Begoña González

A Flexible Evolutionary Agent: cooperation and competition among real-coded evolutionary operators

Since it has currently became essential to design more efficient and robust alternative techniques to solve hard optimisation problems in industry or science, and of easy use for practitioners, here a new way of developing simple Artificial Intelligence based Evolutionary Algorithms will be introduced. Our evolutionary computational implementation is a new idea in optimisation. Any evolutionary operators and their associated parameters from well-established evolutionary methods can be considered in such a way that the entire algorithm or intelligent agent-based software performs with very high efficiency without a prior need to investigate which method will be the best for a given optimisation problem.The implementation presented, named Flexible Evolution (FE), has capacity to adapt the operators, the parameters and the algorithm to the circumstances faced at each step of every optimisation run and is able to take into account lessons learned by different research works in the adaptation of operators and parameters. The FE uses Artificial Intelligence concepts to manage internal procedures to adopt decisions and correct the wrong ones. Our aim in this paper will be to give the keys to design these types of procedures, and more specifically, to find the way of achieving an optimum performance of the operators involved in the search, in our case by means of a function included in our algorithm called Sampling Engine. An early implementation has been already developed and tested in our previous works [66–68], so in this paper, new results of a second software implementation are presented comparing the results with those obtained by other methods, using well-known hard test functions.

Viscosity of Liquid Water via Equilibrium Molecular Dynamics Simulations

Molecular dynamics simulations were carried out for liquid water in the NVE ensemble for calculating shear and bulk viscosities. We used two different intermolecular potential functions for the water dimer: the empirical SPCE model and the ab initio NCC one. The results obtained are compared with the available experimental values, and show that for a more accurate description of these macroscopic liquid properties a polarizable (rigid or flexible) interaction potentials should be employed. Such models, based on ab initio data, have been recently developed, and their incorporation for the viscosity calculations is discussed.

Numerical Simulation of Transonic Flows by a Double Loop Flexible Evolution

Publisher Summary The chapter presents a methodology that uses evolutionary algorithms to obtain fluid velocities for potential flows inside a nozzle. A genetic algorithm with real encoding which, with the technique of partial sampling of the solution nodes, presents a number of advantages over other evolutionary algorithms is proposed in the chapter. These include the avoidance of a rigid connectivity to discretize the domain, thus making it a meshless method. The chapter also proposes a double-loop strategy with EAs which enables improvement of the methodology. The amount of computer storage is low and convergence behavior is good as the qualitative characteristics of the solution are taken into account in the algorithm. The artificial intelligence based on evolutionary algorithms can bring together lessons learned by developers and decision tools into advanced software developments, which can facilitate their use to solve complex challenging problems in industry.

New lightweight optimisation method applied in parts made by selective laser sintering and Polyjet technologies

The continuous evolution of materials and technologies of additive manufacturing has led to a competitive production process even for functional parts. The capabilities of these technologies for manufacturing complex geometries allow the definition of new designs that cannot be obtained with any other manufacturing processes. An application where this capability can be exploited is the lightening of parts using internal structures. This allows to obtain more efficient parts and, at the same time, reduce the costs of material and manufacturing time. A new lightweight optimisation method to optimise the design of these structures and minimise weight while keeping the minimal mechanical properties is presented in this paper. This method is based on genetic algorithms, metamodels and finite element analysis (FEA). This combination reduces the number of FEA simulations required during the optimisation process, thereby reducing the design time. This methodology is experimentally applied to a reference geometry oriented both for selective laser sintering (SLS) and Polyjet technologies. In both cases, an optimised and a non-optimised design are manufactured and tested in order to experimentally compare the stiffness results between them. The optimum design achieved a specific stiffness 72.82% higher than the non-optimised design in the SLS case study, and 3.14 times higher in the Polyjet case study.

Lightweight Optimization for Additive Manufacturing Parts Based on Genetic Algorithms, Metamodels and Finite Element Analysis

Additive manufacturing (AM) has become in a competitive method for short series production and high flexibility applications even for functional parts. Few constraints in the manufacturing process involve a great design freedom, allowing minimization of weight by using internal cellular and lattice structures, while minimal mechanical requirements are kept. Weight minimization implies a lower use of material and hence a reduction in manufacturing time, leading to a cost reduction. However, design optimization requires a greater effort in the design process, which also results in more costs. In order to reduce the design process, an optimization method based on genetic algorithms (GAs) and computer aided design/finite element method (CAD/FEM) simulations is proposed to optimize the cellular structure design and minimize the weight for AM parts. New optimization strategies based on GAs combined with surrogate models are evaluated and compared to reduce as much as possible the number of FEM simulations.

