Expert System for Aerial Vehicle Deployment System Selection
thesisposted on 22.05.2021, 18:03 by Ming Yui Edwin Lau
An expert system is a programmable device developed to provide automation for engineering problem solving. It is composed of artificial intelligence modules, subroutine functions, and databases. Under this framework, a design process is proposed to assist the conceptual design of aerial vehicles' deployment systems. The problem is first defined by a set of design requirements for take-off, landing, and cruise. The values are then translated to a set of performance parameters needed for the design process via a newly developed parametric search algorithm. Such parameters are categorised by a fuzzy inference module to determine the most suitable deployment-propulsion system, for conventional and V/STOL vehicles. Through the use of linear and neural network regression, a number of aerodynamic terms are estimated to support flight mechanics analyses, where the optimal take-off and landing thrust vectors are determined. Engine specifications are deduced in terms of unit thrust, weight, bypass ratio and dimension. The design process demonstrates effectiveness in sizing engines for V/STOL operations.