Search NASA⌕ Search

SEARCH · Search NASA

Results for “optimization problems”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 775 records · Page 43

Leveraging Multitime Hamilton–Jacobi PDEs for Certain Scientific Machine Learning Problems

Hamilton-Jacobi partial differential equations (HJ PDEs) have deep connections with a wide range of fields, including optimal control, differential games, and imaging sciences. By considering the time variable to be a higher dimensional quantity, HJ PDEs can be extended to the multi-time case. In this paper, we establish a novel theoretical connection between specific optimization problems arising in machine learning and the multi-time Hopf formula, which corresponds to a representation of the solution to certain multi-time HJ PDEs. Through this connection, we increase the interpretability of the training process of certain machine learning applications by showing that when we solve these learning problems, we also solve a multi-time HJ PDE and, by extension, its corresponding optimal control problem. As a first exploration of this connection, we develop the relation between the regularized linear regression problem and the Linear Quadratic Regulator (LQR). We then leverage our theoretical connection to adapt standard LQR solvers (namely, those based on the Riccati ordinary differential equations) to design new training approaches for machine learning. Lastly, we provide some numerical examples that demonstrate the versatility and possible computational advantages of our Riccati-based approach in the context of continual learning, post-training calibration, transfer learning, and sparse dynamics identification.

97 MATHEMATICS AND COMPUTING↗

Improve earthquake hypocenter using adaptive simulated annealing inversion in regional tectonic, volcano tectonic, and geothermal observation

Observation of earthquakes is routinely used widely in tectonic activity observation, and also in local scale such as volcano tectonic and geothermal activity observation. It is necessary for determining the location of precise hypocenter which the process involves finding a hypocenter location that has minimum error between the observed and the calculated travel times. When solving this nonlinear inverse problem, simulated annealing inversion method can be applied to such global optimization problems, which the convergence of its solution is independent of the initial model. In this study, we developed own program codeby applying adaptive simulated annealing inversion in Matlab environment. We applied this method to determine earthquake hypocenter using several data cases which are regional tectonic, volcano tectonic, and geothermal field. The travel times were calculated using ray tracing shooting method. We then compared its results with the results using Geiger’s method to analyze its reliability. Our results show hypocenter location has smaller RMS error compared to the Geiger’s result that can be statistically associated with better solution. The hypocenter of earthquakes also well correlated with geological structure in the study area. Werecommend using adaptive simulated annealing inversion to relocate hypocenter location in purpose to get precise and accurate earthquake location.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Optimal trajectories for an aerospace plane. Part 1: Formulation, results, and analysis

The optimization of the trajectories of an aerospace plane is discussed. This is a hypervelocity vehicle capable of achieving orbital speed, while taking off horizontally. The vehicle is propelled by four types of engines: turbojet engines for flight at subsonic speeds/low supersonic speeds; ramjet engines for flight at moderate supersonic speeds/low hypersonic speeds; scramjet engines for flight at hypersonic speeds; and rocket engines for flight at near-orbital speeds. A single-stage-to-orbit (SSTO) configuration is considered, and the transition from low supersonic speeds to orbital speeds is studied under the following assumptions: the turbojet portion of the trajectory has been completed; the aerospace plane is controlled via the angle of attack and the power setting; the aerodynamic model is the generic hypersonic aerodynamics model example (GHAME). Concerning the engine model, three options are considered: (EM1), a ramjet/scramjet combination in which the scramjet specific impulse tends to a nearly-constant value at large Mach numbers; (EM2), a ramjet/scramjet combination in which the scramjet specific impulse decreases monotonically at large Mach numbers; and (EM3), a ramjet/scramjet/rocket combination in which, owing to stagnation temperature limitations, the scramjet operates only at M approx. less than 15; at higher Mach numbers, the scramjet is shut off and the aerospace plane is driven only by the rocket engines. Under the above assumptions, four optimization problems are solved using the sequential gradient-restoration algorithm for optimal control problems: (P1) minimization of the weight of fuel consumed; (P2) minimization of the peak dynamic pressure; (P3) minimization of the peak heating rate; and (P4) minimization of the peak tangential acceleration.

Miele, Angelo↗

Singular linear-quadratic control problem for systems with linear delay

A singular linear-quadratic optimization problem on the trajectories of non-autonomous linear differential equations with linear delay is considered. The peculiarity of this problem is the fact that this problem has no solution in the class of integrable controls. To ensure the existence of solutions is required to expand the class of controls including controls with impulse components. Dynamical systems with linear delay are used to describe the motion of pantograph from the current collector with electric traction, biology, etc. It should be noted that for practical problems fact singularity criterion of quality is quite commonly occurring, and therefore the study of these problems is surely important. For the problem under discussion optimal programming control contained impulse components at the initial and final moments of time is constructed under certain assumptions on the functional and the right side of the control system.

