Multi-quadrotor master-slave type guaranteed-performance neural adaptive cooperative formation control method

Abstract

The invention discloses a multi-quadrotor master-slave type guaranteed-performance neural adaptive cooperative formation control method, and relates to the field of multi-quadrotor formation automaticcontrol. The method has the characteristics of the low online calculation complexity and capability of quickly and smoothly achieving interference identification and compensation. Specifically, a preset performance function, an error conversion technology and a consistency technology are introduced into a four-rotor trajectory loop, an originally constrained formation error control problem is converted into an error stabilization problem of an unconstrained system, and it is ensured that a formation synchronization error meets a preset transient performance and the steady-state precision; a predicted minimum parameter learning neural network approximator which is low in online learning dimension and high in calculation real-time performance is constructed for the trajectory and attitude loop, so that smooth and rapid online learning and compensation of unknown non-linear interference are achieved, the coupling influence of a control loop and a neural learning loop is eliminated, and the real-time performance and the rapid response capacity of formation control of the multi-quadrotor system can be remarkably improved.

Description

technical field [0001] The invention relates to the field of automatic control of multi-quadrotor formations, specifically a multi-quadrotor master-slave type performance-guaranteed neural adaptive cooperative formation control method, which is applied to multi-quadrotor formations with strong interference, unmodeled parts and uncertain parameters Task. Background technique [0002] The multi-quadrotor cooperative formation refers to the multi-quadrotors maintaining the predetermined spatial geometric topology, with the help of global or local information interaction and sharing, to form a multi-agent system's collaborative transcendence ability, in order to perform complex tasks that cannot be performed by a single Provide effective solutions. For example, multi-quadrotors can use reasonable formation flight to replace soldiers to efficiently perform dangerous military tasks such as target detection and enemy information collection. The disadvantages of incomplete acquisi...

AI Technical Summary

Problems solved by technology

[0004] The problem solved by the technology of the present invention is to provide an online low-computational A multi-quadrotor master-slave performance-guaranteed neural adaptive cooperative formation control method that can quickly and smoothly realize interference identification and compensation. Specifically, a preset performance function, error conversion, and consistency technology are introduced into the quadrotor trajectory loop. Transform the original constrained formation error control problem into the error stabilization problem of an unconstrained system to ensure that the formation synchronization error meets the preset transient performance and steady-state accuracy; for the trajectory and attitude loops, construct online learning with low dimensionality and calculation Real-time predictive minimum parameter learning neural network approximator to achieve smooth and fast online learning and compensation for unknown nonlinear disturbances, and at the same time eliminate the coupling effect of control loops and neural learning loops

Images