Estimation of the Electricity Demand of La Palma Island (Spain)

Historical data of electricity demand in La Palma island (Spain) were collected and electricity demand estimates conducted by different organizations were sought. Some factors that could affect these data were studied and its predictions by the next years were looked for. The idea was to use these factors as explanatory variables in order to predict the values of electricity demand in the next years. Moreover, with the aim of minimizing the limitation of predicting the future based only on relationships between variables that occurred in the past, it has been considered the annual demand forecast for various scenarios, taking into account, for each of them, different variations of the explanatory variables. All that with the goal that the estimate band of the demand for each year includes the real future demand with high probability. This provided a prediction model that takes into account population and gross domestic product. Results and their graphical representation along with the other estimates found are presented. A similar approach was carried out to predict peak powers.

Testing Some Alpha-Models of Turbulence on Wing Profiles

In this paper some numerical simulations of the Navier-Stokes Equations (NSE) to test the novel NS-α and NS-ω turbulence models [1, , which conserve energy, enstrophy, and helicity, are presented. These algorithms verify more conservation properties than other implementations of the NSE, however their rotational form [ makes the scaling study of the coupling between the velocity and pressure errors with respect to the Reynolds number, a very interesting research line. Nowadays we are designing a wing profile in the context of Unmanned Aerial Vehicle (UAV) on incompressible flow conditions [. First a genetic algorithm (GA) is used to obtain the optimized design geometry and then the NS-α and NS-ω turbulence models are run to study its performance for different attack angles. The GA objective function evaluates the general potential theory of each wing section considered, because that requires less computational cost than the alternative of solving the NSE, and a wing design method proposed in [ is applied. Thus the optimized design geometry was found by evaluating the potential flow of all candidate solutions generated from the selection, crossover and mutation operators in each GA iteration. It takes the order of hundreds of simulations per iteration to evaluate all candidate solutions. Summarizing, two practical applications for a UAV are presented: the optimized design of an airfoil for environmental purposes, named CEANI airfoil, and the application of relevant turbulence models as NS-α and NS-ω in order to evaluate with accuracy the lift, drag and maximum angle of attack.

Numerical Simulation of Transonic Flows by a Flexible and Parallel Evolutionary Computation

In this work we propose a Flexible and Parallel Evolutionary Computation to solve a problem of Aeronautic interest, which is the non-linear full potential flow problem inside a nozzle. As the target to verify a numerical scheme for nodal points is equivalent to minimize a corresponding objective function, our methodology consists in using Evolutionary Algorithms as a numerical simulation method. Moreover, as Parallel Evolutionary Algorithms are demanded for an effective reduction of computation time, an Island or Coarse-Grained Parallel model is considered. The methodology proposed opens the possibility to solve the problem from a more local point of view as domain decomposition and permits to incorporate easily theoretical mathematical foundations of the problem. These are powerful characteristics of the method.

Chapter 24 – Optimal placement of wastewater outfalls using an internet-based distributed computing environment and a 3D finite volume model of pollutant dispersion in offshore waters

Publisher Summary This chapter presents solution for a state constrained optimal control problem using an intelligent system based on evolutionary self-learning computation. The problem is solved under a parallel implementation linked to flexible evolution agent (FEA) using an Internet-based distributed computing environment. The aim of the FEA is to secure the highest benefit possible from the exploitation of the stored information, and in addition, to incorporate new procedures to facilitate internal decision-making to be automatically made in the optimization process. This chapter tries to find the optimal locations of liquid waste dumping sites in offshore waters to verify that the concentration of a considered pollutant does not exceed a pre-established limitation, either at a specific distance from the shore or in different protected areas such as fish hatcheries. The results obtained demonstrate that it is possible to use advanced evolutionary optimization tools placed on different computer platforms on the Internet to solve a real complex problem in a coordinated way.