Professional Documents
Culture Documents
Average Consensus
THE PROBLEM, ITS APPLICATIONS, AND THE ALGORITHMS
T
echnological advances in ad hoc networking and the system to (in a decentralized fashion) fuse information, com-
availability of low-cost reliable computing, data stor- pute common estimates of unknown quantities, and agree on
age, and sensing devices have made scenarios pos- a common view of the world is critical. These problems can be
sible where the coordination of many subsystems formulated as agreement problems on linear combinations of
extends the range of human capabilities. Smart grid dynamically changing reference signals or local parameters.
operations, smart transportation, smart health care, and sens- This dynamic agreement problem corresponds to dynamic
ing networks for environmental monitoring and exploration average consensus, which, as discussed in “Summary,” is the
in hazardous situations are just a few examples of such net- problem of interest of this article. The dynamic average con-
work operations. In these applications, the ability of a network sensus problem is for a group of agents to cooperate to track
the average of locally available time-varying reference signals,
Digital Object Identifier 10.1109/MCS.2019.2900783 where each agent is capable only of local computations and
Date of publication: 17 May 2019 communicating with local neighbors (see Figure 1).
4 4
2 2
xi
xi
0 0
0 2 4 6 8 10 12 14 16 18 20 0 2 4 6 8 10 12 14 16 18 20
Time Time
(a) (b)
4
4
2
2
xi
xi
0
0
–2
0 2 4 6 8 10 12 14 16 18 20 0 2 4 6 8 10 12 14 16 18 20
Time Time
(c) (d)
FIGURE 2 A comparison of performance between a static average consensus algorithm reinitialized at each sampling time versus a
dynamic average consensus algorithm. The solid lines: red curves (respectively, blue curves) represent the time history of the agree-
ment state of each agent generated by the Laplacian static average consensus approach [respectively, the dynamic average consensus
of (S15)]; # : sampling points at mDt; & : the average at mDt; + : the average of reference signals at kd. The dynamic consensus algo-
rithm very closely tracks the average over time as the static consensus does not have enough time between sampling times to converge.
This trend is preserved even if the frequency of the communication between the agents increases. In these simulations, a = b = 1 in
(S15). (a) Static algorithm; three communications in t ! [m, m + 1] . (b) Static algorithm; 20 communications in t ! [m, m + 1] . (c) Dynamic
algorithm; three communications in t ! [m, m + 1] . (d) Dynamic algorithm; 20 communications in t ! [m, m + 1] .
Further Reading
N umerous works have studied the robustness of dynamic dom link failures [S16] and optimizing the edge weights for fast
average consensus algorithms against a variety of distur- consensus [S17], [7].
bances and sources of error present in practical scenarios.
These include fixed communication delays [S1], additive input
REFERENCES
disturbances [S2], time-varying communication graphs [S3], [S1] H. Moradian and S. S. Kia, “On robustness analysis of a dynamic
and driving command saturation [19]. Variations of the dy- average consensus algorithm to communication delay,” IEEE Trans.
namic average consensus problems explore scenarios where Control Netw. Syst. Aug. 6, 2018. doi: 10.1109/TCNS.2018.2863568.
[S2] G. Shi and K. H. Johansson, “Robust consensus for continuous-time multi-
the algorithm design depends on the specific agent dynamics agent dynamics,” SIAM J. Control Optim., vol. 51, no. 5, pp. 3673–3691, 2013.
[S4], [S5], [71] or incorporates different agent roles, such as in [S3] B. Van Scoy, R. A. Freeman, and K. M. Lynch, “Asymptotic mean
leader–follower networks of mobile agents [15], [S6], [S7]. ergodicity of average consensus estimators,” in Proc. American Con-
trol Conf., 2014, pp. 4696–4701.
When dealing with directed agent interactions, a common [S4] F. Chen, G. Feng, L. Liu, and W. Ren, “Distributed average tracking
assumption in solving the average consensus problem is that of networked Euler–Lagrange systems,” IEEE Trans. Autom. Control,
the communication graph is weight balanced, which is equiv- vol. 60, no. 2, pp. 547–552, 2015.
[S5] S. Ghapania, W. Ren, F. Chen, and Y. Song, “Distributed aver-
alent to the graph consensus matrix W: = I - L being doubly age tracking for double-integrator multi-agent systems with reduced
stochastic. In [S8], it is shown that calculating an average over requirement on velocity measurements,” Automatica, vol. 81, no. 7,
a network requires either explicit or implicit use of either 1) the pp. 1–7, 2017.
[S6] G. Shi, Y. Hong, and K. H. Johansson, “Connectivity and set track-
out-degree of each agent, 2) global node identifiers, 3) random- ing of multi-agent systems guided by multiple moving leaders,” IEEE
ization, or 4) asynchronous updates with specific properties. In Trans. Autom. Control, vol. 57, no. 3, pp. 663–676, 2012.
particular, the balanced assumption is necessary for scalable, [S7] Z. Meng, D. V. Dimarogonas, and K. H. Johansson, “Leader-follow-
er coordinated tracking of multiple heterogeneous Lagrange systems
deterministic, synchronous algorithms. In general, agents may using continuous control,” IEEE Trans. Robot., vol. 30, no. 3, pp. 739–
not have access to their out-degree (for example, agents that 745, 2014.
use local broadcast communication). If each agent knows its [S8] J. M. Hendrickx and J. N. Tsitsiklis, “Fundamental limitations for
anonymous distributed systems with broadcast communications,” in Proc.
out-degree, however, then distributed algorithms may be used Allerton Conf. Communication, Control, and Computing, 2015, pp. 9–16.
to generate weight-balanced and doubly stochastic digraphs [S9] B. Gharesifard and J. Cortés, “Distributed strategies for generating
[S9], [S10]. weight-balanced and doubly stochastic digraphs,” Eur. J. Control, vol.
18, no. 6, pp. 539–557, 2012.
