This paper investigates the problem of leader-following consensus and multiconsensus of multiagent systems. A leader-following and bounded consensus protocol via impulsive control for multiagent systems is proposed by using sampled position and velocity data. Consensus and multiconsensus commutative evolution stimulated by varying intelligence degrees of each agent can be achieved to avoid obstacles. Especially, the multiconsensus leader-following can be obtained without grouping the multiagent networks in advance. The necessary and sufficient condition is given to the leader-following bounded consensus tracking of the system by using the Hurwitz criterion and properties of the Laplacian matrix. A simulation is provided to verify the availability of the proposed impulsive control protocol. Furthermore, the result can be applied in obstacle avoidance and round up of target by regulating the intelligence degrees.
National Natural Science Foundation of China6206301161922037Scientific Research Fund of Yunnan Education Department2020J00672019J00461. Introduction
In recent years, multiagent systems have attracted a great deal of attention from researchers because agents can communicate and coordinate with each other for achieving various complex tasks, such as wireless sensor navigation [1], distributed strikes of missiles [2], and coordination of robots [3–7]. Olfati-saber Murray [8] proposed the concept of consensus protocol for multiagent systems, and the consensus of multiagent systems with directed topology, undirected topology, and switching topology was studied. Ren and Beard [9] studied the directed switching topology, and their results showed that if the network topology contains a directed spanning tree, then the multiagent systems will achieve consensus. At present, tons of research studies have been conducted about the consensus of multiagent systems, such as the delay [10], heterogeneous [11], and aperiodic sampled [12] studies.
As a special case of consensus, multiconsensus/group consensus has received much attention for its multiconvergent states and promising potential applications. Pu et al. [13] investigated the group consensus for discrete-time heterogeneous multiagent systems with input time delays. Ji et al. [14] investigated group consensus based on linearly coupled multiagent networks, and they extended their work to study the group consensus for the multiagent network with generally connected topology which neither needs to be strongly connected nor needs to contain a directed spanning tree. Shang [15] studied the L-1 group consensus problem of discrete-time multiagent systems with external stochastic inputs. L-1 group consensus criteria were obtained for multiagent systems with switching topologies. Furthermore, a large number of results were produced, such as multiconsensus of continuous-time systems [16], tracking control [17], and event-driver control [18]. Considering that most of the studies were performed by presenting groupings, Han et al. [19] proposed the concept of intelligence degree and also gave necessary and sufficient conditions of multiconsensus with fixed intelligence degree. By adjusting the intelligent degree of each agent, it is possible to regulate the dynamic behaviour of the multiagent system.
All agents following the trajectory of the leader is called leader-following consensus in multiagent systems. Leader-following consensus is one of the hot topics of multiagent systems, which are widely applied in military field, such as striking a known dynamic target [20] and UAV formation [21]. Considering a nonlinear multiagent system, He et al. [22] proposed a distributed consensus protocol with probabilistic sampling in two sampling periods and solved the problem of sampled-data leader-following consensus of a group of agents with nonlinear characteristics. Zhu et al. [23] achieved the leader-following consensus by adaptive event control. Ren et al. [24] designed a fuzzy adaptive control mechanism for nonlinear multiagent systems with unknown control directions. However, as far as I am concerned, almost all works mentioned above paid attention to how to obtain a consensus protocol or better consensus performances under the given communication constraints. Currently, some works discussed how to avoid static obstacles and dynamic obstacles during the process of following the trajectory of the leader under given environment constraints. Douthwaite et al. [25] presented a critical analysis of some of the most promising approaches to geometric collision avoidance in multiagent systems, and the respective features and limitations of each algorithm were discussed and presented through examples. Guo et al. [26] used model predictive control to realize collision avoidance of multiagent systems. This technique has strong practical significance especially in dynamic path planning, round up, and battle among USV or UAV. On the one hand, the leader tries to plan an available path in order to overcome static obstacles [27] and dynamic obstacles [28]. On the other hand, all group members should build an effective mechanism to avoid obstacles when the preplanned path is disrupted by moving obstacles.
