Disturbance-parameterised indirect adaptive resilient prescribed-performance control for ship heading tracking

This paper develops an indirect adaptive heading tracking control scheme for ships with unknown model coefficients and equivalent rudder disturbances. Stable disturbance-parameterisation filters convert structured forcing into unknown constant coefficients, which are collected with the physical model parameters in a measurable regression. Identical second-order filters provide the regression and its derivative without a yaw-acceleration measurement. A projected, normalised indirect adaptive law is developed using the filtered model-prediction error. To accommodate temporary violations of the nominal performance specification, smooth algebraic resilient boundaries and a completed error coordinate are proposed. The construction has explicit positive distance and Jacobian lower bounds and returns exactly to conventional prescribed performance control inside an alarm core. A two-step dynamic-surface design establishes regional practical boundedness of the closed-loop system. Numerical comparisons demonstrate algebraic envelope recovery and prediction-based disturbance reconstruction, and characterise the trade-offs in requested input and sensitivity to measurement noise.