Particle filtering for obstacle tracking in UAS sense and avoid applications.

Area of Science:

• Aerospace Engineering
• Robotics
• Control Systems

Background:

• Accurate obstacle detection and tracking are crucial for unmanned aircraft system (UAS) sense and avoid capabilities.
• Collision risk assessment relies on predicting the Distance at Closest Point of Approach (DCP).
• Traditional methods may suffer accuracy loss due to nonlinearities in tracking scenarios.

Purpose of the Study:

• To develop and evaluate a particle filter-based algorithm for obstacle tracking in UAS.
• To assess the performance improvements offered by particle filters in estimating target states for collision avoidance.
• To investigate the impact of particle filtering on radar-based tracking within a single-sensor framework.

Main Methods:

• Algorithm development for an obstacle tracking system utilizing particle filters.
• Off-line simulations using flight test data to evaluate algorithm performance.
• Focus on radar-based tracking to analyze particle filter effectiveness in a practical sensor setup.

Main Results:

• Particle filters demonstrated improved accuracy in estimating the Distance at Closest Point of Approach.
• Enhanced estimation led to a reduction in the delay associated with collision detection.
• The study confirmed the benefits of advanced filtering for nonlinear tracking problems.

Conclusions:

• Particle filters offer a viable solution for improving the accuracy of obstacle tracking in UAS.
• The developed algorithm shows potential for enhancing the safety and reliability of sense and avoid systems.
• Accurate DCP estimation through particle filtering is key to mitigating collision risks in dynamic flight environments.