Therefore, combined with leader-following mechanism, obstacle avoidance, and consensus of multiagent systems, the problem of leader-following consensus and multiconsensus of second-order continuous-time multiagent systems is considered in this paper. A leader-following bounded consensus protocol with different intelligence degrees via impulsive control is proposed. All grouped members follow the trajectory of the virtual leader in the dynamic process, and then, dynamic obstacles which hinder the preplanned path are avoided by regulating suitable intelligence degree of agents which results in varying grouping and different dynamic behaviours of the system. Sufficient and necessary conditions of multiconsensus leader-following of multiagent systems via impulsive control are given. Some simulation results show that the impulsive control protocol with different intelligent degrees can realize the consensus and multiconsensus leader-following.
The rest of paper is organized as follows. In Section 2, we provide some preliminary notions including lemmas in graph theory and matrix theory. The impulsive consensus leader-following protocol is also introduced, and the main result of this paper, that is, the bounded convergence of the proposed protocol, is presented in Section 3. In Section 4, an illustrative numerical example is given. The paper is concluded in Section 5.
2. Preliminaries and Problem Formulation2.1. Notions from Graph Theory
Let G=V,E,A be a digraph with the set of nodes V=1,…,n, the set of edges E⊆V×V, and the weighted adjacency matrix A=aij∈Rn×n in which aij≠0 if j,i∈E and aij=0 otherwise. Then, weight reuse di=∑j=1naij for node i, and the diagonal matrix D=diagd1,…,dn represents the input weight matrix of the directed graph G.
The Laplace matrix L=lij∈Rn×n of G is defined as(1)lij=−aij,j≠i,∑k=1,k≠inaik,j=i.
A directed path is a sequence of edges in a digraph of the form i1,i2…in−1,in with all distinct ij. A directed tree is a digraph, where there exists an agent, called the root, such that any other agent of the digraph can be reached by one and only one path starting at the root [29].
2.2. Preliminaries
Assume that agent i evolves according to the multiagent system dynamics:(2)x˙it=vit,v˙it=uit,∀i∈V,where xit and vit∈Rn are the position and velocity states of the agent i, respectively, and uit∈Rn is its control input. The communications between agents are given by a digraph G=V,E,W.
Definition 1.
Multiagent system (2) is said to achieve bounded consensus if for any initial state values xit0 and vit0, it follows that(3)suplimt⟶∞xit−xjt≤ε,suplimt⟶∞vit−vjt≤ξ,where ε and ξ are constants.
Definition 2.
Multiagent system (2) is said to achieve dynamic following consensus if for any initial state values xit0 and vit0, it follows that(4)limt⟶∞xit=xrt,limt⟶∞vit=vrt,where xrt and vrt are the position and velocity of the leader.
Lemma 1 (see [30]).
Consider a polynomial Qz=z2+m+inz+p+iq, where m, n, p, and q are real constants; then, Qz is Hurwitz stable if and only if m>0 and mnq+m2p−q2>0.
Lemma 2 (see [31]).
All the roots of the equation s2+as+b=0, where a,b∈C, are within the unit circle if and only if all the roots of the equation a+b+1z2+21−bz+b−a+1=0 are in the open left half plane.
Lemma 3 (see [13]).
Laplacian matrix L∈Rn×n has a simple eigenvalue 0, and all the nonzero eigenvalues have positive real parts if and only if the digraph has a spanning tree. In addition, L1n=0n, and there exists a nonnegative vector q∈Rn×1 satisfying qTL=0nT and qT1n=1.
Lemma 4 (see [32]).
Let A11,A12,A21,A22∈Rn×n and A=A11A12A21A22. Then, detA=detA11A22−A12A21 if A11,A12,A21, and A22 commute pairwise.
3. Problem Illuminations
In the section, a dynamic leader-following consensus protocol via impulsive control with different intelligence degrees is designed and necessary and sufficient conditions for its convergence are also given.
3.1. Leader-Following Consensus Protocol via Impulsive Control with Different Intelligence Degrees
A leader-following and bounded consensus protocol is designed as(5)uit=−∑l=1∞p1∑j=1nωitωjtlijxjt+p2∑j=1nωitωjtlijvjt+p1bitxit−xrt+p2bitvit−vrtδt−tl,where δ⋅ is Dirac impulsive function and the agents intelligence degrees is a set of ω1t,…,ωnt. The sampling interval is denoted as h with tl−tl−1≡h>0.p1 and p2 are feedback gains of position error and speed error, respectively. The bt is a time-varying parameter, which is denoted as bt=b1t,b2t,…,bnt with adjacency element bit>0 if agent i is a neighbor of the leader; otherwise, bit=0. Especially, xrt and vrt are the position and speed of the virtual leader at t time instance, respectively.
