A linear finite-difference scheme for approximating randers distances on cartesian grids
ESAIM: Control, Optimisation and Calculus of Variations, Tome 28 (2022), article no. 45

Randers distances are an asymmetric generalization of Riemannian distances, and arise in optimal control problems subject to a drift term, among other applications. We show that Randers eikonal equation can be approximated by a logarithmic transformation of an anisotropic second order linear equation, generalizing Varadhan’s formula for Riemannian manifolds. Based on this observation, we establish the convergence of a numerical method for computing Randers distances, from point sources or from a domain’s boundary, on Cartesian grids of dimension 2 and 3, which is consistent at order 2/3, and uses tools from low-dimensional algorithmic geometry for best efficiency. We also propose a numerical method for optimal transport problems whose cost is a Randers distance, exploiting the linear structure of our discretization and generalizing previous works in the Riemannian case. Numerical experiments illustrate our results.

DOI : 10.1051/cocv/2022043
Classification : 65N06, 65N12, 49L25
Keywords: Randers metric, Varadhan’s formula, Hamilton-Jacobi equation, viscosity solutions, finite-difference scheme, convergence analysis
@article{COCV_2022__28_1_A45_0,
     author = {Bonnans, J. Fr\'ed\'eric and Bonnet, Guillaume and Mirebeau, Jean-Marie},
     title = {A linear finite-difference scheme for approximating randers distances on cartesian grids},
     journal = {ESAIM: Control, Optimisation and Calculus of Variations},
     year = {2022},
     publisher = {EDP-Sciences},
     volume = {28},
     doi = {10.1051/cocv/2022043},
     mrnumber = {4448807},
     zbl = {1504.65235},
     language = {en},
     url = {https://www.numdam.org/articles/10.1051/cocv/2022043/}
}
TY  - JOUR
AU  - Bonnans, J. Frédéric
AU  - Bonnet, Guillaume
AU  - Mirebeau, Jean-Marie
TI  - A linear finite-difference scheme for approximating randers distances on cartesian grids
JO  - ESAIM: Control, Optimisation and Calculus of Variations
PY  - 2022
VL  - 28
PB  - EDP-Sciences
UR  - https://www.numdam.org/articles/10.1051/cocv/2022043/
DO  - 10.1051/cocv/2022043
LA  - en
ID  - COCV_2022__28_1_A45_0
ER  - 
%0 Journal Article
%A Bonnans, J. Frédéric
%A Bonnet, Guillaume
%A Mirebeau, Jean-Marie
%T A linear finite-difference scheme for approximating randers distances on cartesian grids
%J ESAIM: Control, Optimisation and Calculus of Variations
%D 2022
%V 28
%I EDP-Sciences
%U https://www.numdam.org/articles/10.1051/cocv/2022043/
%R 10.1051/cocv/2022043
%G en
%F COCV_2022__28_1_A45_0
Bonnans, J. Frédéric; Bonnet, Guillaume; Mirebeau, Jean-Marie. A linear finite-difference scheme for approximating randers distances on cartesian grids. ESAIM: Control, Optimisation and Calculus of Variations, Tome 28 (2022), article no. 45. doi: 10.1051/cocv/2022043

[1] S. Alama, L. Bronsard and J. A. Montero, On the Ginzburg-Landau model of a superconducting ball in a uniform field. Annales de l'IHP Analyse non linéaire (2006) 237–267. | MR | Zbl | Numdam | DOI

[2] D. Bao, C. Robles, Z. Shen et al., Zermelo navigation on Riemannian manifolds. J. Differ. Geometry 66 (2004) 377–435. | MR | Zbl

[3] M. Bardi and I. Capuzzo Dolcetta, Optimal Control and Viscosity Solutions of Hamilton-Jacobi-Bellman Equations, Modern Birkhäuser Classics, Birkhauser, Basel (1997). | Zbl | DOI

[4] G. Barles and B. Perthame, Exit time problems in optimal control and vanishing viscosity method. SIAM J. Control Optim. 26 (1988) 1133–1148. | MR | Zbl | DOI

[5] G. Barles and E. Rouy, A Strong Comparison Result for the Bellman equation arising in stochastic exit time control problems and its applications. Comm. Partial Differ. Equ. 23 (1998) 1995–2033. | MR | Zbl | DOI

[6] G. Barles and P. E. Souganidis, Convergence of approximation schemes for fully nonlinear second order equations. Asymptotic Anal. 4 (1991) 271–283. | MR | Zbl | DOI

[7] J.-D. Benamou, G. Carlier and R. Hatchi, A numerical solution to Monge’s problem with a Finsler distance as cost. ESAIM: MMNA 52 (2018) 2133–2148. | MR | Zbl | Numdam | DOI

[8] R. J. Berman, The Sinkhorn algorithm, parabolic optimal transport and geometric Monge-Ampère equations. Numer. Math. 145 (2020) 771–836. | MR | Zbl | DOI

[9] F. Bonnans and S. Gaubert, Recherche opérationnelle. Aspects mathématiques et applications. Ellipse (2016).

