Improving Computational Cost of Bayesian Optimization for Controller Tuning with a Multi-stage Tuning Framework

TL;DR

This paper introduces a multi-stage framework that decomposes control tuning tasks into smaller subtasks, significantly reducing the computational cost and sample complexity of Bayesian optimization for high-dimensional control parameter spaces.

Contribution

The paper proposes a novel multi-stage control tuning framework that decomposes the problem, reducing sample complexity and computational time, validated on an underwater vehicle benchmark.

Findings

86% reduction in computational time

36% decrease in sample complexity

Effective for tuning multiple PID controllers

Abstract

Control auto-tuning for industrial and robotic systems, when framed as an optimization problem, provides an excellent means to tune these systems. However, most optimization methods are computationally costly, and this is problematic for high-dimension control parameter spaces. In this paper, we present a multi-stage control tuning framework that decomposes control tuning into subtasks, each with a reduced-dimension search space. We show formally that this framework reduces the sample complexity of the control-tuning task. We empirically validate this result by applying a Bayesian optimization approach to tuning multiple PID controllers in an unmanned underwater vehicle benchmark system. We demonstrate an 86\% decrease in computational time and 36\% decrease in sample complexity.