Font Size: a A A

Self_repairing Control Of Flighter Based On Neural Networks Inverse System And Study Of The Visualization

Posted on:2003-01-29Degree:MasterType:Thesis
Country:ChinaCandidate:Q L ZhangFull Text:PDF
GTID:2168360062950282Subject:Control theory and control engineering
Abstract/Summary:PDF Full Text Request
This paper integrates inverse system method and adaptive control method with artificial neural networks technology and applies them into the study of the self-repairing flight control system.The self-repairing control laws with better real-time and robustness are provided and a set of software framework for the control system is also developed.In the paper,a feedback linearization method-inverse system method is applied to eliminate nonlinearities of flighter nonlinear dynamical system .and the approximate inversion of the flighter is provided by off-line trained neural networks;a direct adaptive self-repairing control reconfiguration approach based on inverse system is pretented,and the RBF NN on-line adaptive architecture is applied to compensate the inversion error due to modeling uncertainties and failures;a stable weight adjustment rule for the RBF NN is derived using a Lyapunov-like function.For the need of real-time ,a a -modification method is used in adaptive control reconfiguration;for the actuator dynamics,the dynamic nonlinear damping method is applied to robustify the controller architecture.Based on the above direct adaptive reconfigurable control method,the weights,centers and widths of the Gaussian function of RBF NN adptive adjust method,full adaptive control,is augmented;then the approach is extended to the large-scale interconnected systems and an additional robustifying adaptive term is applied to eliminate the interconnection effects.At last based on the technology of 3D_Visualization theory of the Object-Oriented of Visual C++ programming and OpenGL Graphic Library,a module of realization 3D graphics arithmetic is developed to the study of the self-repairing flight control system.
Keywords/Search Tags:self-repairing control, inverse system, adaptive control, artificial neural netwoks, visualization
PDF Full Text Request
Related items