Remark 1.
The system runs without any control input if t≠tl, and the control input is available at t=tl to update the information from its neighbors.
Remark 2.
If there are multiple leaders in the multiagent system, which means position and speed xrt and vrt are replaced by xrqt and vrqt,q>1, then the system will achieve multiobject tracking under the leader-following protocol.
System (2) and consensus protocol (5) can be rewritten as follows:(6)x˙it=vit,v˙it=0,Δvitl=−p1∑j=1nωitlωjtllijxjt−p2∑j=1nωitlωjtllijvjtl−p1bitxiti−xrt−p2bitvit−vrt,where Δvitl=vitl+−vitl and vit is left-continuous at t=tl. Let x^it=xit−xrt and v^it=vit−vrt. Since ∑i=1nlijxrt=0, Lxrt=0n, and ∑i=1nlijvrt=0, Lvrt=0n, the system (6) could be written as(7)x^⋅it=v^it,v^⋅it=0,Δv^itl=−p1∑j=1nωitlωjtllijx^jtl−p2∑j=1nωitlωjtllijv^jtl−p1bitlx^itl−p2bitlv^it.
From (7), one has(8)x^itl+1=x^itl+hv^itl+,v^itl+1=v^itl+,v^itl+=v^itl−p1∑j=1nωitlωjtllijx^jtl−p2∑j=1nωitlωjtllijv^jtl−p1bitlx^itl−p2bitlv^itl,where h is the sampling interval.
Therefore, states x^it and x^jt under protocol (5) can be described, respectively, as follows:(9)x^itl+1=x^itl+v^itl−p1∑j=1nωitlωjtllijx^jtl−p2∑j=1nωitlωjtllijv^jtl−p1bitlx^itl−p2bitlv^itlh,v^itl+1=v^itl−p1∑j=1nωitlωjtllijx^jtl−p2∑j=1nωitlωjtllijv^jtl−p1bitlx^itl−p2bitv^itl.
Let x^tl=x^1tl,…,x^ntlT,v^tl=v^1tl,…,v^ntl and(10)Btl=b1tlb2tl⋱bntl,then equation (9) can be equivalently expressed as follows:(11)x^tl+1v^tl+1=Ptlx^tlv^tl,where(12)Ptl=In−hp1L^tl+BtlhIn−hp2L^tl+Btl−p1L^tl+BtlIn−p2L^tl+Btl.
Let(13)Xtl=ω1tl⋱ωntl,Xtl−1=1ω1tl⋱1ωntl.
The linear transformation of matrix L^tl=XtlLXtl−1 is as follows:(14)L^tl=l11ω1tlω2tll12⋯ω1tlωntll1nω2tlω1tll21l22⋯ω2tlωntll2n⋮⋮⋱⋮ωntlω1tlln1ωntlω2tlln2⋯lnn.
Matrix Ptl has two eigenvalues 1, and the other eigenvalues lie in the unit circle, if and only if the digraph G contains a spanning tree and(16)0<p2<2Reλitlλitl2,(17)0<h<minλitl≠0−2p23λitl4+4p22Reλitlλitlp22λitl4+4Imλitl2p12,where the eigenvalues of matrix L^tl+Btl are λitl and the real and imaginary parts of the matrix L^tl+Btl are Reλitl and Imλitl.
Proof 1.
The characteristic polynomial of Ptl is given by(18)detsI2n−Ptl=detsIn−In+hp1L^tl+Btl−hIn+hp2L^tl+Btlp1L^tl+BtlsIn−In+p2L^tl+Btl=dets2In+shp1L^tl+Btl+p2L^tl+Btl−2In+In−p2L^tl+Btl=∏i=1ns2+shp1λitl+p2λitl−2+1−p2λitl.
Let(19)s2+shp1λitl+p2λitl−2+1−p2λitl.
By Lemma 2, the eigenvalues of matrix Ptl lie in the unit circle if and only if all the roots of(20)z2+2p2hp1z+41/λitl−2p2−hp1hp1=0,are in the open left half plane.
Define Qiz as(21)Qiz=z2+2p2hp1z+41/λitl−2p2−hp1hp1=z2+2p2hp1z+−2p2−hp1λitl2+4Reλitlhp1λitl2−4Imλitlhp1λitl2i.