Another approach is to explicitly use the out-degree in the [S10] A. Rikos, T. Charalambous, and C. N. Hadjicostis, “Distributed
algorithm by having agents share their out-weights and use weight balancing over digraphs,” IEEE Trans. Control Netw. Syst., vol.
them to adjust for the imbalances in the graph. This approach 1, no. 2, pp. 190–201, 2014.
[S11] F. Bénézit, V. Blondel, P. Thiran, J. Tsitsiklis, and M. Vetterli, “Weight-
is referred to as the push-sum protocol and has been applied to ed gossip: Distributed averaging using non-doubly stochastic matrices,”
the static average consensus problem (see [S11]–[S14]). Both in Proc. IEEE Int. Symp. Information Theory, 2010, pp. 1753–1757.
of these approaches of dealing with unbalanced graphs require [S12] A. D. Domínguez-García and C. N. Hadjicostis, “Distributed ma-
trix scaling and application to average consensus in directed graphs,”
each agent to know its out-degree. Furthermore, when com- IEEE Trans. Autom. Control, vol. 58, no. 3, pp. 667–681, 2013.
munication links are time varying, these approaches work only [S13] A. Nedic and A. Olshevsky, “Distributed optimization over time-
if the time varying graph remains weight balanced (see [19] and varying directed graphs,” IEEE Trans. Autom. Control, vol. 60, no. 3, pp.
601–615, 2015.
[S15]). If communication failures caused by limited communi- [S14] P. Rezaienia, B. Gharesifard, T. Linder, and B. Touri. Push-sum on
cation ranges or external events, such as obstacle blocking, random graphs. 2017. [Online]. Available: arXiv:1708.00915
destroy the weight-balanced character of the graph, then it is [S15] S. S. Kia, J. Cortés, and S. Martínez, “Distributed event-triggered
communication for dynamic average consensus in networked sys-
still possible to solve the dynamic average consensus problem tems,” Automatica, vol. 59, pp. 112–119, Sept. 2015.
if the expected graph is balanced [S3]. Another set of works has [S16] S. Kar and J. M. F. Moura, “Sensor networks with random links:
explored the question of how to optimize the graph topology Topology design for distributed consensus,” in IEEE Trans. Signal Pro-
cess., vol. 56, no. 7, pp. 3315–3326, 2008.
to endow consensus algorithms with better properties. These [S17] L. Xiao and S. Boyd, “Fast linear iterations for distributed averag-
include designing the network topology in the presence of ran- ing,” in Proc. IEEE Conf. Decision and Control, 2003, pp. 4997–5002.
j
discrete time: x i (t k + 1) = c i ( J i (t k), {I j (t k)} j ! Nout
i
), i ! {1, f, N}, such that x i (t) " u avg (t) as t " 3. Here, t k j ! R is the k jth
(2) transmission time of agent j, which is not necessarily
.
x = Ax + Bu, x ! R n, u ! R m, (S3) A tighter exponential bound of
l B x k + 1 = Ax k + Bu k, x k ! R n, u k ! R m (S11)
# le - m t x (0) + sup u (x) , 6t ! R $ 0.(S6)
m 0#x#t
The bound shows that the zero-input response decays to zero with initial condition x 0 ! R n satisfies the bound
exponentially fast, whereas the zero-state response is bound-
e tk x0 + B sup u j o, (S12)
v max (P) 1 - tk
ed for every bounded input, indicating an input-to-state stabil- xk #
v min (P) 1-t 0#j1k
ity behavior. Note that the ultimate bound on the system state
is proportional to the bound on the input. where P ! R n # n and t ! R satisfy
Next, how to compute the parameters l, m ! R 2 0 is shown
in (S5). Recall that [S19, Fact 11.15.5] for any matrix A ! R n # n. A < PA - t 2 P # 0, P 2 0, t $ 0. (S13)
Therefore,
REFERENCES
e A t # e mmax (Sym (A)) t . 6t ! R $ 0, (S7) [S19] D. S. Bernstein, Matrix Mathematics: Theory, Facts, and For-
mulas with Application to Linear System Theory. New York: Springer-
where Sym (A) = ^1/2h (A + A <). Therefore, for a Hurwitz matrix Verlag, 2005.
[S20] D. Hinrichsen and A. J. Pritchard, Mathematical Systems Theory
A whose Sym (A) is also Hurwitz, the exponential bound pa- I: Modeling, State Space Analysis, Stability and Robustness. Princ-
rameters can be set to eton, NJ: Princeton Univ. Press, 2005.
Cyber Layer
N
1
Cyber Layer Computes ∑ xTi (t )
Ni=1
Physical Layer
FIGURE 3 A two-layer consensus-based formation for tracking a team of mobile targets. The larger triangle robots are the mobile agents,
and the smaller round robots are the mobile moving targets. The physical layer shows the situational distribution of the mobile agents
and the moving targets. The cyber layer shows which mobile agent has a computational capability and the interagent communi-
cation topology.
n
atio
m unic
Com ange
R 1 1
Smart Camera (y , Y )
Node 1
Sens
ing Z
one 2 2
Smart Camera (y , Y )
Node 2
Fusion Center
Neighbors
FIGURE 4 A networked smart camera system that monitors and estimates the position of moving targets.
x4
x5
x3
x1 + x2 + x3 + x4 + x5 = d
x2
x1
FIGURE 5 A network of five generators with connected undirected topology works together to meet a demand of x 1 + x 2 + x 3 + x 4 + x 5 = d
in a manner in which the overall cost R 5i = 1 f i (x i) for the group is minimized.
