Brain-like pulse neural network control method applied to morphing aircraft

Abstract

The invention relates to a brain-like pulse neural network control method applied to a morphing aircraft, which is based on a morphing aircraft model considering strong coupling dynamic relationship between various aerodynamic coefficients and deformation rates in the morphing aircraft model. A certain number of spiking neuron models are introduced to form a small-range network for simulating a biological nervous system to give control information. At any flight moment, attitude and trajectory information received from the variant aircraft is processed by a pulse coding layer of the pulse neural network system and is input into the pulse neural network liquid state machine; and the internal topological structure of the pulse neural network is mapped and converted into a control instruction by a decoding layer formed by a memory-free decoding function after the connection weight is correspondingly changed. And finally, according to the current state quantity of the aircraft, online scheduling is carried out to calculate gain so as to generate a control command, instability caused by a variant structure in the real-time flight process of the variant aircraft is effectively eliminated, and a tracking instruction of a key state is completed.

Description

technical field [0001] The invention belongs to the field of flight control simulation of high-speed morphing aircraft, and more specifically relates to a method for controlling a morphing aircraft based on a brain-like impulse neural network, which can quickly and stably control the structure switching process of the morphing aircraft. Background technique [0002] The spiking neural network is a neural network model with biological significance and the operating mechanism most similar to the brain. In the research of brain-like science in recent years, SNN has always occupied a core position. Generally speaking, spiking neurons receive external information through dendrites, then process the received information inside the cell, and transmit it to the dendrites of other neurons through axons. Dendrites can thus be considered as input ports and axons as output ports. Due to the transmission of spatiotemporal information inside SNN, it not only has better biological fidelit...

AI Technical Summary

Problems solved by technology

However, due to the large differences in the aerodynamic structure of the morphing aircraft in different shapes and states, the tradi

Images