Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “convex optimization”

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 541 records · Page 30

Memory-efficient nonsmooth dynamic optimization using adaptive randomized compression

Dynamic optimization problems arise in many applications including flow control, full waveform inversion, and medical imaging. These problems are plagued by significant computational challenges. One such challenge — and the focus of this work — is the memory limitation induced by the size of the underlying dynamical system. In particular, the entire dynamic trajectory is required for derivative computation and therefore must be stored or recomputed using, e.g., checkpointing. Although recent work demonstrated the use of adaptive randomized sketching to overcome the memory challenge, that work only applies to smooth unconstrained problems, prohibiting its use for nonsmooth regularized and constrained problems. The inclusion of nonsmooth regularizers and constraints is critical as they often arise in an attempt to preserve certain physical properties or to promote sparsity. To solve these problems, we introduce a trust-region algorithm for minimizing the sum of a smooth nonconvex function and a nonsmooth convex function that leverages randomized sketching to compress the dynamical system trajectories and adaptively adjust the sketch rank to satisfy a gradient inexactness condition. We prove convergence of this algorithm and demonstrate that it achieves substantial memory reduction on three discretized PDE-constrained optimization applications.

97 MATHEMATICS AND COMPUTING↗

Formulations and Valid Inequalities for Optimal Black Start Allocation in Power Systems

The restoration of a power system after a blackout starts around units with enhanced technical capabilities, referred to as black start units (BSUs). We examine the planning problem of optimally allocating these units on the grid subject to a budget constraint. We present a mixed integer programming model based on current literature in power systems. Binary variables are associated with the allocation of BSUs and with the energization state of buses, branches, and generators of the power system over a time horizon. We extract a substructure of the feasible region which imposes the requirement that each island that appears during the restoration process must have at least one operational generator. We discuss three equivalent reformulations for this requirement. We introduce a family of exponentially many, polynomially separable, valid inequalities to strengthen the formulation. Under simplifying assumptions, we show that the convex hull of the feasible region is a full-dimensional polyhedron, and prove that some of the constraints we introduced are facet-defining. We perform experiments to examine the difference in strength between the formulations as well as the computational times to solve the problem to near optimality for synthetic instances of the IEEE-39, IEEE-118, Illinois-200, WECC-225, IEEE-300, South Carolina-500, and Texas-2000 power systems. We illustrate a use case of the model. We conclude by suggesting extensions of the current work for future research.

Black Start Allocation↗

Divide and conquer: Learning chaotic dynamical systems with multistep penalty neural ordinary differential equations

Forecasting high-dimensional dynamical systems is a fundamental challenge in various fields, such as geosciences and engineering. Neural Ordinary Differential Equations (NODEs), which combine the power of neural networks and numerical solvers, have emerged as a promising algorithm for forecasting complex nonlinear dynamical systems. However, classical techniques used for NODE training are ineffective for learning chaotic dynamical systems. In this work, we propose a novel NODE-training approach that allows for robust learning of chaotic dynamical systems. Here, our method addresses the challenges of non-convexity and exploding gradients associated with underlying chaotic dynamics. Training data trajectories from such systems are split into multiple, non-overlapping time windows. In addition to the deviation from the training data, the optimization loss term further penalizes the discontinuities of the predicted trajectory between the time windows. The window size is selected based on the fastest Lyapunov time scale of the system. Multi-step penalty(MP) method is first demonstrated on Lorenz equation, to illustrate how it improves the loss landscape and thereby accelerates the optimization convergence. MP method can optimize chaotic systems in a manner similar to least-squares shadowing with significantly lower computational costs. Our proposed algorithm, denoted the Multistep Penalty NODE, is applied to chaotic systems such as the Kuramoto-Sivashinsky equation, the two-dimensional Kolmogorov flow, and ERA5 reanalysis data for the atmosphere. It is observed that MP-NODE provide viable performance for such chaotic systems, not only for short-term trajectory predictions but also for invariant statistics that are hallmarks of the chaotic nature of these dynamics.