Combined with Lemma 1 that Qiz is Hurwitz stable, if and only if 2p2/hp1>0, and 4p22λitl2−2p2−hp1λitl2+4Reλitl−16hp1Imλitl2/h3p13λitl4>0, it is indicated that(22)4p22λitl2−2p2−hp1λitl2+4Reλitl−16hp1Imλitl2>0.
Then, one has(23)0<h<minλitl≠0−2p23λitl4+4p22Reλitlλitlp22λitl4+4Imλitl2p12.
By Lemma 3, Ptl has two eigenvalues 1, and the other eigenvalues lie in the unit circle if and only if the digraph has a spanning tree, and(25)0<p2<2Reλitlλitl2,0<h<minλitl≠0−2p23λitl4+4p22Reλitlλitlp22λitl4+4Imλitl2p12.
This completes the proof.
If there is not a leader-following mechanism in system (2), the multiagent system becomes a general consensus system. That is, letting(26)B0tl=00⋱0,L^tl+B0tl=L^tl,then(27)P0tl=IN−hp1L^tlhIN−hp2L^tl−p1L^tlIN−p2L^tl,where P0t is the state space matrix such that there is no leader-following system (2).
Remark 3.
L^tl has one right and left eigenvector p^tl=ω1tl,…,ωntlT and q^tl∈R1×n associated with the eigenvalue 0, respectively, such that L^tlp^tl=0 and q^tlL^tl=0T with q^tlp^tl=1 based on Lemma 3.
If there is no virtual leader, this system can only achieve bounded consensus. Then, we can achieve obstacle avoidance by changing the intelligent degree of agents. For illuminating the multiconsensus convergence of multiagent systems, Theorem 2 is given to prove the final convergent states.
Theorem 2.
Under impulsive control protocol (5), as tl⟶∞ and there is no leader-following, multiagent system (2) of the states of agents is(28)xitl⟶ωitlq^tlxt0+vt0tl,(29)vitl⟶ωitlq^tlvt0,if and only if the digraph contains a spanning tree and equations (16) and (17) hold.
Proof 2.
Sufficiency: the matrix P0tl has one right and generalized right eigenvector φ1tl=1,0T⊗p^tl and φ2tl=0,1/hT⊗p^tl associated with the eigenvalue 1, respectively. Moreover, P0tl has one left and generalized left eigenvector γ1tl=1,0⊗q^tl and γ2tl=0,h⊗q^tl associated with the eigenvalue 1, respectively. There exist nonsingular matrices Φtl=φ1tl,…,φ2ntl and Γtl=γ1Ttl,…,γ2nTtlT such that P0tl can be written in a Jordan canonical form as(30)P0tl=ΦtlJ0JΓtl,where J0=111 and ΦtlΓtl=I2n.
Hence, one can obtain that(31)P0tll=ΦtlJ0lJlΓtl,where J0l=1l1. Denote that(32)∏l≜φ1tl,φ2tlJ0lγ1Ttl,γ2TtlT=ω1tltlω1tl⋮⋮ωntltlωntl0ω1tl⋮⋮0ωntl⊗q^tl,where tl=lh.
Letting Φ¯tl=φ3tl,…,φ2ntl and Γ¯tl=γ3tl,…,γ2ntl, then(33)liml⟶∞P0tll−∏l=liml⟶∞Φ¯tlJlΓ¯tl=0.
As tl⟶∞, we have(34)xitl⟶ωitlq^tlxt0+vt0tl,vitl⟶ωitlq^tlvt0.
Necessity: if the dynamic bounded consensus is solved in protocol (5), it holds that liml⟶∞P0tll has rank two; that is, P0tll has two eigenvalues 1 and the other eigenvalues lie in the unit circle. Based on Theorem 1, the digraph has a spanning tree, and (16) and (17) hold.
This completes the proof.
Remark 4.
From equations (28) and (29), we know that if we change the dynamic behaviour of agents by regulating the intelligent degree of agents, the dynamic characteristic of the agents at each moment can be obtained by equations (28) and (29).
4. Numerical Example
In this section, an illustrative example is given to demonstrate the correctness of the theoretical results.
Example 1.
Consider the multiagent system with 30 agents as shown in Figure 1.
The directed topology with 30 agents, where agent 0 is the virtual leader.
