This paper describes an evolutionary hybrid model linking two algorithms: Particle Swarm Optimization (PSO) and Simulated Annealing (SA). The basic idea behind using a hybrid model is improving the reliability of the obtained results from our first model, namely MPSO (Modified PSO) based on PSO algorithm, by adding SA algorithm which is quite popular for its powerful feature of effective escaping from the trap of local minima. MPSO model uses the concept of evolutionary neighborhoods associated to parallel computation, to overcome to the two essential disadvantages of PSO: high running time and premature convergence. The presented algorithm has two essential operations: first running PSO algorithm in parallel using the new concept of evolutionary neighborhood to obtain a global best solution, then improving the results with SA algorithm to get the global optimal solution. By testing this hybrid algorithm (H-MPSO-SA) on a set of standard benchmark functions and according to the obtained results, the program have given satisfactory results of the hybrid model compared to the basic PSO and MPSO algorithms.
RÉSUMÉ.Reconnue pendant plusieurs années comme une métaheuristique stochastique performante dans la résolution des problèmes d'optimisation difficiles, la méthode d'optimisation par essaim particulaire « PSO » présente cependant des points faibles : le temps de calcul considérable et la convergence prématurée. Plusieurs études ont été menées pour trouver le jeu de paramètres qui conduit à de bonnes performances de l'algorithme. Dans ce papier, nous proposons une version de l'algorithme PSO, permettant d'améliorer ses performances en introduisant sa parallélisation associée à la notion du voisinage évolutif. L'algorithme proposé a été testé afin d'améliorer la performance et la fiabilité des structures mécaniques « le problème de transport d'électricité » ; plus précisément l'optimisation de la durée de vie du pylône d'une ligne de transport d'électricité ; l'objectif est de maximiser la résistance à la charge tout en réduisant le coût « la minimisation de l'utilisation des matériaux ». Dans nos expérimentations, les tests effectués sur le programme ont donné des résultats satisfaisants du modèle parallèle par rapport au modèle séquentiel. ABSTRACT. In this paper, we suggest a new version of PSO algorithm, that allows the amelioration of its performance by introducing its parallelization associated to the concept of evolutionary neighborhood. The main objective of our approach is to overcome to the two essential disadvantages of PSO: high running time and premature convergence. The proposed algorithm was tested in order to improve the performance and reliability of mechanical structures; more precisely on the electricity pylon example; the objective is to maximize resistance to load while reducing material usage and cost. Experimental results demonstrate that the proposed method is effective and outperforms basic PSO in terms of solution quality, accuracy, constraint handling, and time consuming.
In this paper, we present a probability study about spring of clutch structure. In the structure problems, the randomness and the uncertainties of the distribution of the structural parameters are a crucial problem. In the case of Reliability Based Design Optimization (RBDO), it is the objective is to play a dominant role in the structural optimization problem introducing the reliability concept. The RBDO problem is often formulated as a minimization of the initial structural cost under constraints imposed on the values of elemental reliability indices corresponding to various limit states. This paper proposes a new method for a modified particle swarm optimization algorithm (MPSO) combined with a simulated annealing algorithm (SA) and RBDO. MPSO is known as an efficient approach with a high performance of solving optimization problems in many research fields. It is a population intelligence algorithm inspired by social behavior simulations of bird flocking. Numerical results show the robustness of the MPSO-SA algorithm and RBDO.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.