Optimization-Based Actuator Allocation for Underwater Vehicles with Variable Buoyancy Systems

Abstract

Underwater vehicles are often equipped with variable buoyancy systems (VBS) for buoyancy variation. This also applies to the novel unmanned underwater vehicle (UUV) class MUM (Large Modifiable Underwater Mothership) which features a modular structure and variable shape, thus requiring a generalized motion control approach that adapts to the specific vehicle configuration. Regarding MUM's maneuvering devices, the control system generalization is achieved by means of actuator allocation which distributes the motion controller output to the available actuators. Consequently, this paper proposes an actuator allocation approach that treats conventional propulsion units and VBS jointly. Inspired by optimization-based solutions known from dynamically positioning ships, the selected allocation method employs a problem formulation based on model predictive control (MPC) in order to optimize usage of the actuators subject to their characteristics and physical limits. The application of the allocation algorithm to an actuator configuration with VBS is presented and results from early trials in simulation are shown.