Use this URL to cite or link to this record in EThOS: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.512734
Title: Desaid : the development of an expert system for aircraft initial design
Author: Nah, Seung-Hyeog
Awarding Body: Cranfield University
Current Institution: Cranfield University
Date of Award: 1991
Availability of Full Text:
Access through EThOS:
Access through Institution:
Abstract:
As all engineering works are a blend of theory and empiricism, aircraft design, by its nature, represents a mixture of aircraft designer's knowledge obtained from aeronautical engineering disciplines and its usage combined with his experience. This means not only the application but also the integration of all the fundamental knowledge of aerodynamics, structure, propulsion, stability and control, operational and economic aspects, etc., based upon the designer's judgements and experiences. Thus the tasks involved in designing an aircraft configuration, without exception, show complex characteristics, considering the fact that aircraft configuration design means the integration of components such as lifting surfaces ( wing ), fuselage, power-plant, control surfaces ( tail or canard ), and undercarriage. The discrepancies and mismatches among the aircraft components make the configuration design iterative, repetitive, and thus time - consuming. Such complexities of configuration design processes often require compromise, through trial and error, to resolve conflicts between the major design areas. Moreover, it takes tens of years to become a experienced design expert whose sound judgement, based upon experience and profound knowledge, influences greatly the aircraft configuration design. The differences in judgements depend upon the designers' imagination and experience, and they are the cause of variations in aircraft configurations. Therefore, the efforts were made to overcome those difficulties which hinder the aircraft designer from making the task of configuration design more efficient, and further to assist the aircraft designer in getting an easy and interactive preliminary aircraft configuration without always relying upon design experts. Hence the current research project is directed at the development of an expert system for aircraft design. This involves the use of Artificial Intelligence and its programming language called PROLOG ( PROgramming in LOGic ). The research started from a thorough analysis of the major component design areas and has constructed an EXPERT SYSTEM to find out the efficient Control Mechanism which can search intensively for the solutions to design problems for all types of aircraft; civil and military, subsonic and supersonic, conventional and unconventional, etc. In addition, users can have access to the explanations of important items such as a design process, terminology, equations, and results. The explanation facility is one of the most important functions of Expert Systems. Partly due to the limit of computer capacity and partly due to the magnitude of laborious program execution at this stage, the system implementation has focused on the high - subsonic, conventional and jet transport aircraft categories. The approach taken was to find an efficient and effective control mechanism (i.e. an Inference Engine ), which integrated the Parametric Study, Wing Design, Fuselage Design, Engine Design, Tail Design, Undercarriage Design, Weight Analysis and Cost Analysis into a whole configuration system. The comparison between Expert System results and existing aircraft such as Boeing 747, Airbus 300 series, BAe 146 series, McDonnell Douglas MD series, etc., showed the permissible ranges of error to be within about 10 %. Such results enable the Expert System to claim that it can act as a useful design tool for the aircraft designer in the initial stage of aircraft configuration design. Finally, the author believes that the control mechanism devised for this Expert System can be used as a sound basis for extending the Expert System to include other types of aircraft and further to encompass spacecraft design, as the designer wishes.
Supervisor: Morris, Alan J. Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.512734  DOI: Not available
Share: