Safety-critical Receding-horizon Motion Planning and Containment Control of Autonomous Surface Vehicles via Neurodynamic Optimization

Lyu Guanghao, Zhouhua Peng, Wang Dan

Abstract:

This paper addresses the safety-certified motion planning and containment control of under-actuated autonomous
surface vehicles subject to model uncertainties, external disturbances, and input constraints in the presence of stationary and
moving obstacles. A three-level modular control architecture is proposed with a trajectory generation module at its planning
level, an adaptive guidance module at its guidance level, and a kinetic control module at its control level. Specifically, at the
planning level, a safety-certified containment trajectory generator is designed to generate safe trajectories over a rolling time
window to achieve containment formation and collision avoidance with neighboring ASVs, stationary obstacles, and moving
obstacles via dynamic control barrier functions and two-timescale neurodynamic optimization models. At the guidance level, an
adaptive line-of-sight guidance law is developed based on a finite-time predictor to estimate unknown sideslip angles and
generate guidance commands. At the control level, an optimal control law is designed based on finite-time neural predictors
and control Lyapunov functions for the autonomous surface vehicle with input constraints to follow the desired guidance
commands. The effectiveness and characteristics of the proposed method are demonstrated via simulations and hardware-in-theloop
experiments for cooperative exploration.

Submission Number: 123