Enhancing Flight Control using Fuzzy Logic

Abstract

An increasing amount of attention has been paid in recent years to the application of intelligent control methodologies in Aeronautical systems (Stengel, 1993; Steinberg, 1992). The integration of qualitative knowledge (heuristics) and quantitative knowledge (analytical models) through fuzzy modeling and control offers in particular many new possibilities. For example, the aggregation of human pilot heuristics and analytical aircraft models has yielded simple and transparent flight control designs (Chui and Chand, 1992; (Chand and Chui, 1991; Larkin, 1985; Steinberg, 1993). Besides aircraft, fuzzy controllers have been developed for helicopters (Jiang et al., 1996; Phillips et al., 1996; Sugeno et al., 1993) and spacecraft (Berenji et al., 1993; Schram et al., 1996b; Woodard, 1996). Fuzzy modeling on the other hand has been applied to the identification of control system failures (Laukonen and Passino, 1994; Kwong et al., 1994; Youssef et al., 1996) and to the modeling of the human pilot (KrishnaKumar et al., 1995). However, it may be wondered why fuzzy logic could play a role in the control of aircraft. The dynamics of the aircraft are well described by first principles modeling and, although of nonlinear nature, the adaptation of controllers to varying dynamics has been successfully applied in practice. Therefore (classical) control structures are found in all present-day, flight-by-wire control systems (McRuer et al., 1973; Tischler, 1996). This is however only possible when extensive design efforts are executed for the various circumstances to be expected. Moreover, unexpected situations such as changing weather conditions and system failures are difficult to model and thus difficult to translate into appropriate classical control designs. For these reasons, the application of fuzzy logic can be justified as will be explored in this Chapter.