Processing math: 90%
A journal of IEEE and CAA , publishes high-quality papers in English on original theoretical/experimental research and development in all areas of automation
W. C. Huang, H. L. Liu, and J. Huang, “Distributed robust containment control of linear heterogeneous multi-agent systems: An output regulation approach,” IEEE/CAA J. Autom. Sinica, vol. 9, no. 5, pp. 864–877, May 2022. doi: 10.1109/JAS.2022.105560
Citation: W. C. Huang, H. L. Liu, and J. Huang, “Distributed robust containment control of linear heterogeneous multi-agent systems: An output regulation approach,” IEEE/CAA J. Autom. Sinica, vol. 9, no. 5, pp. 864–877, May 2022. doi: 10.1109/JAS.2022.105560

Distributed Robust Containment Control of Linear Heterogeneous Multi-Agent Systems: An Output Regulation Approach

doi: 10.1109/JAS.2022.105560
Funds:  This work was supported by the National Science Foundation of China (51977040)
More Information
  • In this paper, we consider the robust output containment problem of linear heterogeneous multi-agent systems under fixed directed networks. A distributed dynamic observer based on the leaders’ measurable output was designed to estimate a convex combination of the leaders’ states. First, for the case of followers with identical state dimensions, distributed dynamic state and output feedback control laws were designed based on the state-coupled item and the internal model compensator to drive the uncertain followers into the leaders’ convex hull within the output regulation framework. Subsequently, we extended theoretical results to the case where followers have nonidentical state dimensions. By establishing virtual errors between the dynamic observer and followers, a new distributed dynamic output feedback control law was constructed using only the states of the compensator to solve the robust output containment problem. Finally, two numerical simulations verified the effectiveness of the designed schemes.

     

  • RECENTLY, multi-agent techniques have received increasing attention due to their potential applications in industrial, agricultural and military fields [1]–[4]. The cooperative problem is one of the basic research directions of multi-agent systems, and considerable research on this topic have been presented in the literature [5]–[11]. The main focus of cooperative control is the design of distributed strategies under which the agents reach control objectives accessing only local information from itself and its neighbors rather than global information. Distributed cooperative control has been successfully applied in large-scale complex system control problems that are difficult or even impossible to solve via centralized or decentralized strategies.

    Containment control, as a typical distributed cooperative control problem, has also attracted substantial attention. The control purpose is to drive followers to asymptotically converge to the convex hull spanned by the leaders [8], [10]-[14]. The containment control problem originates from natural phenomena and has broad practical applications, for example, when a team of autonomous vehicles or robots has to move from one area to another but only some individuals have the ability to detect dangerous areas. The method adopted is to treat autonomous vehicles or robots with detection capability as leaders and the rest as followers to drive the followers to enter the convex hull spanned by the leaders and move together, so as to ensure group security in the transfer process [12], [13].

    Many theoretical studies on the containment problem have been conducted, and the integrator system has played an important role. Among them, containment problems with single-integrator and double-integrator agents were investigated in [8], and the necessary and sufficient conditions were provided with continuous-time and sampled-data based protocols under a fixed directed topology. Moreover, the final output of the follower systems were found to be completely dependent on topology structure and the initial states of the leaders. Subsequently, the containment problem of integrator agents with switching topology was considered in [15]–[17], and sufficient conditions for containment control were given utilizing different control approaches. Moreover, the containment problem of the integrator system subject to complex situations has begun to receive attention, with many meaningful results, including nonlinear system containment [18], [19], event-triggered containment [20], [21], finite-time containment [22]–[24], communication-delay containment [25]–[28], and input-saturation containment [29]-[31]. Due to the simplicity of the integrator system structure, it can give full play to the mathematical operation function based on the topology network, and we can obtain many meaningful results. However, it is difficult to directly extend this to problem solving of high-order systems.

    Compared with the integrator system, the high-order system has complex system dynamics, so it is more difficult to solve the containment problem, but it is more in line with the needs of practical applications. The forementioned achievements have been reconsidered for high-order systems. For example, the distributed output and state feedback approaches were adopted for state containment [12], [32], and the distributed adaptive approach with agent output was proposed for output containment [33]. The containment problem for communication delay was investigated in [34], [35]. The containment problems for switching topology and performance optimization were considered in [11] simultaneously. To explore the impact of parameter uncertainty of system plants, the robust containment problem was discussed utilizing diverse control ideas [36]–[39]. In [37], a composite feedback containment controller was synthesized to solve the containment problem of unknown nonlinear dynamics based on neural network approximation theory and the Lyapunov theory. In [38], the adaptive factor was embedded in the robust containment protocol to address uncertain leaders utilizing an adaptive internal model and a recursive stabilization approach. In [39], the robust containment problem of fractional-order system dynamics was investigated, and the sufficient condition to state containment was provided according to the linear matrix inequalities technique. The containment problem with input saturation was considered in [40]. The bipartite containment problem was studied in [41] and [42] for linear multi-agent systems and descriptor multi-agent systems, respectively.

    Notably, a common drawback associated with most of the forementioned results is that each agent must have the relative values of the state with respect to its neighbors. One major reason is that many containment methods involve the differential term of containment error, and containment error itself is an expression of variable coupling [15], [32], [39]–[42]. Therefore, these methods can only be applied in the case where the agents have identical dynamics or state dimensions and absolutely rules out the possibility of agents with nonidentical state dimensions. However, multi-agent systems with nonidentical state dimensions are common in practical applications. For example, different types of robots must work cooperatively to complete complex production tasks in industry.

    In addition, the method in this paper is based upon output regulation theory, which can realize asymptotic tracking of the reference input [43] and has been broadly used in the cooperative control problem of multi-agent systems with leaders [5], [12], [13], [38], [44]–[46]. Under the framework of output regulation, the design of the compensator is an important issue. For example, a cooperative output regulation protocol was introduced to solve the containment problem of linear heterogeneous multi-agent systems using a dynamic compensator based on the leaders’ state information [12] (Details can be found in Remark 3). Output regulation protocols with similar compensators can be found in [5], [13], [44], [47]. However, the dynamic compensators are associated with the leaders’ state information: However, the state information is not always available due to various practical factors, such as high acquisition cost and difficult acquisition.

    Recently, an interesting design idea was adopted in [11] to realize output containment by driving all agents to track embedded compensator systems. Inspired by [11], we have solved the robust output containment control problem of linear multi-agent systems with a directed network. The main differences are that the problem we considered involves robust containment, and the control strategies do not take measures against the leaders. The main contributions of this paper include the following three points. First, we modify the conventional state observer to produce an estimate of the convex combination of the leaders’ states by applying a directed network, which lends itself to the design of the distributed protocols. Moreover, the distributed observer can also be viewed as an extension of the compensators associated with the leaders’ states in [5], [12], [13], [44], [47] as these compensators can be regarded as a special case of our observer when the output matrix of the leader systems is of full column rank.

    Second, based on the internal model principle and the compensator technique, distributed dynamic state and output feedback control laws were introduced to drive the containment errors to converge to the origin asymptotically, such that uncertain followers with identical nominal dynamics entered the convex hull spanned by the leaders under the output regulation framework. In addition, for the closed-loop stability analysis, the closed-loop system was divided into multiple subsystems via nonsingular transformation, and a Lyapunov inequality method was used to make multiple subsystems reach a stable state simultaneously, thereby stabilizing the closed-loop system.

    Finally, we extended the theoretical results to a more general case where the followers have nonidentical state dimensions, where the robust containment problem was converted into a new tracking problem between the distributed observer systems and the follower systems by constructing a virtual error vector. A distributed dynamic output feedback control law was further devised by modifying the state-coupled item in the previous method to drive virtual error to converge to the origin asymptotically, such that multi-agent systems achieved output containment control. In this way, we have avoided the dependence of most existing containment protocols on the relative states of followers, such that the distributed control law is capable of solving the robust output containment problem for linear heterogeneous multi-agent systems with nonidentical state dimensions.

    The remainder of this paper is organized as follows. Section II presents the preliminary knowledge and the problem formulation. Section III provides the main result, where the robust output containment control problem was solved for multi-agent systems with followers with identical state dimensions. Section IV presents further results about heterogeneous followers with nonidentical state dimensions. Section V discusses two numerical simulation examples to verify the effectiveness of the designed schemes. Finally, Section VI concludes this paper.

    Notation: denotes the kronecker product. Given any matrices A, B, C, D of compatible dimensions, the following hold: (A+B)C=AC+BC, (AB)T=ATBT, (AB)×(CD)=ACBD. IN denotes the N-dimensional identity matrix. I denotes the dimensionally-compatible identity matrix. 1N denotes the N-dimensional column vector with all 1 elements. 0N×M denotes a N×M-dimensional matrix with all 0 elements. 0 denotes a dimensionally-compatible matrix or vector with all 0 elements. ∅ denotes the null set. σ(S) denotes the spectrum of the square matrix S. Given a vector set {x1,,xN}, denote col(x1,,xN)=[xT1,,xTN]T. Given any matrix ARm×n, denote Vec(A)=col(A1,,An) where Ai, i=1,,n is the i-th column of A. R denotes the set of real numbers, and C denotes the set of complex numbers. x denotes the Euclidean norm of the finite dimensional vector x. dist(x,C) denotes the distance from xRn to the set CRn in the sense of Euclidean norm, that is, dist(x,C)=infyCxy.

    Definition 1 (Convex hull): A vector set CRN is convex if (1λ)x+λyC for any x,yC and any λ[0,1]. The convex hull of a finite set of points X={x1,,xq} is defined as Co(X)={qi=1αixi|xiX,αiR,αi0,qi=1αi=1}.

    A group of agents, including N followers marked as 1,2,,N and M leaders marked as N+1,N+2,,N+M, is considered. The information communication relationship between all agents can be denoted by a directed graph G=(V,E,A), where V={V1,V2,,VN+M} is the node set, E{(i,j)|i,jV} is the edge set, and A=[aij]R(N+M)×(N+M) is the weight adjacency matrix. If an edge (i,j)E, then node j can access the information from node i, and aji>0; otherwise, aji=0. Assume no self-loops exist, i.e., aii=0,iV. The neighbor set of node j is denoted as Nj={i|i,jV,(i,j)E}. The edge sequence (i1,i2),,(ik1,ik) represents a directed path from node i1 to ik, and node ik is said to be reachable from node i1. The graph G is said to contain a spanning tree if and only if there exists at least one node from which every other node is reachable. The graph G is said to have a united spanning tree if all the leaders have no neighbors and each follower has at least one directed path from the leaders. The Laplacian matrix L=[lij] R(N+M)×(N+M) of the graph G is defined as lii=N+Mj=1aij, and lij=aij, ij. The Laplacian matrix of G has the following form:

    L=(L1L20M×N0M×M)

    where L1RN×N, L2RN×M satisfy L1N+M=0, L11N+L21M=0. The matrix L11L2 is nonnegative and satisfy L11L21M=1N when the graph G has a united spanning tree [8].

    For convenience, denote VF={1,,N} as the follower set, VL={N+1,,N+M} as the leader set.

    In this paper, we consider the robust output containment problem of linear heterogeneous multi-agent systems with uncertain followers. Assume the N follower subsystems have the following dynamics:

    ˙xi=ˉAixi+ˉBiuiyi=ˉCixiymi=ˉCmixi,iVF (1)

    where xiRn, uiRm, yiRp, and ymiRpm are the state, control input, regulated output and measurable output of the i-th follower, respectively. ˉAi, ˉBi, ˉCi, and ˉCmi are uncertain constant matrices with compatible dimensions

    ˉAi=A+ΔAi,ˉBi=B+ΔBiˉCi=C+ΔCi,ˉCmi=Cm+ΔCmi

    where A, B, C, and Cm are the nominal values of these matrices and ΔAi, ΔBi, ΔCi, and ΔCmi are the uncertain parts. For convenience, denote

    Δ=(Vec(ΔA1,,ΔAN)Vec(ΔB1,,ΔBN)Vec(ΔC1,,ΔCN)Vec(ΔCm1,,ΔCmN))RNn(n+m+p+pm).

    Assume the M leader subsystems have the following dynamics:

    ˙wk=Swkyk=Qwkymk=Qmwk,kVL (2)

    where wkRq, ykRp, and ymkRpm are the state, tracked output and measurable output of the k-th leader, respectively. S, Q, and Qm are constant matrices with compatible dimensions.

    The problem of robust output containment control is: Under appropriate control laws ui, iVF, multi-agent systems (1) and (2) subject to uncertainty Δ are said to have achieved robust output containment control if the output of each follower asymptotically convergences to the convex hull spanned by the leaders’ output; that is, limtdist(yi,Co(yN+1,,yN+M))=0, iVF.

    Here, we introduce the output containment error vectors for the followers:

    ei=Nj=1aij(yiyj)+N+Mk=N+1aik(yiyk),iVF. (3)

    We can rewrite (3) in a compact form

    e=(L1Ip)yF+(L2Ip)yL (4)

    where e=col(e1,,eN), yF=col(y1,,yN), yL=col(yN+1,,yN+M).

    If digraph G has a united spanning tree and the containment error e converges to the origin asymptotically, the output of each follower is contained in the convex hull spanned by the leaders’ output. The final outputs of the followers are determined by both the topology structure and the leaders’ initial state and satisfy limtyF=(L11L2Ip)yL, which is in line with Definition 1 with the nonnegative matrix L11L2 satisfying L11L21M=1N [8].

    Remark 1: In the research on multi-agent cooperative control, the leader-following consensus is a special situation of the containment problem with only one leader. Therefore, the robust tracking problem considered in [46] can be viewed as a special case of our problem. In [46], the virtual output error vectors were established for followers:

    evi=jNiaij(yiyj),i=1,2,,N. (5)

    Additionally, virtual errors (5) can be regarded as a special containment error with a single leader. When the digraph G has a directed spanning tree, the leader-following consensus of [46] can be achieved if and only if the virtual errors converge to the origin asymptotically. However, this result is not valid when it is extended to a containment problem with more than one leader. Here, if the digraph G has a united spanning tree, the containment errors (3) converge to the origin asymptotically is only a sufficient condition for containment control.

    To solve the robust output containment problem of multi-agent systems (1) and (2) via the output regulation approach, some standard assumptions are necessary.

    Assumption 1: Digraph G has a united spanning tree.

    Assumption 2: (A,B,Cm) is stabilizable and detectable.

    Assumption 3: S has no eigenvalue with a negative real part.

    Assumption 4: (S,Qm) is detectable.

    Assumption 5: rank(AλiIBC0)=n+p, for all λiσ(S).

    These assumptions are loose for the output regulation problem, and similar entries can be found in [5], [6], [45], [46]. Here, if S has eigenvalues with a negative real part, the partial states of the leaders will converge to the origin asymptotically without any control command and have no influence on the final result. For a more interesting situation, consider Assumption 3. Assumption 5 is a transmission zero condition, which is a necessary condition to ensure that robust output regulation problem is solvable [43].

    Definition 2 (Internal model): Given a square matrix S, the pair (G1,G2) incorporates a p-copy internal model of S if it admits

    G1=Γ(Γ1Γ20G1)Γ1,G2=Γ(Γ3G2) (6)

    where Γ is any nonsingular matrix with compatible dimensions and Γi,i=1,2,3 are any constant matrices with compatible dimensions. The pair (G1,G2) can be described as

    G1= diag{β1,β2,,βpp-tuple}G2= diag{σ1,σ2,,σpp-tuple}

    where the pairs (βi,σi),i=1,2,,p satisfy: 1) βi is a constant square matrix with its characteristic polynomial divisible by the minimum polynomial of S. 2) σi is a column vector such that the pair (βi,σi) is controllable. Particularly, we say the pair (G1,G2) incorporates a minimum p-copy internal model of S when dim(Γi)=0,i=1,2,3 and Γ=I.

    The following lemmas will be used to solve the robust containment problem.

    Lemma 1: Under Assumption 1, all nonzero eigenvalues of Laplacian matrix L are eigenvalues of matrix L1, and all have positive real parts. Furthermore, for the same eigenvalue, the eigenvector to L1 is a subvector of the eigenvector to L.

    Proof: See Appendix.

    Remark 2: There is a similar lemma for leader-following consensus (see Lemma 1 in [5]), but the proof here is more concise and clear and has given the connection between their eigenvectors.

    Lemma 2 (See Lemma 1 in [48]): Given matrices ARn×n,BRn×m, (A,B) is stabilizable, and the algebraic Riccati equation:

    ATP+PAPBBTP+I=0 (7)

    admits a unique solution P=PT>0. Then, AνBBTP is Hurwitz for any νC with Re(ν)1.

    Lemma 3 (See Proposition A.2 in [43]): Given matrices ARn×n, BRm×m and CRm×n, the Sylvester equation

    XA=BX+C (8)

    provides a unique solution XRm×n, if and only if σ(A)σ(B)=.

    Lemma 4 (See Lemma 1.26 in [43]): Under Assumptions 2, 3, and 5, (G1,G2) incorporates a minimum p-copy internal model of S,

    A=(A0G2CG1),B=(B0)

    then the pair (A,B) is stabilizable.

    Lemma 5 (See Lemma 1.27 in [43]): If (G1,G2) incorporates a minimum p-copy internal model of S and the Sylvester equation

    XS=G1X+G2Π (9)

    provides a solution X, we have Π=0.

    Lemma 6 (See Chapter 7.2.1 in [49]): Given matrices ARn×n and BRn×m, (A,B) is stabilizable if and only if there exists a positive definite symmetric solution P to the Lyapunov inequality

    AP+PAT2BBT<0. (10)

    It is well-known that the internal model principle can eliminate the steady-state error of the uncertain system by embedding a model of the exogenous signals [50]–[52]. It is also a key part of the solution to the cooperative robust output regulation problem.

    The robust output regulation for a single plant has been provided in detail in [43] via the internal model principle, and it is theoretically proven that pure state feedback protocol cannot solve this problem. Subsequently, the robust leader-following consensus and containment problem of multi-agent systems were further considered in [6], [46], [47], [53] via the internal model principle and the compensator technique. In this paper, we continue to consider the robust output containment problem of linear multi-agent systems within the output regulation framework.

    In practice, on the one hand, partial followers can not directly access information from the leader due to the lack of existence of an edge from the leaders to the followers in the communication network; on the other hand, the leaders’ state variate may not be available due to measurement difficulties, no actual physical meaning, etc. To address these problems, we design a distributed dynamic compensator for the followers based on the leaders’ measurable output:

    ˙ˆwi=Sˆwi+ˉLQmNj=1aij(ˆwiˆwj)+ˉLN+Mk=N+1aik(Qmˆwiymk), iVF (11)

    where ˆwiRq is the state of the compensator and ˉL is a constant matrix to be designed. Treat ˆyi=Qˆwi, iVF as the output term of the compensator (11), and rewrite them in a compact form:

    ˙ˆw=(INS+L1ˉLQm)ˆw+(L2ˉLQm)wˆyL=(INQ)ˆw (12)

    where ˆw=col(ˆw1,,ˆwN),ˆyL=col(ˆy1,,ˆyN),w= col(wN+1,,wN+M).