If the local cost functions are strictly convex, then every tra- for i ! {1, f, N}, with x i (0) = u i constant for both algo-
jectory t7x (t) converges to the optimal solution x *. The rithms. Here, [a ij] N # N is the adjacency matrix of the com-
source of coupling in (7) is the demand mismatch that munication graph (see “Basic Notions from Graph Theory”).
appears in the right-hand side of (7a). However, the dynamic By stacking the agent variables into vectors, the static aver-
average consensus can be employed to estimate this age consensus algorithms can be written compactly using a
quantity online and feed it back into the algorithm. This graph Laplacian as
approach is taken in [46] and [47]. This can be accomplished,
.
for instance, by having agent i use the reference signal continuous time: x (t) = - Lx (t), (9a)
x i (t) - d/N (this assumes that every agent knows the
demand and number of agents in the network, but other discrete time: x (k + 1) = (I - L) x (k), (9b)
A First Design for Dynamic Average Consensus where t 0 is the initial time. Using the ISS bound on the
Because the reference signals enter the static average con- trajectories of LTI systems (see “Input-to-State Stability of
sensus algorithms (8) as initial conditions, they cannot Linear Time-Invariant Systems”), the tracking error of
track time-varying signals. Looking at the frequency-
domain representation in Figure 6 of the static average
consensus algorithms (8), it is clear that what is needed
. 1
instead is to continuously inject the signals as inputs u(t ) x(t )
s IN
into the dynamical system. This allows the system to nat- −
urally respond to changes in the signals without any need x(0)
for reinitialization. This basic observation is made in [49],
resulting in the systems shown in Figure 7.
L
More precisely, [49] argues that considering the static
inputs as a dynamic step function, the algorithm (a)
. .
x (t) =-Lx (t) + u (t), x i (0) = u i (0), u(t ) x(t )
u i (t) = u i h (t), i ! {1, f, N}, −
N 2
/ (x j(t 0) - u j(t 0))
sup t0 # x # t < Puo (x) < o f j = 1 p,
. 2
as (17b) shows for the case of the alternative algorithm (16), is a jointly strongly connected digraph, goes to infin-
R Ni = 1 p i (t) = 0 is preserved in time as long as the algorithm ity as t " 3.
is initialized such that R Ni = 1 p i (t 0) = 0, which can be easily When the switching signal belongs to the admissible
done by setting p i (t 0) = 0 for i ! {1, f, N } . Consequently, set S admis, [19] shows that there always exists m ! R 2 0 and
-R< L R
when the perturbations are removed, then (16) recovers the l ! R $ 1 such that < e v
< # l e - mt, t ! R $ 0 . Implementing
FIGURE 8 A simple dynamic average consensus-based containment and tracking of a team of mobile targets. (a) The triangle robots
cooperatively want to contain the moving round robots by making a formation around the geometric center of the round robots that they
are observing. At (b) (after, for example, 10 s from the start of the operation), one of the triangle robots leaves the team. At (c) (after, for
example, 20 s from the start of the operation), a new triangle robot joins the group to take over tracking the abandoned round robot.
i + i 2
xT (t ) Dynamic x i (t ) A xT (t ) – x i (t )B Dynamic
(.)2 Consensus
Consensus –
N
1 N
∑i = 1 xT (t )
i 1
∑i = 1 A xT (t ) – x i (t )B
i 2
x i (t ) → y i (t ) →
N N
N N
1
∑i = 1 (xT (t ) – N1 ∑i = 1 xT (t ))2
i i
→
N
FIGURE 9 A group of N mobile agents uses a set of dynamic consensus algorithms to asymptotically track the geometric center
x T (t) = (1/N) / Ni = 1 x iT (t) and its variance (1/N) / Ni = 1 (x i (t) - x T (t)) 2. In this setup, each mobile agent i ! {1, f, N} is monitoring its
avg avg
6
4
4
2
2
xi
xi
0
0
–2 uavg(t) x1 x2 x3 x4 x5 uavg(t) x1 x2 x3 x4
–2
0 5 10 15 20 25 30 0 2 4 6 8 10
t t
(a) (b)
3 6
1 2 1 2 1 2 1 2 1 2
2
4 3 4 3 4 3 4 3 4 3 4
[0,1) [2,3) [3,4) [4,5) [5,10)
1
2
xi
xi
0
0
–1
uavg(t) x1 x2 x3 x4 uavg(t) x1 x2 x3 x4
–2 –2
0 2 4 6 8 10 0 2 4 6 8 10
t t
(c) (d)
FIGURE 10 A performance evaluation of dynamic average consensus algorithms (11) and (16) for a group of four agents with reference
inputs (18) for a tracking scenario described in Figure 8. In the simulation in plot (a), the agents are using (16). As guaranteed in Lemma 1
avg
[under proper initialization p i (0) = 0, i ! {1, 2, 3, 4}] during the time interval [0,10] s, the agents are able to track x T (t) with a small error.
3
The challenge presents itself when agent 4 leaves the operation at t = 10 s . Because after agent 4 leaves R i = 1 p i (10 +) ! 0, the remain-
ing agents fail to follow the average of their reference values, which now is (1/3) R 3l = 1 u l. Similarly, even with initialization of p 5 (20) = 0
for the new agent 5, because p 1 (20) + p 2 (20) + p 3 (20) + p 5 (20) ! 0, the agents track the average (1/4) R 4l = 1 x lT (t) with a steady-state
error. In the simulation in plot (b), the agents are using (16), and at time interval [0, 10] s, agent 1’s reference input is subject to a mea-
surement perturbation according to u 1 (t) = x 1T (t) + w 1 (t), where w 1 (t) = - 4 cos (t) at t ! [0, 2], and t ! [3, 5] and w 1 (t) = 0 at other times.
As guaranteed by Lemma 1, despite the perturbation, including the initial measurement error of u 1 (0) = x 1T (0) - 4, (16) has robustness
to the measurement perturbation and recovers its performance after the perturbation is removed. A large perturbation error was used,
so that its effect is observed more visibly in the simulation plots. In the simulation in plot (c), the agents are using (11). Agent 1’s reference
input has an initial measurement error of u 1 (0) = x 1T (0) - 4. Because the measurement error directly affects the initialization condition of
the algorithm, it fails to preserve R 4l = 1 x l (t) = R 4l = 1 u l (t). As a result, the effect of initialization error persists, and the algorithm maintains a
significant tracking error. In the simulation in plot (d), the agents are using (16) [similar results are also obtained for (11)]. The network
communication topology is a switching graph, where the graph topology at different time intervals is shown on the plot. Because the
switching signal v belongs to S admis, as predicted by Lemma 2, the trajectories of the algorithm stay bounded, and once the topology
becomes fully connected, the agents follow their respective dynamic average closely. (a) Agent departure and arrival. (b) Perturbation
of input signals. (c) Initialization error. (d) Switching topology.
