-
Constrained Bayesian optimization with merit functions
Authors:
J. Wang,
C. G. Petra,
J. L. Peterson
Abstract:
Bayesian optimization is a powerful optimization tool for problems where native first-order derivatives are unavailable. Recently, constrained Bayesian optimization (CBO) has been applied to many engineering applications where constraints are essential. However, several obstacles remain with current CBO algorithms that could prevent a wider adoption. We propose CBO algorithms using merit functions…
▽ More
Bayesian optimization is a powerful optimization tool for problems where native first-order derivatives are unavailable. Recently, constrained Bayesian optimization (CBO) has been applied to many engineering applications where constraints are essential. However, several obstacles remain with current CBO algorithms that could prevent a wider adoption. We propose CBO algorithms using merit functions, such as the penalty merit function, in acquisition functions, inspired by nonlinear optimization methods, e.g., sequential quadratic programming. Merit functions measure the potential progress of both the objective and constraint functions, thus increasing algorithmic efficiency and allowing infeasible initial samples. The acquisition functions with merit functions are relaxed to have closed forms, making its implementation readily available wherever Bayesian optimization is. We further propose a unified CBO algorithm that can be seen as extension to the popular expected constrained improvement (ECI) approach. We demonstrate the effectiveness and efficiency of the proposed algorithms through numerical experiments on synthetic problems and a practical data-driven engineering design problem in the field of plasma physics.
△ Less
Submitted 19 March, 2024;
originally announced March 2024.
-
Structure of Periodic Orbit Families in the Hill Restricted 4-Body Problem
Authors:
Gavin M. Brown,
Luke T. Peterson,
Damennick B. Henry,
Daniel J. Scheeres
Abstract:
The Hill Restricted 4-Body Problem (HR4BP) is a coherent time-periodic model that can be used to represent motion in the Sun-Earth-Moon (SEM) system. Periodic orbits were computed in this model to better understand the periodic orbit family structures that exist in these types of systems. First, periodic orbits in the Circular Restricted 3-Body Problem (CR3BP) representation of the Earth-Moon (EM)…
▽ More
The Hill Restricted 4-Body Problem (HR4BP) is a coherent time-periodic model that can be used to represent motion in the Sun-Earth-Moon (SEM) system. Periodic orbits were computed in this model to better understand the periodic orbit family structures that exist in these types of systems. First, periodic orbits in the Circular Restricted 3-Body Problem (CR3BP) representation of the Earth-Moon (EM) system were identified. A Melnikov-type function was used to identify a set of candidate points on the EM CR3BP periodic orbits to start a continuation algorithm. A pseudo-arclength continuation scheme was then used to obtain the corresponding periodic orbit families in the HR4BP when including the effect of the Sun. Bifurcation points were identified in the computed families to obtain additional orbit families.
△ Less
Submitted 29 February, 2024;
originally announced February 2024.
-
Dynamics Around the Earth-Moon Triangular Points in the Hill Restricted 4-Body Problem
Authors:
Luke T. Peterson,
Gavin Brown,
Àngel Jorba,
Daniel Scheeres
Abstract:
This paper investigates the motion of a small particle moving near the triangular points of the Earth-Moon system. The dynamics are modeled in the Hill restricted 4-body problem (HR4BP), which includes the effect of the Earth and Moon as in the circular restricted 3-body problem (CR3BP), as well as the direct and indirect effect of the Sun as a periodic time-dependent perturbation of the CR3BP. Du…
▽ More
This paper investigates the motion of a small particle moving near the triangular points of the Earth-Moon system. The dynamics are modeled in the Hill restricted 4-body problem (HR4BP), which includes the effect of the Earth and Moon as in the circular restricted 3-body problem (CR3BP), as well as the direct and indirect effect of the Sun as a periodic time-dependent perturbation of the CR3BP. Due to the periodic perturbation, the triangular points of the CR3BP are no longer equilibrium solutions; rather, the triangular points are replaced by periodic orbits with the same period as the perturbation. Additionally, there is a 2:1 resonant periodic orbit that persists from the CR3BP into the HR4BP. In this work, we investigate the dynamics around these invariant objects by computing families of 2-dimensional invariant tori and their linear normal behavior. We identify bifurcations and relationships between families. Mechanisms for transport between Earth, L4, and Moon are discussed. Comparisons are made between the results presented here and in the bicircular problem (BCP).
