Wind Turbine Vertical Wind Profile Determination Using Trained Data Driven Model
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for determining a vertical wind speed profile of a wind field approaching a wind turbine are inadequate, relying on limited point measurements that fail to capture the complex behavior of the wind field, leading to under-complex control strategies and poor performance analysis.
Innovation Solution
A computer-implemented method using a trained data-driven model, such as a Deep Neural Network, processes measurement values from wind turbines to determine the wind speed profile, including the shear coefficient, which describes the vertical wind profile, enabling improved control strategies and monitoring.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If only one or two point measurements of wind speed are used, then the measurement system remains simple, but the accuracy of determining the vertical wind speed profile deteriorates
Solution Approach 1:
The patent introduces an intermediate computational model that processes existing measurement data to infer the vertical wind speed profile. Instead of directly measuring at multiple points, the system uses a trained model as an intermediary to derive profile information from limited measurements, resolving the contradiction between simple measurement systems and accurate profile determination.
Solution Approach 2:
The patent creates a virtual copy of the vertical wind speed profile through computational modeling. Rather than physically measuring at multiple heights, the system generates a replicated representation of the wind profile using trained models, achieving accurate profile determination without complex multi-point measurement infrastructure.
2Device complexity
If limited point measurements are used, then the control strategy remains simple, but the ability to capture complex wind field behavior deteriorates
Solution Approach 1:
The trained computational model serves as an intermediary that bridges simple measurements and complex wind field understanding. It processes limited measurement inputs and generates comprehensive wind profile information, enabling sophisticated control strategies without requiring complex measurement systems.
Solution Approach 2:
The patent transitions from one-dimensional point measurements to a multi-dimensional vertical profile representation through computational modeling. The trained model expands the information dimensionality, deriving vertical distribution characteristics from single-point or few-point measurements, thus capturing complex wind field behavior without proportionally increasing measurement complexity.
3Measurement precision
If more measurement points are installed to capture vertical wind profile, then the wind speed profile accuracy improves, but the device complexity and cost increase
Solution Approach 1:
The patent replaces the mechanical approach of installing multiple physical measurement points with a computational model-based system. The trained data-driven model substitutes for additional sensors and measurement infrastructure, achieving accurate vertical profile determination through information processing rather than physical expansion of the measurement system.
Solution Approach 2:
Instead of physically replicating measurement points at different heights, the system creates a computational copy of the vertical wind profile. The trained model generates a virtual representation of the wind speed distribution, achieving profile accuracy without the cost and complexity of multiple physical sensors.
Data Source
Figure 1
Figure 2~3
AI summary
A method for computer-implemented determination of a wind speed profile information (WSPI) characterizing a vertical wind profile of a wind field (WF) approaching a wind turbine (1) is described. At each time point of one or more time points during the operation of the wind farm the following steps are performed: obtaining a number of measurement values (MV) of the wind turbine (1), the number of measurement values (MV) being captured by respective sensors of the wind turbine (1) to monitor and/or control the wind turbine (1); determining a wind speed profile information (WSPI) by processing the number of measurement values (MV) by a trained data driven model (MO), where the number of measurement values (MV) is fed as a digital input to the trained data driven model (MO) and the trained data driven model (MO) provides the wind speed profile information (WSPI) as a digital output, where the wind speed profile information (WSPI) at least comprises as a target value a shear coefficient of the wind field (WF) in direction to the wind turbine (1).