βL
e / x j (t 0) - / u j (t 0) o,
N N N N
/ x j(t) = / u j (t) + e - a (t - t 0)
q(t )
j=1 j=1 j=1 j=1
αβ
L
s
for t ! 6t 0, 3) . q(0)
Figure 12 shows the performance of (19) in the distributed
formation control scenario represented in Figure 8. This plot (b)
illustrates how the property of robustness to the initializa- FIGURE 11 A block diagram of continuous-time dynamic average con-
tion error of (19) allows it to accommodate the addition and sensus algorithms. These dynamic algorithms naturally adapt to
deletion of agents with satisfactory tracking performance. changes in the reference signals, which are applied as inputs to the
Although the convergence guarantees of (19) are valid for system. Continuous-time algorithm (19) is robust to initialization. To see
why the algorithm is robust, consider multiplying the input signal on the
strongly connected and weight-balanced digraphs, from
left in plot (a) by 1 <N . The output of the integrator block (1/s) is multiplied
an implementation perspective, the use of this strategy over by zero (because L I 1 N = 0) and therefore does not affect the output.
directed graphs may not be feasible. In fact, the presence of Although the output is affected by the initial state of the 1/ (s + a) block,
the transposed integral Laplacian L <I in (20b) requires each this term decays to zero and therefore does not affect the steady state.
agent i ! {1, f, N } to know not only the entries in row i but Also, the requirement of needing the derivative of the input uo (t) can be
removed by a change of variable. The continuous-time algorithm in
also the column i of L I and receive information from the cor-
(25) is not robust to initialization. In this algorithm, the parameter b
responding agents. However, for undirected graph topolo- may be used to control the tracking error size, and a may be used to
gies, this requirement is satisfied trivially as LI< = L I . control the rate of convergence. Furthermore, this algorithm is robust
to reference signal measurement perturbations and naturally pre-
Controlling the Rate of Convergence serves the privacy of the input signals against adversaries [19]. (a)
Continuous-time algorithm (19). (b) Continuous-time algorithm (25).
A common feature of the dynamic average consensus
algorithms presented in the “A First Design for Dynamic
Average Consensus” and “Robustness to Initialization and 4
Permanent Agent Dropout” sections is that the rate of conver-
gence is the same for all agents and dictated by network topol- 2
ogy as well as some algorithm parameters [see (14) and (23)].
xi
a nd e z = T T ^z + ^ b N h R j = 1 u j 1 N + ^1 N h R j = 1 uo j 1 N h are ap
N N x i (t 0), q i (t 0) ! R s.t. / q j (t 0) = 0, (25c)
j=1
plied to write the FOI-DC algorithm as
i ! {1, f, N}, where a, b ! R 2 0 . Equation (25) in compact
wo 1 = 0, form can be equivalently written as
0 60 –R T L I R@
= . G = e -1 > H; E
.
y y
x = - a (x - u) - bLx - ab # Lx (x) dx - q (t 0) + u,
. t .
; T T E ; E ez
0 –1 0
ez T
R L I R 0 –I –R L p R
t0
144444442 44444443
r A which demonstrates the proportional and integral agree-
– (R T L TI R) –1 60 (N–1) # 1 I N–1@
+> Hf (t),
ment feedback structure of this algorithm. As was done
61 0 1 # N–1@
= G for (11), a change of variables p i = u i - x i can be used to
0 (N–1) # N
14444444244444443 write this algorithm in a form whose implementation
r
B
. does not require the knowledge of the derivative of the
e = –be – e z,
reference signals.
Note an interesting connection between (25) and (16).
. ..
where f (t) = T T (b u + u). Using the ISS bound on the trajec- Writing the transfer function from the reference input to the
tories of LTI systems (see “Input-to-State Stability of Linear tracking error state (25), there is a pole-zero cancellation that
Time-Invariant Systems”), the tracking error of each agent reduces (25) to (11) and (16). Despite this close relationship,
i ! {1, f, N} while implementing the FOI-DC algorithm there are some subtle differences. For example, unlike (11),
with e ! R 2 0 is, (25) enjoys robustness to reference signal measurement per-
turbations and naturally preserves the privacy of the input
e i (t) # e - b (t - t0) e i (t 0) of each agent against adversaries. Specifically, an adversary
+ l sup e e - e m (t - t 0) ; E
-1 y (t 0) with access to the time history of all network communica-
b t0 # x # t e z (t 0) tion messages cannot uniquely reconstruct the reference
r signal of any agent [19], which is not the case for (16).
+ e B sup b u (x) + u (x) m ,
. ..
m t0 # x # t Figure 11(b) shows the block diagram representation of
this algorithm. The next result states the convergence prop-
r
where e A t # le - m t. From this error bound, it is observed erties of (25). See [19] for the proof of this statement, which
that, for dynamic signals with bounded first and second is established using the time-domain analysis implemented
derivatives, the FOI-DC algorithm is guaranteed to track to analyze the algorithms reviewed so far.
the dynamic average with an ultimately bounded error.