△ Less
Submitted 28 February, 2024;
originally announced February 2024.
-
Learning reduced-order Quadratic-Linear models in Process Engineering using Operator Inference
Authors:
Ion Victor Gosea,
Luisa Peterson,
Pawan Goyal,
Jens Bremer,
Kai Sundmacher,
Peter Benner
Abstract:
In this work, we address the challenge of efficiently modeling dynamical systems in process engineering. We use reduced-order model learning, specifically operator inference. This is a non-intrusive, data-driven method for learning dynamical systems from time-domain data. The application in our study is carbon dioxide methanation, an important reaction within the Power-to-X framework, to demonstra…
▽ More
In this work, we address the challenge of efficiently modeling dynamical systems in process engineering. We use reduced-order model learning, specifically operator inference. This is a non-intrusive, data-driven method for learning dynamical systems from time-domain data. The application in our study is carbon dioxide methanation, an important reaction within the Power-to-X framework, to demonstrate its potential. The numerical results show the ability of the reduced-order models constructed with operator inference to provide a reduced yet accurate surrogate solution. This represents an important milestone towards the implementation of fast and reliable digital twin architectures.
△ Less
Submitted 27 February, 2024;
originally announced February 2024.
-
A multifidelity Bayesian optimization method for inertial confinement fusion design
Authors:
J. Wang,
N. Chiang,
A. Gillette,
J. L. Peterson
Abstract:
Due to their cost, experiments for inertial confinement fusion (ICF) heavily rely on numerical simulations to guide design. As simulation technology progresses, so too can the fidelity of models used to plan for new experiments. However, these high-fidelity models are by themselves insufficient for optimal experimental design, because their computational cost remains too high to efficiently and ef…
▽ More
Due to their cost, experiments for inertial confinement fusion (ICF) heavily rely on numerical simulations to guide design. As simulation technology progresses, so too can the fidelity of models used to plan for new experiments. However, these high-fidelity models are by themselves insufficient for optimal experimental design, because their computational cost remains too high to efficiently and effectively explore the numerous parameters required to describe a typical experiment. Traditionally, ICF design has relied on low-fidelity modeling to initially identify potentially interesting design regions, which are then subsequently explored via selected high-fidelity modeling. In this paper, we demonstrate that this two-step approach can be insufficient: even for simple design problems, a two-step optimization strategy can lead high-fidelity searching towards incorrect regions and consequently waste computational resources on parameter regimes far away from the true optimal solution. We reveal that a primary cause of this behavior in ICF design problems is the presence of low-fidelity optima in distinct regions of the parameter space from high-fidelity optima. To address this issue, we propose an iterative multifidelity Bayesian optimization method based on Gaussian Process Regression that leverages both low- and high-fidelity modelings. We demonstrate, using both two- and eight-dimensional ICF test problems, that our algorithm can effectively utilize low-fidelity modeling for exploration, while automatically refining promising designs with high-fidelity models. This approach proves to be more efficient than relying solely on high-fidelity modeling for optimization.
△ Less
Submitted 15 December, 2023;
originally announced December 2023.
-
Nonlinear dynamics of asymmetric bistable energy harvesters
Authors:
João Pedro Norenberg,
Roberto Luo,
Vinicius Goncaalves Lopes,
João Victor L. L. Peterson,
Americo Cunha Jr
Abstract:
The paper investigates asymmetries effects over a nonlinear vibration energy harvester dynamics. The asymmetric system performance is compared with symmetric ones. Different asymmetry levels on restoring force and gravity action are investigated from a system-slo** angle variation. Bifurcation diagrams and basins of attraction are used to examine the local and global characteristics underlying d…
▽ More
The paper investigates asymmetries effects over a nonlinear vibration energy harvester dynamics. The asymmetric system performance is compared with symmetric ones. Different asymmetry levels on restoring force and gravity action are investigated from a system-slo** angle variation. Bifurcation diagrams and basins of attraction are used to examine the local and global characteristics underlying dynamical systems under different excitation energy. The results show the adverse effects of asymmetries on system dynamics. They also reveal ways to overcome them by canceling asymmetric influence from optimal slo** angle values and improving asymmetric system performance over symmetrical ones. This comprehensive numerical study provides novel valuable insights into asymmetrical energy harvester dynamics, a wide and still less explored topic.