    The following will show that the compensator (11) can be regarded as a state observer of the convex combination of the leaders.

    Lemma 7: Under Assumption 4, for any initial states wk(0),kVL and ˆwi(0),iVF, the states of compensator (11) can asymptotically converge to the convex hull spanned by the leaders’ states if and only if the digraph G has a united spanning tree.

    Proof: See Appendix.

    Remark 3: Reference [12] has considered the state containment problem of linear multi-agent systems, in which a distributed dynamic compensator (13), based on the leaders’ states, was adopted to observe the convex combination of the leaders’ states. A similar compensator can be found in the research on containment problems [13], [40], [47] and leader-following consensus problems [5], [44]:

    ˙ζi=Sζi+β[jNiaij(ζjζi)+n+mk=n+1δki(wkζi)]. (13)

    Despite playing the same role, the observer (11) differs from (13) in the following two aspects:

    1) The distributed observer (11) is more realistic than (13) because each follower needs to know the leaders’ state information in (13) but only the measurable output information in (11). Generally, the system output information is easier to obtain than state information. Considering the same problem, [13] introduced a classical state observer for each leader to be suited for situations where the leaders’ states are not available. However, this approach increases the state dimension of the control protocol and makes the closed-loop system more complicated.

    2) The distributed observer (11) contains the compensator (13) as a special case, where the measurable output matrix Qm is of full column rank. This scenario is equivalent to the state variate of leader systems (2) being available. We further make ˉL satisfy βIq=ˉLQm, where β is a constant to be designed. Then, observer (11) is simplified to (13). At this time, let β>max{Re(λi),λiσ(S)}/min{Re(μj),μjσ(L1)}; then, matrix Ξ is Hurwitz. As a whole, the compensator (11) is a generalized case that includes the compensator (13) as a special case.

    In this paper, we considered the robust output containment problem for linear heterogeneous multi-agent systems. First, for the case where the followers have identical nominal dynamics, the distributed dynamic state and output feedback protocols were introduced to drive uncertain followers into a convex hull spanned by the leaders. Subsequently, to consider a more realistic case where the followers may have nonidentical nominal dynamics and dimensions, with the aid of compensator (11), a new distributed dynamic output feedback protocol was established.

    In the following part, two distributed dynamic control laws are introduced to solve the robust output containment problem using the internal model principle.

    For convenience, the function G() associated with the communication network G was defined as

    G(xi)=Nj=1aij(xixj)+N+Mk=N+1aik(xiΦiwk)iVF (14)

    where ΦiRn×q,iVF are any constant matrices.

    The following forms of control protocols are constructed:

    1) Distributed dynamic state feedback control law

    ui=K1G(xi)+K2zi˙zi=G1zi+G2ei, iVF (15)

    where ziRnz is the state of the internal model compensator, (K1,K2) are the constant matrices to be designed, and (G1,G2) incorporates a minimum p-copy internal model of S.

    In practice, the state of followers is not always available, but the outputs are usually measurable.

    2) Distributed dynamic output feedback control law

    ui=KxG(ˆxi)+Kzzi˙ˆxi=Aˆxi+Bui+H(CmG(ˆxi)emi)˙zi=G1zi+G2ei, iVF (16)

    where ˆxiRn, ziRnz are the states of the compensators, (Kx,Kz,H) are the constant matrices to be designed, the pair (G1,G2) incorporates a minimum p-copy internal model of S, and

    emi=Nj=1aij(ymiymj)+N+Mk=N+1aik(ymiymk).

    Substituting the control law (15) or (16) into the follower system (1), and combining the leader system (2) and containment error (4) yields the closed-loop system:

    ˙xc=ˉAcxc+ˉBcw˙w=ˉSwe=ˉCcxc+ˉDcw (17)

    where xc is the closed-loop state, ˉS=IMS, and the details of ˉAc, ˉBc, ˉCc, and ˉDc are provided in Theorems 1 and 2. Denote ˉAc=Ac, ˉBc=Bc, ˉCc=Cc, and ˉDc=Dc for Δ=0. Thus, the robust output containment problem can be equivalently converted into a cooperative robust output regulation problem.

    The problem of cooperative robust output regulation is: Substitute the control protocol (15) or (16) into linear multi-agent systems (1) and (2) under digraph G, such that the closed-loop system (17) has the following properties:

    1) When Δ=0, the origin of the system ˙xc=Acxc is exponentially stable, i.e., Ac is Hurwitz.

    2) For any initial states xi(0), ˆxi(0), zi(0), iVF, and wk(0), kVL, there exists an open neighborhood W of Δ=0, for all ΔW such that:

    limtei=0,iVF. (18)

    In this section, two main results are presented under control laws (15) and (16).

    Under control law (15), the closed-loop state is xc=col(x,z), x=col(x1,,xN), z=col(z1,,zN). The dynamic matrices of closed-loop system (17) are as follows:

    ˉAc=(ˉA+ˉB(L1K1)ˉB(INK2)(L1G2)ˉCING1)ˉBc=(ˉB(INK1)Φ(L2Iq)L2G2Q)ˉCc=((L1Ip)ˉC0), ˉDc=L2Q (19)

    where ˉA=diag{ˉA1,,ˉAN}, ˉB= diag{ˉB1,,ˉBN}, ˉC= diag{ˉC1,,ˉCN}, Φ= diag{Φ1,,ΦN}.

    Theorem 1: Under Assumptions 1–3 and 5, (G1,G2) incorporates a minimum p-copy internal model of S, and the robust output containment problem of multi-agent systems (1) and (2) can be solved by the distributed dynamic state feedback control law (15).

    Proof: First, consider the first property of cooperative robust output regulation:

    Ac, the nominal value of the uncertain system matrix ˉAc of (19), is as follows:

    Ac=(INA+L1BK1INBK2L1G2CING1). (20)

    Let

    Ψ=(L11In0Nn×Nnz0Nnz×NnINInz). (21)

    Nonsingular transformation yields

    ˊAc=Ψ1AcΨ=(INA+L1BK1L1BK2ING2CING1). (22)

    Let U1= diag{UIn,UInz}. Nonsingular transformation yields:

    ˜Ac=U1ˊAcU1=(INA+ΛBK1ΛBK2ING2CING1). (23)

    Thus, ˜Ac is Hurwitz if and only if

    ˜Aci=(A+λiBK1λiBK2G2CG1),iVF (24)

    are Hurwitz. Rewrite (24) as follows:

    ˜Aci=A+λiBK, λiσ(L1) (25)

    where A=(A0G2CG1),B=(B0),K=(K1K2). According to Assumptions 2, 3, and 5 and Lemma 4, the pair (A,B) is stabilizable. Furthermore, according to Lemma 6, there exists a P=PT>0 such that

    PAT+AP2BBT<0. (26)

    Let

    K=βBTP1 (27)

    where βR satisfies β1/min{Re(λi),λiσ(L1)}.

    Let λi=σi+jwi with σi, wiR. Then

    ˜AciP+P˜Aci=(A+λiBK)P+P(A+λiBK)=(A+(σi+jwi)BK)P+P(A+(σi+jwi)BK)=(Aβ(σi+jwi)BBTP1)P+P(Aβ(σijwi)BBTP1)T=AP+PAT2βσiBBT=AP+PAT2BBT2(βσi1)BBTAP+PAT2BBT. (28)

    Therefore, ˜Aci,iVF are Hurwitz; thus, Ac is Hurwitz.

    We now continue to analyze the second property of cooperative robust output regulation:

    Based on the above analysis, there exists an open neighborhood W of Δ=0, for all ΔW, such that ˉAc is Hurwitz. On the basis of the closed-loop system (17) and dynamic matrices (19), we establish a Sylvester equation:

    XcˉS=ˉAcXc+ˉBc. (29)

    Since ˉAc is Hurwitz, σ(ˉS)σ(ˉAc)=. By Lemma 5, (29) admits a unique solution

    Xc=(X1X2). (30)

    Substituting (30) into (29) yields:

    X1ˉS=(ˉA+ˉB(L1K1))X1+ˉB(INK2)X2+ˉB(INK1)Φ(L2Iq),X2ˉS=(ING2)((L1Ip)ˉCX1+L2Q)+(ING1)X2.

    From Lemma 5,

    0=(L1Ip)ˉCX1+L2Q. (31)

    Let ˜xc=xcXcw; taking the derivative of ˜xc yields:

    ˙˜xc=˙xcXc˙w=ˉAcxc+ˉBcwXcˉSw=ˉAcxc+ˉBcw(ˉAcXc+ˉBc)w=ˉAcxcˉAcXcw=ˉAc˜xc. (32)

    Since ˉAc is Hurwitz, limt˜xc=0. For the containment errors:

    e=(L1Ip)ˉCx+(L2Q)w=((L1Ip)ˉC0)xc+(L2Q)w=((L1Ip)ˉC0)(˜xc+Xcw)+(L2Q)w=((L1Ip)ˉC0)˜xc+((L1Ip)ˉCX1+L2Q)w=((L1Ip)ˉC0)˜xc. (33)

    Since limt˜xc=0, and ((L1Ip)ˉC0) is bounded, hence limte=0.

