Premium
A model‐based method for the synthesis and optimization of systems architectures
Author(s) -
ALBARELLO Nicolas,
WELCOMME JeanBaptiste
Publication year - 2012
Publication title -
incose international symposium
Language(s) - English
Resource type - Journals
ISSN - 2334-5837
DOI - 10.1002/j.2334-5837.2012.tb01456.x
Subject(s) - component (thermodynamics) , computer science , consistency (knowledge bases) , set (abstract data type) , constraint (computer aided design) , pareto principle , process (computing) , mathematical optimization , genetic algorithm , function (biology) , mathematics , artificial intelligence , machine learning , programming language , evolutionary biology , biology , thermodynamics , physics , geometry
The architecture design process requires to define several design alternatives and to compare them in order to choose the most relevant system architecture given a set of objectives. Nevertheless, designers are generally constraint to restrict their studies to a small set of alternatives due to time constraints and combinatorial aspects of the problem. The objective of our method is to assist them by automatically generating a larger number of design alternatives. The proposed algorithm will first generate alternatives by adding components to the architecture and allocate them to functions. The originality of our approach is that it takes into account two rules that ensure the viability (component–to‐component consistency) and the validity (function‐to‐component consistency) of the generated architectures. Once a set of consistent alternatives are generated, we use them as an input of a multi‐objective genetic algorithm to propose a set of Pareto‐optimal alternatives.