DocumentCode :
2548044
Title :
Gain-scheduled flight control law design using a new fuzzy clustering technique
Author :
Mehrabian, Ali Reza ; Hashemi, S. Vahid ; Roshanian, Jafar
Author_Institution :
Concordia Univ., Montreal
fYear :
2007
fDate :
7-10 Oct. 2007
Firstpage :
795
Lastpage :
800
Abstract :
In this paper a gain scheduling methodology is proposed, which exploits a fuzzy modeling technology based on quasi-linear parameter varying (QLPV) description of a nonlinear model of a missile, to resolve two main deficiencies of classical gain scheduling approaches (restriction to near equilibrium operation and lack of a satisfactory interpolation mechanism). A new clustering approach (designated as fuzzy clustering based classification trees: FC2T) is employed in fuzzy modeling. It is shown that employment of FC2T in the development of fuzzy model for the system produces a lower estimation error than Gustafson-Kessel fuzzy clustering and a tree partitioning algorithm, while it does not confront by randomly initialization problem. Merits of the proposed fuzzy model-based gain scheduling technique are demonstrated in a demanding application. Simulation studies are reported to demonstrate the stability, the performance and the robustness of the designed fuzzy controller.
Keywords :
control system synthesis; fuzzy control; missile control; pattern clustering; scheduling; trees (mathematics); classification trees; fuzzy clustering; fuzzy model-based gain scheduling; fuzzy modeling; gain-scheduled flight control; missile; quasilinear parameter varying description; Aerospace control; Classification tree analysis; Clustering algorithms; Employment; Estimation error; Fuzzy control; Fuzzy systems; Interpolation; Missiles; Robust stability;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Systems, Man and Cybernetics, 2007. ISIC. IEEE International Conference on
Conference_Location :
Montreal, Que.
Print_ISBN :
978-1-4244-0990-7
Electronic_ISBN :
978-1-4244-0991-4
Type :
conf
DOI :
10.1109/ICSMC.2007.4414083
Filename :
4414083
Link To Document :
بازگشت