Chaotic dynamical systems↗

Predictively Coordinated Vehicle Acceleration and Lane Selection Using Mixed Integer Programming

Autonomous vehicle technology provides the means to optimize motion planning beyond human capacity. In particular, the problem of navigating multi-lane traffic optimally for trip time, energy efficiency, and collision avoidance presents challenges beyond those of single-lane roadways. For example, the host vehicle must simultaneously track multiple obstacles, the drivable region is non-convex, and automated vehicles must obey social expectations. Furthermore, reactive decision-making may result in becoming stuck in an undesirable traffic position. This paper presents a fundamental approach to these problems using model predictive control with a mixed integer quadratic program at its core. Lateral and longitudinal movements are coordinated to avoid collisions, track a velocity and lane, and minimize acceleration. Vehicle-to-vehicle connectivity provides a preview of surrounding vehicles’ motion. Simulation results show a 79% reduction in congestion-induced travel time and an 80% decrease in congestion-induced fuel consumption compared to a rule-based approach.

Dollar, Robert Austin↗

Efficient proximal subproblem solvers for a nonsmooth trust-region method

In [R. J. Baraldi and D. P. Kouri, Mathematical Programming, (2022), pp. 1-40], we introduced an inexact trust-region algorithm for minimizing the sum of a smooth nonconvex and nonsmooth convex function. The principle expense of this method is in computing a trial iterate that satisfies the so-called fraction of Cauchy decrease condition—a bound that ensures the trial iterate produces sufficient decrease of the subproblem model. In this paper, we expound on various proximal trust-region subproblem solvers that generalize traditional trust-region methods for smooth unconstrained and convex-constrained problems. We introduce a simplified spectral proximal gradient solver, a truncated nonlinear conjugate gradient solver, and a dogleg method. Finally, we compare algorithm performance on examples from data science and PDE-constrained optimization.

97 MATHEMATICS AND COMPUTING↗

A Chaos Theoretic Approach to Animal Activity Recognition

Animal activity is a descriptor that can be potentially used to track the health and well-being, which is obviously very important to improve the management process and productivity of farms. This paper deals with an animal behavior recognition problem using a chaos theory approach where we adopt such a technique for automatic classification of calves behavioral states. Two main mutually exclusive behaviors are of interest, namely, lying and standing/walking with six possible activities: feeding, drinking water, drinking milk, playing, rumination, and neutral. The time series generated by ear-tags with a 3D-accelerometer after a wavelet denoising transformation and frequency stabilization are treated as representations of the nonlinear dynamical system. The dynamical system of a certain animal state exhibits specific strange attractor in a phase space. A characterization of such an attractor is performed through metric, dynamic, and topological invariants including Lyapunov exponent, correlation dimension, length of a phase trajectory, sum of edges forming a convex hull and others. These measures are used as a feature vector for the subsequent classification. In a cross validation scheme, six classifiers are built on each training set, and the hyper-parameters are optimized using an inner validation set. The classifier that reaches the highest accuracy on the inner validation set is used to classify the outer validation set. It is shown that this approach can be useful at predicting activity states as an alternative methodology for the animal behavior state recognition problem with acceptable classification accuracy. Furthermore it is possible to include this procedure as part of an ensemble method in machine learning where a combination of different models is used.

97 MATHEMATICS AND COMPUTING↗

Semilinear (topological) spaces and applications

Semivector spaces are defined and some of their algebraic aspects are developed including some structure theory. These spaces are then topologized to obtain semilinear topological spaces for which a hierarchy of local convexity axioms is identified. A number of fixed point and minmax theorems for spaces with various local convexity properties are established. The spaces of concern arise naturally as various hyperspaces of linear and semilinear (topological) spaces. It is indicated briefly how all this can be applied in socio-economic analysis and optimization.

