Application of AI for Automatic Aimpoint Selection for HEL
Loading...
Authors
Agrawal, Brij
Subjects
Artificial Intelligence
High Energy Laser
Prediction Accuracy
Synthetic Data
Real Data
High Energy Laser
Prediction Accuracy
Synthetic Data
Real Data
Advisors
Date of Issue
2025-03-31
Date
Publisher
Monterey, CA; Naval Postgraduate School
Language
Abstract
The objective of this project was to improve reliability and accuracy for the application of artificial intelligence (AI) for autonomous aimpoint selection for High Energy Laser (HEL) systems. Naval Postgraduate School (NPS) has done considerable development work in this area including development of AI algorithms, development of target image data, and testing on different target trajectories, both by simulations and experiments. Research was performed in four areas: (1) Quantify model prediction reliability. Monte Carlo node dropout concept in AI model was used to quantify model prediction reliability. It was found that for target which was not included in training the AI model, the mean error and error variance were significantly higher for different node dropouts in comparison to targets included in training the model. (2) Improve accuracy of aimpoint selection prediction, by comparing two AI algorithms. HRNet and Keypoint R-CNN were evaluated. It was found that HRNet has pose ambiguity issues, resulting in higher prediction error in comparison to Keypoint R-CNN. (3) Solve the ambiguity of pose that is a major source of error in predicting UAV aimpoint selection. It was found that the predictive estimation filter developed under this project was effective to solve the pose ambiguity problem. (4) Compare accuracy of predicting aimpoint selection for training the AI model for the same UAVs, but with synthetic image data versus real image data. It was found that using synthetic
Type
Technical Report
Description
Series/Report No
Department
Mechanical and Aerospace Engineering (MAE)
Organization
Naval Research Program (NRP)
Identifiers
NPS Report Number
NPS-MAE-25-001
Sponsors
Naval Postgraduate School, Naval Research Program
Office of Naval Research
Office of Naval Research
Funding
This research is supported by funding from the Naval Postgraduate School, Naval Research Program (PE0605853N/2098). https://nps.edu/nrp
Chief of Naval Operations (CNO)
Chief of Naval Operations (CNO)
Format
51 p.
Citation
Distribution Statement
Approved for public release; distribution is unlimited.
Rights
This publication is a work of the U.S. Government as defined in Title 17, United States Code, Section 101. Copyright protection is not available for this work in the United States.