Robust tracking control of uncertain autoloaders by implicit Lyapunov method and scleronomic Lagrangian mechanics-informed neural network.

Abstract

The autoloader is a key subsystem in modern main battle tanks, mainly responsible for ammunition transfer, loading, and resupply. However, it often suffers from uncertainties induced by base oscillations, leading to potential instability. While various control strategies have been proposed, most rely on prior knowledge of such oscillations. Additionally, model inaccuracies further challenge precise trajectory tracking. To address these issues, this paper proposes a novel trajectory tracking control strategy based on the computed torque method (CTM). A scleronomic Lagrangian mechanics-informed neural network is developed to approximate the inverse dynamics required by CTM. An implicit Lyapunov-based stabilizer is then designed to handle uncertainties from base oscillations. Furthermore, Lyapunov theory is used to prove the asymptotic stability of the closed-loop system. Several simulations and hardware experiments are conducted to demonstrate the effectiveness and robustness of the proposed control strategy, as well as its superiority over conventional approaches.