97 MATHEMATICS AND COMPUTING↗

Aerospace applications on integer and combinatorial optimization

Research supported by NASA Langley Research Center includes many applications of aerospace design optimization and is conducted by teams of applied mathematicians and aerospace engineers. This paper investigates the benefits from this combined expertise in formulating and solving integer and combinatorial optimization problems. Applications range from the design of large space antennas to interior noise control. A typical problem. for example, seeks the optimal locations for vibration-damping devices on an orbiting platform and is expressed as a mixed/integer linear programming problem with more than 1500 design variables.

Padula, S. L.↗

Aerospace Applications of Integer and Combinatorial Optimization

Research supported by NASA Langley Research Center includes many applications of aerospace design optimization and is conducted by teams of applied mathematicians and aerospace engineers. This paper investigates the benefits from this combined expertise in formulating and solving integer and combinatorial optimization problems. Applications range from the design of large space antennas to interior noise control. A typical problem, for example, seeks the optimal locations for vibration-damping devices on an orbiting platform and is expressed as a mixed/integer linear programming problem with more than 1500 design variables.

Padula, S. L.↗

An Optimization-based Atomistic-to-Continuum Coupling Method

In this paper, we present a new optimization-based method for atomistic-to-continuum (AtC) coupling. The main idea is to cast the latter as a constrained optimization problem with virtual Dirichlet controls on the interfaces between the atomistic and continuum subdomains. The optimization objective is to minimize the error between the atomistic and continuum solutions on the overlap between the two subdomains, while the atomistic and continuum force balance equations provide the constraints. Separation, rather then blending of the atomistic and continuum problems, and their subsequent use as constraints in the optimization problem distinguishes our approach from the existing AtC formulations. Finally, we present and analyze the method in the context of a one-dimensional chain of atoms modeled using a linearized two-body potential with next-nearest neighbor interactions.

97 MATHEMATICS AND COMPUTING↗

Quantum approximate multi-objective optimization

The goal of multi-objective optimization is to understand optimal trade-offs between competing objective functions by finding the Pareto front, that is, the set of all Pareto-optimal solutions, where no objective can be improved without degrading another one. Multi-objective optimization can be challenging classically, even if the corresponding single-objective optimization problems are efficiently solvable. Thus, multi-objective optimization represents a compelling problem class to analyze with quantum computers. Here we use a low-depth quantum approximate optimization algorithm to approximate the optimal Pareto front of certain multi-objective weighted maximum-cut problems. We demonstrate its performance on an IBM Quantum computer, as well as with matrix product state numerical simulation, and show its potential to outperform classical approaches.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

An integrated controls-structures design methodology for a flexible spacecraft

This paper proposes an approach for the design of flexible spacecraft, wherein the structural design and the control system design are performed simultaneously. The integrated design problem is posed as an optimization problem in which both the structural parameters and the control system parameters constitute the design variables, which are used to optimize a common objective function, thereby resulting in an optimal overall design. The approach is demonstrated by application to the integrated design of a geostationary platform, and to a ground-based flexible structure experiment. The numerical results obtained indicate that the integrated design approach generally yields spacecraft designs that are substantially superior compared to the conventional approach, wherein the structural design and control design are performed sequentially.

Maghami, Peiman G.↗

Aerospace applications of integer and combinatorial optimization

Research supported by NASA Langley Research Center includes many applications of aerospace design optimization and is conducted by teams of applied mathematicians and aerospace engineers. This paper investigates the benefits from this combined expertise in solving combinatorial optimization problems. Applications range from the design of large space antennas to interior noise control. A typical problem, for example, seeks the optimal locations for vibration-damping devices on a large space structure and is expressed as a mixed/integer linear programming problem with more than 1500 design variables.

Padula, S. L.↗

Integrated Controls-Structures Design Methodology for Flexible Spacecraft

This paper proposes an approach for the design of flexible spacecraft, wherein the structural design and the control system design are performed simultaneously. The integrated design problem is posed as an optimization problem in which both the structural parameters and the control system parameters constitute the design variables, which are used to optimize a common objective function, thereby resulting in an optimal overall design. The approach is demonstrated by application to the integrated design of a geostationary platform, and to a ground-based flexible structure experiment. The numerical results obtained indicate that the integrated design approach generally yields spacecraft designs that are substantially superior to the conventional approach, wherein the structural design and control design are performed sequentially.

Maghami, P. G.↗

Best-Fit Conic Approximation of Spacecraft Trajectory

A computer program calculates a best conic fit of a given spacecraft trajectory. Spacecraft trajectories are often propagated as conics onboard. The conic-section parameters as a result of the best-conic-fit are uplinked to computers aboard the spacecraft for use in updating predictions of the spacecraft trajectory for operational purposes. In the initial application for which this program was written, there is a requirement to fit a single conic section (necessitated by onboard memory constraints) accurate within 200 microradians to a sequence of positions measured over a 4.7-hour interval. The present program supplants a prior one that could not cover the interval with fewer than four successive conic sections. The present program is based on formulating the best-fit conic problem as a parameter-optimization problem and solving the problem numerically, on the ground, by use of a modified steepest-descent algorithm. For the purpose of this algorithm, optimization is defined as minimization of the maximum directional propagation error across the fit interval. In the specific initial application, the program generates a single 4.7-hour conic, the directional propagation of which is accurate to within 34 microradians easily exceeding the mission constraints by a wide margin.

