A universal approach for continuous or discrete nonlinear programmings with multiple variables and constraints

The two-dimensional knapsack problem with splittable items in stacks

Any ultrafast optical spectroscopy experiment is usually accompanied by the necessary routine of ultrashort-pulse characterization. The majority of pulse characterization approaches solve either a one-dimensional (e.g., via interferometry) or a two-dimensional (e.g., via frequency-resolved measurements) problem. Solution of the two-dimensional pulse-retrieval problem is generally more consistent due to the problem's over-determined nature. In contrast, the one-dimensional pulse-retrieval problem, unless constraints are added, is impossible to solve unambiguously as ultimately imposed by the fundamental theorem of algebra. In cases where additional constraints are involved, the one-dimensional problem may be possible to solve, however, existing iterative algorithms lack generality, and often stagnate for complicated pulse shapes. Here we use a deep neural network to unambiguously solve a constrained one-dimensional pulse-retrieval problem and show the potential of fast, reliable and complete pulse characterization using interferometric correlation time traces determined by the pulses with partial spectral overlap.

Greedy algorithms for a class of knapsack problems with binary weights

This paper presents the results of an empirical study of the effects of coefficient correlation structure and constraint slackness settings on the performance of solution procedures on synthetic two-dimensional knapsack problems (2KP). The population correlation structure among 2KP coefficients, the level of constraint slackness, and the type of correlation (product moment or rank) are varied in this study. Representative branch-and-bound and heuristic solution procedures are used to investigate the influence of these problem parameters on solution procedure performance. Population correlation structure, and in particular the interconstraint component of the correlation structure, is found to be a significant factor influencing the performance of both the algorithm and the heuristic. In addition, the interaction between constraint slackness and population correlation structure is found to influence solution procedure performance.

This paper presents a subgradient-based algorithm for constrained nonsmooth convex optimization that does not require projections onto the feasible set. While the well-established Frank–Wolfe algorithm and its variants already avoid projections, they are primarily designed for smooth objective functions. In contrast, our proposed algorithm can handle nonsmooth problems with general convex functional inequality constraints. It achieves an ϵ\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\epsilon $$\\end{document}-suboptimal solution in O(ϵ-2)\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\mathcal {O}(\\epsilon ^{-2})$$\\end{document} iterations, with each iteration requiring only a single (potentially inexact) Linear Minimization Oracle call and a (possibly inexact) subgradient computation. This performance is consistent with existing lower bounds. Similar performance is observed when deterministic subgradients are replaced with stochastic subgradients. In the special case where there are no functional inequality constraints, our algorithm competes favorably with a recent nonsmooth projection-free method designed for constraint-free problems. Our approach utilizes a simple separation scheme in conjunction with a new Lagrange multiplier update rule.

Discontinuity solutions to plate and beam problems

This paper proposes a level-set based topology optimization method incorporating a boundary tracking mesh generating method and nonlinear programming. Because the boundary tracking mesh is always conformed to the structural boundary, good approximation to the boundary is maintained during optimization; therefore, structural design problems are solved completely without grayscale material. Previously, we introduced the boundary tracking mesh generating method into level-set based topology optimization and updated the design variables by solving the level-set equation. In order to adapt our previous method to general structural optimization frameworks, the incorporation of the method with nonlinear programming is investigated in this paper. To successfully incorporate nonlinear programming, the optimization problem is regularized using a double-well potential. Furthermore, the sensitivities with respect to the design variables are strictly derived to maintain consistency in mathematical programming. We expect the investigation to open up a new class of grayscale-free topology optimization. The usefulness of the proposed method is demonstrated using several numerical examples targeting two-dimensional compliant mechanism and metallic waveguide design problems. Copyright © 2014 John Wiley & Sons, Ltd.

