Bio-inspired Scalability Analysis Of Multi-UAVs Formation Control

In recent years,people realize that single UAV is often limited facing many complicated missions,and the effectiveness of single UAV is usually not satisfied,either.There are many subjects about UAV formation,one of them is the scalability problem,in formation control systems,high scalability can guarantee the formation stability when UAVs join in or leave the system.However existing methods cannot handle the scale change of the UAV formation.This article investigates the scalability problem for UAV formation control,whose solution method is inspired by pigeon flocks.The main contributions are:Firstly,A method inspired by pigeon formations to build interaction topology is proposed,and we call it the Veteran Rule.It is shown that when using the Veteran Rule,the system with double-integrator dynamics can always have the formation stability with system scalability,and the control parameter in each UAV is fixed.Moreover,the convergence rate of the formation system under the Veteran Rule is analyzed,and the optimal way to build connections is obtained through our analysis.Simulations on MATLAB corroborate the effectiveness of our method.Secondly,to guarantee the robustness of the formation system,we study the tolerance on undesired communication links(which break our proposed Veteran Rule),i.e.,reverse edges.An upper bound for the coupling strength of the reverse edges is provided by using Gershgorin circle theorem.Based on this upper bound,we discuss how the reverse edges change the communication loading and scalability of the formation system in different cases.Finally,we studied the propagation of uncertainty along communication topology from the perspective of expectation and covariance,and Kalman filtering is used to suppress the propagation of uncertainty.Necessary and sufficient conditions for the system state expectation converging to the desired pattern and the system state covariance maintaining bounded are given.Moreover the relationship between the state covariance and the communication topology is analyzed.The theoretical results are verified by the simulations on MATLAB.