Saved in:
Bibliographic Details
Main Authors: Jadbabaie, Ali, Shah, Devavrat, Sinclair, Sean R.
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2408.04488
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916566537338880
author Jadbabaie, Ali
Shah, Devavrat
Sinclair, Sean R.
author_facet Jadbabaie, Ali
Shah, Devavrat
Sinclair, Sean R.
contents The framework of decision-making, modeled as a Markov Decision Process (MDP), typically assumes a single objective. However, practical scenarios often involve tradeoffs between multiple objectives. We address this in the Linear Quadratic Regulator (LQR), a canonical continuous, infinite horizon MDP. First, we establish that the Pareto front for LQR is characterized by linear scalarization: a convex combination of objectives recovers all tradeoff points, making multi-objective LQR reducible to single-objective problems. This highlights an important instance where linear scalarization suffices for a non-convex problem. Second, we show the Pareto front is smooth, in that an $ε$ perturbation of a scalarization parameter yields an $ε$ approximation to the objective. These results inspire a simple algorithm to approximate the Pareto front via grid search over scalarization parameters, where each optimization problem retains the computational efficiency of single-objective LQR. Lastly, we extend the analysis to certainty equivalence, where unknown dynamics are replaced with estimates.
format Preprint
id arxiv_https___arxiv_org_abs_2408_04488
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Multi-Objective LQR with Linear Scalarization
Jadbabaie, Ali
Shah, Devavrat
Sinclair, Sean R.
Optimization and Control
The framework of decision-making, modeled as a Markov Decision Process (MDP), typically assumes a single objective. However, practical scenarios often involve tradeoffs between multiple objectives. We address this in the Linear Quadratic Regulator (LQR), a canonical continuous, infinite horizon MDP. First, we establish that the Pareto front for LQR is characterized by linear scalarization: a convex combination of objectives recovers all tradeoff points, making multi-objective LQR reducible to single-objective problems. This highlights an important instance where linear scalarization suffices for a non-convex problem. Second, we show the Pareto front is smooth, in that an $ε$ perturbation of a scalarization parameter yields an $ε$ approximation to the objective. These results inspire a simple algorithm to approximate the Pareto front via grid search over scalarization parameters, where each optimization problem retains the computational efficiency of single-objective LQR. Lastly, we extend the analysis to certainty equivalence, where unknown dynamics are replaced with estimates.
title Multi-Objective LQR with Linear Scalarization
topic Optimization and Control
url https://arxiv.org/abs/2408.04488