It is sometimes necessary to consider more general nonlinear programs than the ones we have already studied: linear, quadratic, or conic programs. We give a brief introduction to this vast topic, and we discuss an application to estimating a volatility surface. Nonlinear Programming Consider a very general optimization problem of the form or the equivalent more concise form where. In the special case when all functions , hi are linear, problem (20.1) is a linear program as discussed in Chapter 2. When some of the functions , are nonlinear, problem (20.1) is a nonlinear program . Many practical problems are naturally formulated as nonlinear programs. We already saw quadratic programming and conic programming in earlier chapters. However, the family of problems that can be formulated as nonlinear programs is enormous, as many, if not most, imaginable kinds of constraints and objectives can be cast in terms of nonlinear functions. (For some noteworthy examples, see the exercises at the end of the chapter.) The immense modeling power of nonlinear programming comes at a cost: unlike linear, quadratic, and conic programming, which have a solid theory and are solvable via a few algorithmic templates, both the theory and methods to solve general nonlinear programs are far more complicated. Different types of nonlinear programs, determined by structural properties of the objective and constraint functions, are amenable to different types of algorithms. The subsequent sections sketch the main theory and most popular algorithmic ideas. A comprehensive treatment of this vast topic is beyond the scope of this textbook. We refer the reader to the excellent references Bertsekas (1999), Guler (2010), and Nocedal and Wright (2006) for more details. Numerical Nonlinear Programming Solvers There are numerous software packages for solving nonlinear programs. The following are some popular ones. We list them according to the class of algorithms (discussed in Section 20.4) that they are based on: (1) CONOPT, GRG2, Excel SOLVER. These solvers are based on the generalized reduced-gradient method . (2) MATLAB optimization toolbox, SNOPT, NLPQL. These solvers are based on sequential quadratic programming. (3) MINOS, LANCELOT. These solvers are based on an augmented Lagrangian approach. (4) MOSEK, LOQO, IPOPT. These solvers are based on interior-point methods.

To solve structural optimization problems, it is necessary to integrate a structural analysis package and an optimization package. Since most structural analysis packages suffer from closeness of system, it is very difficult to integrate it with an optimization package. To overcome the difficulty, we propose a possible alternative, DAMDO, which integrate Design, Analysis, Modeling, Definition, and Optimization phases into an integration environment as follows. (1) Design first generate many possible structural design alternatives. Each design alternative consists of many design variables X. (2) Analysis employ the structural analysis software to analyze all structural design alternatives to obtain their internal forces and displacements. They are the response variables Y. (3) Modeling employ artificial neural networks to build model Y = f(X) to obtain the relationship functions between the design variables X and the response variables Y. (4) Definition employ the design variables X and the response variables Y to define the objective function and constraint functions. (5) Optimization employ the optimization software to solve the optimization problem consisting of the objective function and the constraint functions to produce the optimum design variables X*. Optimization of truss structures was used to validate the DAMDO approach. The empirical results show that the truss optimization problems can be solved by the DAMDO approach, which employ neural networks to integrate the structural analysis package and optimization package without requiring direct integration of the two packages. This approach is promising in many engineering optimization domains which need to couple an analysis package and an optimization one to obtain the optimum solutions.

The nonaxisymmetric problem of natural vibrations of radially polarized hollow cylinders made of functionally gradient piezoelectric materials is solved. The properties of the material change continuously along a radial coordinate according to an exponential law. The lateral surfaces of the cylinder are free of external tractions and short circuited by electrodes. After separation of variables and representation of the components of the displacement vector in the form of standing circumferential waves, the initially three-dimensional problem is reduced to a two-dimensional partial differential equation problem. By using the method of spline-collocations with respect to a longitudinal coordinate, this two-dimensional problem is reduced to a boundary-value problem for the eigenvalues expressed in terms of ordinary differential equations. This problem is solved by the stable discrete-orthogonalization technique in combination with a step-by-step search method with respect to the radial coordinate. Results were obtained numerically and subsequently analyzed in this paper.

The Karush–Kuhn–Tucker (KKT) optimality conditions are necessary and sufficient for a convex programming problem under suitable constraint qualification. Recently, several papers (Dutta and Lalitha in Optim Lett 7(2):221–229, 2013; Lasserre in Optim Lett 4(1):1–5, 2010; Suneja et al. Am J Oper Res 3(6):536–541, 2013) have appeared wherein the convexity of constraint function has been replaced by convexity of the feasible set. Further, Ho (Optim Lett 11(1):41–46, 2017) studied nonlinear programming problem with non-convex feasible set. We have used this modified approach in the present paper to study vector optimization problem over cones. The KKT optimality conditions are proved by replacing the convexity of the objective function with convexity of strict level set, convexity of feasible set is replaced by a weaker condition and no condition is assumed on the constraint function. We have also formulated a Mond–Weir type dual and proved duality results in the modified setting. Our results directly extend the work of Ho (2017) Suneja et al. (2013) and Lasserre (2010).

