TABLE OF CONTENTS This chapter describes a selection of what are sometimes referred to as AI techniques (in fact these methods are, in general, empirically/numerically based rather than being analytically/theoretically based like the bulk of conventional control theory).
Neural networks are sets of interconnected artificial neurons that, very simplistically, imitate some of the logical functioning of the brain. After training, they can represent any algebraic non-linearity. They have to be trained by being presented with sufficient examples of the input-output behaviour that is desired, so to a large extent they can only represent existing data-generating phenomena by empirical equivalents.
Fuzzy logic emulates the reliable but approximate reasoning of humans, who, it is said, distinguish only six or seven different levels of any variable during decision making. Fuzzy logic algorithms can represent this style of reasoning by easily understood curves that are ideal for implementing those many control systems that are based on 'having a feel' or on 'rules of thumb' rather than on equations.
Genetic algorithms and genetic programming are powerful evolutionary search methods that can search for structures as well as numerical parameters. These qualities allow the methods to synthesise solutions to a wide variety of problems. The approaches rely heavily on imitating the methods of animal/human reproduction followed by natural selection. Because the methods can search amongst many alternative structures, they can also be regarded as design or synthesis methods.
Learning systems aim to emulate the human learning-by-experience mechanism so that a system can potentially learn to perform a...