This tracking error can be made small using a small Theorem 4: Convergence of (25) Over Strongly Connected and
e ! R 2 0 . Use of small e ! R 2 0 also results in dynamics Weight-Balanced Digraphs for Dynamic Input Signals [19]
(25a) to have a higher decay rate. Therefore, the domi- Let G be a strongly connected and weight-balanced digraph.
nant rate of convergence of the FOI-DC algorithm is Let sup x ! [t, 3) < ^I N - (1/N ) 1 N 1 <N h uo (x) < = c (t) 1 3. For any
determined by b, which can be prespecified regardless a, b ! R 2 0, the trajectories of (25) satisfy
provided R Nj = 1 q j (t 0) = 0. The convergence rate to the error Nonrobust Dynamic Average Consensus Algorithms
bound is min {a, b Re (m 2)} . First consider the discretized version of the continuous-time
The inverse relation between b and the tracking error in dynamic average consensus algorithm in (16) (“Euler Dis-
(26) indicates that the parameter b can be used to control cretizations of Continuous-Time Dynamic Average Consen-
the tracking error size, and a can be used to control the rate sus Algorithms” elaborates on the method for discretization
of convergence. and the associated range of admissible step sizes). This algo-
rithm has the iterations
DISCRETE-TIME DYNAMIC N
AVERAGE CONSENSUS ALGORITHMS p ik + 1 = p ik + k I / a ij (x ik - x kj ), p i0 ! R,
i ! {1, f, N}, (27a)
Although the continuous-time dynamic average consensus j=1
T he continuous-time algorithms described in the article can where x ! R n and u ! R m are, respectively, state and input
also give rise to discrete-time strategies. Here, we describe vectors, and d ! R 2 0 is the discretization step size. Let the
how to discretize them so that they are implementable over system matrix A = [a ij] ! R n # n be a Hurwitz matrix with eigen-
wireless communication channels. This can be done by using values " n i ,in= 1, and the difference of the input signal be bound-
the (forward) Euler discretization of the derivatives ed, Du 1 t 1 3. For any d ! (0, dr ) where
x (k + 1) - x (k) n
dr = min ) - 2 2 3
. Re (n i)
x (t) . , , (S17)
d ni i=1
where d ! R 2 0 is the step size. To illustrate the discussion, we the eigenvalues of (I + dA) are all located inside the until circle
develop this approach for (25) over a connected graph topol- in the complex plane. Moreover, starting from any x (0) ! R n,
ogy. The following discussion can also be extended to include the trajectories of (S16) satisfy
iterative forms of the other continuous-time algorithms studied
lt B
in the article. Using the Euler discretization in (25) leads to lim x (k + 1) # , (S18)
k"3 1-~
N
v i (k + 1) = v i (k) + dab / a ij (x i (k) - x j (k)), (S14a)
k
where ~ ! (0, 1), and l ! R 2 0 such that I + dA # l~ k .
j=1 k
The bounds ~ ! (0, 1) and l ! R 2 0 in I + dA # l~ k when
x i (k + 1) = x i (k) + Du i (k) - da (x i (k) - u i (k)) all the eigenvalues of I + dA are located in the unit circle of
N
- db / a ij (x i (k) - x j (k)) - dv i (k), (S14b) the complex plane can be obtained from the following linear
j=1
matrix inequality optimization problem (see [S21, Theorem
where Du i (k) = u i (k + 1) - u i (k). To implement this iterative form 23.3] for details):
at each time step k, access to the future value of the reference
(~, l, Q) = argmin ~ 2, subject to (S19)
input at time step k + 1 is needed. Such a requirement is not
1 I # Q # I, 0 1 ~ 2 1 1, l 2 1,
practical when the reference input is sampled from a physical l
process or is a result of another online algorithm. This require- (I + dA) < Q (I + dA) - Q # - (1 - ~ 2) I.
ment can be circumvented using a backward Euler discretiza-
Building on Lemma S1, the next result characterizes the
tion, but the resulting algorithm tracks the reference dynamic
admissible discretization step size for (S15) and its ultimate
average with one-step delay. A practical solution that avoids
tracking behavior.
requiring the future values of the reference input is obtained
by introducing an intermediate variable z i (k) = x i (k) - u i (k) and
Theorem S2: Convergence of (S15) Over
representing the iterative algorithm (S14) in the form
Connected Graphs [19]
N
v i (k + 1) = v i (k) + dab / a ij (x i (k) - x j (k)), (S15a) Let G be a connected, undirected graph. Assume that the
j=1 differences of the inputs of the network satisfy max k ! Z $ 0 < (I -
N
z (k + 1) = z (k) - daz (k) - db / a ij (x i(k) - x j(k)) - dv i(k), (S15b)
i i i (1/N ) 1 N 1 <N) Du (k) < = c 1 3. Then, for any a, b 2 0, (S15) over
j=1
G initialized at z i (0) ! R and v i (0) ! R such that R Ni = 1 v j (0) = 0
x i (k) = z i (k) + u i (k), (S15c)
has bounded trajectories that satisfy
for i ! " 1, f, N , . Equation (S15) is then implementable without
, i ! " 1, f, N , (S20)
dlc
lim x i (k) - u avg (k) #
the use of future inputs. k"3 1-~
The question then is to characterize the adequate step siz-
provided d ! ^0, min " a -1, 2b -1 (m N) -1 ,h . Here, m N is the largest
es that guarantee that the convergence properties of the con-
eigenvalue of the Laplacian, and ~ ! (0, 1) and l ! R 2 0 satisfy
tinuous-time algorithm are retained by its discrete implemen- k
I - d bR < LR # l~ k, k ! Z $ 0.
tation. Intuitively, the smaller the step size, the better for this
Note that the characterization of the step size requires knowl-
purpose. However, this also requires more communication. To
edge of the largest eigenvalue m N of the Laplacian. Because such
ascertain this issue, the following result is particularly useful.
knowledge is not readily available to the network unless dedicated
distributed algorithms are introduced to compute it, [19] provides
Lemma S1 : Admissible Step Size for the Euler Discretized
the sufficient characterization d ! ^0, min " a -1, b -1 (d max out )
-1
,h
Form of Linear Time-Invariant Systems and a Bound on
along with the ultimate tracking bound
Their Trajectories
, i ! " 1, f, N , .
Consider dc
lim x i (k) - u avg (k) #
k"3 bm 2
.
x = Ax + Bu, t ! R $ 0,
− −
kp
kI I L
I L z−ρ
z−1
kI
I L
p0 z−1
(a) (b)
uk xk uk xk
− −
kp z
kI z I L
I L (z − ρ)2
(z − ρ2)(z − 1) kI z
p0 I L
(z − ρ2)(z − 1)
(c) (d)
FIGURE 13 A block diagram of discrete-time dynamic average consensus algorithms. The algorithms in (b) and (d) use proportional-integral
(PI) dynamics to obtain robustness to initial conditions, whereas those in (c) and (d) use extra dynamics to accelerate the convergence rate.