    Under control law (16), the closed-loop state of (17) is xc=col(x,ˆx,z), x=col(x1,,xN), ˆx=col(ˆx1,,ˆxN), z=col(z1,,zN). Denote Ω=INA+L1(BKx+HCm). The dynamic matrices are as follows:

    ˉAc=(ˉAˉB(L1Kx)ˉB(INKz)(L1H)ˉCmΩINBKz(L1G2)ˉC0ING1)ˉBc=(ˉB(INKx)Φ(L2Iq)(IN(BKx+HCm))Φ(L2Iq)L2HQmL2G2Q)ˉCc=((L1Ip)ˉC00), ˉDc=L2Q (34)

    where ˉCm= diag{ˉCm1,,ˉCmN}.

    Theorem 2: Under Assumptions 1–3 and 5, (G1,G2) incorporates a minimum p-copy internal model of S; then, the cooperative robust output containment problem of multi-agent systems (1) and (2) can be solved by the distributed dynamic output feedback control law (16).

    Proof: First, consider the first property of cooperative robust output regulation:

    Ac, the nominal value of ˉAc within (34), is as follows:

    Ac=(INAL1BKxINBKzL1HCmΩINBKzL1G2C0ING1). (35)

    Let

    T=(INn00INn0INn0INnz0). (36)

    Denote Π=INA+L1HCm. Nonsingular transformation yields:

    ˊAc=T1AcT=(INA+L1BKxINBKzL1BKxL1G2CING1000Π) (37)

    where (Kx,Kz)=(K1,K2). Ac is Hurwitz if and only if Π is Hurwitz. The pair (A,Cm) is detectable, so the pair (AT,CTm) is stabilizable. By Lemma 6, there exists a P=PT>0 such that the Lyapunov inequality

    PA+ATP2CTmCm<0. (38)

    Let H=εP1CTm with ε1/min{Re(λi),λiσ(L1)}, so Π is Hurwitz. The convergence analysis of output containment error e is similar to that in Theorem 1 and is thus omitted here.

    Based on the above analysis, we have achieved robust output containment for linear heterogeneous multi-agent systems (1) and (2) using output regulation theory. Control laws (15) and (16) ensure that an agent accesses only local information from itself and its neighbors instead of global information, so they follow full information distributed strategies. For the stability analysis of the closed-loop system, the Lyapunov inequality method provided in [49] can make multiple subsystems reach a stable state simultaneously. However, a shortcoming remains; that is, the control protocols are associated with the relative state information between followers, so they can only cope with the case where followers have identical state dimensions. Moreover, our research framework is similar to that in [46]; the main differences include the following: 1) Different problems: The problem we consider is containment control with multiple leaders, while the problem in [46] is leader-following consensus with a single leader. 2) Different stability analysis methods for closed-loop system: Our method is based on a Lyapunov inequality, while [46] is based on an algebraic Riccati equation.

    In this section, we further generalize the theoretical results to a more general case, where the followers have nonidentical state dimensions and the above strategies cannot be applied directly. Here, with the aid of the compensator (11), we introduce a new type of distributed dynamic output feedback control law by replacing the state-coupled item with the state variate of the compensator in the previous control law that lends itself to the avoidance of dependence on relative states, such that the new control protocol can cope with the case where the followers have nonidentical dynamic dimensions.

    Consider N heterogeneous follower subsystems with the following dynamics:

    ˙xi=ˆAixi+ˆBiuiyi=ˆCixiymi=ˆCmixi,iVF (39)

    where xiRni, yiRp, ymiRpmi and uiRmi are the state, regulated output, measurable output and control input of the i-th follower, respectively. ˆAi, ˆBi, ˆCi, and ˆCmi are constant matrices with compatible dimensions

    ˆAi=Ai+ˆΔAi,ˆBi=Bi+ˆΔBiˆCi=Ci+ˆΔCi,ˆCmi=Cmi+ˆΔCmi

    where Ai, Bi, Ci, and Cmi are the nominal values of these matrices, and ˆΔAi, ˆΔBi, ˆΔCi, and ˆΔCmi are the uncertain parts. For convenience, denote

    ˆΔ=(Vec(ˆΔA1,,ˆΔAN)Vec(ˆΔB1,,ˆΔBN)Vec(ˆΔC1,,ˆΔCN)Vec(ˆΔCm1,,ˆΔCmN))RNi=1ni(ni+mi+p+pmi).

    Two necessary assumptions must be added.

    Assumption 6: (Ai,Bi,Cmi),iVF are stabilizable and detectable.

    Assumption 7: rank(AjλiIBjCj0)=nj+p, jVF, for all λiσ(S).

    Based on the state observer (12), establish the virtual error vectors for the followers:

    ˆei=yiˆyi, iVF. (40)

    Lemma 8: Under Assumptions 1 and 4, limtei=0,iVFlimtˆei=0,iVF. ˉL is given in Lemma 7.

    Proof: Rewrite (40) in a compact form

    ˆe=yFˆyL (41)

    where ˆe=col(ˆe1,,ˆeN).

    limtˆe=0limtyF=ˆyLlimtyF=(L11L2Ip)yLlimt(L1Ip)yF+(L2Ip)yL=0limte=0. (42)

    At this point, our goal can be transformed from driving the containment error to zero to driving the virtual error to zero. A new distributed protocol of the following form is constructed:

    Distributed dynamic output feedback control law

    ui=Ki1ˆxi+Ki2zi˙ˆxi=Aiˆxi+Biui+Hi(Cmiˆxiymi)˙zi=Gi1zi+Gi2ˆei, iVF (43)

    where ziRnzi, (Gi1,Gi2) incorporate a minimum p-copy internal model of S. (Ki1,Ki2,Hi) are the constant matrices to be designed.

    Under control law (43), the closed-loop system can be described as

    ˙xc=ˆAcxc+ˆBcw˙w=ˉSwˆe=ˆCcxc+ˆDcw (44)

    where the closed-loop state is xc=col(x,ˆx,z,ˆw), x=col(x1,,xN),ˆx=col(ˆx1,,ˆxN),z=col(z1,,zN),ˆw=col(ˆw1,,ˆwN). The dynamic matrices are as follows:

    ˆAc=(ˆAˆBˆK1ˆBˆK20ˆHˆCmˇA+ˇBˆK1+ˆHˇCmˇBˆK20ˆG2ˆC0ˆG1ˆG2ˆQ000Ξ)ˆBc=(000L2ˉLQm), ˆCc=(ˆC00ˆQ),ˆDc=0 (45)

    where ˆA= diag{ˆA1,,ˆAN}, ˆB= diag{ˆB1,,ˆBN}, ˆC= diag{ˆC1,,ˆCN}, ˆCm= diag{ˆCm1,,ˆCmN}, ˇA= diag{A1,,AN}, ˇB= diag{B1,,BN}, ˇC= diag{C1,,CN}, ˇCm= diag{Cm1,,CmN}, ˆG1= diag{G11,,GN1}, ˆG2= diag{G12,,GN2}, ˆK1= diag{K11,,KN1}, ˆK2= diag{K12,,KN2}, ˆH= diag{H1,,HN}, ˆQ=INQ.

    Theorem 3: Under Assumptions 1, 3, 4, 6, and 7, the pair (Gi1,Gi2) incorporates a minimum p-copy internal model of S; thus, the robust output containment problem of linear heterogeneous multi-agent systems (39) and (2) can be solved by the distributed dynamic output feedback control law (43).

    Proof: First, consider the first property of cooperative robust output regulation:

    Ac, the nominal value of system matrix ˆAc, is defined as follows:

    Ac=(ˇAˇBˆK1ˇBˆK20ˆHˇCmˇA+ˇBˆK1+ˆHˇCmˇBˆK20ˆG2ˇC0ˆG1ˆG2ˆQ000Ξ). (46)

    Let

    ˆT=(INi=1ni000INi=1ni0INi=1ni00INnz00000INq). (47)

    Nonsingular transformation yields:

    ˇAc=ˆT1AcˆT=(ˇA+ˇBˆK1ˇBˆK2ˇBˆK10ˆG2ˇCˆG10ˆG2ˇQ00ˇA+ˆHˇCm0000Ξ). (48)

    Denote ˇAc1=(ˇA+ˇBˆK1ˇBˆK2ˆG2ˇCˆG1), and rewrite it in the following form:

    ˇAc1=(ˇA0ˆG2ˇCˆG1)+(ˇB0)(ˆK1ˆK2). (49)

    By Lemma 4, the pair ((ˇA0ˆG2ˇCˆG1),(ˇB0)) is stabilizable, so there exist (Ki1,Ki2), iVF such that ˇAc1 is Hurwitz. According to Assumption 6, there exists Hi such that ˇA+ˆHˇCm is Hurwitz, iVF. Thus, Ac is Hurwitz.

    We now continue to analyze the second property of cooperative robust output regulation:

    There exists an open neighbor W of ˆΔ=0, and ˆAc is Hurwitz for all ˆΔW. According to Lemma 3, the Sylvester equation

    ˆXcˉS=ˆAcˆXc+ˆBc (50)

    admits a unique solution

    ˆXc=(ˆX1ˆX2ˆX3ˆX4). (51)

    Substituting (51) into (50) yields:

    ˆX1ˉS=ˆAˆX1+ˆBˆK1ˆX2+ˆBˆK2ˆX3ˆX2ˉS=ˆHˆCmˆX1+(ˇA+ˇBˆK1+ˆHˇCm)ˆX2+ˇBˆK2ˆX3ˆX3ˉS=ˆG2(ˆCˆX1ˆQˆX4)+ˆG1ˆX3ˆX4ˉS=(INS+L1ˉLQm)ˆX4+L2ˉL2Qm. (52)

    According to the third term of (52) and Lemma 5,

    0=ˆCˆX1ˆQˆX4. (53)

    From the fourth term of (52),

    ˆX4=L11L2Iq. (54)

    Let ˆxc=xcˆXcw; taking the derivative of ˆxc yields:

    ˙ˆxc=˙xcˆXc˙w=ˆAcxc+ˆBcwˆXcˉSw=ˆAcxc+ˆBcw(ˆAcˆXc+ˆBc)w=ˆAcˆxc. (55)

    Since ˆAc is Hurwitz, limtˆxc=0.

