Wind Turbine Control Upsampling Using Predicted Signal Trajectories
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional upsampling techniques in wind turbine control systems often result in aliasing and phase delay, which are undesirable in dynamically changing systems, as they fail to accurately convert low-rate control signals to higher rates required by actuator systems.
Innovation Solution
A wind turbine control system that includes a control module generating a first control signal with a current sample value and a predicted control trajectory, and an upsampling module that calculates a second control signal with a higher frequency by using interpolation functions based on the current sample value and predicted trajectory, effectively addressing the limitations of conventional methods.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If conventional upsampling techniques (zero-order hold or zero-stuffing) are used to convert low-rate control signals to high-rate signals, then the signal rate is increased, but aliasing occurs and phase delay is introduced
Solution Approach 1:
The control module generates a predicted control trajectory in advance that anticipates future control signal values. This preliminary action provides the upsampling module with forward-looking information, enabling it to generate high-rate control signals without aliasing or phase delay by interpolating between predicted future values rather than relying on past values alone.
Solution Approach 2:
The predicted control trajectory acts as an intermediary between the low-rate control signal from the control module and the high-rate control signal required by the actuator system. This intermediary contains pre-calculated future control values that the upsampling module uses to generate smooth, accurate high-rate signals without introducing aliasing or phase delay.
2Reliability
If filtering is applied to smooth discontinuities in upsampled signals, then aliasing is reduced, but phase delay is introduced which is undesirable in dynamically changing systems
Solution Approach 1:
The predicted control trajectory is calculated in advance using a model of the actuator system dynamics. This preliminary calculation incorporates the expected system response, allowing the upsampling to produce smooth signals without requiring post-filtering that would introduce phase delay. The prediction inherently accounts for system dynamics, eliminating the need for corrective filtering.
3Measurement precision
If the control signal rate is increased to match actuator system requirements, then control accuracy is improved, but the control module must operate at higher frequencies which increases computational load
Solution Approach 1:
The control function is segmented into two distinct modules: the control module that operates at low frequency to generate control decisions and predicted trajectories, and the upsampling module that operates at high frequency to generate the actual control signals for the actuator. This segmentation allows the computationally intensive control algorithm to run slowly while still providing high-rate control signals through intelligent interpolation of the predicted trajectory.
Solution Approach 2:
Instead of having the control module generate every high-rate control signal, the upsampling module creates copies and interpolations of the control trajectory generated by the control module. The predicted control trajectory serves as a template that the upsampling module replicates at the required high rate, significantly reducing the computational burden on the control module while maintaining control accuracy.
Data Source
AI summary
A wind turbine control unit includes an upsampling module that receives a first control signal that includes a current control sample value and a predicted control trajectory. The upsampling module also calculates a second control signal in dependence on the current control sample value and the predicted control trajectory. The second control signal has a higher frequency than the first control signal. The upsampling module further outputs the second control signal for controlling an actuator.