△ Less
Submitted 9 June, 2023; v1 submitted 20 August, 2022;
originally announced September 2022.
-
A conformally invariant Yang-Mills type energy and equation on 6-manifolds
Authors:
A. Rod Gover,
Lawrence J. Peterson,
Callum Sleigh
Abstract:
We define a conformally invariant action S on gauge connections on a closed pseudo-Riemannian manifold M of dimension 6. At leading order this is quadratic in the gauge connection. The Euler-Lagrange equations of S, with respect to variation of the gauge connection, provide a higher-order conformally invariant analogue of the (source-free) Yang-Mills equations.
For any gauge connection A on M, w…
▽ More
We define a conformally invariant action S on gauge connections on a closed pseudo-Riemannian manifold M of dimension 6. At leading order this is quadratic in the gauge connection. The Euler-Lagrange equations of S, with respect to variation of the gauge connection, provide a higher-order conformally invariant analogue of the (source-free) Yang-Mills equations.
For any gauge connection A on M, we define S(A) by first defining a Lagrangian density associated to A. This is not conformally invariant but has a conformal transformation analogous to a Q-curvature. Integrating this density provides the conformally invariant action.
In the special case that we apply S to the conformal Cartan-tractor connection, the functional gradient recovers the natural conformal curvature invariant called the Fefferman-Graham obstruction tensor. So in this case the Euler-Lagrange equations are exactly the "obstruction-flat" condition for 6-manifolds. This extends known results for 4-dimensional pseudo-Riemannian manifolds where the Bach tensor is recovered in the Yang-Mills equations of the Cartan-tractor connection.
△ Less
Submitted 18 July, 2021;
originally announced July 2021.
-
Conformal boundary operators, T-curvatures, and conformal fractional Laplacians of odd order
Authors:
A. Rod Gover,
Lawrence J. Peterson
Abstract:
We construct continuously parametrised families of conformally invariant boundary operators on densities. These may also be viewed as conformally covariant boundary operators on functions and generalise to higher orders the first-order conformal Robin operator and an analogous third-order operator of Chang-Qing. Our families include operators of critical order on odd-dimensional boundaries. Combin…
▽ More
We construct continuously parametrised families of conformally invariant boundary operators on densities. These may also be viewed as conformally covariant boundary operators on functions and generalise to higher orders the first-order conformal Robin operator and an analogous third-order operator of Chang-Qing. Our families include operators of critical order on odd-dimensional boundaries. Combined with the (conformal Laplacian power) GJMS operators, a suitable selection of the boundary operators yields formally self-adjoint elliptic conformal boundary problems. Working on a conformal manifold with boundary, we show that the operators yield odd-order conformally invariant fractional Laplacian pseudo-differential operators. To do this, we use higher-order conformally invariant Dirichlet-to-Neumann constructions. We also find and construct new curvature quantities associated to our new operator families. These have links to the Branson Q-curvature and include higher-order generalisations of the mean curvature and the T-curvature of Chang-Qing. In the case of the standard conformal hemisphere, the boundary operator construction is particularly simple; the resulting operators provide an elementary construction of families of symmetry breaking intertwinors between the spherical principal series representations of the conformal group of the equator, as studied by Juhl and others. We use our constructions to shed light on some conjectures of Juhl.
△ Less
Submitted 22 February, 2018;
originally announced February 2018.