    Furthermore,

    ˆe=ˆCxˆQˆw=ˆCcxc=ˆCc(ˆxc+ˆXcw)=ˆCcˆxc+(ˆCˆX1ˆQˆX4)w=ˆCcˆxc. (56)

    Since limtˆxc=0 and ˆCc is bounded, limtˆe=0. According to Lemma 8, limte=0.

    Remark 4: In this section, we have achieved cooperative robust output containment control for linear heterogeneous multi-agent systems with followers having nonidentical dynamic dimensions. The core is the design of the distributed dynamic compensator (11), which can provide the followers with an estimate of the convex combination of the leaders’ states. By constructing the virtual error vectors, the containment problem was further transformed into an equivalent tracking problem between the follower system (39) and the compensator system (12). In this way, it helps the control protocol (43) avoid relying on the relative states of followers. Furthermore, under Assumption 6, we can choose an appropriate (Ki1,Ki2,Hi), iVF to stabilize the nominal closed-loop system, such that the tracking problem can be easily solved via the output regulation method. Finally, the objective of containment control was achieved by driving the virtual errors to converge to zero. However, compared to protocol (16), although, the relative states of followers are not needed for protocol (43), a shortcoming may remain; that is, the dynamic compensator (11) must to be added, which increases the dynamic dimensions of the control protocol and makes the closed-loop system complicated. This scenario may increase the difficulty of practical applications.

    In this section, two numerical simulation examples are adopted to illustrate the validity of the results of Theorems 2 and 3.

    Consider a team of agents consisting of three followers (i=1,2,3) and three leaders (k=4,5,6). The information communication network is depicted in Fig 1. The dynamic matrices of the leaders are as follows:

    Figure  1.  The topology network.
    S=(0.10.30.30.1),Q=(1001).

    The nominal values of the dynamic matrices of the followers are as follows:

    A=(123011100),B=(011001),C=(100001).

    The uncertain parts

    ΔA1=(0.50000.40000),ΔA2=(00000.500.400)ΔA3=(00.50000000.5),ΔB1=(000000.5)ΔB2=(000000.5),ΔB3=(000000.5)ΔC1=(00000.40),ΔC2=(000100)ΔC3=(000000).

    Assume Qm=Q, ˉCmi=ˉCi, i=1,2,3.

    The pair (G1,G2) are as follows:

    G1=(01000.080000001000.080),G2=(00500005).

    By solving the Riccati equation and Lyapunov inequality:

    P1=(0.91820.02740.02741.0964),P2=(P21P22)P21=(0.74990.50740.12650.60060.50741.41060.00190.42531.02650.00190.55010.28010.60060.42530.280110.14360.66590.73270.10742.10750.27510.09760.28861.52300.06501.33760.76580.0574)P22=(0.66590.27510.06500.73270.09761.33760.10740.28860.76582.10751.52300.05744.56900.26231.16240.262310.88221.97101.16241.97104.2977)P3=(0.16530.11450.08530.11450.83500.07730.08530.07730.8222).

    Then, obtain the feedback matrices:

    ˉL=2.5P1QT=(2.28190.06850.06852.7410)K=(K1 K2)=2.5(BT0)P12K1=(8.20588.80483.929910.04024.27596.5614)K2=(0.52752.13710.81813.35490.75422.43760.07600.3907)H=2.5P13CT=(2.80200.25680.36030.01240.25680.5143).

    Let Φi=(100100),i=1,2,3.

    In the simulation results, yi1 and yi2 are the components of the agent outputs yi, i=1,2,3,4,5,6. Their output trajectories are depicted in Fig 2, where the convex hull spanned by the leaders is displayed at several time instants. The followers asymptotically enter the leaders’ convex hull, which is consistent with our expectation. Subsequently, the trajectories of the followers and the leaders over time are depicted in Fig 3. It can be found that the followers output components yi1, i=1,2,3 and yi2, i=1,2,3 converge asymptotically into the convex hull spanned by the leaders output components yk1, k=4,5,6 and yk2, k=4,5,6, respectively. Finally, the containment error and the compensator state trajectories are depicted in Figs. 4 and 5, respectively. The containment error curves asymptotically converge to zero, and the compensator state curves asymptotically converge to the convex hull spanned by the leaders’ states. Overall, the above analysis is consistent with the results in Theorem 2.

    Figure  2.  The output trajectories of the followers and the leaders for the closed loop multi-agent system in Example 1.
    Figure  3.  The output trajectories of the followers and the leaders over time for the closed loop multi-agent system in Example 1
    Figure  4.  The containment error of the followers over time for the closed loop multi-agent system in Example 1.
    Figure  5.  The state trajectories of the compensators and leaders over time.

    Consider a team of agents consisting of three heterogeneous followers (i=1,2,3) and three leaders (k=4,5,6). The topology graph, leader dynamics are the same as in Example 1. Matrix pairs (G_{i1},G_{i2}) = (G_{1},G_{2}),\;i = 1,2,3 . The nominal value of the followers dynamics matrices are as follows:

    \begin{split} &A_{1} = \left( \begin{matrix} 1 & 2 & 3 \\ 0 & -1 & 1 \\ 1 & 0 & 0 \end{matrix} \right),\;\; A_{2} = \left( \begin{matrix} 0 & 1 \\ 1 & 0 \end{matrix} \right),\;\; A_{3} = \left( \begin{matrix} 1 & 0 \\ -1 & 1 \end{matrix} \right) \\ &B_{1} = \left( \begin{matrix} 0 & 1 \\ 1 & 0 \\ 0 & 1 \end{matrix} \right), \;\;B_{2} = \left( \begin{matrix} 0 & 1 \\ 1 & 0 \end{matrix} \right), \;\;B_{3} = \left( \begin{matrix} 1 & 0 \\ 0 & 1 \end{matrix} \right) \\ &C_{1} = \left( \begin{matrix} 1 & 0 & 0 \\ 0 & 0 & 1 \\ \end{matrix} \right),\;\; C_{2} = \left( \begin{matrix} 1 & 0 \\ 0 & 1 \end{matrix} \right), \;\;C_{3} = \left( \begin{matrix} 1 & 1 \\ 0 & 1 \end{matrix} \right). \end{split}

    The uncertain parts are as follows:

    \begin{split} &\hat{\Delta} A_{1} = \left( \begin{matrix} 0 & 0 & -1 \\ 0 & 0 & 0 \\ 0 & 0 & 0 \end{matrix} \right),\;\; \hat{\Delta} A_{2} = \left( \begin{matrix} 0 & 0 \\ 0 & 0.6 \end{matrix} \right),\;\; \hat{\Delta} A_{3} = \left( \begin{matrix} 0 & 0 \\ -0.5 & 0 \end{matrix} \right) \\ &\hat{\Delta} B_{1} = \left( \begin{matrix} 0 & 0 \\ 0.5 & 0 \\ 0 & 0 \end{matrix} \right), \;\;\hat{\Delta} B_{2} = \left( \begin{matrix} 0 & 1 \\ 0 & 0 \end{matrix} \right),\;\; \hat{\Delta} B_{3} = \left( \begin{matrix} 0 & -2 \\ 0 & 0 \end{matrix} \right) \\ &\hat{\Delta} C_{1} = \left( \begin{matrix} 0 & 0 & 1 \\ 0 & 0 & 0 \\ \end{matrix} \right), \;\;\hat{\Delta} C_{2} = \left( \begin{matrix} 0 & 0.5 \\ 0 & 0 \end{matrix} \right),\;\; \hat{\Delta} C_{3} = \left( \begin{matrix} 0 & 0 \\ 0 & 1 \end{matrix} \right). \end{split}

    Assume \hat{C}_{mi} = \hat{C}_{i},\ i = 1,2,3 .

    Under Assumptions 6, 7 and Lemma 4, by choosing some stable closed-loop poles for (48), we can obtain the feedback matrices as follows:

    \begin{split} &H_{1} = \left( \begin{matrix} -7 & -3 \\ -3 & -1 \\ -1 & -5 \end{matrix} \right),\;\; H_{2} = \left( \begin{matrix} -6 & -0.5 \\ -1.5 & -6 \end{matrix} \right) \\ &H_{3} = \left( \begin{matrix} -7 & 7 \\ 1 & -10 \end{matrix} \right) \\ &K_{11} = \left( \begin{matrix} -23.0999 & -10.2533 & 18.7199 \\ -9.2401 & -1.8311 & -3.5066 \end{matrix} \right) \\ &K_{12} = \left( \begin{matrix} -5.0384 & -8.1686 & 12.0224 & 15.9446 \\ -4.2629 & -4.7463 & -0.3934 & -0.4377 \end{matrix} \right) \end{split}
    \begin{split} &K_{21} = \left( \begin{matrix} -1.6736 & -8.5435 \\ -8.9565 & -2.3657 \end{matrix} \right) \\ &K_{22} = \left( \begin{matrix} -1.5966 & -0.8628 & -3.2972 & -4.4391 \\ -5.5551 & -5.4549 & -2.3725 & -1.6594 \end{matrix} \right) \\ &K_{31} = \left( \begin{matrix} -6.8272 & -1.0492 \\ 1.1391 & -6.6728 \end{matrix} \right) \\ &K_{32} = \left( \begin{matrix} -6.1703 & -3.7826 & 5.5077 & 2.6414 \\ 1.8385 & 0.7134 & -5.7390 & -3.3083 \end{matrix} \right). \end{split}

    From the results of the simulation, it is demonstrated that the followers enter the leaders convex hull as t\rightarrow \infty in Figs. 6 and 7. The trajectories are consistent with the expected results. Subsequently, the containment error and virtual error that decay to zero asymptotically are presented in Figs. 8 and 9, respectively. It can be seen that even if the followers have nonidentical state dimensions, multi-agent systems can still achieve output containment control. Overall, the above analysis results show the correctness of Theorem 3.

