Font Size: a A A

Human-Computer Interactive Evolutionary Design Method And Its Application To Spacecraft Vessel Layout Scheme Design

Posted on:2002-04-09Degree:DoctorType:Dissertation
Country:ChinaCandidate:Z Q QianFull Text:PDF
GTID:1102360065956010Subject:Mechanical design and theory
Abstract/Summary:PDF Full Text Request
The concern of this paper is the optimization theory and strategy of layout scheme design of complex engineering systems and their applications to layout optimization of the vessels of complex spacecraft.The layout scheme design of spacecraft vessel is to study how to arrange the instruments and equipment inside (or outside) the vessel while satisfying all kinds of engineering constraints. It is an engineering subject of great importance and urgency and the researches on it are far from enough. Great challenges and barriers exist in both theory and practice due to the NP-hard of layout problems in computation complexity and alarming difficulties in the design of spacecraft themselves.The layout scheme design of spacecraft vessel belongs to 3D packing problem with behavioral constraints, and it is also a complex engineering system problems with inscrutability, uncertainty, high nonlinearity, quality and quantity. It is almost impossible to solve it completely by human or computer alone. The idea of "Man-Machine Synergy" originated from Lenat and Feigenbaum in 1991 and the similar concept of "Man-Machine Cooperation" proposed by Xueshen Qian and Ruwei Dai at the very year turn out to be a promising approach to complex engineering problems and automatic design in 21st century. But there is a key problem that has not been commendably solved yet. That is how to practically implement the cooperation between man and computer. Against the background of layout optimization of spacecraft vessels, frist this paper proposes a human-computer interaction genetic algorithm (HCIGA) which combines the advantages of human and computer at a level of algorithm. Then a strategy of human-computer cooperation design based on HCIGA is developed to solve combinatoral explosion hi optimization problems and layout scheme design in complex engineer system. The proposed methodologies demonstrate superior performance on numerical examples and their application to spacecraft vessel scheme design.In this dissertation, some original concepts and theories are proposed which are summarized in the following.1. A optimization algorithm plays a dominant role in solving layout optimization problems and is also an important factor in evaluating a research work. This paper proposes a hybrid coarse-grained genetic algorithm (HCGGA) in which all chromosomes are ranked according to their fitness values and divided into several subgroups. Each subgroup, whose values of punishmentcoefficient, crossover and mutation operators are different from other subgroups, operates independently. Otherwise the strategies of "the same exclude" and "the best live" are adopted. Enough examples indicate that HCGGA is better than coarse-grained genetic algorithm. It is very HCGGA that human-computer interactive genetic algorithm (HCIGA) in this paper bases on.2. To make the best of human and computer and practice the ideas of "man-machine synergy" and "man-machine cooperation", a human-computer interactive genetic algorithm is put forward. In HCIGA, the artificial individuals (AIs) are taken as a part of the original chromosome population, and then the worse individuals are replaced by new AIs real-timely during computation process. The cooperation between human and computer is carried out by genetic operations, including selection, crossover and mutation. The HCIGA is just a cell in the following cooperation design method.3. In order to realize multi-person design and decision, as well as to ensure the characteristic of cooperative operation can be fully exerted, the thought of cooperative design is combined with HCIGA. Consequently human-computer interactive evolutionary cooperation design (HCIECD) is constructed, which can simplify the conformation of database and enforce flexible cooperation among many persons in conventional collaboration design. There are two modules in the method, which are design layer and decision layer modules. The socket technology of network programming, and client/server mode are introduced to constitute a...
Keywords/Search Tags:genetic algorithm, human-computer interaction, evolutionary computation, cooperation design, spacecraft, layout, scheme design
PDF Full Text Request
Related items