Wind Turbine Tower-Top Acceleration Estimation with Kalman Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing wind turbine acceleration measurement systems are susceptible to inaccuracies due to yaw-induced vibrations and DC offset, and are limited to use in control domains, lacking robustness for safety domain applications.
Innovation Solution
A method for determining tower top acceleration of a wind turbine using a plurality of acceleration sensors, where the acceleration data is processed using a Kalman filter algorithm to update predicted accelerations based on kinematic models and sensor measurements, thereby compensating for torsional acceleration disturbances and improving measurement accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a single accelerometer is located at the top of the wind turbine tower, then the device complexity is reduced, but the measurement precision deteriorates due to yaw-induced vibrations and DC offset
Solution Approach 1:
The single accelerometer measurement task is segmented into multiple sensors distributed in the nacelle. Each sensor captures local acceleration data, and through coordinate transformations and fusion algorithms, the system reconstructs tower top acceleration with higher precision while compensating for yaw-induced vibrations and DC offset effects
Solution Approach 2:
A Kalman filter algorithm acts as an intermediary that processes measurements from multiple nacelle-mounted accelerometers. The filter fuses these measurements with a kinematic model of the wind turbine to produce accurate tower top acceleration estimates, effectively mediating between imperfect sensor readings and the desired measurement outcome
2Productivity
If acceleration measurements are used in control domain applications, then the productivity is improved through optimized power generation, but the reliability deteriorates because the system cannot withstand component faults and inaccuracies
Solution Approach 1:
The system implements beforehand cushioning by using multiple redundant accelerometers and a robust estimation algorithm that anticipates potential sensor failures. The Kalman filter is designed to handle measurement uncertainties and can continue providing reliable estimates even when individual sensors fail, thus cushioning against the impact of component faults before they compromise system reliability
Solution Approach 2:
The system changes parameters by transitioning from direct single-sensor measurements to a multi-sensor fusion approach with adjustable weighting factors in the Kalman filter. This allows the system to adapt to varying operational conditions and sensor quality, maintaining high reliability across different scenarios while enabling both control and safety domain applications
Data Source
AI summary
A method of determining tower top acceleration of a wind turbine is provided. The method includes receiving acceleration data from a plurality of acceleration sensors positioned in a nacelle of the wind turbine, including data indicative of a measured acceleration in a direction along at least one measurement axis of each respective acceleration sensor at a current time step. The method includes determining a predicted tower top acceleration of the wind turbine tower at the current time step, the predicted tower top acceleration being determined in dependence on a kinematic model of the wind turbine, and on a determined estimation of tower top acceleration at a previous time step. The method includes determining an estimated tower top acceleration of the wind turbine tower at the current time step by updating the predicted tower top acceleration based on the measured acceleration from each of the acceleration sensors.