Autori: Mario G. Perhinschi

Editorial: Proceedings of the AIAA Guidance, Navigation, and Control Conference, Portland OR, USA, August 1999, vol. 2, p. 790-797, 1999.

Rezumat:

In this paper, an optimal controller for the longitudinal channel of an autonomous helicopter model is designed by blending together two artificial intelligence techniques, genetic algorithms and fuzzy control. An evaluation index that captures the complex, constrained, multiple objective character of the problem was built based on several design requirements expressed in terms of the time response of the controlled system. The parameters of the fuzzy controller are optimized to maximize the evaluation index using a genetic algorithm. The parameters subject to optimization are: the shape and width of the membership functions, number of linguistic values, defuzzification method and scaling factors. The genetic algorithm is based on binary genetic representation, a roulette wheel selection technique with elitist selection strategy and classic genetic operators: mutation and crossover. The performance of the resulting optimal controller is compared with performance obtained with standard design. Observations are made regarding influences of fuzzy controller parameters on the general performance of the controlled system.

Cuvinte cheie: fuzzy control, genetic algorithms