The Ground Cost for Optimal Transport of Angular Velocity

Karthik Elamvazhuthi, Abhishek Halder

TL;DR

This paper tackles the problem of optimally steering the statistical distribution of a rigid body's angular velocity from an initial state to a desired final state within a fixed time. The authors derive a crucial component, the "ground cost," for this complex optimal transport problem, which has direct applications in spacecraft guidance and stabilization. They achieve this by leveraging a powerful analytical technique that bypasses the difficulties of traditional optimal control methods for nonlinear systems.

What Problem Does This Paper Address?

Imagine you need to precisely control the spin of a spacecraft, but you don't know its exact initial spin state – you only have a statistical description (a probability distribution). Furthermore, you need to reach a specific target spin state (also described statistically) by a hard deadline. This is the essence of the generalized optimal mass transport (GOMT) problem over nonlinear dynamics, specifically the controlled Euler equation governing rigid body angular velocity. While previous work has established the existence and numerical solutions for this problem, a key piece of the puzzle – the "ground cost" within the Kantorovich formulation of optimal transport – remained elusive. This ground cost is fundamental to defining the optimal "coupling" between initial and final probability distributions.

Methods

The core of this paper lies in deriving the ground cost for the GOMT problem. The authors frame this as solving a deterministic nonlinear optimal control problem. Instead of relying on Pontryagin's Minimum Principle, which becomes intractable for these nonlinear dynamics, they employ an analytical technique pioneered by Athans et al. This method cleverly uses variants of the Cauchy-Schwarz inequality to bound the optimal control cost.

A key step involves a change of coordinates to simplify the dynamics, transforming the problem into steering a state to zero. They then construct a specific feasible control that guides the system to the desired endpoint. Crucially, they then derive a cost inequality that bounds the cost of any feasible control. By showing that their constructed control saturates this bound under specific conditions (translated norm-invariant systems), they prove its optimality and thus determine the ground cost.

Datasets

Not discussed in the paper.

Key Insights and Ideas

  • The "Ground Cost" is Key: The paper highlights that the ground cost is the fundamental building block for the Kantorovich formulation of optimal transport, essentially defining the cost of moving "unit mass" between specific initial and final states.
  • Analytical Power of Cauchy-Schwarz: The authors demonstrate the remarkable power of Cauchy-Schwarz inequality variants in solving complex optimal control problems for nonlinear systems, bypassing the limitations of more traditional methods like Pontryagin's.
  • Generalizability to Norm-Invariant Systems: The analytical techniques developed are shown to be applicable not just to angular velocity dynamics but to a broader class of nonlinear systems exhibiting "translated norm-invariance," suggesting a wider impact.
  • Optimal Control as a Proxy for Ground Cost: The paper elegantly connects the ground cost of the optimal transport problem to the optimal value of a related deterministic optimal control problem.
  • Geodesics Aren't Always Straight Lines: For these nonlinear systems, the "optimal paths" (geodesics) in state space are not simple straight lines, even though the ground cost might have a similar form to classical optimal transport.

Key Results

  • The paper derives a bound for the Eulerian ground cost of the angular velocity dynamics.
  • They prove that the static GOMT problem (using the Kantorovich formulation) is well-posed under standard assumptions on endpoint probability distributions.
  • A specific feasible control is constructed that steers the system to the desired endpoint.
  • For a class of "translated norm-invariant" nonlinear systems, the authors derive a closed-form expression for the ground cost, which surprisingly matches the scaled Euclidean distance of classical optimal transport, but with nonlinear geodesics.

Limitations and Future Work

  • Suboptimality of General Feasible Controls: The paper notes that the constructed feasible control is not always optimal for general nonlinear systems, but it serves as a crucial tool for deriving cost inequalities.
  • Focus on Ground Cost: While the paper focuses on the ground cost, the full Kantorovich formulation and its numerical implementation for these nonlinear systems would be a natural next step.
  • Broader Applicability: The authors suggest that their analytical techniques could be applied to a wider range of GOMT problems, and exploring these further would be valuable.
  • Practical Implementation: While the theoretical framework is laid out, translating these findings into practical guidance laws for real-world systems like spacecraft would require further development.

Summary updated

Abstract

We revisit the optimal transport problem over angular velocity dynamics given by the controlled Euler equation. The solution of this problem enables stochastic guidance of spin states of a rigid body (e.g., spacecraft) over a hard deadline constraint by transferring a given initial state statistics to a desired terminal state statistics. This is an instance of generalized optimal transport over a nonlinear dynamical system. While prior work has reported existence-uniqueness and numerical solution of this dynamical optimal transport problem, here we present structural results about the equivalent Kantorovich a.k.a. optimal coupling formulation. Specifically, we focus on deriving the ground cost for the associated Kantorovich optimal coupling formulation. The ground cost is equal to the cost of transporting unit amount of mass from a specific realization of the initial or source joint probability measure to a realization of the terminal or target joint probability measure, and determines the Kantorovich formulation. Finding the ground cost leads to solving a structured deterministic nonlinear optimal control problem, which is shown to be amenable to an analysis technique pioneered by Athans et al. We show that such techniques have broader applicability in determining the ground cost (thus Kantorovich formulation) for a class of generalized optimal mass transport problems involving nonlinear dynamics with translated norm-invariant drift.
BibTeX
@misc{elamvazhuthi2025ground,
  author = {Karthik Elamvazhuthi and Abhishek Halder},
  title = {{The Ground Cost for Optimal Transport of Angular Velocity}},
  year = {2025},
  eprint = {2504.03190},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2504.03190},
}

Loading related content...

Figure