Fleet Control using Coregionalized Gaussian Process Policy Iteration

TL;DR

This paper introduces a novel reinforcement learning approach using coregionalized Gaussian processes to improve control in fleet systems with similar but slightly different members, enhancing sample efficiency and performance.

Contribution

It presents a new method for transfer learning across fleet members by modeling member-specific dynamics with Gaussian processes, differing from traditional task-based transfer learning.

Findings

Significantly outperforms individual learning methods.

Achieves better median and variance in results.

Effective in wind farm control scenarios.

Abstract

In many settings, as for example wind farms, multiple machines are instantiated to perform the same task, which is called a fleet. The recent advances with respect to the Internet of Things allow control devices and/or machines to connect through cloud-based architectures in order to share information about their status and environment. Such an infrastructure allows seamless data sharing between fleet members, which could greatly improve the sample-efficiency of reinforcement learning techniques. However in practice, these machines, while almost identical in design, have small discrepancies due to production errors or degradation, preventing control algorithms to simply aggregate and employ all fleet data. We propose a novel reinforcement learning method that learns to transfer knowledge between similar fleet members and creates member-specific dynamics models for control. Our algorithm uses Gaussian processes to establish cross-member covariances. This is significantly different from standard transfer learning methods, as the focus is not on sharing information over tasks, but rather over system specifications. We demonstrate our approach on two benchmarks and a realistic wind farm setting. Our method significantly outperforms two baseline approaches, namely individual learning and joint learning where all samples are aggregated, in terms of the median and variance of the results.