Learning References with Gaussian Processes in Model Predictive Control applied to Robot Assisted Surgery
This work addresses the challenge of improving control accuracy in robot-assisted surgery, but it is incremental as it applies an existing learning method to a specific domain.
The paper tackles the problem of model predictive controllers being shortsighted due to unknown external references by proposing to learn these references using Gaussian processes, applied to robot-assisted surgery where a robotic manipulator follows a learned reference position based on optical tracking measurements.
One of the key benefits of model predictive control is the capability of controlling a system proactively in the sense of taking the future system evolution into account. However, often external disturbances or references are not a priori known, which renders the predictive controllers shortsighted or uninformed. Adaptive prediction models can be used to overcome this issue and provide predictions of these signals to the controller. In this work we propose to learn references via Gaussian processes for model predictive controllers. To illustrate the approach, we consider robot assisted surgery, where a robotic manipulator needs to follow a learned reference position based on optical tracking measurements.