When the graph is connected and balanced and upper and lower bounds on the nonzero eigenvalues of the graph Laplacian are known,
closed-form solutions for the parameters that optimize the convergence rate are known (see Theorem 5). (a) The nonrobust, nonaccelerated
dynamic average consensus algorithm (27). (b) The robust, nonaccelerated, PI dynamic average consensus algorithm (30). (c) The nonro-
bust, accelerated, dynamic average consensus algorithm (29). (d) The robust, accelerated, PI dynamic average consensus algorithm (31).
kI = 2 and t = m N - m 2 .
m2 + mN mN + m2
(a) (b)
While the previous choice of parameters optimizes the
convergence rate, even faster convergence can be achieved
FIGURE 14 The root locus design of dynamic average consensus
by introducing extra dynamics into the dynamic average
algorithms. The dynamic average consensus algorithm poles are
consensus algorithm. Consider the accelerated dynamic the points on the root locus at gains m i for i ! {1, f, N}, where m i
average consensus algorithm in Figure 13(c), given by are the eigenvalues of the graph Laplacian. To optimize the con-
vergence rate, the parameters are chosen to minimize t such that
N all poles corresponding to eigenvalues m i for i ! {2, f, N} are
p ik + 1 = (1 + t 2) p ik - t 2 p ik - 1 + k I / a ij (x ik - x kj), inside the circle centered at the origin of radius t. The dynamic
j=1 average consensus algorithm then converges linearly with rate t.
p 0i ! R, i ! {1, f, N}, (29a) (a) The accelerated, dynamic average consensus algorithm in
Figure 13(a). (b) The nonaccelerated, dynamic average consen-
x ik = u ik - p ik .(29b) sus algorithm in Figure 13(c).
Instead of a simple integrator, the transfer function in the consensus algorithm (29). By adding an open-loop pole at
feedback loop now has two poles (one of which is still at z = t 2 and zero at z = 0, the root locus now goes around the
z = 1) . To implement the dynamic average consensus algo- t circle. Similar to the previous case, the convergence rate is
rithm, each agent must track two internal state variables optimized when there is a repeated pole at z = t when
(p ik and p ik - 1) . This small increase in memory, however, can m = m 2 and a repeated pole at z = - t when m = m N . This
result in a significant improvement in the rate of conver- gives the optimal parameter k I and convergence rate t
gence, as discussed next. given by
Once again, root locus techniques can be used to design
the parameters to optimize the convergence rate. Figure 14(b) kI = 4 and t =
mN - m2
.
^ m2 + mN h
2
shows the root locus of the accelerated dynamic average mN + m2
0.8
design involves a quadratic root locus. Although this
complicates the design process, closed-form solutions
0.6 for the algorithm parameters can still be found [57],
0.4 even for the accelerated version using extra dynamics,
given by
0.2
N
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 q ik + 1 = 2tq ik - t 2 q ik - 1 + k p / a ij ((x ik - x kj) + (p ik - p kj)), (31a)
j=1
λ 2/λN
N
The convergence rates of both the standard (27) and Theorem 5: Optimal Convergence Rates of
accelerated (29) versions of the dynamic average consen- Discrete-Time Dynamic Average Consensus Algorithms
sus algorithm are plotted in Figure 15 as a function of the Let G be a connected, undirected graph. Suppose the refer-
ratio m 2 /m N . ence signal u i at each agent i ! {1, f, N} is a constant scalar.
Consider the dynamic average consensus algorithms in
Robust Dynamic Average Consensus Algorithms Figure 13, with the parameters chosen according to Table 3
Although the previous dynamic average consensus algo- [the algorithms in Figure 13(a) and (c) are initialized such
rithms are not robust to initial conditions, root locus that the average of the initial integrator states is zero, that
techniques can also be used to optimize the convergence is, R Ni = 1 p i0 = 0@ . The agreement states x ik, i ! {1, f, N} con-
rate of dynamic average consensus algorithms that are verge to u avg exponentially with rate t.
robust to initial conditions. Consider the discrete-time
version of the PI estimator from (19), whose iterations are PERFECT TRACKING USING A PRIORI KNOWLEDGE
given by OF THE INPUT SIGNALS
The design of the dynamic average consensus algorithms
N described in the discussion so far does not require prior
q i
k+1
i
= tq + k p
k / a ij ((x i
k
j
k
i
k
j
- x ) + (p - p )), (30a)
k
knowledge of the reference signals and is therefore
j=1
N broadly applicable. This also comes at a cost. The conver-
p ik + 1 = p ik + k I / a ij (x ik - x kj), (30b) gence guarantees of these algorithms are strong only
j=1
when the reference signals are constant or slowly varying.
x ik = u ik - q ik, p i0, q i0 ! R, i ! {1, f, N}, (30c)
The error of such algorithms can be large, however, when
the reference signals change quickly in time. This section
with parameters t, k p, k I ! R . The block diagram of this describes dynamic average consensus algorithms, which
algorithm is shown in Figure 13(b). are capable of tracking fast time-varying signals with
Because the dynamic average consensus algorithms either zero or small steady-state error. In each case, their
(27) and (29) have only one Laplacian block in the block design assumes some specific information about the
diagram, the resulting root loci are linear in the Lapla- nature of the reference signals. In particular, consider refer-
cian eigenvalues. For the PI dynamic average consensus ence signals that 1) have a known model, 2) are band lim-
algorithm, however, the block diagram contains two ited, or 3) have bounded derivatives.
