TA186 : Optimization of Structures by Combinatorial mexta-heuristic Methods
Thesis > Central Library of Shahrood University > Civil & Architectural Engineering > PhD > 2013
Authors:
Abstarct: Reach to maximum efficiency and minimum expenditure by choosing a proper system is an art in engineering. This issue results in appearance and development of optimization in different fields of engineering. In this regard, this thesis is established on five investigative lines to develop the optimization of structures using mexta-heuristic methods and subsequently operation improvement of these methods. Novel achievements are obtained by trying purposive steps. Therefore, some of most common mexta-heuristic methods in optimization problems of structures are studied in first investigative line. In this part, different dimensions and challenges of each method and sensitivity of algorithms to current parameters are trying to be determined and relevant suggestions are also proposed to improve the operation and to use the algorithms properly. In second investigative line, Multi Search Method (MSM) is offered to advance the operation of mexta-heuristic algorithms using the results of previous line. Proficiency of this idea is evaluated in Genetic Algorithm baxsed on various current structures. Afterwards, Multi Heuristic Search Method (MHSM) is presented to develop MSM takes advantage of different algorithms of mexta-heuristic as the structure of each island. This method is one of the most successful algorithms for searching design spaces with remarkable precision and it could be beneficial in important fields of engineering. This method is studied in various domains and issues. Third investigative line proposes novel idea of sampling the searching space inspired by meshing process in finite element method. This concept, utilizing design space partitioning and seek the search space in two steps (global and local), can improve the operation speed and obtain desirably precise answers. Hybrid algorithm of HACOHSGA, which is resulted by combination of ACO, HS and GA methods, is offered as forth line of investigation. This method also has suitable speed for process of optimum design and employs the advantages of each method appropriately. Finally, at the last investigative step of this thesis, Moving Points (MP) method is suggested after a proper observation to mexta-heuristic optimization methods. This method carries out the optimization operation as an evolutionary process by mapping members and designs of each population to vector space and moving baxsed on current population features. Characteristic of MP methods is stability in different performances, accordingly, after one performance can rely on obtaining suitable answer of optimization.
Keywords:
#Optimization #mexta-heuristic algorithms #Multi search method #Multi heuristic search method #Sampling search space method #HACOHSGA method #Moving points method. Link
Keeping place: Central Library of Shahrood University
Visitor:
Keeping place: Central Library of Shahrood University
Visitor: