Hybrid Metaheuristic and Fuzzy Optimisation for Complex Engineering Applications
Abstract
Fuzzy modelling for engineering systems is one of the crucial techniques for modelling, analysis and control of complex systems in which uncertainty, imprecision, vagueness, nonlinear relations and incomplete information make it difficult or inadequate to describe the systems using conventional mathematics. In many engineering systems, however, it is difficult to define the relationship between input and output variables exactly, due to variations in the environment, inaccuracies in measurement, human manipulation, component deterioration, manufacturing variations, or lack of understanding of the system dynamics. Typically, the elements of a fuzzy modelling system are stated variables, output variables, membership functions, rule base, fuzzy inference mechanism, and defuzzification method. Fuzzification is the process of converting the input variables into degrees of membership from precise numerical values. The membership functions specify the degree of membership of a given numerical value in a linguistic category. Triangular membership functions, trapezoidal membership functions, Gaussian, bell-shaped, sigmoid and generalised bell functions are common ones. The selection of membership function is based on the features of an engineering problem, the information at hand, the precision required, and the modelling aim. The fuzzy inference system processes a set of if–then rules, which are relations between system variables, after the fuzzification. The final stage of a conventional fuzzy inference process is the defuzzification which is the conversion of the fuzzy output to a numerical value suitable for engineering actions. Depending on the structure of the fuzzy system, some most widely used methods are centroid, weighted average, mean of maxima and maximum membership. In mechanical engineering, electrical engineering, civil engineering, chemical engineering, aerospace engineering, manufacturing, robotics, automotive engineering, energy sector and environmental engineering, fuzzy modelling has thus become applicable.