Many cutting problems on two- or three-dimensional objects require that the cuts be orthogonal and of guillotine type. However, there are applications in which the cuts must be orthogonal but need not be of guillotine type. In this paper we focus on the latter type of cuts on rectangular bins. We investigate the so-called \(L\)-approach, introduced by Lins et al. (J Oper Res Soc 54:777–789, 2003), which has been used to tackle, among others, the pallet loading problem and the two-dimensional unconstrained knapsack problem. This approach concerns a method to generate two \(L\)-shaped pieces from an \(L\)-shaped piece. More recently, Birgin et al. (J Oper Res Soc 63(2):183–200, 2012), raised two questions concerning this approach. The first question is whether one may restrict only to cuts on raster points (rather than on all discretization points), without loss in the quality of the solution. We prove that the answer to this question is positive. The second question is whether the \(L\)-approach is an optimal method to solve the unconstrained knapsack problem. We show an instance of the problem for which this approach fails to find an optimum solution.

Recent studies on thermoelectric (TE) systems indicate that the existence of high figure of merit (ZT) materials alone is not sufficient for superior system performance and an integrated system level analysis is necessary to attain such performance. This is because there are numerous design parameters at various levels of the system that are randomly variable in nature that could affect the overall system performance. In this work the effect of stochasticity in design variables at various levels of a TE system has been studied and analyzed to attain optimal design solutions. Starting with stochasticity in one of the environmental variables, a progression was made towards studying the coupled effects of stochasticity in multiple variables at environmental and heat exchanger levels of a thermoelectric generator (TEG) system. Research and analysis tools were developed to incorporate stochasticities in single or multiple variables individually or simultaneously to study both the individual and coupled affects of input design variable stochasticities (probabilities) on output performance variables. Results indicate that normal or Gaussian distribution in input design parameters may not produce Gaussian output parameters. Also when the stochasticities in multiple variables are coupled, the standard deviations in performance parameters are magnified, and their means/averages deviate moremore » from the deterministic values. Although more studies are required to quantify the parameters for design modifications, the studies presented in this paper affirm that incorporating stochastic variability not only aids in understanding the effects of system design variable randomness on expected output performance, but also serves to guide design decisions for optimal TE system design solutions that provide more robust system designs with improved reliability and performance across a range of off-nominal conditions.« less

In this paper, we will give a p-version finite element method for the one-dimensional nonlinear elastoplasticity problem. A family of admissible constitutive laws based on the so-called gauge function method is introduced first, and then a p-version semi-discretization scheme is presented. We show that the existence and uniqueness of the solution for the semi-discrete problem can be guaranteed by using some special properties of the constitutive law. Finally, we show that as the polynomial degree of the elements ${\bf p} \to \infty $ oo the solution of the semi-discrete problem will converge to the solution of the continuous problem. The p-version convergence analysis given here is only for one-dimensional problems with a linear hardening constitutive law hypothesis. However, it also gives us a very useful idea for the study of the hp-version finite element method for the one-dimensional problem with more general constitutive laws or even two-dimensional problems.

The two-dimensional rectangular cutting problem with defects is discussed. The goal is to cut a small rectangular block of a given height and width from a large rectangular object containing a plurality of defects on the premise of satisfying several constraints, so that the sum of the area of the cut small rectangular blocks are maximized. The constraint is that each cutting operation must be guillotine, and the number of small rectangular blocks of each type is not limited and maintains a given direction. The problem is regarded as covering the original plate with small rectangular block, and a Quick Heuristic-Dynamic Programming (QHDP) algorithm is proposed. Firstly, a one-dimensional knapsack problem is established according to the height and width of small rectangular blocks, respectively, and an efficient discretization sets is generated respectively. Then, each value in the sets is used as a possible cutting line coordinate for sub-problem division. The algorithm calculates 14 typical examples. The experimental results show that it has obtained the optimal solution of all the examples, and the calculation time is less than one tenth of the best algorithm in the current literature. The algorithm complexity is analyzed and proved.