a method based on the ARMA model known as the Box-Jenkins [3] and a method based on neural networks are being actively studied [4,5]. In Japan, Taniguchi et al [6] and Kakuta et al [7] are engaged in wind speed prediction and power generation prediction using neural networks.…”

Short-Term Wind Speed and Output Power Forecasting Based on WT and LSSVM

“…Such as: Physical method [3], it uses physical considerations to predict the future speed and direction of wind, so the input variables will be the physical or meteorology information, such as description of orography, roughness, obstacles, meteo and so on. Statistical method [4], it has a high precision about short-term wind speed forecast. But, the precision of prediction for long term will be decreased obviously.…”