t kI kp
Figure 13(a) mN - m2 2 N/A
mN + m2 m2 + mN
Z 1-t
Figure 13(b)
2
] 8 - 8m r + m r , 1 t (1 - t) m r
] 8 - m r2 0 1 mr # 3 - 5 m2 mN t + mr - 1
[ 2
] (1 - m r) (4 + m r (5 - m2 r)) - m r (1 - m r) , 3 - 5 1 m r # 1
] 2 (1 + m r )
\
Figure 13(c) mN - m2 4 N/A
mN + m2 ( m2 + mN ) 2
Z (1 - t) 2 (2 + 2 1 - m r - m r) k I
Figure 13(d) ] 6 - 2 1 - m r + m r - 4 2 - 2 1 - m r + m r) , 0 1 m # 2 ( 2 - 1)
] 2 + 2 1 - mr - mr
r m2
[
] - 3 - 2 1 - m r + m r + 2 2 + 2 1 - m r - m r , 2 ( 2 - 1) 1 m r # 1
] - 1 - 2 1 - mr + mr
\
Signals With a Known Model (Discrete Time) where the mth divided difference is defined recursively
The discrete-time dynamic average consensus algorithms as D (m) u ik = D (m - 1) u ik - D (m - 1) u ik - 1 for m $ 2 with D (1) u ik =
discussed previously are designed with the idea of tracking u ik - u ik - 1 . The estimate of the average, however, is delayed
constant reference signals with zero steady-state error. To do by m iterations due to the transfer function having a factor
this, the algorithms contain an integrator in the feedback of z -m between the input and output. This problem is fixed
loop. This concept generalizes to time-varying signals by the dynamic average consensus algorithm in Figure 16(b),
with a known model using the internal model principle. given by
Consider reference signals whose z-transform has the form
N
u i (z) = n i (z) /d (z), where n i (z) and d (z) are polynomials in p i1, k + 1 = p i1, k + / a ij ((u ik - u kj ) - (p i1,k - p 1j ,k )), (33a)
z for i ! {1, f, N} . Dynamic average consensus algorithms j=1
d (z ) = )
(z - 1) m, polynomial of degree m - 1 p im, k + 1 = p im, k + / a ij ((u ik - u kj) - / (p ,i,k - p ,j,k)), (33c)
z 2 - 2z cos (~) + 1, sinusoid with frequency ~. m
j=1 ,=1
x =u -/p ,
i
k
i
k
i
,, k p i
,, 0 ! R, , ! {1, f, m}, i ! {1, f, N},
This section focuses on dynamic average consensus algo- ,=1
(33d)
rithms that track degree m - 1 polynomial reference sig-
nals with zero steady-state error. which tracks degree m - 1 polynomial reference signals
Consider the dynamic average consensus algorithms in with zero steady-state error without delay. Note, however,
Figure 16. The transfer function of each algorithm has m that the communication graph is assumed to be constant to
zeros at z = 1, so the algorithms track degree m - 1 polyno- use frequency-domain arguments; although the output of
mial references signals with zero steady-state error. The the dynamic average consensus algorithm in Figure 16(a) is
time-domain equations for the dynamic average consensus delayed, it also has nice tracking properties when the com-
algorithm in Figure 16(a) are munication graph is time varying, whereas the dynamic
average consensus algorithm in Figure 16(b) does not.
N To track degree m - 1 polynomial reference signals,
x i1, k + 1 = x i1, k - / a ij (x i1,k - x 1j ,k ) + D (m) u k, (32a) each dynamic average consensus algorithm in Figure 16
j=1
N cascades m dynamic average consensus algorithms, each
x i
2, k + 1 =x i
2, k - / a ij (x 2i ,k - x 2j ,k) + x i1,k, (32b) with a pole at z = 1 in the feedback loop. The dynamic
j=1
average consensus algorithm (27) is cascaded in Figure 16(b),
h
N
but any of the dynamic average consensus algorithms
x im, k + 1 = x im, k - / a ij (x im,k - x mj ,k) + x im - 1,k, (32c) from the previous section could also be used. For example,
j=1
the PI dynamic average consensus algorithm could be cas-
x ik = x im, k, x i,, 0 ! R, , ! {1, f, m}, i ! {1, f, N}, (32d) caded m times to track degree m - 1 polynomial reference
Band-Limited Signals (Discrete Time) so the error can be made small by making m large enough
To use algorithms designed using the model of the reference (so long as L is scaled such that < I - L - 11 T /N < 2 1 1) .
signals, the signals must be composed of a finite number of Specifically, the prefilter is designed such that h (z) is
known frequencies. When either the frequencies are unknown proper and h (z) . z m for z = exp (ji) for all i ! [0, i c] [note
or there are infinitely many frequencies, dynamic average that h (z) = z m cannot be used because it is not causal]. An
consensus algorithms can still be designed if the reference m-step filter can be obtained by cascading a one-step filter m
uk 1 ... 1 xk
∆(m) I I
z−1 N z−1 N
− −
L L
m Times
(a)
uk ... xk
− −
1 1
I L I L
z−1 N z−1 N
m Times
(b)
FIGURE 16 A block diagram of dynamic average consensus algorithms that track polynomial signals of degree m - 1 with zero steady-state
error when initialized correctly (neither algorithm is robust to initial conditions). The indicated section is repeated in series m times. (a) The
performance of (32) does not degrade when the graph is time varying, but the estimate is delayed by m iterations. Furthermore, the algo-
rithm is numerically unstable when m is large and eventually diverges from tracking the average when implemented using finite precision
arithmetic. (b) The estimate of the average by (33) is not delayed, and the algorithm is numerically stable, but the tracking performance
degrades when the communication graph is time varying. (a) Dynamic average consensus algorithm (32) in [76] where D (m) = (1 - z -1) m is
the mth divided difference (see also [77] for a step-size analysis). (b) Dynamic average consensus algorithm (33), which is the algorithm
in [53] cascaded in series m times.
reference signals.
po i = k p / sgn (x i - x j), / p j (0) = 0, (35a)
j ! N iout j=1
i i i
x = u - p , i ! {1, f, N} . (35b)
Signals With Bounded Derivatives (Continuous Time)
Stronger tracking results can be obtained using algorithms
implemented in continuous time. Here, a number of contin- T he cor respondi ng block diag ra m is show n i n
uous-time dynamic average consensus algorithms are pre- Figure 18(b).
