TY - GEN
T1 - Towards a general tool chain integration platform for multi-disciplinary analysis and optimization in wind energy
AU - Feng, Ju
AU - Sørensen, Jens Nørkær
PY - 2024
Y1 - 2024
N2 - To enable easier uptake and flexible usage for the wind energy community, a general tool chain integration platform for multi-disciplinary analysis and optimization, TopDesign, is proposed and developed. This work describes its concept and the initial implementation. The platform is written in Python and provided as an open source framework. The design of this platform puts emphasis on flexibility, usability and extensibility. General model classes with standard interfaces, which can wrap custom tools/models in different formats based on different programming languages, will be provided together with examples that are easy to understand and modify. Coupling tools/models into a system thus becomes an easy task of connecting models by inputs and outputs and defining workflow with a directed graph. To demonstrate the usage of the proposed platform, we implement a wind farm evaluation system by coupling the PyWake model, the FLORIS model and self-written constraint models in the platform and apply it in solving wind farm layout optimization problems with two algorithms, as implemented in an in-house optimization Python library. A case study for a wind farm composed of 9 turbines shows the potential of the proposed platform and methodology.
AB - To enable easier uptake and flexible usage for the wind energy community, a general tool chain integration platform for multi-disciplinary analysis and optimization, TopDesign, is proposed and developed. This work describes its concept and the initial implementation. The platform is written in Python and provided as an open source framework. The design of this platform puts emphasis on flexibility, usability and extensibility. General model classes with standard interfaces, which can wrap custom tools/models in different formats based on different programming languages, will be provided together with examples that are easy to understand and modify. Coupling tools/models into a system thus becomes an easy task of connecting models by inputs and outputs and defining workflow with a directed graph. To demonstrate the usage of the proposed platform, we implement a wind farm evaluation system by coupling the PyWake model, the FLORIS model and self-written constraint models in the platform and apply it in solving wind farm layout optimization problems with two algorithms, as implemented in an in-house optimization Python library. A case study for a wind farm composed of 9 turbines shows the potential of the proposed platform and methodology.
U2 - 10.1088/1742-6596/2767/8/082014
DO - 10.1088/1742-6596/2767/8/082014
M3 - Article in proceedings
T3 - Journal of Physics: Conference Series
BT - The Science of Making Torque from Wind (TORQUE 2024): System design and multi-fidelity/multi-disciplinary modeling
PB - IOP Publishing
T2 - The Science of Making Torque from Wind (TORQUE 2024)
Y2 - 29 May 2024 through 31 May 2024
ER -