Event-triggered prescribed performance control for safety-aware and resource-efficient isoline tracking of autonomous underwater vehicle

In complex and GPS-denied marine environments, autonomous underwater vehicles (AUVs) face significant safety risks that can lead to mission failure. Existing methods often struggle to balance tracking accuracy, obstacle-avoidance safety, and resource efficiency under dynamic uncertainties. Traditional trajectory tracking control approaches lack adaptive safety constraints, while prescribed performance control (PPC) imposes strict initial state requirements and high communication overhead. To address these challenges, this paper proposes an event-triggered modified prescribed performance control (ET-MPPC) scheme for isoline tracking control. This scheme combines a gradient-free isoline tracker that relies solely on real-time odometry with a hyperbolic-cosecant-based modified prescribed performance control law, thereby relaxing the need for precise initial state knowledge. To handle dynamic uncertainty, a modified echo state neural network (MESNN) is introduced to achieve low-cost disturbance compensation. Additionally, a relative threshold-based event-triggered control (RTETC) dynamically activates control updates, eliminating Zeno behavior while reducing communication overhead. Finally, the effectiveness of the proposed control scheme is verified through theoretical analysis and numerical simulations.