Adaptive control method for discrete neural network of hypersonic aerocraft on basis of equivalence model

Abstract

The invention discloses an adaptive control method for a discrete neural network of a hypersonic aerocraft on the basis of an equivalence model. The adaptive control method is used for solving the technical problem that discrete adaptive control for an existing hypersonic aerocraft is difficult to implement in engineering. The adaptive control method includes converting a height subsystem model of the hypersonic aerocraft into a strict-feedback form at first, and creating a discrete strict-feedback form of an original system by an Eulerian method; constructing the equivalence model of the original system under the consideration of causal relationship of the system; and sufficiently utilizing the equivalence model for future output so as to realize a prediction function. Besides, large quantities of system states and future output information are involved in model analysis, so that necessary and rich information is provided for designing the neural network. Unknown portions of the system are estimated owing to a nominal system, error feedback and the neural network, and a controller is designed according to back-stepping strategies; and non-causal problems are effectively avoided by control characteristics of a computer and the controller obtained by model conversion.

Description

technical field [0001] The invention relates to a hypersonic aircraft control method, in particular to an equivalent model-based discrete neural network adaptive control method for a hypersonic aircraft, which belongs to the field of aircraft control. Background technique [0002] Due to its outstanding flight capability, hypersonic aircraft makes global real-time strike possible, so it has attracted widespread attention at home and abroad; NASA X-43A test flight successfully confirmed the feasibility of this technology; affected by its own complex dynamic characteristics and the airframe With the integrated design of the engine, the coupling between the elastic body, propulsion system and structural dynamics of the hypersonic vehicle is stronger, and the nonlinearity of the model is also higher; sensitive. [0003] The control of hypersonic vehicles is mostly concentrated in the continuous domain; with the development of computer technology, the control system of hypersoni...

AI Technical Summary

Problems solved by technology

[0004] The article "Back-stepping-Based Discrete Controller Design for Hypersonic Vehicle" (Gao Daoxiang, Sun Zengqi, Du Tianrong, "Control and Decision-Making", Volume 24, Issue 3, 2009) adopts the second method to transform the altitude subsystem into a four-dimensional First-order model, by designing virtual control variables (track angle, pitch angle, and pitch angular velocity) to control the previous state quantity, and finally use the rudder angle to control the pitch angular velocity; this method only uses the information of the current moment and the next moment , the nominal system is used to approximate the future information of the required virtual control quantity; due to the uncertainty of the system dynamics parameters, the future information of the system state cannot be known, and the future of the virtual control quantity cannot be obtained according to the relevant expression Information, there are non-causal problems, it is difficult to realize engineering

Images