[Paper Review] Mean-field limit for collective behavior models with sharp sensitivity regions

José A. Carrillo, Young-Pil Choi+2 more|arXiv (Cornell University)|Oct 8, 2015

Mathematical Biology Tumor GrowthMathematics28 references44 citations

TL;DR

This paper rigorously establishes the mean-field limit for collective behavior models with sharp sensitivity regions—such as vision cones or discontinuous communication weights—by introducing a differential inclusion framework to handle non-smooth particle dynamics. It proves quantitative convergence of particle systems to kinetic equations via optimal transport and weak-strong stability estimates, with Lipschitz continuity of the velocity field under geometric assumptions on sensitivity sets.

ABSTRACT

We rigorously show the mean-field limit for a large class of swarming individual based models with local sharp sensitivity regions. For instance, these models include nonlocal repulsive-attractive forces locally averaged over sharp vision cones and Cucker-Smale interactions with discontinuous communication weights. We construct global-in-time defined notion of solutions through a differential inclusion system corresponding to the particle descriptions. We estimate the error between the solutions to the differential inclusion system and weak solutions to the expected limiting kinetic equation by employing tools from optimal transport theory. Quantitative bounds on the expansion of the 1-Wasserstein distance along flows based on a weak-strong stability estimate are obtained. We also provide different examples of realistic sensitivity sets satisfying the assumptions of our main results.

Motivation & Objective

• To rigorously derive the mean-field limit for individual-based models of collective motion with local, sharp sensitivity regions such as vision cones.
• To address the challenge of non-smooth, discontinuous interaction regions that prevent classical mean-field limit arguments.
• To establish a well-defined solution concept for particle systems with discontinuous right-hand sides via Filippov's differential inclusion theory.
• To prove quantitative convergence of the empirical measure of N particles to the solution of the limiting kinetic equation.
• To generalize the result to both Cucker-Smale-type alignment and repulsive-attractive potential models with sharp sensing regions.

Proposed method

• Formalize particle dynamics with sharp sensitivity regions using a differential inclusion system to handle discontinuous forces.
• Employ optimal transport theory and 1-Wasserstein distance to measure the error between particle and kinetic solutions.
• Establish weak-strong stability estimates for the kinetic equation by proving the velocity field is locally Lipschitz under geometric assumptions on sensitivity sets.
• Use regularization techniques and flow maps to prove existence and compact support of weak solutions to the kinetic equation.
• Apply geometric assumptions (H1)-(H2) on sensitivity sets K(v), including uniform regularity and controlled variation with velocity.
• Generalize the framework to both Cucker-Smale alignment and repulsive-attractive potential models via a unified force formulation.

Experimental results

Research questions

• RQ1 Can the mean-field limit be rigorously justified for particle systems with discontinuous, velocity-dependent sensitivity regions?
• RQ2 How can a well-defined solution concept be constructed for particle systems with non-Lipschitz, discontinuous dynamics?
• RQ3 What conditions on the shape and variation of sensitivity regions ensure the velocity field in the kinetic limit is sufficiently regular?
• RQ4 Can quantitative convergence rates be established between the empirical measure of N particles and the solution of the limiting kinetic equation?
• RQ5 To what extent can the framework be extended to general locally averaged interaction models with sharp sensing zones?

Key findings

• The paper establishes the mean-field limit for collective behavior models with sharp sensitivity regions, proving convergence of particle systems to the expected kinetic equation.
• A global-in-time solution concept for the particle system is constructed via Filippov's differential inclusion theory, resolving the issue of discontinuous dynamics.
• The velocity field in the kinetic equation is shown to be locally Lipschitz continuous under assumptions (H1)-(H2) on the sensitivity sets, enabling weak-strong stability estimates.
• A quantitative bound on the 1-Wasserstein distance between the particle system and the kinetic solution is derived: d₁(f(t), µ_N(t)) ≤ e^{Ct} d₁(f(0), µ_N(0)), with C depending on T, f₀, and d.
• The results are extended to both Cucker-Smale-type alignment and repulsive-attractive potential models with sharp sensing zones.
• The framework applies to realistic sensitivity sets such as conical vision zones and lateral line sensing in fish, satisfying the required geometric assumptions.