Use status feedback and neural networks to compound control methods

Abstract

The invention discloses a hypersonic aircraft composite control method utilizing state feedback and neural network. The hypersonic aircraft composite control method utilizing state feedback and neural network is characterized by measuring the attack angle speed signal and the pitch angle speed signal of a hypersonic aircraft, designing a big gain state feedback controller, and then on the above basis, utilizing a neural network structure taking the sine function as the primary function, aiming at the strong nondeterminacy of a pneumatic parameter of the hypersonic aircraft, and designing an adaptive adjusting rule of the neural network weight, and finally forming a composite controller of state feedback and neural network of the hypersonic aircraft to realize tracking of an expected attack angle signal. The hypersonic aircraft composite control method utilizes state feedback and neural network utilizes the state feedback controller to guarantee the basic stability of the system while the neural network is used for compensating a modeling error which is caused by nondeterminacy of the pneumatic parameter, and finally can select the Lyapunov function to perform closed loop stability analysis on the whole system so as to guarantee the theoretical correctness of system design.

Description

technical field [0001] The invention belongs to the technical field of control of hypersonic aircraft, and relates to a composite control method for hypersonic aircraft using state feedback and a neural network. Background technique [0002] In recent years, the control of hypersonic vehicles has aroused widespread interest of researchers from various countries, mainly because the fast time-varying and strong nonlinear characteristics of the model have brought great challenges to the design of the control system. Especially the coupling between the engine and the elastic form and the control system increases the difficulty of the control design. [0003] However, due to the current experimental level, accurate modeling is difficult to obtain, so the model inevitably has unmodeled dynamics, and at the same time, the hypersonic vehicle is also faced with system uncertainties during real flight, so the controller must have strong robustness. Rodness, this is the main problem f...

AI Technical Summary

Problems solved by technology

[0002] In recent years, the control of hypersonic vehicles has aroused extensive interest of researchers from various countries, mainly because the fast time-varying and strong nonlinear characteristics of the model have brought great challenges to the design of the control system.

Especially the coupling between the engine and the elastic form and the control system increases the difficulty of the control design

[0003] However, due to the current experimental level, accurate modeling is difficult to obtain, so the model inevitably has unmodeled dynamics, and at the same time, the hypersonic vehicle is also faced with system uncertainties during real flight, so the controller must have strong robustness. Rodness, which is the main problem facing hypersonic control

Images