[10] J. Bonnans, G. Bonnet and J.-M. Mirebeau, Monotone and second order consistent scheme for the two dimensional Pucci equation (2020). | MR | Zbl

[11] J. F. Bonnans, G. Bonnet and J.-M. Mirebeau, Second order monotone finite differences discretization of linear anisotropic differential operators. Math. Comput. 90 (2021) 2671–2703. | MR | Zbl

[12] O. P. Bruno and A. Prieto, Spatially dispersionless, unconditionally stable FC-AD solvers for variable-coefficient PDEs. J. Sci. Comput. 58 (2014) 331–366. | MR | Zbl | DOI

[13] E. Casas, Control of an elliptic problem with pointwise state constraints. SIAM J. Control Optim. 24 (1986) 1309–1318. | MR | Zbl | DOI

[14] D. Chen, J.-M. Mirebeau and L. D. Cohen, Global minimum for a Finsler Elastica minimal path approach. Int. J. Comput. Vis. 122 (2017) 458–483. | MR | Zbl | DOI

[15] Y. Chen, E. Fan and M. Yuen, The Hopf-Cole transformation, topological solitons and multiple fusion solutions for the n -dimensional Burgers system. Phys. Lett. A 380 (2016) 9–14. | MR | Zbl | DOI

[16] X. Cheng and Z. Shen, Finsler geometry, An approach via Randers spaces (2012). | MR | Zbl

[17] L. Chizat, P. Roussillon, F. Léger, F. X. Vialard and G. Peyré, Faster wasserstein distance estimation with the Sinkhorn divergence. Adv. Neural Inf. Process. Syst. 33 (2020).

[18] L. D. Cohen, D. Chen and J.-M. Mirebeau, Finsler geodesics evolution model for region based active contours, in Proceedings of the British Machine Vision Conference (BMVC), edited by E. R. H. Richard C. Wilson and W. A. P. Smith. BMVA Press (2016) 22.1–22.12.

[19] M. B. Cohen, J. Kelner, R. Kyng, J. Peebles, R. Peng, A. B. Rao and A. Sidford, Solving directed laplacian systems in nearly- linear time through sparse LU factorizations, in 2018 IEEE 59th Annual Symposium on Foundations of Computer Science (FOCS), IEEE (2018) 898–909. | MR | Zbl | DOI

[20] J. H. Conway and N. J. A. Sloane, Low-dimensional lattices. VI. Voronoi reduction of three-dimensional lattices. Proc. R. Soc. A 436 (1992) 55–68. | MR | Zbl

[21] M. G. Crandall, H. Ishii and P.-L. Lions, User’s guide to viscosity solutions of second order partial differential equations. Bull. Amer. Math. Soc. 27 (1992) 1–67. | MR | Zbl | DOI

[22] K. Crane, M. Livesu, E. Puppo and Y. Qin, A Survey of Algorithms for Geodesic Paths and Distances. Preprint (2020). | arXiv

[23] K. Crane, C. Weischedel and M. Wardetzky, Geodesics in heat: a new approach to computing distance based on heat flow. ACM Trans. Graph. 32 (2013) 152:1–152:11. | DOI

[24] K. Crane, C. Weischedel and M. Wardetzky, The heat method for distance computation. Commun. ACM 60 (2017) 90–99. | DOI

[25] M. Cuturi, Sinkhorn distances: lightspeed computation of optimal transport, in Proc. 26th International Conference on Neural Information Processing Systems — Volume 2 (2013) 2292–2300.

[26] C. De Lellis, F. Otto and M. Westdickenberg, Minimal entropy conditions for Burgers equation. Quart. Appl. Math. 62 (2004) 687–700. | MR | Zbl | DOI

[27] R. Duits, S. P. Meesters, J.-M. Mirebeau and J. M. Portegies, Optimal paths for variants of the 2D and 3D Reeds-Shepp car with applications in image analysis. J. Math. Imag. Vision (2018) 1–33. | MR | Zbl

[28] A. Ern and J.-L. Guermond, Theory and practice of finite elements, vol. 159. Springer Science and Business Media (2013). | MR | Zbl

[29] J. Fehrenbach and J.-M. Mirebeau, Sparse non-negative stencils for anisotropic diffusion. J. Math. Imag. Vision 49 (2014) 123–147. | MR | Zbl | DOI

[30] M. Feldman and R. J. Mccann, Uniqueness and transport density in Monge’s mass transportation problem. Calc. Variat. Partial Differ. Equ. 15 (2002) 81–113. | MR | Zbl | DOI

[31] E. Hopf, The partial differential equation u t + u u x = μ x x . Commun. Pure Appl. Math. 3 (1950) 201–230. | MR | Zbl | DOI

[32] P. Houston, I. Muga, S. Roggendorf and K.G. Van Der Zee, The convection-diffusion-reaction equation in non-Hilbert Sobolev spaces: a direct proof of the inf-sup condition and stability of Galerkin’s method. Comput. Methods Appl. Math. 19 (2019) 503–522. | MR | Zbl | DOI

