Primal/dual descent methods for dynamics
Research output: Contribution to journal › Conference article › Research › peer-review
Standard
Primal/dual descent methods for dynamics. / Macklin, M.; Erleben, K.; Müller, M.; Chentanez, N.; Jeschke, S.; Kim, T. Y.
In: Computer Graphics Forum, Vol. 39, No. 8, 2020, p. 89-100.Research output: Contribution to journal › Conference article › Research › peer-review
Harvard
APA
Vancouver
Author
Bibtex
}
RIS
TY - GEN
T1 - Primal/dual descent methods for dynamics
AU - Macklin, M.
AU - Erleben, K.
AU - Müller, M.
AU - Chentanez, N.
AU - Jeschke, S.
AU - Kim, T. Y.
N1 - Publisher Copyright: © 2020 ACM. All rights reserved.
PY - 2020
Y1 - 2020
N2 - We examine the relationship between primal, or force-based, and dual, or constraint-based formulations of dynamics. Variational frameworks such as Projective Dynamics have proved popular for deformable simulation, however they have not been adopted for contact-rich scenarios such as rigid body simulation. We propose a new preconditioned frictional contact solver that is compatible with existing primal optimization methods, and competitive with complementarity-based approaches. Our relaxed primal model generates improved contact force distributions when compared to dual methods, and has the advantage of being differentiable, making it well-suited for trajectory optimization. We derive both primal and dual methods from a common variational point of view, and present a comprehensive numerical analysis of both methods with respect to conditioning. We demonstrate our method on scenarios including rigid body contact, deformable simulation, and robotic manipulation.
AB - We examine the relationship between primal, or force-based, and dual, or constraint-based formulations of dynamics. Variational frameworks such as Projective Dynamics have proved popular for deformable simulation, however they have not been adopted for contact-rich scenarios such as rigid body simulation. We propose a new preconditioned frictional contact solver that is compatible with existing primal optimization methods, and competitive with complementarity-based approaches. Our relaxed primal model generates improved contact force distributions when compared to dual methods, and has the advantage of being differentiable, making it well-suited for trajectory optimization. We derive both primal and dual methods from a common variational point of view, and present a comprehensive numerical analysis of both methods with respect to conditioning. We demonstrate our method on scenarios including rigid body contact, deformable simulation, and robotic manipulation.
KW - Contact
KW - Friction
KW - Numerical optimization
KW - Robotics
UR - http://www.scopus.com/inward/record.url?scp=85097342057&partnerID=8YFLogxK
U2 - 10.1111/cgf.14104
DO - 10.1111/cgf.14104
M3 - Conference article
AN - SCOPUS:85097342057
VL - 39
SP - 89
EP - 100
JO - Computer Graphics Forum
JF - Computer Graphics Forum
SN - 1467-8659
IS - 8
T2 - 2020 ACM SIGGRAPH/Eurographics Symposium on Computer Animation, SCA 2020
Y2 - 6 October 2020 through 9 October 2020
ER -
ID: 307086130