Optimal Control of Dual Buoyancy Controllers for Underwater Walking on Various Slopes

This study proposes an optimal control methodology for dual active buoyancy controllers (ABCs) to achieve stable slope walking using an underwater legged robot. Utilizing an underwater walking-capable robot platform equipped with multiple ABCs, we simplified the leg models based on walking phases and developed models for walking moments and slip conditions according to the slope angle. Based on these models, we defined and solved optimization problems and proposed a control strategy that integrates the optimized results. The proposed buoyancy control method was quantitatively validated in a simulation environment under various slope terrains. The robot simulation results in the margin zone demonstrated stable slope walking trajectories comparable to those on flat terrain.