Control-oriented modelling and disturbance-adaptive prescribed-performance control for turtle-inspired underwater robots

Turtle-inspired underwater robots exhibit excellent manoeuvrability and propulsion efficiency, but their flipper-driven locomotion introduces strong nonlinear coupling, hydrodynamic uncertainty, and external disturbance sensitivity. This paper develops a control-oriented six-degree-of-freedom dynamic model that captures propulsion-induced oscillatory motion and underactuated coupling characteristics. Based on the developed model, a prescribed-performance disturbance-adaptive control framework is proposed by integrating a disturbance-aware integral bidirectional fuzzy brain emotional learning (DA-IBFBEL) controller with a prescribed-performance integral disturbance observer. The observer simultaneously estimates system states and lumped disturbances, while the controller incorporates disturbance information into the emotional-learning process to improve adaptation and robustness. Simulation studies under ocean-current disturbances, modelling uncertainties, measurement noise, and actuator degradation demonstrate that the proposed approach achieves superior tracking accuracy, disturbance rejection, and robustness compared with PID and current BFBEL plus linear extended state observer controllers. The results verify the effectiveness of the proposed modelling and control framework for turtle-inspired underwater robots operating in uncertain underwater environments.