Trajectory tracking control of robotic fish in offshore disturbance environments via disturbance observer-based inverse sliding mode

Considering the complex offshore environment, where small robotic fish are exposed to surface disturbances from wave forces, suspended particles, and model uncertainties, as well as deep-water disturbances from bottom currents, this paper proposes an inverse sliding mode control strategy based on a disturbance observer to enhance trajectory tracking robustness. The research focuses on a tail-fin-driven bionic robotic fish. Utilizing its kinematic and dynamic models, a virtual control law based on an agent dynamics model is proposed, with the tail fin integrated as a real control input. A nonlinear disturbance observer with controllable convergence characteristics is developed to estimate and compensate for disturbances, including wave forces and model uncertainties. Additionally, a velocity error correction function is introduced to mitigate the impact of strong disturbances. Based on Lyapunov theory, an adaptive sliding mode control law is derived to ensure system stability. The control law for the caudal fin swing angle and bias is obtained by inverting the virtual control inputs, applied to the robotic fish’s accurate model. Numerical simulations show that the disturbance observer’s tracking error remains below 5%, and the trajectory tracking error is within 0.1 meters, representing only 2.2% of the robotic fish’s body length. Compared to mainstream control methods, the proposed approach significantly enhances robustness in contrast to the conventional sliding mode control with observers, and exhibits substantially smaller tracking errors, especially during trajectory transitions.