Kangle Bai, Kun Chen, Shengwei Sun, Hao Liu
Robotic fish, model predictive control, trajectory tracking, active deceleration strategy, random disturbances, sharp turn
This paper investigates the two-dimensional trajectory tracking control problem of a robotic fish under disturbances and sharp turn conditions. Aiming at the limitation that the traditional absolute error metric is greatly affected by the geometric scale of the trajectory, this paper adopts the normalised error metric. A novel active deceleration strategy-based model predictive control (ADS-MPC) method is proposed to solve the robustness challenge of trajectory tracking in complex environments. On the one hand, the method introduces control increment constraints in the roll optimisation framework to smooth the inputs, dynamically adjusts the reference speed with the trajectory curvature threshold, and prolongs the sharp turn response time to suppress trajectory offset. On the other hand, in order to confirm the tracking ability of the robotic fish in both disturbances-free and random disturbances environments, simulation experiments are conducted under square and two-dimensional spiral trajectories. The simulation results validate the efficacy of the control method.
Important Links:
Go Back
