Toward Federated DeePC: borrowing data from similar systems

TL;DR

This paper introduces a federated extension of Data-enabled Predictive Control (DeePC) that utilizes data from multiple similar systems to enhance control performance, highlighting benefits and potential drawbacks through numerical examples.

Contribution

It proposes a novel federated DeePC framework that leverages data from similar systems, expanding the applicability of data-driven predictive control methods.

Findings

Federated DeePC can improve control accuracy using data from similar systems.

Sharing data across systems offers potential benefits but also introduces privacy and compatibility challenges.

Numerical examples demonstrate the effectiveness and limitations of the proposed approach.

Abstract

Data-driven predictive control approaches, in general, and Data-enabled Predictive Control (DeePC), in particular, exploit matrices of raw input/output trajectories for control design. These data are typically gathered only from the system to be controlled. Nonetheless, the increasing connectivity and inherent similarity of (mass-produced) systems have the potential to generate a considerable amount of information that can be exploited to undertake a control task. In light of this, we propose a preliminary federated extension of DeePC that leverages a combination of input/output trajectories from multiple similar systems for predictive control. Supported by a suite of numerical examples, our analysis unveils the potential benefits of exploiting information from similar systems and its possible downsides.