Prakash, P.↗

A vectorized Lanczos eigensolver for high-performance computers

The computational strategies used to implement a Lanczos-based-method eigensolver on the latest generation of supercomputers are described. Several examples of structural vibration and buckling problems are presented that show the effects of using optimization techniques to increase the vectorization of the computational steps. The data storage and access schemes and the tools and strategies that best exploit the computer resources are presented. The method is implemented on the Convex C220, the Cray 2, and the Cray Y-MP computers. Results show that very good computation rates are achieved for the most computationally intensive steps of the Lanczos algorithm and that the Lanczos algorithm is many times faster than other methods extensively used in the past.

Bostic, Susan W.↗

Efficient Automated Driving Strategies Leveraging Anticipation and Optimal Control

Automated vehicles and advanced driver assistance systems bring computation, sensing, and communication technologies that exceed human abilities in some ways. For example, automated vehicles may sense a panorama all at once, do not suffer from human impairments and distractions, and could wirelessly communicate precise data with neighboring vehicles. Prototype and commercial deployments have demonstrated the capability to relieve human operators of some driving tasks up to and including fully autonomous taxi rides in some areas. The ultimate impact of this technology’s large-scale market penetration on energy efficiency remains unclear, with potential negative factors like road use by empty vehicles competing with positive ones like automatic eco-driving. Fundamentally enabled by historic and look-ahead data, this dissertation addresses the use of automated driving and driver assistance to optimize vehicle motion for energy efficiency. Facets of this problem include car following, co-optimized acceleration and lane change planning, and collaborative multi-agent guidance. Optimal control, especially model predictive control, is used extensively to improve energy efficiency while maintaining safe and timely driving via constraints. Techniques including chance constraints and mixed integer programming help overcome uncertainty and non-convexity challenges. Extensions of these techniques to tractor trailers on sloping roads are provided by making use of linear parameter-varying models. To approach the wheel-input energy eco-driving problem over generally shaped sloping roads with the computational potential for closed-loop implementation, a linear programming formulation is constructed. Distributed and collaborative techniques that enable connected and automated vehicles to accommodate their neighbors in traffic are also explored and compared to centralized control. Using simulations and vehicle-in-the-loop car following experiments, the proposed algorithms are benchmarked against others that do not make use of look-ahead information.

Dollar, Robert Austin↗

Certifiably Correct Range-Aided SLAM

We present the first algorithm capable of efficiently computing certifiably optimal solutions to range-aided simultaneous localization and mapping (RA-SLAM) problems. Robotic navigation systems are increasingly incorporating point-to-point ranging sensors, leading state estimation which takes the form of RA-SLAM. However, the RA-SLAM problem is more difficult to solve than traditional pose-graph SLAM; ranging sensor models introduce additional non-convexity, unlike pose-pose or pose-landmark measurements, a single range measurement does not uniquely determine the relative transform between the involved sensors, and RA-SLAM inference is highly sensitive to initial estimates. Our approach relaxes the RA-SLAM problem to a semidefinite program (SDP), which we show how to solve efficiently using the Riemannian staircase methodology. The solution of this SDP provides a high-quality initialization for our original RA-SLAM problem, which is subsequently refined via local optimization, as well as a lower-bound on the RA-SLAM problem's optimal value. Our algorithm, named certifiably correct RA-SLAM (CORA), applies to problems comprised of arbitrary pose-pose, pose-landmark, and ranging measurements. Evaluation on simulated and real-world marine examples shows that our algorithm frequently produces certifiably optimal RA-SLAM solutions; moreover, even suboptimal estimates are typically within 1-2\% of the optimal value.

Papalia, Alan↗

Structural optimization via a design space hierarchy

Mathematical programming techniques provide a general approach to automated structural design. An iterative method is proposed in which design is treated as a hierarchy of subproblems, one being locally constrained and the other being locally unconstrained. It is assumed that the design space is locally convex in the case of good initial designs and that the objective and constraint functions are continuous, with continuous first derivatives. A general design algorithm is outlined for finding a move direction which will decrease the value of the objective function while maintaining a feasible design. The case of one-dimensional search in a two-variable design space is discussed. Possible applications are discussed. A major feature of the proposed algorithm is its application to problems which are inherently ill-conditioned, such as design of structures for optimum geometry.

