Wind Turbine Yaw Control via Statistical Power Offset Calibration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing wind turbine control methods face inaccuracies in wind direction measurements due to rotor interference, requiring complex and sensitive calibration methods that are time-consuming and sensitive to fluctuating wind speeds, leading to misalignment and reduced power production.
Innovation Solution
A method that pre-sets intervals of wind power parameters, collects data on wind direction and power, determines statistical representations of wind power as a function of relative wind direction, estimates wind direction offsets, and adjusts the rotor alignment based on these offsets to improve accuracy and reduce sensitivity to wind speed fluctuations, allowing for calibration without physical adjustments or comparative measurements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If wind direction sensor is used to measure wind direction for rotor alignment, then the rotor can be oriented to face the wind, but the rotor disturbs the free flow wind before reaching the sensor causing measurement inaccuracy
Solution Approach 1:
The patent introduces an intermediary calibration process that uses statistical analysis of power distribution curves to determine correction parameters. Instead of directly trusting the raw sensor measurements, a mathematical model (power distribution histogram) is used as an intermediary to calculate the actual wind direction offset, thereby compensating for the rotor's disturbance effect on the wind flow reaching the sensor.
Solution Approach 2:
The patent changes the parameter being measured from direct wind direction angle to power distribution statistics. By analyzing how power varies with apparent wind direction and fitting statistical models (histograms, Gaussian distributions), the system derives corrected wind direction parameters that account for the rotor interference effect without requiring physical modification to the sensor or rotor position.
2Measurement precision
If basic Wind Correction Parameters are obtained through prototype calibration with Met Mast comparison, then correction parameters can be established, but the process takes several weeks and requires special alignment tools
Solution Approach 1:
The patent enables the wind turbine to self-calibrate using its own operational data. Instead of requiring external reference equipment (Met Mast) and manual intervention with alignment tools, the system uses its own power production measurements combined with sensor readings to automatically compute correction parameters through statistical analysis of power distribution curves, making the calibration process self-sufficient and dramatically reducing time requirements.
Solution Approach 2:
The patent replaces the mechanical/physical calibration approach (using alignment tools and physical comparison with Met Mast) with a computational approach. Statistical models and algorithms process operational data to determine correction parameters, substituting mechanical measurement and adjustment processes with mathematical computation, thereby eliminating the need for special tools and reducing calibration time from weeks to minutes or hours.
3Productivity
If Wind Correction Parameters are transferred from prototype to production turbines, then calibration can be applied across the fleet, but the parameters do not fit individual turbines well due to variations in airflow
Solution Approach 1:
The patent implements a dynamic calibration approach where correction parameters are not fixed but can be individually determined for each turbine based on its specific operational characteristics and environmental conditions. The statistical analysis adapts to each turbine's unique power distribution patterns, allowing the system to account for variations in airflow caused by different installation sites, turbine configurations, and operational modes, thereby achieving both scalability and individualized accuracy.
4Manufacturing precision
If calibration is performed by physically aligning wind sensor with main shaft using alignment tool, then sensor alignment can be achieved, but the process is time-consuming and not always accurate when carried out in 100m height
Solution Approach 1:
The patent replaces the mechanical alignment process (using physical alignment tools to position the sensor relative to the main shaft) with a computational alignment approach. Statistical analysis of operational data automatically determines the correct alignment parameters, eliminating the need for manual physical adjustment at height. This substitution of mechanical alignment with mathematical computation achieves both higher precision and dramatically reduced time requirements.
5Measurement precision
If correction parameters are obtained from power distribution curves at each incoming wind speed, then wind direction can be calibrated, but the method is very sensitive to fluctuating or fast changing wind speeds
Solution Approach 1:
The patent performs preliminary statistical analysis to establish stable baseline characteristics of power distribution before using them for calibration. By accumulating sufficient operational data and establishing robust statistical models (histograms, mean, standard deviation) that represent typical operating conditions, the system creates a stable reference framework that is not easily disrupted by short-term wind fluctuations, thereby improving reliability while maintaining accuracy.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The invention relates to a method of controlling a wind turbine comprising a wind direction sensor, a yawing system, and a control system for turning the wind turbine rotor relative to the wind. The method comprises firstly obtaining at time intervals a data set comprising a direction of the wind relative to the wind turbine as measured by the wind direction sensor and a wind power parameter determined as one of a power, a torque, or a blade load of the wind turbine. The data sets are sorted into a number of bins of different intervals of wind power parameter. Based on the obtained data sets is then over time and for each power bin determined a statistical representation of the wind power parameter as a function of the relative wind direction which is then used in estimating a wind direction offset corresponding to the relative wind direction where the wind power parameter attains a peak value. The relative wind direction of the wind turbine is then adjusted as a function of the set of wind direction offsets to yield more accurate wind direction data which can be used in the controlling of the turbine. The invention further relates to a control system for a wind turbine for performing a control method as mentioned above.