-
Future Directions of Research in Geometry: A Summary of the Panel Discussion at the 2007 Midwest Geometry Conference
Authors:
Lawrence J. Peterson
Abstract:
The 2007 Midwest Geometry Conference included a panel discussion devoted to open problems and the general direction of future research in fields related to the main themes of the conference. This paper summarizes the comments made during the panel discussion.
The 2007 Midwest Geometry Conference included a panel discussion devoted to open problems and the general direction of future research in fields related to the main themes of the conference. This paper summarizes the comments made during the panel discussion.
△ Less
Submitted 16 August, 2007;
originally announced August 2007.
-
The ambient obstruction tensor and the conformal deformation complex
Authors:
A. Rod Gover,
Lawrence J. Peterson
Abstract:
The (Fefferman-Graham) ambient obstruction tensor is a conformally invariant symmetric trace-free 2-tensor on even-dimensional Riemannian and pseudo-Riemannian manifolds. The conformal deformation complex is a differential complex related to infinitesimal deformations of conformal structure. We construct a conformally invariant differential operator on algebraic Weyl tensors that gives special c…
▽ More
The (Fefferman-Graham) ambient obstruction tensor is a conformally invariant symmetric trace-free 2-tensor on even-dimensional Riemannian and pseudo-Riemannian manifolds. The conformal deformation complex is a differential complex related to infinitesimal deformations of conformal structure. We construct a conformally invariant differential operator on algebraic Weyl tensors that gives special curved analogues of certain operators related to the deformation complex and that, upon application to the Weyl curvature, yields the obstruction tensor. This new definition of the obstruction tensor leads to simple direct proofs that the obstruction tensor is divergence free and vanishes for a class of metrics which (strictly) includes the class of conformally Einstein metrics. Our main constructions are based on the ambient metric of Fefferman-Graham and its relation to the conformal tractor connection. We prove that the obstruction tensor is an obstruction to the formal problem of obtaining a Ricci-flat ambient metric and that it may equivalently be viewed as the obstruction to finding an ambient metric with curvature harmonic for a certain (ambient) form Laplacian. This leads to a new ambient formula for the obstruction in terms of a power of this form Laplacian acting on the ambient curvature. This result leads us to construct Laplacian type operators that generalise the conformal Laplacians of Graham-Jenne-Mason-Sparling. As background to these issues, we give an explicit construction of the deformation complex in dimensions n>3, construct two related (detour) complexes, and establish essential properties of the operators in these.
△ Less
Submitted 19 May, 2005; v1 submitted 17 August, 2004;
originally announced August 2004.
-
Conformally invariant powers of the Laplacian, Q-curvature, and tractor calculus
Authors:
A. Rod Gover,
Lawrence J. Peterson
Abstract:
We describe an elementary algorithm for expressing, as explicit formulae in tractor calculus, the conformally invariant GJMS operators due to C.R. Graham et alia. These differential operators have leading part a power of the Laplacian. Conformal tractor calculus is the natural induced bundle calculus associated to the conformal Cartan connection. Applications discussed include standard formulae…
▽ More
We describe an elementary algorithm for expressing, as explicit formulae in tractor calculus, the conformally invariant GJMS operators due to C.R. Graham et alia. These differential operators have leading part a power of the Laplacian. Conformal tractor calculus is the natural induced bundle calculus associated to the conformal Cartan connection. Applications discussed include standard formulae for these operators in terms of the Levi-Civita connection and its curvature and a direct definition and formula for T. Branson's so-called Q-curvature (which integrates to a global conformal invariant) as well as generalisations of the operators and the Q-curvature. Among examples, the operators of order 4, 6, and 8 and the related Q-curvatures are treated explicitly. The algorithm exploits the ambient metric construction of Fefferman and Graham and includes a procedure for converting the ambient curvature and its covariant derivatives into tractor calculus expressions. This is partly based on "Standard tractors and the conformal ambient metric construction" (A. Cap and A.R. Gover, math.DG/0207016), where the relationship of the normal standard tractor bundle to the ambient construction is described.
△ Less
Submitted 1 February, 2003; v1 submitted 15 January, 2002;
originally announced January 2002.