The dynamic equation of the virtual leader are x0t=0.1∗t and v0t=0.001 in simulation. Choose p1=0.4 and p2=0.3, and the result is 0<h<0.5035 via Theorem 1. Therefore, choose h=0.5 and h=0.504, and the intelligent degrees belong to ωt∈5,5.5; the results are shown in Figures 2 and 3.
Trajectory of leader-following bounded consensus with h=0.5.
Trajectory of leader-following bounded consensus with h=0.504.
Choose the intelligence degrees from ωt∈3,3.5, ωt∈10,10.5, and ωt∈16,16.5 as initial values, respectively. Meanwhile, there is a virtual leader in the system. When time is 30 seconds, the intelligence degrees of agents become ωt∈3,3.5. The multiconsensus to consensus dynamic evolution is realized, as shown in Figure 4.
Trajectory of multiconsensus to consensus dynamic evolution by changed intelligence degrees of some agents with h=0.1, p1=0.4, and p2=0.3.
For showing obstacle avoidance through the dynamic evolution from consensus to multiconsensus when initial intelligence degrees are switched from ωt∈3,3.5 to choose the intelligence degrees from ωt∈3,3.5, it is shown that ωt∈10,10.5 and ωt∈16,16.5. Meanwhile, there is no virtual leader within t∈100,200, and details are shown in Figure 5.
Trajectory of consensus to multiconsensus dynamic evolution and obstacle avoidance by changed intelligences degrees with h=0.1, p1=0.4, and p2=0.3, and the red point is the obstacle in the trajectory of the virtual leader.
In Figure 5, if the anticipate trajectory of the virtual leader is hindered by the red and round obstacle, agent 1 gives up to connect the virtual leader. Then, all agents no longer follow the virtual leader and change their intelligence degrees. As a result, the multiagent system will run according to multiconsensus behaviour. The virtual leader will not reestablish communication with agent 1 until all agents have avoided the obstacle. Then, the system achieves leader-following and bounded consensus again.
5. Conclusions
In this paper, the leader-following consensus and multiconsensus problem has been considered for continuous-time second-order multiagent systems. The system could achieve consensus and multiconsensus commutative evolution by regulating the intelligence degrees. An impulsive control protocol with different intelligence degrees is proposed to perform obstacle avoidance, and some necessary and sufficient conditions for leader-following bounded consensus and multiconsensus have been obtained. It is found that the consensus and multiconsensus leader-following are critically dependent on the sampling interval, control gains, and the intelligence degrees. In the future, one of our works will consider multiagent systems achieving obstacle avoidance automatically.
Data Availability
The data used to support the findings of the study are available from the corresponding author upon request.
Conflicts of Interest
The authors declare that there are no conflicts of interest regarding the publication of this paper.
Acknowledgments
This work was supported by the National Nature Science Foundation of China under Grants 62063011 and 61922037 and the Scientific Research Fund of Yunnan Education Department under Grants 2020J0067 and 2019J0046.
LeeH. M.MinS. D.ChoiM. H.Multi-agent system for fault tolerance in wireless sensor networks201610313211332ZouL.DingQ.ZhouR.Distributed cooperative guidance for multiple heterogeneous networked missiles2010361214321435XiaoH.ChenC. L. P.Leader-follower consensus multi-robot formation control using neurodynamic-optimization-based nonlinear model predictive control20197435814359010.1109/access.2019.29079602-s2.0-85064624710WangS. B.NaJ.XingY. S.Adaptive optimal parameter estimation and control of servo mechanisms: theory and experiment201910.1109/TIE.2019.2962445WangS. B.NaJ.Parameter estimation and adaptive control for servo mechanisms with friction compensation2020161168166825WangS. B.TaoL.ChenQ.NaJ.RenX. M.USDE-Based sliding model control for servo mechanisms with unknown system dynamics202025210561066ChenQ.YuX. Q.SunM. X.WuC.FuZ. J.Adaptive repetitive learning control of PMSM servo systems with bounded