On Dynamic Adaptive Surface Containment Control For Nonlinear Multi-agent Systems

Because multi-agent systems have the advantages that a single system does not have and can complete complex tasks that a single system cannot complete,the research on adaptive control of multi-agent systems in the fields of UAV,formation control,sensor network and so on has become one of the research hotspots of control discipline.In addition,the stability of nonlinear systems is easily affected by unmodeled dynamics and input dead zone.Therefore,the research on the control of nonlinear multi-agent systems with unmodeled dynamics and input dead zone has important theoretical significance and application value.In this paper,for multi-agent systems with unmodeled dynamics and input dead zone in the presence of multiple leaders,three adaptive control schemes are proposed by combining with time-varying gain method,command filter technology,dynamic surface control method and decentralized sliding mode estimator.The main research contents are as follows:Firstly,an adaptive containment control scheme is put forward for uncertain non-strict feedback multi-agent systems(MASs)under directed graphs.The dynamic surface control(DSC)approach is extended to the containment control of a class of non-strict feedback MASs,and time-varying gains are used to replace constant gains in the design process.The property of radial basis function is used to overcome the design obstacle resulting from non-strict feedback structure.An available dynamic signal produced by an auxiliary linear system is introduced to deal with unmodeled dynamics.The designed controller can realize that the outputs of all the followers belong to a convex hull spanned by all the outputs of the leaders,and all the signals are bounded.Finally,compared with the constant gain control method,the validity of the time-varying gain control scheme is illustrated by numerical and actual simulation results.Secondly,a command-filter-based containment control is put forward for a class of non-strict feedback multi-agent systems(MASs)with input dead-zone.The filtering errors are counteracted by the compensation signals,and the assumption related to upper limit of the gain is reduced.The property of radial basis function is used to overcome the design obstacle resulting from non-strict feedback structure.The decentralized sliding-mode estimators are utilized to acquire estimations of leaders' trajectories for each follower.The proposed controller can guarantee that outputs of all the followers converge to convex hull spanned by leaders,and all the internal signals are bounded.Finally,the effectiveness of the control scheme is illustrated by two simulation examples.Thirdly,This article is involved with the containment control problem that relate to a class of non-strict feedback multi-agent systems(MASs)with input dead-zone and prescribed performance.The utilization of the dynamic surface control(DSC)method is to simplify design,and the filtering errors are counteracted by the compensation signals.The inherent difficulties of non-strict feedback agents are overcome by the property of radial basis function.The decentralized sliding-mode estimators are utilized to acquire estimations of leaders' trajectories for each follower.The proposed controller can guarantee that outputs of all the followers converge to convex hull spanned by leaders,the synchronization errors can drive into the range of preset performance and all the internal signals are bounded.Finally,the effectiveness of the control scheme is illustrated by two simulation examples.