Pasar al contenido principal

Resultados relacionados

Resultados totales (Incluyendo duplicados): 2
Encontrada(s) 1 página(s)

RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

A particle swarm optimization algorithm with novelty search for combustion systems with ultra-low emissions and minimum fuel consumption

  • Martínez-Rodríguez, David|||0000-0002-2093-7840
  • Novella Rosa, Ricardo|||0000-0002-5123-6924
  • Bracho Leon, Gabriela|||0000-0002-9198-7044
  • Gómez-Soriano, Josep|||0000-0002-2742-9224
  • Villanueva Micó, Rafael Jacinto|||0000-0002-0131-0532
  • Fernandes, Cássio
  • Lucchini, Tommaso
  • Della Torre, Augusto
  • Hidalgo, J. Ignacio
[EN] The particle swarm optimization algorithm is primarily inspired by the natural behaviour of swarms and achieves important results in different applications. However, it is not exempt from stagnation in local optima and has a tendency to prematurely converge to them. Novelty Search is a concept that appeared recently in different fields of computational intelligence. It aims at exploring non-visited areas of the search space through solutions that bring novelty to already discovered solutions. The novelty of this work can be divided into two steps: on one side, this article proposes a variant of the particle swarm optimization algorithm which uses Novelty Search concepts to improve the algorithm¿s performance. Our proposal is first checked and compared using the CEC 2005 benchmark suite and then, we apply it to solve a real-world optimization problem: the design of a combustion system targeting the reduction of pollutant emissions and fuel consumption. The combustion chamber design phase usually is a complex and time-consuming process even with advanced supercomputers, since it depends on several input variables which are highly non-linear and with crossed interaction. Then, the second contribution of this work is to develop a methodology that couples a computational fluid dynamics (CFD) simulation tool with the new optimization algorithm for minimizing the specific fuel consumption of a compression-ignited engine, while constraining the NOx and soot emissions. A 3D-CFD model of the combustion system was built to predict and analyse the performance of the combustion system and hence, select the parameters with a higher impact on the system. The method reduces the computational time and includes tools for the automatic preparation of the input parameters and geometry of the system. The input parameters correspond to geometrical variables that control the bowl shape, the number of holes in the injector, the injection pressure, the swirl number and the exhaust gas recirculation rate. Results show how the simulation tool and the new PSO with Novelty Search algorithm allow us to obtain a new combustion system that minimizes the fuel consumption by 3%, simultaneously reducing NOx and soot emissions., This work has been partially supported by This work has been supported by the grant PID2020-1152 70GB I00 funded by MCIN/AEI/10.13039/501100011033; European Union FEDER funds; Spanish Ministerio de Economía y Competitividad through Grants PID2021-125549OB-I00 and PDC2022-133429-I00. The author C. S. Fernandes thanks the Universitat Politecnica de Valencia for his predoctoral contract (FPI-2019-S2- 20-555), which is included within the framework of Programa de Apoyo para la Investigacion y Desarrollo (PAID).
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Building uncertainty-aware mathematical models based on evidence from datasets using grammatical evolution optimization techniques: the case of the obesity dynamics

  • Parra, Daniel
  • Hidalgo, J. Ignacio
  • Velasco, J.M.
  • Villanueva Micó, Rafael Jacinto|||0000-0002-0131-0532
[EN] This paper introduces a methodology to build mathematical models based on evidence and data sets, considering data and model uncertainty. We study the evolution of obesity in the population, being obesity a consequence of the transmission of unhealthy lifestyle habits and behavioral patterns influenced by social networks (family, friends, peers, etc.). We propose a three-step methodology. First, we create a synthetic data set based on a previous model with real data. Then, we search for dynamic models based on difference equations that best fit the dynamics described by the dataset and their uncertainty (uncertainty-aware). To do this, we use a dynamic structured grammatical evolution algorithm (an algorithm that builds possible models) on which we have defined a grammar (set of possible expressions that can be part of the model). The definition of appropriate grammar is crucial because it allows us to build models that do not contradict the knowledge of the phenomenon studied. However, the data may suggest introducing new terms that indicate the influence of unknown factors. Finally, from among all the models obtained, we will algorithmically search for a selection of them that best describes the uncertainty of the data. This methodology can be applied to various scenarios with available datasets and a limited understanding of the phenomenon. It aims to generate models that not only achieve precision but also incorporate terms that correspond to identifiable processes, which can be explained within the context of the study problem., This work has been supported by the grants PID2020-115270GB-I00, PDC2022-133429-I00 and PID2021-125549OB-I00 funded by MCIN / AEI/ 10.13039 / 501100011033 and by European Union Next GenerationEU / PRTR.