Physics-informed Gaussian Processes as Linear Model Predictive Controller with Constraint Satisfaction

TL;DR

This paper introduces a physics-informed Gaussian Process-based Model Predictive Control method that guarantees constraint satisfaction through Hamiltonian Monte Carlo sampling and Gaussian Process optimization, maintaining the ODE structure.

Contribution

It provides a novel approach combining physics-informed Gaussian Processes with sampling-based guarantees for safe control in MPC, including the use of the Matérn kernel.

Findings

Guarantees of constraint satisfaction are achieved.

Closed-form Gaussian Process optimization is demonstrated.

Incorporation of the Matérn kernel enhances inference.

Abstract

Model Predictive Control evolved as the state of the art paradigm for safety critical control tasks. Control-as-Inference approaches thereof model the constrained optimization problem as a probabilistic inference problem. The constraints have to be implemented into the inference model. A recently introduced physics-informed Gaussian Process method uses Control-as-Inference with a Gaussian likelihood for state constraint modeling, but lacks guarantees of open-loop constraint satisfaction. We mitigate the lack of guarantees via an additional sampling step using Hamiltonian Monte Carlo sampling in order to obtain safe rollouts of the open-loop dynamics which are then used to obtain an approximation of the truncated normal distribution which has full probability mass in the safe area. We provide formal guarantees of constraint satisfaction while maintaining the ODE structure of the Gaussian Process on a discretized grid. Moreover, we show that we are able to perform optimization of a quadratic cost function by closed form Gaussian Process computations only and introduce the Mat\'ern kernel into the inference model.