    Figure  6.  The output trajectories of the followers and the leaders for the closed loop multi-agent system in Example 2.
    Figure  7.  The output trajectories of the followers and the leaders over time for the closed loop multi-agent system in Example 2.
    Figure  8.  The containment error of the followers over time.
    Figure  9.  The virtual error of the followers over time.

    In this paper, we have solved the robust output containment control problem for linear multi-agent systems with a directed fixed network {\cal{G}} . A novel distributed dynamic compensator with measurable leaders’ output was designed to estimate the convex combination of the leaders’ states in combination with the communication network. First, based on the state-coupled item, the internal model compensator and the follower dynamic compensator, the distributed dynamic state and output feedback laws were adopted to drive uncertain followers with identical state dimensions into the convex hull spanned by the leaders. Subsequently, based on the leader compensator, the robust containment problem was converted into an equivalent tracking problem. A new distributed dynamic output feedback control law was designed to drive uncertain followers into the leaders’ convex hull under the output regulation framework. Unlike most existing control approaches, our method addressed the dependence on the relative states; therefore, it is capable of coping with containment problems where the followers have nonidentical state dimensions.

    Proof: First, let all the leaders be a single node, where the Laplacian matrix can be expressed as

    L' = \left( \begin{matrix} L_{1} & L_{2}\cdot{\bf{1}}_{M} \\ {\bf{0}}_{1\times N} & 0 \end{matrix} \right). (57)

    The eigenvalues of L' satisfy the equation det (\lambda I-L_{1})\lambda = 0 , and by Lemma 3.3 in [54], we have that the eigenvalues of L_{1} have positive real parts. The eigenvalues of L satisfy the equation det (\lambda I-L_{1})\lambda^{M} = 0 , so 0 is the M-copy eigenvalues of L and all the nonzero eigenvalues of L have positive real parts.

    Thus, \mathrm{Re}(\lambda_{1})\geq \cdots\geq \mathrm{Re}(\lambda_{N}) > \lambda_{N+1} = \cdots = \lambda_{N+M} = 0, where \lambda_{i}\in \sigma(L),\ i\in {\cal{V}}_{{\cal{F}}}\bigcup {\cal{V}}_{{\cal{L}}} . Assume \alpha_{i} is the right eigenvector of the non-zero eigenvalue \lambda_{i} , then L\alpha_{i} = \lambda_{i}\alpha_{i} hold. Let \alpha_{i} = [\alpha_{ i1(1\times N)}^{T}\quad {\bf{0}}_{1\times M}^{{T}}]^T, then L_{1}\alpha_{i1(1\times N)} = \lambda_{i}\alpha_{i1(1\times N)},\ i\in {\cal{V}}_{\cal{F}} .

    Proof (If part): Let \tilde{w} = \hat{w}-(-L_{1}^{-1}L_{2}\otimes I_{q})w , and denote \Xi = I_{N}\otimes S+L_{1}\otimes \bar{L}Q_{m} ; taking the derivative of \tilde{w} yields:

    \begin{split} \dot{\tilde{w}} =\; &\dot{\hat{w}}-(-L_{1}^{-1}L_{2}\otimes I_{p})\dot{w}\\ =\; &(I_{N}\otimes S+L_{1}\otimes \bar{L}Q_{m})\hat{w} \\ &+(L_{2}\otimes \bar{L}Q_{m})w-(-L_{1}^{-1}L_{2}\otimes S)w \\ =\;&(I_{N}\otimes S+L_{1}\otimes \bar{L}Q_{m})\tilde{w} \\ =\; &\Xi\tilde{w}.\\ \end{split} (58)

    By Lemma 1, L_{1} is nonsingular and there exists a unitary matrix U such that \Lambda = U^{*}L_{1}U is an upper-triangular matrix with \lambda_{i},\;i\in {\cal{V}}_{{\cal{F}}} as the i-th diagonal item. Let

    \begin{split} \bar{\Xi}& = (U^{*}\otimes I_{q})\Xi(U\otimes I_{q}) \\ & = I_{N}\otimes S+\Lambda \otimes \bar{L}Q_{m}. \end{split} (59)

    \bar{\Xi} is Hurwitz if and only if S+\lambda_{i}\bar{L}Q_{m},\ i\in {\cal{V}}_{{\cal{F}}} are Hurwitz. Let

    \bar{L} = -\mu PQ_{m}^{{T}} (60)

    where \mu\in \mathbb{R} satisfies \mu\geq{1}/{\mathrm{min}\{\mathrm{Re}(\lambda_{i}),\lambda_{i}\in\sigma(L_{1})\}}, and P is the solution to SP+PS^{{T}}-PQ^{{T}}_{m}Q_{m}P+I_{q} = 0. Thus, \mathrm{Re}\{\mu\lambda_{i}\}\geq 1 for all \lambda_{i}\in\sigma(L_{1}) . Since (S,Q_{m}) is detectable, (S^{{T}},Q_{m}^{{T}}) is stabilizable. According to Lemma 2, (S+\lambda_{i}\bar{L}Q_{m})^{{T}} = S^{{T}}- \mu\lambda_{i}Q_{m}^{{T}}Q_{m}P,\;\lambda_{i}\in\sigma(L_{1}) are Hurwitz. That is, S+\lambda_{i}\bar{L}Q_{m}, \lambda_{i}\in \sigma(L_{1}) are Hurwitz; thus, Ξ is Hurwitz. Therefore,

    \begin{aligned} \mathop{\mathrm{lim}}\limits_{t\rightarrow \infty}\hat{w} = -(L_{1}^{-1}L_{2}\otimes I_{p})w. \end{aligned} (61)

    Proof (Only if part): If the digraph {\cal{G}} has no united spanning tree, at least one follower cannot receive information from the leaders through the information exchange network. Assuming that the initial states of these followers are all zero, they will always remain zero. Therefore, it is impossible to achieve control for any initial states w_{k}(0),\;k\in{\cal{V}}_{{\cal{L}}}.

