Hierarchical Data Fusion Algorithm for Multiple Wind Speed Sensors in Anemometer Tower.

Area of Science:

• Engineering
• Data Science
• Environmental Science

Background:

• Accurate wind speed measurement is crucial for wind power generation and meteorological monitoring.
• Multi-sensor data fusion from anemometers on wind towers is a primary method for high-precision wind speed data.
• Existing fusion methods face challenges in quality and efficiency.

Purpose of the Study:

• To propose a hierarchical data fusion strategy to enhance the quality and efficiency of multi-sensor wind speed data fusion.
• To improve the accuracy and speed of data processing in wind measurement systems.

Main Methods:

• A two-stage fusion approach: local fusion using a fuzzy logic and robustness factor-enhanced unscented Kalman filter (FLR-UKF) for denoising and fusion.
• Global fusion using an extreme learning machine (ELM) optimized by a Q-learning-improved Aquila optimizer (QLIAO-ELM) with enhanced search capabilities.

Main Results:

• The FLR-UKF reduced Root Mean Square Error (RMSE) by 26.46%–28.6% compared to traditional unscented Kalman filters (UKF).
• The QLIAO-ELM achieved RMSE reductions of 27.1% and 14.0% compared to standard ELM and ISSA-ELM, respectively.
• The proposed method demonstrated improved accuracy and efficiency in wind speed data fusion.

Conclusions:

• The hierarchical data fusion strategy effectively enhances both the accuracy and efficiency of multi-sensor wind speed measurements.
• The novel FLR-UKF and QLIAO-ELM methods offer significant improvements over existing techniques for wind data processing.
• This approach provides a reliable solution for high-precision wind speed information vital for renewable energy and meteorology.