Vanderplaats, G. N.↗

A two-level approach to large mixed-integer programs with application to cogeneration in energy-efficient buildings

We study a two-stage mixed-integer linear program (MILP) with more than 1 million binary variables in the second stage. We develop a two-level approach by constructing a semi-coarse model that coarsens with respect to variables and a coarse model that coarsens with respect to both variables and constraints. We coarsen binary variables by selecting a small number of prespecified on/off profiles. We aggregate constraints by partitioning them into groups and taking convex combination over each group. With an appropriate choice of coarsened profiles, the semi-coarse model is guaranteed to find a feasible solution of the original problem and hence provides an upper bound on the optimal solution. We show that solving a sequence of coarse models converges to the same upper bound with proven finite steps. This is achieved by adding violated constraints to coarse models until all constraints in the semi-coarse model are satisfied. We demonstrate the effectiveness of our approach in cogeneration for buildings. Here, the coarsened models allow us to obtain good approximate solutions at a fraction of the time required by solving the original problem. Extensive numerical experiments show that the two-level approach scales to large problems that are beyond the capacity of state-of-the-art commercial MILP solvers.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

Relating Lexicographic Smoothness and Directed Subdifferentiability

Lexicographic derivatives developed by Nesterov and directed subdifferentials developed by Baier, Farkhi, and Roshchina are both essentially nonconvex generalized derivatives for nonsmooth nonconvex functions and satisfy strict calculus rules and mean-value theorems. This article aims to clarify the relationship between the two generalized derivatives. In particular, for scalar-valued functions that are locally Lipschitz continuous, lexicographic smoothness and directed subdifferentiability are shown to be equivalent, along with the necessary optimality conditions corresponding to each. For such functions, the visualization of the directed subdifferential-the Rubinov subdifferential-is shown to include the lexicographic subdifferential, and is also shown to be included in its closed convex hull. As a result, various implications of these results are discussed.

96 KNOWLEDGE MANAGEMENT AND PRESERVATION↗

Throughput Analytics of Cloud Networks

A network of virtual machines at cloud server sites connected over virtual IO connections is a flexible, easily deployable, and cost-effective alternative to a physical network infrastructure with dedicated servers connected over leased fiber lines. We study the throughput performance of such a cloud network by collecting measurements over Google Cloud infrastructure spanning multiple continents. To study its ideal performance and impact of packet losses, we utilize its emulation using dedicated servers and connection hardware emulation devices. We compare the measurements over the cloud network to those over its emulation on a testbed. We examine the throughput profiles of both networks as a function of the round trip time and their utilization-concavity coefficients, estimated using measurements for common TCP versions. The throughput profile's concave-convex shape and its coefficient are critical indicators of the network performance, qualitatively and quantitatively, respectively. The results indicate their overall agreement between the production cloud network and its emulation using dedicated connections, and a near optimal throughput performance of the former except for a few under-performing connections. Also, the number of parallel flows is found to be a dominant factor in optimizing the throughput across various conditions and TCP versions.

Phanekham, Derek↗

WE-AB-209-10: Optimizing the Delivery of Sequential Fluence Maps for Efficient VMAT Delivery