  • 1 This paper has two types of dynamic output feedback law, the footnote is to distinguish so as not to confuse.
    2 This footnote is for the same purpose as Footnote 1.
  • [1]
    G. A. Kaminka, R. Schechter-Glick, and V. Sadov, “Using sensor morphology for multirobot formations,” IEEE Trans. Robotics, vol. 24, no. 2, pp. 271–282, 2008. doi: 10.1109/TRO.2008.918054
    [2]
    P. Hernandez-Leon, J. Dvila, S. Salazar, and X. Ping, “Distance-based formation maneuvering of non-holonomic wheeled mobile robot multi-agent system,” IFAC Proceedings Volumes, vol. 53, no. 2, pp. 5665–5670, 2020.
    [3]
    B. Alberto, V. G. Gabriel, D. P. Juan, L. Alvaro, and B. Javier, “Combination of multi-agent systems and wireless sensor networks for the monitoring of cattle,” Sensors, vol. 18, no. 2, pp. 108–124, 2018. doi: 10.3390/s18010108
    [4]
    S. Wang, W. Yang, and H. Shi, “Consensus-based filtering algorithm with packet-dropping,” Acta Automatica Sinica, vol. 36, no. 12, pp. 1689–1696, 2010.
    [5]
    Y. Su and J. Huang, “Cooperative output regulation of linear multi-agent systems,” IEEE Trans. Automatic Control, vol. 57, no. 4, pp. 1062–1066, 2012. doi: 10.1109/TAC.2011.2169618
    [6]
    X. Wang, Y. Hong, J. Huang, and Z. Jiang, “A distributed control approach to a robust output regulation problem for multi-agent linear systems,” IEEE Trans. Automatic Control, vol. 55, no. 12, pp. 2891–2895, 2010. doi: 10.1109/TAC.2010.2076250
    [7]
    X. Liu, Y. Xie, F. Li, P. Shi, W. Gui, and W. Li, “Formation control of singular multi-agent systems with switching topologies,” Int. Journal of Robust and Nonlinear Control, vol. 2, no. 30, pp. 1–13, 2020.
    [8]
    H. Liu, G. Xie, and L. Wang, “Necessary and sufficient conditions for containment control of networked multi-agent systems,” Automatica, vol. 48, no. 7, pp. 1415–1422, 2012. doi: 10.1016/j.automatica.2012.05.010
    [9]
    A. T. Koru, S. B. Sarsilmaz, Selahattin, T. Yucelen, and E. N. Johnson, “Cooperative output regulation of heterogeneous multi-agent systems: A global distributed control synthesis approach,” IEEE Trans. Automatic Control, vol. 66, no. 9, pp. 4289–4296, 2021. doi: 10.1109/tac.2020.3032496
    [10]
    T. Liu, J. Qi, and Z. Jiang, “Distributed containment control of multi-agent systems with velocity and acceleration saturations,” Automatica, vol. 117, pp. 1–10, 2020.
    [11]
    J. Qin, Q. Ma, X. Yu, and Y. Kang, “Output containment control for heterogeneous linear multi-agent systems with fixed and switching topologies,” IEEE Trans. Cybernetics, vol. 49, no. 12, pp. 4117–4128, 2019. doi: 10.1109/TCYB.2018.2859159
    [12]
    H. Haghshenas, M. A. Badamchizadeh, and M. Baradarannia, “Containment control of heterogeneous linear multi-agent systems,” Automatica, vol. 54, pp. 210–216, 2015. doi: 10.1016/j.automatica.2015.02.002
    [13]
    H. Chu, L. Gao, and W. Zhang, “Distributed adaptive containment control of heterogeneous linear multi-agent systems: An output regulation approach,” IET Control Theory and Applications, vol. 10, no. 1, pp. 95–102, 2016. doi: 10.1049/iet-cta.2015.0398
    [14]
    W. Wang, S. Tong, and D. Wang, “Adaptive fuzzy containment control of nonlinear systems with unmeasurable states,” IEEE Trans. Cybernetics, vol. 49, no. 3, pp. 961–973, 2019. doi: 10.1109/TCYB.2018.2789917
    [15]
    Y. Cao, W. Ren, and M. Egerstedt, “Distributed containment control with multiple stationary or dynamic leaders in fixed and switching directed networks,” Automatica, vol. 48, no. 8, pp. 1586–1597, 2012. doi: 10.1016/j.automatica.2012.05.071
    [16]
    J. Shao, L. Shi, M. Cao, and H. Xia, “Distributed containment control for asynchronous discrete-time second-order multi-agent systems with switching topologies,” Applied Mathematics &Computation, vol. 336, pp. 47–59, 2018.
    [17]
    K. Chen, J. Wang, Y. Zhang, and Z. Liu, “Consensus of second-order nonlinear multi-agent systems under state-controlled switching topology,” Nonlinear Dynamics, vol. 81, no. 4, pp. 1871–1878, 2015. doi: 10.1007/s11071-015-2112-3
    [18]
    P. Wang and Y. Jia, “Distributed containment control of second-order multi-agent systems with inherent non-linear dynamics,” IET Control Theory and Applications, vol. 8, no. 4, pp. 277–287, 2014. doi: 10.1049/iet-cta.2013.0686
    [19]
    F. Wang, Z. Liu, Z. Chen, and S. Wang, “Containment control for second-order nonlinear multi-agent systems with intermittent communications,” Int. Journal of Systems Science, vol. 50, no. 5, pp. 919–934, 2019. doi: 10.1080/00207721.2019.1585997
    [20]
    T. Li, Z. Li, S. Fei, and Z. Ding, “Second-order event-triggered adaptive containment control for a class of multi-agent systems,” ISA Transactions, vol. 96, pp. 132–142, 2020. doi: 10.1016/j.isatra.2019.06.003
    [21]
    H. Xia, W. X. Zheng, and J. Shao, “Event-triggered containment control for second-order multi-agent systems with sampled position data,” ISA Transactions, vol. 73, pp. 91–99, 2018. doi: 10.1016/j.isatra.2017.11.001
    [22]
    Y. Zheng and L. Wang, “Containment control of heterogeneous multi-agent systems,” Int. Journal of Control, vol. 87, no. 1, pp. 1–8, 2014. doi: 10.1080/00207179.2013.814074
    [23]
    R. Liao, L. Han, X. Dong, Q. Li, and Z. Ren, “Finite-time formation-containment tracking for second-order multi-agent systems with a virtual leader of fully unknown input,” Neurocomputing, vol. 415, pp. 234–246, 2020. doi: 10.1016/j.neucom.2020.07.067
    [24]
    X. He, Q. Wang, and W. Yu, “Distributed finite-time containment control for second-order nonlinear multi-agent systems,” Applied Mathematics and Computation, vol. 268, pp. 509–521, 2015. doi: 10.1016/j.amc.2015.06.101
    [25]
    F. Wang, Y. Ni, Z. Liu, and Z. Chen, “Fully distributed containment control for second-order multi-agent systems with communication delay,” ISA Transactions, vol. 99, pp. 123–129, 2020. doi: 10.1016/j.isatra.2019.09.009
    [26]
    L. Han, X. Dong, Q. Li, and Z. Ren, “Formation-containment control for second-order multi-agent systems with time-varying delays,” Neurocomputing, vol. 218, pp. 439–447, 2016. doi: 10.1016/j.neucom.2016.09.001
    [27]
    K. Liu, G. Xie, and L. Wang, “Containment control for second-order multi-agent systems with time-varying delays,” Systems &Control Letters, vol. 67, pp. 24–34, 2014.
    [28]
    D. Wang, D. Wang, and W. Wei, “Necessary and sufficient conditions for containment control of multi-agent systems with time delay,” Automatica, vol. 103, pp. 418–423, 2019. doi: 10.1016/j.automatica.2018.12.029
    [29]
    Q. Song, F. Liu, H. Su, and A. V. Vasilakos, “Semi-global and global containment control of multi-agent systems with second-order dynamics and input saturation,” Int. Journal of Robust &Nonlinear Control, vol. 26, no. 16, pp. 3460–3480, 2016.
    [30]
    C. Xu, B. Li, and L. Yang, “Semi-global containment of discrete-time high-order multi-agent systems with input saturation via intermittent control,” IET Control Theory and Applications, vol. 14, no. 16, pp. 2303–2309, 2020. doi: 10.1049/iet-cta.2020.0110
    [31]
    T. Liu, J. Qi and Z. P. Jiang, “ Distributed containment control of multi-agent systems with velocity and acceleration saturations,” Automatica, vol. 117, p. 108992, 2020.
    [32]
    G. Wen, Y. Zhao, and Z. Duan, “Containment of higher-order multi-leader multi-agent systems: A dynamic output approach,” IEEE Trans. Automatic Control, vol. 61, no. 4, pp. 1135–1140, 2016. doi: 10.1109/TAC.2015.2465071
    [33]
    S. Zuo, Y. Song, F. Lewis, and A. Davoudi, “Adaptive output containment control of heterogeneous multi-agent systems with unknown leaders,” Automatica, vol. 92, pp. 235–239, 2018. doi: 10.1016/j.automatica.2018.02.004
    [34]
    J. Zhang and H. Su, “Formation-containment control for multi-agent systems with sampled data and time delays,” Neurocomputing, vol. 424, pp. 125–131, 2021. doi: 10.1016/j.neucom.2019.11.030
    [35]
    X. Dong, Q. Li, Z. Ren, and Y. Zhong, “Formation-containment control for high-order linear time-invariant multi-agent systems with time delays,” Journal of the Franklin Institute, vol. 352, pp. 3564–3584, 2015. doi: 10.1016/j.jfranklin.2015.05.008
    [36]
    G. Wen, G. Hu, Z. Zuo, Y. Zhao, and J. Cao, “Robust containment of uncertain linear multi-agent systems under adaptive protocols,” Int. Journal of Robust and Nonlinear Control, vol. 27, no. 12, pp. 2053–2069, 2017. doi: 10.1002/rnc.3670
    [37]
    G. Wen, P. Wang, T. Huang, W. Yu, and J. Sun, “Robust neuro-adaptive containment of multileader multi-agent systems with uncertain dynamics,” IEEE Trans. Systems,Man,and Cybernetics:Systems, vol. 99, pp. 1–12, 2017.
    [38]
    X. Wang, Y. Hong, and H. Ji, “Adaptive multi-agent containment control with multiple parametric uncertain leaders,” Automatica, vol. 50, no. 9, pp. 2366–2372, 2014. doi: 10.1016/j.automatica.2014.07.019
    [39]
    J. Chen, Z. Guan, C. Yang, T. Li, D. He, and X. Zhang, “Distributed containment control of fractional-order uncertain multi-agent systems,” Journal of the Franklin Institute, vol. 353, no. 7, pp. 1672–1688, 2016. doi: 10.1016/j.jfranklin.2016.02.002
    [40]
    P. Li, F. Jabbari and X. M. Sun, “Containment control of multi-agent systems with input saturation and unknown leader inputs,” Automatica, vol. 130, p. 109677, 2021.
    [41]
    L. Wang, T. Han, X. S. Zhan, J. Wu, and H. Yan, “Bipartite containment for linear multi-agent systems subject to unknown exogenous disturbances,” Asian Journal of Control, DOI: 10.1002/asjc.2580, 2021.
    [42]
    T. Han, B. Xiao, X. S. Zhan, and H. Yan, “Bipartite containment of descriptor multi-agent systems via an observer-based approach,” IET Control Theory and Applications, vol. 14, no. 9, pp. 3047–3051, 2020.
    [43]
    J. Huang, Nonlinear Output Regulation: Theory and Applications, Philadelphia, PA: SIAM, 2004.
    [44]
    H. Liang, H. Zhang, Z. Wang, and J. Wang, “Output regulation of state-coupled linear multi-agent systems with globally reachable topologies,” Neurocomputing, vol. 123, pp. 337–343, 2014. doi: 10.1016/j.neucom.2013.07.028
    [45]
    Y. Su and J. Huang, “Cooperative output regulation of linear multi-agent systems by output feedback,” Systems &Control Letters, vol. 61, pp. 1248–1253, 2012.
    [46]
    Y. Su, Y. Hong, and J. Huang, “A general result on the robust cooperative output regulation for linear uncertain multi-agent systems,” IEEE Trans. Automatic Control, vol. 58, no. 5, pp. 1275–1279, 2013. doi: 10.1109/TAC.2012.2229837
    [47]
    S. Li, J. Zhang, M. Er, X. Luo, Z. Yang, and N. Wang, “Robust containment control of heterogeneous non-linear multi-agent systems via power series approach,” IET Control Theory and Applications, vol. 13, no. 4, pp. 496–505, 2019. doi: 10.1049/iet-cta.2018.5385
    [48]
    S. E. Tuna, “LQR-based coupling gain for synchronization of linear systems”, 2008. [Online]. Available: http://arxiv.org/abs/0801.3390. Accessed on: May 21, 2019.
    [49]
    S. Boyd, L. E. Ghaoui, E. Feron, and V. Balakrishnan, Linear Matrix Inequality in Systems and Control Theory, Philadelphia, PA: SIMA, 1994.
    [50]
    B. A. Francis and W. M. Wonham, “The internal model principle for linear multivariable regulators,” Applied Mathematics and Optimization, vol. 2, no. 2, pp. 170–194, 1975. doi: 10.1007/BF01447855
    [51]
    B. A. Francis and W. M. Wonham, “The internal model principle of control theory,” Automatica, vol. 12, no. 5, pp. 457–465, 1976. doi: 10.1016/0005-1098(76)90006-6
    [52]
    E. J. Davison, “The robust control of a servomechanism problem for linear time-invariant multivariable systems,” IEEE Trans. Automatic Control, vol. 21, no. 1, pp. 25–34, 1976. doi: 10.1109/TAC.1976.1101137
    [53]
    H. Liang, H. Zhang, Z. Wang, and J. Zhang, “Output regulation for heterogeneous linear multi-agent systems based on distributed internal model compensator,” Applied Mathematics &Computation, vol. 242, pp. 736–747, 2014.
    [54]
    W. Ren and R. W. Beard, “Consensus seeking in multi-agent systems under dynamically changing interaction topologies,” IEEE Trans. Automatic Control, vol. 50, no. 5, pp. 655–661, 2005. doi: 10.1109/TAC.2005.846556
  • Related Articles

