PENC: a predictive-estimative nonlinear control framework for robust target tracking of fixed-wing UAVs in complex...

Area of Science:

• Robotics and Control Systems
• Computer Vision and Target Tracking
• Aerospace Engineering

Background:

• Fixed-wing unmanned aerial vehicles (UAVs) face significant challenges in tracking targets within complex urban environments, primarily due to occlusions from vertical structures.
• Existing tracking algorithms often struggle with target state loss and require robust estimation and prediction capabilities to maintain operational effectiveness.

Purpose of the Study:

• To introduce a novel Predictive-Estimative Nonlinear Control (PENC) framework designed to improve UAV target tracking performance in obstructed urban settings.
• To enhance the robustness and continuity of target tracking, particularly during periods of lost target measurements.

Main Methods:

• The PENC framework optimizes UAV-gimbal control inputs in real-time to maintain target visibility within the camera's field of view (FOV).
• It employs a unique weight-switching mechanism to dynamically adjust noise covariance matrices (R and Q) in the estimator when measurements are lost, prioritizing prediction based on the target's dynamic model and historical data.
• Simulations were conducted in a virtual urban environment with representative vertical obstacles.

Main Results:

• PENC significantly outperforms conventional Nonlinear Model Predictive Control (NMPC) and NMPC with Extended Kalman Filtering (NMPC-EKF) in UAV target tracking.
• The framework demonstrates superior performance during target measurement loss, ensuring tracking continuity and robustness.
• Quantitative improvements include an increase in Target Visibility Percentage (TVP) by up to 14 percentage points, a reduction in Mean Recovery Time (MRT) to 0.03 seconds, and a decrease in prediction Root Mean Square Error (RMSE) during state loss.

Conclusions:

• The proposed PENC framework offers a substantial advancement in UAV target tracking for complex urban environments.
• Its adaptive nature in handling measurement loss provides robust and continuous tracking capabilities, crucial for mission success.
• PENC represents a significant improvement over existing methods, particularly in challenging scenarios involving occlusions and intermittent data.