MPC-guided Imitation Learning of Bayesian Neural Network Policies for the Artificial Pancreas

Abstract

Even though model predictive control (MPC) is currently the main algorithm for insulin control in the artificial pancreas (AP), it usually requires complex online optimizations, which are infeasible for resource-constrained medical devices. MPC also typically relies on state estimation, an error-prone process. In this paper, we introduce a novel approach to AP control that uses Imitation Learning to synthesize neural-network insulin policies from MPC-computed demonstrations. Such policies are computationally efficient and, by instrumenting MPC at training time with full state information, they can directly map measurements into optimal therapy decisions, thus bypassing state estimation. We apply Bayesian inference via Monte Carlo Dropout to learn policies, which allows us to quantify prediction uncertainty and thereby derive safer therapy decisions. We show that our control policies trained under a specific patient model readily generalize (in terms of model parameters and disturbance distributions) to patient cohorts, consistently outperforming traditional MPC with state estimation.

An STL-based Approach to Resilient Control for Cyber-Physical Systems

“…The controller presented in this paper can be run repeatedly in simulation mode to provide the training data for the NC. Approaches of this nature can be found in [7,19]. Such an NC is expected to improve upon the execution time of the MPC controller (which has to solve a multi-objective MILP problem at every time step) for the package-delivery case study by orders of magnitude.…”