Trajectory optimization for target tracking in UUV-based multistatic sonar systems with position offset correction

This article presents a comprehensive framework for unmanned underwater vehicle (UUV) trajectory optimization to enhance target-tracking performance. Our work addresses a critical gap in existing methods by establishing a unified architecture that systematically accounts for common time offsets, UUV position drift, motion constraints, and safety. First, to mitigate these issues, an extended Kalman filter (EKF) is designed to fuse hybrid measurements of time difference of arrival (TDOA) and differential time delay (DTD), enabling simultaneous common time offset compensation and UUV position correction. Second, we derive the predicted conditional Cramér-Rao lower bound (PC-CRLB) for tracking errors, which serves as a performance criterion to directly guide waypoint selection. Third, for motion-constrained planning, we propose the spectral projected gradient method with nonmonotone line search and gradient direction evolution strategy (SPG-NLS-GDES), which allows for dynamic speed adjustment within the UUV’s physical limits and offers enhanced robustness. Finally, safety is enforced via the proposed alternating direction penalty method (ADPM) with soft margin (ADPM-SM), which leverages ADPM and incorporates a soft margin to dynamically adjust the obstacle warning zone. Extensive simulations demonstrate that the proposed framework improves both tracking accuracy and safety in UUV-based multistatic sonar target-tracking systems.