Unable to log in or get member pricing? Having trouble changing your password?

Please review our Frequently Asked Questions for complete information on these and other common situations.
 

Vertical Flight Library & Store

Ensemble Integration Methods for Load Estimation

Catherine Cheung, Zouhair Adam Hamaimou, National Research Council Canada

May 10, 2022

https://doi.org/10.4050/F-0078-2022-17553

Abstract:
Helicopter component load estimation can be achieved through a variety of machine learning techniques and algorithms. To increase confidence in the load estimation process, ensemble methods are employed combining multiple individual load estimators that increase predictive stability across flights and add robustness to noisy data. In this work, several load estimation methods are applied to a variety of machine learning algorithms to build a large library of individual load estimation models for main rotor yoke loads from 28 flight state and control system parameters. This paper explores several ensemble integration methods including simple averaging, weighted averaging using rank sum, and forward selection. From the 426 individual models, 25 top models were selected based on four ranking metrics, root mean squared error (RMSE), correlation coefficient, and interquartile ranges of these two metrics. All ensembles achieved improved performance for these four metrics compared to the best individual model, with the forward selection ensemble obtaining the lowest RMSE, highest correlation, and closest load signal prediction visually of all models.


Ensemble Integration Methods for Load Estimation

  • Presented at Forum 78
  • 9 pages
  • SKU # : F-0078-2022-17553
  • HUMS I

  • Your Price : $30.00
  • Join or log in to receive the member price of $15.00!


VFS member?
Don't add this to your cart just yet!
Be sure to log in first to receive the member price of $15.00!

 
Add To Cart

Add to Wish List

Reward Value:
(60) Member Points

Ensemble Integration Methods for Load Estimation

Authors / Details:
Catherine Cheung, Zouhair Adam Hamaimou, National Research Council Canada