[33] R. Kannan and Z. J. Wang, A high order spectral volume solution to the Burgers’ equation using the Hopf-Cole transformation. Int. J. Numer. Methods Fluids 69 (2012) 781–801. | MR | Zbl | DOI

[34] H. Komiya, Elementary proof for Sion’s minimax theorem. Kodai Math. J. 11 (1988) 5–7. | MR | Zbl | DOI

[35] F. Labelle and J. R. Shewchuk, Anisotropic voronoi diagrams and guaranteed-quality anisotropic mesh generation, in Proceedings of the nineteenth annual symposium on Computational geometry (2003) 191–200. | Zbl | DOI

[36] C. Léonard, A survey of the Schrodinger problem and some of its connections with optimal transport. Discr. Continu. Dyn. Syst. 34 (2014) 1533. | MR | Zbl | DOI

[37] T.-T. Lu and S.-H. Shiou, Inverses of 2 x 2 block matrices. Comput. Math. Appl. 43 (2002) 119–129. | MR | Zbl | DOI

[38] X.-N. Ma, N. S. Trudinger and X.-J. Wang, Regularity of potential functions of the optimal transportation problem. Arch. Ratl. Mech. Anal. 177 (2005) 151–183. | MR | Zbl | DOI

[39] J.-M. Mirebeau, Efficient fast marching with Finsler metrics. Numer. Math. 126 (2014) 515–557. | MR | Zbl | DOI

[40] J.-M. Mirebeau, Minimal stencils for discretizations of anisotropic PDEs preserving causality or the maximum principle. SIAM J. Numer. Anal. 54 (2016) 1582–1611. | MR | Zbl | DOI

[41] J.-M. Mirebeau, Fast-marching methods for curvature penalized shortest paths. J. Math. Imag. Vision 60 (2018) 784–815. | MR | Zbl | DOI

[42] J.-M. Mirebeau, Riemannian fast-marching on cartesian grids, using voronoi’s first reduction of quadratic forms. SIAM J. Numer. Anal. 57 (2019) 2608–2655. | MR | Zbl | DOI

[43] A. M. Oberman, Convergent difference schemes for degenerate elliptic and parabolic equations: Hamilton-Jacobi equations and free boundary problems. SIAM J. Numer. Anal. 44 (2006) 879–895. | MR | Zbl | DOI

[44] S.-I. Ohta and K.-T. Sturm, Heat flow on Finsler manifolds. Commun. Pure Appl. Math. 62 (2009) 1386–1433. | MR | Zbl | DOI

[45] T. Ohwada, Cole-Hopf transformation as numerical tool for the Burgers equation. Appl. Comput. Math 8 (2009) 107–113. | MR | Zbl

[46] G. Randers, On an asymmetrical metric in the four-space of general relativity. Phys. Rev. 59 (1941) 195–199. | MR | Zbl | DOI

[47] E. Selling, Uber die Binaren und Ternären Quadratischen Formen. J. Reine Angew. Math. 77 (1874) 143–229. | MR | JFM

[48] J. A. Sethian and A. B. Vladimirsky, Ordered upwind methods for static Hamilton-Jacobi equations. Proc. Natl. Acad. Sci. USA 98 (2001) 11069–11074. | MR | Zbl | DOI

[49] R. Sinkhorn, A relationship between arbitrary positive matrices and doubly stochastic matrices. Ann. Math. Stat. 35 (1964) 876–879. | MR | Zbl | DOI

[50] J. Solomon, F. De Goes, G. Peyré, M. Cuturi, A. Butscher, A. Nguyen, T. Du and L. Guibas, Convolutional Wasserstein distances: efficient optimal transportation on geometric domains. ACM Trans. Graph. 34 (2015) 66:1–66:11. | Zbl | DOI

[51] J. Solomon, R. Rustamov, L. Guibas and A. Butscher, Earth mover’s distances on discrete surfaces. ACM Trans. Graphics 33 (2014) 1–12. | Zbl | DOI

[52] S. R. S. Varadhan, On the behavior of the fundamental solution of the heat equation with variable coefficients. Comm. Pure Appl. Math. 20 (1967) 431–455. | MR | Zbl | DOI

[53] C. Villani, Optimal transport: old and new, vol. 338. Springer (2009). | MR | Zbl | DOI

[54] F. Yang, L. Chai, D. Chen and L. D. Cohen, Geodesic via asymmetric heat diffusion based on Finsler metric, in Asian Conference on Computer Vision. Springer (2018) 371–386.

[55] F. Yang and L. D. Cohen, Geodesic distance and curves through isotropic and anisotropic heat equations on images and surfaces. J. Math. Imag. Vision 55 (2016) 210–228. | MR | Zbl | DOI

Cité par Sources :

The first author was partially supported by the FiME Lab Research Initiative (Institut Europlace de Finance).