Periodic Reporting for period 1 - TWAIN (Integrated, Value-based and Multi-objective wind farm control powered by Artificial Intelligence)
Période du rapport: 2023-11-01 au 2025-02-28
The overall objective of TWAIN project is to develop an open-source, AI-driven decision support environment for wind power asset management to support WF owners/operators to make better decisions for system-wide optimised performance. Accordingly, TWAIN’s concept pivots on a full integration of Wind Farm Control (WFC) at five different levels:
1) Integration of multi-source and multi-format data of varied nature from WFs in different life stages.
2) AI-enabled Integration of multi-disciplinary processes and phenomena affecting the WF operation.
3) Integration of multi-objective prospects of WFC to assess the true added value of a certain operation mode.
4) Integration of multi-level controllers and scenario analyses in decision support provision for harmonious co-existence of wind power plants with their environment and society via optimised operation and design.
5) Integration of wider audience to TWAIN outcomes.
TWAIN will operate on a digital environment architected for multi-source data integration and optimised computing, which contains a set of toolboxes with the critical analytical steps to define and assess an effective and efficient wind farm operation.
1) Integration of multi-source and multi-format data of varied nature from WFs in different life stages.
2) AI-enabled Integration of multi-disciplinary processes and phenomena affecting the WF operation.
3) Integration of multi-objective prospects of WFC to assess the true added value of a certain operation mode.
4) Integration of multi-level controllers and scenario analyses in decision support provision for harmonious co-existence of wind power plants with their environment and society via optimised operation and design.
5) Integration of wider audience to TWAIN outcomes.
TWAIN will operate on a digital environment architected for multi-source data integration and optimised computing, which contains a set of toolboxes with the critical analytical steps to define and assess an effective and efficient wind farm operation.
TWAIN will:
• Create an open-source data management toolbox including containerised experimental and numerical data;
• Establish AI-driven WFC-oriented tools for multi-objective WFC by digitalising the existing models, building new ones and incorporating them into model ensembles;
• Assess the holistic socio-economic and environmental impact of WF operation modes, reinforced by experiments and experience in the field; and
• Develop a secure-by-design, open-source TWAIN integrated WF controller and decision support environment for wind power asset management, including wider testing by the external users towards its earlier adoption.
• Create an open-source data management toolbox including containerised experimental and numerical data;
• Establish AI-driven WFC-oriented tools for multi-objective WFC by digitalising the existing models, building new ones and incorporating them into model ensembles;
• Assess the holistic socio-economic and environmental impact of WF operation modes, reinforced by experiments and experience in the field; and
• Develop a secure-by-design, open-source TWAIN integrated WF controller and decision support environment for wind power asset management, including wider testing by the external users towards its earlier adoption.
TWAIN will support Europe in maintaining its global leadership in wind energy technology by generating knowledge for industrial innovation and offering open-source tools and training to the wind energy industry. TWAIN decision support environment together with its submodules and the associated training on data types, data transformations, and capabilities to (re)build and (re)train algorithms will improve the skills of engineers and applied computational scientists for novel AI-driven workflows, and within the next decade more than 1000 engineers are expected to use the tools developed within TWAIN. The architectural design and overall data-driven framework of TWAIN are expected to generate ‘best practices’ for several digitalisation activities within wind energy, to be used universally by academia and industry. They include concrete analysis through quantified social, economic, and environmental metrics for industrially critical system-wide optimisation scenarios up to 2050, promising further acceleration and growth of wind as the most affordable, available, and efficient energy source – with up to 7% decrease in LCOE and AI-driven acceleration towards REPowerEU 45% renewable energy by 2030 targets. A wide uptake of the TWAIN outcomes is supported by the involvement of three important owner/operators that have a growing 9% European and 4% global market share of installed wind capacity. With their influence on the market, the TWAIN outcomes are estimated to be used for (at least) 50% of their WFs within the next decade, both for operation of the existing WFs and design of the new ones. TWAIN will have further outreach outside of the consortium via the TWAIN Open Call and Research & Innovation Board (RIB) where, among others, RIB.3 (Ørsted, with 1% additional global market share) will have active interaction with the TWAIN outcomes.