Delegates are invited to meet and discuss with the poster presenters during the poster presentation sessions between 10:30-11:30 and 16:00-17:00 on Thursday, 19 November 2015.
Lead Session Chair:
Stephan Barth, ForWind - Center for Wind Energy Research, Germany
Boris Golub (3) F P Ivan Sturlic (2) Miroslav Sturlan (1)
(1) Faculty of Electrical Engineering and Computing, Zagreb, Croatia (2) Croatian Transmission System Operator Ltd., Zagreb, Croatia (3) KONCAR, Zagreb, Croatia
Printer friendly version: Print
Presenter's biographyBiographies are supplied directly by presenters at EWEA 2015 and are published here unedited
Mr. Golub is head of Custom Software Development at the KONCAR - Power Plant and Electric Traction Engineering Inc. He earned his master’s degree in Computer Science and Mathematics at Department of Mathematics, Faculty of Science in Zagreb 2003.
With more than ten years of experience in software development he changed different positions starting from development engineer, business analyst to project management roles.
His main references include development of Market Management System modules for Croatian Transmission System Operator. During his working experience in KONČAR he was co-author and author of several technical articles dealing with information system implementation.
PosterDownload poster (12.52 MB)
Wind Power Information System with powerful integration capabilities and simple physical model for "out-of-the-box" predicting
In this article we introduce the features of the wind power information system (short WIPS) both from motivation aspect describing business and user requirements and the technical implementation view point. Technical implementation details and reasoning behind choices made will be given on information system architecture as well as statistical processing for forecasting improvement purposes. Results and learnings from the specific implementation will be given as well.
WIPS was initially developed for purposes of integrating different wind power forecasting software’s originating from different vendors into one system responsible for archiving, displaying and end user statistical analysis. The software was developed as a part of SCADA/EMS/AGC Refurbishment and Upgrade Project for Croatian TSO but was subsequently substantially upgraded with support for multiple prognostic core, Market Management Information System integration as well as implementation of simple physical forecasting functions intended for use with wind power plants with no prognostic model available.
Main body of abstract
Transmission System Operators cope with the wind volatility issues while trying to optimise regulating reserve grid requirements. In order to lower the wind forecasting error frequently more than one forecasting model is used with different numerical weather predictions (if available). In order to efficiently evaluate and in the end prepare the final reserve plan TSOs require a tool allowing them to read, preview, compare and combine forecasts as frequently as new forecasts arrive. Information System users must benefit from the software ability to automatically prepare the initial wind power production plan. Initial plan can then be changed manually allowing users to build in their experience and additional field information that can have an effect to a power production forecast but was not (or cannot) be modelled programmatically. Final wind production plan is then being send to a so called Market Management Information System (short MMS) for ancillary services procurement purposes. Most of the processes handled within MMS system are done according to ENTSO-E Implementation Guides enabling information exchange automatisms between different actors (both internal and external information systems). Any redundant manual work is avoided as it increases time needed to react on forecast changes and prepare new wind power plan versions.
Apart from data exchange automatism the software provides algorithms for post factum evaluation of wind power forecasts errors which in the end generate wind power prognostic cores weight factors for future horizons. This way it is possible to provide automatic generation of the foreseen optimal wind power forecast as well as using various combination techniques to calculate optimal forecast.
Another important feature to be presented is the implementation of simple physical model (short SPPC). SPPC is implemented in a way that allows end users to substitute any standard analytical function as a prognostic core implementation. In this way end users can have rough forecasts based merely on manufacturer’s power curves modelled for example as a polynomial even if they don’t have real forecasts available. This is usually the case where artificial neural networks are solely used for forecasting and no historical data is available (e.g. new wind power plant is installed).
Experiences in using the software will be presented through forecasts errors analysis. Some insight will also be given on specific project challenges regarding software development problems coming from the regulatory requirements issues. Some ideas for future development of the WIPS system especially with respect to more close integration with SCADA system for wind power plant management will also be discussed.