nonparametric uncertainties: theory and experiments202010.1109/TIE.2020.3016257Olfati-SaberR.MurrayR. M.Consensus problems in networks of agents with switching topology and time-delays20044991520153310.1109/tac.2004.8341132-s2.0-4644244041RenW.BeardR. W.Consensus seeking in multi-agent systems under dynamically changing interaction topologies2005505655661SunF.ShenY.KurthsJ.ZhuW.Mean-square consensus of multi-agent systems with noise and time delay via event-triggered control202035795317533910.1016/j.jfranklin.2020.02.047CaiY.ZhangH.GaoZ.SunS.The distributed output consensus control of linear heterogeneous multi-agent systems based on event-triggered transmission mechanism under directed topology202035763267329810.1016/j.jfranklin.2019.11.051LiuZ.-W.GuanZ.-H.ShenX.FengG.Consensus of multi-agent networks with aperiodic sampled communication via impulsive algorithms using position-only measurements201257102639264310.1109/tac.2012.22144512-s2.0-84866952371PuX.ZhaoL.XiongC.Weighted group consensus for discrete-time heterogeneous multi-agent systems in the cooperative-competitive network with time delays2019712367912368810.1109/access.2019.2934988JiL.LiuQ.LiaoX.On reaching group consensus for linearly coupled multi-agent networks201428711210.1016/j.ins.2014.07.0242-s2.0-84906855391ShangY.Group consensus of multi-agent systems with switching topologies and stochastic inputs201337725-271582158610.1016/j.physleta.2013.04.0542-s2.0-84877777769AlmeidaJ.SilvestreC.PascoalA. M.Continuous-time consensus with discrete-time communications201261778879610.1016/j.sysconle.2012.04.0042-s2.0-84861203694ChenQ.ShiH. H.SunM. X.Echo state network based backstepping adaptive iterative learning control for strict-feedback systems: an error-tracking approach201950710.1109/TCYB.2019.2931877.2019QianY.-Y.LiuL.FengG.Distributed event-triggered adaptive control for consensus of linear multi-agent systems with external disturbances20205052197220810.1109/tcyb.2018.28814842-s2.0-85058075679HanG.-S.HeD.-X.GuanZ.-H.HuB.LiT.LiaoR.-Q.Multi-consensus of multi-agent systems with various intelligences using switched impulsive protocols2016349-35018819810.1016/j.ins.2016.02.0382-s2.0-84960361598SunJ.LiuC.Distributed fuzzy adaptive backstepping optimal control for nonlinear multi-missile guidance systems with input saturation20183110.1109/tfuzz.2018.28599042-s2.0-85050646781ZhangB.SunX.LiuS.DengX.Adaptive differential evolution-based receding horizon control design for multi-UAV Formation reconfiguration201917123009302010.1007/s12555-018-0421-22-s2.0-85070835872HeW.ZhangB.HanQ. L.QianFKurthsJCaoJLeader-following consensus of nonlinear multiagent systems with stochastic sampling2017472327338ZhuW.WangD.ZhouQ.Leader-following consensus of multi-agent systems via adaptive event-based control201932384685610.1007/s11424-018-7177-x2-s2.0-85056902281RenC.-E.ChenC. L. P.DuT.GuanY.Fuzzy adaptive leader-following consensus control for nonlinear multi-agent systems with unknown control directions20192172066207610.1007/s40815-019-00710-12-s2.0-85070642623DouthwaiteJ. A.ZhaoS.MihaylovaL. S.Velocity obstacle approaches for multi-agent collision avoidance201971556410.1142/s23013850194000652-s2.0-85063295529GuoY.ZhouJ.LiuY.Distributed Lyapunov-based model predictive control for collision avoidance of multi-agent formation201812182569257710.1049/iet-cta.2018.53172-s2.0-85057612836ZhangW.MaY.ZhaoH. D.Obstacle avoidance path planning of intelligent mobile based on improved fireworks-ant colony hybrid algorithm2019342335343GuoH.FuW.-x.FuB.ChenK.YanJ.Suboptimal trajectory programming for unmanned aerial vehicles with dynamic obstacle avoidance2019233103857386910.1177/09544100188032792-s2.0-85061191610DiestelR.Graph theory202017350267128HuangL.WangL.HollotC. V.On robust stability of polynomials and related topics1992514254CaoY.RenW.Multi-vehicle coordination for double-integrator dynamics under fixed undirected/directed interaction in a sampled-data setting2010209987100010.1002/rnc.14952-s2.0-77952976389KovacsI.SilverD. S.WilliamsS. G.Determinants of commuting-block matrices19991061095095210.1080/00029890.1999.12005145