Purpose: To develop an optimization model and solution approach for computing MLC leaf trajectories and dose rates for high quality matching of a set of optimized fluence maps to be delivered sequentially around a patient in a VMAT treatment. Methods: We formulate the fluence map matching problem as a nonlinear optimization problem where time is discretized but dose rates and leaf positions are continuous variables. For a given allotted time, which is allocated across the fluence maps based on the complexity of each fluence map, the optimization problem searches for the best leaf trajectories and dose rates such that the original fluence maps are closely recreated. Constraints include maximum leaf speed, maximum dose rate, and leaf collision avoidance, as well as the constraint that the ending leaf positions for one map are the starting leaf positions for the next map. The resulting model is non-convex but smooth, and therefore we solve it by local searches from a variety of starting positions. We improve solution time by a custom decomposition approach which allows us to decouple the rows of the fluence maps and solve each leaf pair individually. This decomposition also makes the problem easily parallelized. Results: We demonstrate method on a prostate case and a head-and-neck case and show that one can recreate fluence maps to high degree of fidelity in modest total delivery time (minutes). Conclusion: We present a VMAT sequencing method that reproduces optimal fluence maps by searching over a vast number of possible leaf trajectories. By varying the total allotted time given, this approach is the first of its kind to allow users to produce VMAT solutions that span the range of wide-field coarse VMAT deliveries to narrow-field high-MU sliding window-like approaches.

60 APPLIED LIFE SCIENCES↗

An adaptive moments-based interface reconstruction using intersection of the cell with one half-plane, two half-planes and a circle

We present a new adaptive moment-of-fluid (A-MOF) interface reconstruction method. It uses the zeroth, first, and second moments of the fragment of material inside a cell of the mesh to construct a shape that approximates the respective material fragment. The new method requires information about the material moments only for the cell under consideration. The adaptive method chooses between shapes obtained by the intersection of the cell with one half-plane, two half-planes, or a circle. The A-MOF method allows to exactly reproduce several convex shapes: corners, filaments, and their concave cell-complements; as well as pieces of the circles and its cell-compliments. Interface reconstruction is formulated as a local (for each cell), non-linear, equality constrained optimization problem, which does not require additional communication and allows for an efficient parallel implementation. In conclusion, we present an extensive set of test problems, both for interface reconstruction on a single cell, and for reconstruction of a variety of shapes on the entire mesh.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Realizability-preserving DG-IMEX method for the two-moment model of fermion transport

Building on the framework of Zhang & Shu [1], [2], we develop a realizability-preserving method to simulate the transport of particles (fermions) through a background material using a two-moment model that evolves the angular moments of a phase space distribution function f . The two-moment model is closed using algebraic moment closures; e.g., as proposed by Cernohorsky & Bludman [3] and Banach & Larecki [4]. Variations of this model have recently been used to simulate neutrino transport in nuclear astrophysics applications, including core-collapse supernovae and compact binary mergers. We employ the discontinuous Galerkin (DG) method for spatial discretization (in part to capture the asymptotic diffusion limit of the model) combined with implicit-explicit (IMEX) time integration to stably bypass short timescales induced by frequent interactions between particles and the background. Appropriate care is taken to ensure the method preserves strict algebraic bounds on the evolved moments (particle density and flux) as dictated by Pauli's exclusion principle, which demands a bounded distribution function (i.e., f ϵ [0,1]). This realizability-preserving scheme combines a suitable CFL condition, a realizability-enforcing limiter, a closure procedure based on Fermi-Dirac statistics, and an IMEX scheme whose stages can be written as a convex combination of forward Euler steps combined with a backward Euler step. The IMEX scheme is formally only first-order accurate, but works well in the diffusion limit, and — without interactions with the background — reduces to the optimal second-order strong stability-preserving explicit Runge-Kutta scheme of Shu & Osher [5]. Numerical results demonstrate the realizability-preserving properties of the scheme. As a result, we also demonstrate that the use of algebraic moment closures not based on Fermi-Dirac statistics can lead to unphysical moments in the context of fermion transport.

97 MATHEMATICS AND COMPUTING↗

An algorithm for the rapid location of an extreme of a function subject only to geometric restrictions

The requirements of symmetry and convexity in the application of algorithms for the minimization or maximization of a function are discussed. It is argued that if a function of a single variable is convex and symmetric in a neighborhood of an extremum, the extremum may be approximated to the precision that increases by at least a power of two per functional evaluation. The procedure may be used to drive a complex optimization procedure in the multivariate area estimation problem encountered in remote sensing.

Terrell, G. R.↗