Low computational complexity model predictive control for high-speed trimaran motion reduction

To address the high computational complexity of model predictive control (MPC) in trimaran motion reduction, this paper develops a low-complexity predictive control framework that integrates three key mechanisms. First, within the control horizon, the control input is constructed as the sum of an optimal baseline and auxiliary variables, ensuring constraint satisfaction and smooth switching. Second, the auxiliary variables are parameterized by Laguerre functions, thereby compressing the optimization dimension. Third, a double-constraint triggering mechanism based on actuator magnitude and rate adaptively reduces optimization frequency. With appropriately designed terminal constraints, closed-loop input-to-state stability (ISS) is established via Lyapunov theory. Simulations under Sea State 4 demonstrate that the proposed method reduces the root-mean-square (RMS) value of heave displacement by 58.61% and pitch angle by 50.27% compared to uncontrolled baselines. Experimental tests on a 1:10 scaled trimaran further confirm its practical effectiveness. Importantly, the controller achieves real-time feasibility with a computation time of 36 ms for a 100-step horizon, well within the 50 ms control cycle.