In this paper, a new python package (optipoly) is described that solves box-constrained optimization problem over multivariate polynomial cost functions. The principle of the algorithm is described before its performance is compared to three general purpose NLP solvers implemented in the state-of-the-art Gekko and scipy packages. The comparison show statistically better best solution provided by the algorithm with significantly less computation times. The package will be shortly made freely and easily available through the simple (pip install) process.
We use Algorithmic Differentiation (AD) to implement type-generic tangent and adjoint versions of $$ y=\sum_{i=0}^{n-1} x_{2 i} \cdot x_{2 i+1} $$ in C++. We run an instantiation for char-arithmetic and we print the gradient at $(101~77~114~114~32~121~109~88~115~97)^T$ to std::cout, yielding the output ``Merry Xmas''. Similar instantiations of type-generic second-order tangent and second-order adjoint versions of $$ y=\frac{1}{6} \cdot \sum_{i=0}^{n-1} x^3_{i} $$ yield ``Happy 2026'' at $(72~97~112~112~121~32~50~48~50~54)^T.$ Prepend a sufficiently large number of zeros to the input vector to explore the varying run times of the different derivative codes. The entire source code can be found on https://github.com/un110076/SeasonsGreetings.
Many problems in science and engineering can be rigorously recast into minimizing a suitable energy functional. We have been developing efficient and flexible solution strategies to tackle various minimization problems by employing finite element discretization with P1 triangular elements [1,2]. An extension to rectangular hp-finite elements in 2D is introduced in this contribution.
In this short contribution we introduce a new procedure to recover the stress and strain fields for particle systems by mechanical models. Numerical tests for simple loading conditions have shown an excellent match between the estimated values and the reference values. The estimated stress field is also consistent with the so called Quasicontinuum stress field, which suggests its potential application for scale bridging techniques. The estimated stress fields for complicated loading conditions such as defect and indentation are also demonstrated
In this work an efficient strategy for yield optimization with uncertain and deterministic optimization variables is presented. The gradient based adaptive Newton-Monte Carlo method is modified, such that it can handle variables with (uncertain parameters) and without (deterministic parameters) analytical gradient information. This mixed strategy is numerically compared to derivative free approaches.
Stefan Kurz, Sebastian Schöps, Gerhard Unger
et al.
We solve Maxwell's eigenvalue problem via isogeometric boundary elements and a contour integral method. We discuss the analytic properties of the discretisation, outline the implementation, and showcase numerical examples.
The present paper aims at analytically evaluating the natural frequencies of cracked slender masonry elements. The problem is dealt with in the framework of linear perturbation, and the small oscillations of the structure are studied under loaded conditions, after the equilibrium for permanent loads has been achieved. A masonry beam element made of masonry-like material is considered, and some explicit expressions of the beam's fundamental frequency as a function of the external loads and the amplitude of imposed deformations are derived. The analytical results are validated via finite-element analysis.
This paper proposes a data-driven approach for computing elasticity by means of a non-parametric regression approach rather than an optimization approach. The Chebyshev approximation is utilized for tackling the material data-sets non-linearity of the elasticity. Also, additional efforts have been taken to compare the results with several other state-of-the-art methodologies.
Nowadays hydroelectric energy is one of the best energy sources: it is cleaner, safer and more programmable than other sources. For this reason, its manage could not be done in an approssimative way, but advance mathematical models must be use. In this article we consider an overview of the problem: we introduce the problem, then we show its simplest but quite exaustive mathematical formulation and in the end we produce numerical results under the ipothesis that all input are deterministic.
In this paper, attempt is made to solve a few problems using the Polynomial Point Collocation Method (PPCM), the Radial Point Collocation Method (RPCM), Smoothed Particle Hydrodynamics (SPH), and the Finite Point Method (FPM). A few observations on the accuracy of these methods are recorded. All the simulations in this paper are three dimensional linear elastostatic simulations, without accounting for body forces.
In this paper we present applications of different machine learning algorithms in aquaculture. Machine learning algorithms learn models from historical data. In aquaculture historical data are obtained from farm practices, yields, and environmental data sources. Associations between these different variables can be obtained by applying machine learning algorithms to historical data. In this paper we present applications of different machine learning algorithms in aquaculture applications.
In this little vision paper we analyze the human immune system from a computer science point of view with the aim of understanding the architecture and features that allow robust, effective behavior to emerge from local sensing and actions. We then recall the notion of fractionated cyber-physical systems, and compare and contrast this to the immune system. We conclude with some challenges.
In this paper we present OSC, a scientific workflow specification language based on software architecture principles. In contrast with other approaches, OSC employs connectors as first-class constructs. In this way, we leverage reusability and compositionality in the workflow modeling process, specially in the configuration of mechanisms that manage non-functional attributes.
Considering the natural ventilation, the thermal behavior of buildings can be described by a linear time varying model. In this paper, we describe an implementation of model reduction of linear time varying systems. We show the consequences of the model reduction on computing time and accuracy. Finally, we compare experimental measures and simulation results using the initial model or the reduced model. The reduced model shows negligible difference in accuracy, and the computing time shortens.