Singh, Gurkipal↗

$L^1$ penalization of volumetric dose objectives in optimal control of PDEs

This work is concerned with a class of PDE-constrained optimization problems that are motivated by an application in radiotherapy treatment planning. Here the primary design objective is to minimize the volume where a functional of the state violates a prescribed level, but prescribing these levels in the form of pointwise state constraints leads to infeasible problems. We therefore propose an alternative approach based on L 1 penalization of the violation that is also applicable when state constraints are infeasible. We establish well-posedness of the corresponding optimal control problem, derive first-order optimality conditions, discuss convergence of minimizers as the penalty parameter tends to infinity, and present a semismooth Newton method for their efficient numerical solution. Finally, the performance of this method for a model problem is illustrated and contrasted with an alternative approach based on (regularized) state constraints.

97 MATHEMATICS AND COMPUTING↗

Analytical Partial Derivatives of the Q-Law Guidance Algorithm

One method of producing low-thrust trajectories is the closed-loop Q-Law guidance algorithm. This work wraps Q-Law inside a nonlinear programming problem to search for locally optimal gain combinations and enforce nonlinear constraints on the initial state. Gradient-based trajectory optimization has been shown to benefit greatly when analytical partial derivatives are supplied to the optimizer. This paper presents the Q-Law optimization problem setup and derives the Q-Law thrust vector partial derivatives for the State Transition Matrix required by standard nonlinear optimization software.

Jackson L Shannon↗

Applying optimization software libraries to engineering problems

Nonlinear programming, preliminary design problems, performance simulation problems trajectory optimization, flight computer optimization, and linear least squares problems are among the topics covered. The nonlinear programming applications encountered in a large aerospace company are a real challenge to those who provide mathematical software libraries and consultation services. Typical applications include preliminary design studies, data fitting and filtering, jet engine simulations, control system analysis, and trajectory optimization and optimal control. Problem sizes range from single-variable unconstrained minimization to constrained problems with highly nonlinear functions and hundreds of variables. Most of the applications can be posed as nonlinearly constrained minimization problems. Highly complex optimization problems with many variables were formulated in the early days of computing. At the time, many problems had to be reformulated or bypassed entirely, and solution methods often relied on problem-specific strategies. Problems with more than ten variables usually went unsolved.

Healy, M. J.↗

Radiotherapy Dose Fractionation under Parameter Uncertainty

In radiotherapy, radiation is directed to damage a tumor while avoiding surrounding healthy tissue. Tradeoffs ensue because dose cannot be exactly shaped to the tumor. It is particularly important to ensure that sensitive biological structures near the tumor are not damaged more than a certain amount. Biological tissue is known to have a nonlinear response to incident radiation. The linear quadratic dose response model, which requires the specification of two clinically and experimentally observed response coefficients, is commonly used to model this effect. This model yields an optimization problem giving two different types of optimal dose sequences (fractionation schedules). Which fractionation schedule is preferred depends on the response coefficients. These coefficients are uncertainly known and may differ from patient to patient. Because of this not only the expected outcomes but also the uncertainty around these outcomes are important, and it might not be prudent to select the strategy with the best expected outcome.

60 APPLIED LIFE SCIENCES↗

Model Predictive Control of a Wave Energy Converter Using Duality Techniques

In this paper, a method for implementing model predictive control for vibration energy harvesting systems with the goal of maximizing generated energy is presented. This is generally a nonconvex problem, which can be computationally expensive and converging to a global minimum cannot be guaranteed. We introduce a dual-domain method to relax the nonconvex primal problem into a convex dual optimization problem, which can be performed in real time. Based on previous work, we determine when the duality gap is zero and, by extension, whether the global optimal primal solution is found. Here, we demonstrate the algorithm on a single, buoy-type wave energy converter with energy storage constraints and power take-off force constraints.

16 TIDAL AND WAVE POWER↗

Model Predictive Control of a Wave Energy Converter Using Duality Techniques

In this paper, a method for implementing model predictive control for vibration energy harvesting systems with the goal of maximizing generated energy is presented. This is generally a nonconvex problem, which can be computationally expensive and converging to a global minimum cannot be guaranteed. We introduce a dual-domain method to relax the nonconvex primal problem into a convex dual optimization problem, which can be performed in real time. Based on previous work, we determine when the duality gap is zero and, by extension, whether the global optimal primal solution is found. We demonstrate the algorithm on a single, buoy-type wave energy converter with energy storage constraints and power take-off force constraints.

16 TIDAL AND WAVE POWER↗