    [1]Lihong Feng, Bonan Huang, Jiayue Sun, Qiuye Sun, Xiangpeng Xie. Adaptive Event-Triggered Time-Varying Output Group Formation Containment Control of Heterogeneous Multiagent Systems[J]. IEEE/CAA Journal of Automatica Sinica, 2024, 11(6): 1398-1409. doi: 10.1109/JAS.2024.124260
    [2]Chi Ma, Dianbiao Dong. Finite-time Prescribed Performance Time-Varying Formation Control for Second-Order Multi-Agent Systems With Non-Strict Feedback Based on a Neural Network Observer[J]. IEEE/CAA Journal of Automatica Sinica, 2024, 11(4): 1039-1050. doi: 10.1109/JAS.2023.123615
    [3]Jianguo Zhao, Chunyu Yang, Weinan Gao, Linna Zhou, Xiaomin Liu. Adaptive Optimal Output Regulation of Interconnected Singularly Perturbed Systems With Application to Power Systems[J]. IEEE/CAA Journal of Automatica Sinica, 2024, 11(3): 595-607. doi: 10.1109/JAS.2023.123651
    [4]Jingshu Sang, Dazhong Ma, Yu Zhou. Group-Consensus of Hierarchical Containment Control for Linear Multi-Agent Systems[J]. IEEE/CAA Journal of Automatica Sinica, 2023, 10(6): 1462-1474. doi: 10.1109/JAS.2023.123528
    [5]Haihua Guo, Min Meng, Gang Feng. Lyapunov-Based Output Containment Control of Heterogeneous Multi-Agent Systems With Markovian Switching Topologies and Distributed Delays[J]. IEEE/CAA Journal of Automatica Sinica, 2023, 10(6): 1421-1433. doi: 10.1109/JAS.2023.123198
    [6]Yuan Jiang, Jiyang Dai. Adaptive Output Regulation of a Class of Nonlinear Output Feedback Systems With Unknown High Frequency Gain[J]. IEEE/CAA Journal of Automatica Sinica, 2020, 7(2): 568-574. doi: 10.1109/JAS.2020.1003060
    [7]Adedapo Odekunle, Weinan Gao, Yebin Wang. Data-Driven Global Robust Optimal Output Regulation of Uncertain Partially Linear Systems[J]. IEEE/CAA Journal of Automatica Sinica, 2019, 6(5): 1108-1115. doi: 10.1109/JAS.2019.1911678
    [8]Hongyong Yang, Fuyong Wang, Fujun Han. Containment Control of Fractional Order Multi-Agent Systems With Time Delays[J]. IEEE/CAA Journal of Automatica Sinica, 2018, 5(3): 727-732. doi: 10.1109/JAS.2016.7510211
    [9]Long Ma, Haibo Min, Shicheng Wang, Yuan Liu, Zhiguo Liu. Distributed Containment Control of Networked Nonlinear Second-order Systems With Unknown Parameters[J]. IEEE/CAA Journal of Automatica Sinica, 2018, 5(1): 232-239. doi: 10.1109/JAS.2016.7510235
    [10]Zhenhua Wang, Juanjuan Xu, Huanshui Zhang. Consensus Seeking for Discrete-time Multi-agent Systems with Communication Delay[J]. IEEE/CAA Journal of Automatica Sinica, 2015, 2(2): 151-157.
    [11]Jie Huang, Hao Fang, Lihua Dou, Jie Chen. An Overview of Distributed High-order Multi-agent Coordination[J]. IEEE/CAA Journal of Automatica Sinica, 2014, 1(1): 1-9.
    [12]Chenghui Zhang, Le Chang, Xianfu Zhang. Leader-follower Consensus of Upper-triangular Nonlinear Multi-agent Systems[J]. IEEE/CAA Journal of Automatica Sinica, 2014, 1(2): 210-217.
    [13]Hongbin Ma, Yini Lv, Chenguang Yang, Mengyin Fu. Decentralized Adaptive Filtering for Multi-agent Systems with Uncertain Couplings[J]. IEEE/CAA Journal of Automatica Sinica, 2014, 1(1): 101-112.
    [14]Yi Dong, Jie Huang. Leader-following Rendezvous with Connectivity Preservation of Single-integrator Multi-agent Systems[J]. IEEE/CAA Journal of Automatica Sinica, 2014, 1(1): 19-23.
    [15]Wen Qin, Zhongxin Liu, Zengqiang Chen. Formation Control for Nonlinear Multi-agent Systems with Linear Extended State Observer[J]. IEEE/CAA Journal of Automatica Sinica, 2014, 1(2): 171-179.
    [16]Hongjing Liang, Huaguang Zhang, Zhanshan Wang, Junyi Wang. Consensus Robust Output Regulation of Discrete-time Linear Multi-agent Systems[J]. IEEE/CAA Journal of Automatica Sinica, 2014, 1(2): 204-209.
    [17]Hejin Zhang, Zhiyun Zhao, Ziyang Meng, Zongli Lin. Experimental Verification of a Multi-robot Distributed Control Algorithm with Containment and Group Dispersion Behaviors: the Case of Dynamic Leaders[J]. IEEE/CAA Journal of Automatica Sinica, 2014, 1(1): 54-60.
    [18]Chuanrui Wang, Xinghu Wang, Haibo Ji. A Continuous Leader-following Consensus Control Strategy for a Class of Uncertain Multi-agent Systems[J]. IEEE/CAA Journal of Automatica Sinica, 2014, 1(2): 187-192.
    [19]Airong Wei, Xiaoming Hu, Yuzhen Wang. Tracking Control of Leader-follower Multi-agent Systems Subject to Actuator Saturation[J]. IEEE/CAA Journal of Automatica Sinica, 2014, 1(1): 84-91.
    [20]Huiyang Liu, Long Cheng, Min Tan, Zengguang Hou. Containment Control of General Linear Multi-agent Systems with Multiple Dynamic Leaders: a Fast Sliding Mode Based Approach[J]. IEEE/CAA Journal of Automatica Sinica, 2014, 1(2): 134-140.

Catalog

    通讯作者: 陈斌, bchen63@163.com
    • 1. 

      沈阳化工大学材料科学与工程学院 沈阳 110142

    1. 本站搜索
    2. 百度学术搜索
    3. 万方数据库搜索
    4. CNKI搜索

    Figures(9)

    Article Metrics

    Article views (911) PDF downloads(176) Cited by()

    Highlights

    • For the robust output containment control problem with the uncertain followers of identical nominal dynamics, based on the internal model principle and the compensator technique, the distributed dynamic state and output feedback control laws were introduced to drive the uncertain followers to enter the convex hull spanned by the leaders under the output regulation framework. Among them, the nonsingular transformation and a Lyapunov inequality method was used to analysis the closed-loop system stabilization
    • By introducing the distributed observer systems, we extended the theoretical results to a more general case where the followers have nonidentical state dimensions. In this section, the robust containment problem was converted into a new tracking problem between the distributed observer systems and the follower systems by constructing a virtual error. A distributed dynamic output feedback control law was further devised to drive the virtual error to converge to the origin asymptotically, such that multi-agent systems achieved output containment control
    • At present, most of the containment control protocols must know the relative values of the state with respect to its neighbors, so they can only deal with the situation with followers having the same state dimensions. In our research, the distributed control law has avoided the dependence on the relative states of followers by utilizing the virtual error, and is capable of solving the robust output containment problem for linear heterogeneous multi-agent systems with nonidentical state dimensions
    • We modify the conventional state observer to produce an estimate of the convex combination of the leaders’ states by applying a directed network, which lends itself to the design of the distributed protocols. Moreover, the distributed observer can also be viewed as an extension of the compensators associated with the leaders’ states in some relevant literature, because these compensators can be regarded as a special case of our observer when the output matrix of the leader systems is of full column rank

    /

    DownLoad:  Full-Size Img  PowerPoint
    Return
    Return