Application of Model Predictive Control for a Work Class Remotely Operated Vehicle*

The work-class Remotely Operated Vehicles (ROVs) are crucial for a broad range of underwater applications such as Inspection, Maintenance, and Repair (IMR) of subsea infrastructure, pipelines, and offshore wind farms. Despite the recent advances in the field of automation and robotics, the ROV operations are largely dependent on manual control. The common practice in the industry for controlling the ROV’s robotic arm involves using two pilots, one controlling the ROV and the other controlling the arm. To address this challenge, this paper applies nonlinear model predictive control (MPC) methodology to control the position of the tip of the robotic arm. To this effect, two separate MPC controllers are designed, one controlling the ROV and the other controlling the ROV’s manipulator. This configuration allows combining different control methodologies for controlling the ROV or its arm. It also facilitates a single pilot in the loop, controlling an arm or ROV while the other unit is controlled autonomously. Multiple configurations, with different controller types, including a unified MPC based controller for ROV plus its arm, are tested and compared using numerical simulations.