sented that are capable of tracking time-varying reference It is simple to see from the block diagram of Figure 18(a)
signals whose derivatives are bounded with zero error in why (34) is not robust to initial conditions; the integrator
finite time. For simplicity, assume that the communication state is directly connected to the output and therefore
graph is constant, connected, and undirected. Also, the affects the steady-state output in the consensus direction.
reference signals are assumed to be differentiable with This issue is addressed by the dynamic average consensus
bounded derivatives. algorithm in Figure 18(c), given by
In discrete time, zero steady-state error is obtained by
placing the internal model of the reference signals in the po i = k p sgn c / ^x i - x jh m , p i (0) ! R, (36a)
feedback loop. This provides infinite loop gain at the fre- j ! N iout
quencies contained in the reference signals. In continuous xi = ui - / (p i - p j), i ! {1, f, N}, (36b)
time, however, the discontinuous signum function sgn can j ! N iout
L L
m Times
FIGURE 17 The feedforward dynamic average consensus algorithm for tracking the average of band-limited reference signals. The prefilter
h (z) is applied to the reference signals before passing through the graph Laplacian. For an appropriately designed prefilter, the dynamic
average consensus algorithm can track band-limited reference signals with arbitrarily small steady-state error when using exact arithmetic
(and small error for finite precision) [68].
. 1
u(t ) I x(t )
s N
−
x(0)
kp IN B sgn(·) B
(a)
u(t ) x(t )
−
kp
B sgn(·) B
p(t ) s IN
p(0)
(b)
u(t ) x(t )
−
kp
L sgn(·) L
p(t ) s IN
(c)
u(t ) x(t )
−
kp 1
L sgn(·) L I
p(t ) s IN s+α N
(d)
u(t ) x(t )
−
kp
IN B sgn(·) B
p(t ) s+1
(e)
FIGURE 18 A block diagram of discontinuous dynamic average consensus algorithms in continuous time. In each case, the communica-
tion graph is assumed to be constant, connected, and balanced with Laplacian matrix L = BB <. Furthermore, the reference signals are
assumed to have bounded derivatives. The dynamic average consensus algorithm (34), shown in (a), achieves perfect tracking in finite
time and uses one-hop communication, but it is not robust to initial conditions [that is, the steady-state error is zero only if 1 N< x (0) = 1 N< u (0)].
Furthermore, the derivative of the reference signals is required (see [69]). The dynamic average consensus algorithm (35) shown in (b)
is equivalent to the algorithm in (a), although this form does not require the derivative of the reference signals. In this case, the require-
ment on the initial conditions is 1 <N p (0) = 0. The dynamic average consensus algorithm (36) shown in (c) converges to zero error in finite
time and is robust to initial conditions but requires two-hop communication (in other words, two rounds of communication are performed
at each time instant) (see [70]). The dynamic average consensus algorithm (37) shown in (d) is robust to initial conditions and uses one-
hop communication but converges to zero error exponentially instead of in finite time (see [70]). The dynamic average consensus algo-
rithm (38) shown in (e) is robust to initial conditions and uses one-hop communication, although the error converges to zero exponentially
instead of in finite time (see [71]). (a) The dynamic average consensus algorithm (34). (b) The dynamic average consensus algorithm
(35). (c) The dynamic average consensus algorithm (36). (d) The dynamic average consensus algorithm (37). (e) The dynamic average
consensus algorithm (38).
W e discuss here an alternative to the discretization route infinite number of communication rounds are triggered in a fi-
explained in “Euler Discretizations of Continuous-Time nite amount of time), and 3) they have to ensure the network
Dynamic Average Consensus Algorithms” to produce imple- achieves dynamic average consensus, although agents operate
mentable strategies from the continuous-time algorithms with outdated information while inputs are changing with time.
described in the article. This approach is based on the ob- Consider the following event-triggered communication law
servation that, when implementing the algorithms over digital [S15]: each agent is to communicate with its in-neighbors at
platforms, computation can still be reasonably approximated times {t ik} k ! N 1 R $ 0, starting at t i1 = 0, determined by
by continuous-time evolution (given the ever-growing capabili-
ties of modern embedded processors and computers), whereas t ik +1 = argmax $ t ! [t ik, 3) x i (t ik) - x i (t) # e i .. (S22)
communication is a process that still requires proper acknowl-
edgment of its discrete-time nature. The basic idea is to op- Here, e i ! R 2 0 is a constant value that each agent chooses
portunistically trigger, based on the network state, the times for locally to control its inter-event times and avoid Zeno behavior.
information sharing among agents to take place and allow indi- Specifically, the interexecution times of each agent i ! {1, f, N}
vidual agents to determine these autonomously. This has the employing (S22) are lower bounded by
potential to result in more efficient algorithm implementations
` j
because performing communication usually requires more en- xi =
1 ln 1 + ae i , (S23)
a ci
ergy than computation [S22]. In addition, the use of fixed com-
munication step sizes can lead to a wasteful use of the network where c i and h are positive real numbers that depend on the
resources because of the need to select it, taking into account initial conditions and network parameters (we omit for simplici-
worst-case scenarios. These observations are aligned with the ty their specific form, but see [S15] for the explicit expressions).
ongoing research activity [S23], [S24] on event-triggered con- The lower bound (S23) shows that, for a positive nonzero e i,
trol and aperiodic sampling for controlled dynamical systems the interexecution times are bounded away from zero, and it is
that seeks to trade computation and decision making for less guaranteed that, for networks with a finite number of agents,
communication, sensing, or actuator effort while still guaran- the implementation of (S21) with the communication trigger law
teeing a desired level of performance. The surveys [S25], [S26] (S22) is Zeno free. The following result formally describes the
describe how these ideas can be employed to design event- convergence behavior of (S21) under (S22) when the interac-
triggered communication laws for static average consensus. tion topology is modeled by a strongly connected and weight-
Motivated by these observations, [S15] investigates a dis- balanced digraph.
crete-time communication implementation of the continuous-
time algorithm (25) for dynamic average consensus. Under this Theorem S3: Convergence of (S21) Over Strongly
strategy, the algorithm becomes Connected and Weight-Balanced Digraph with Asynchronous
Distributed Event-Triggered Communication [S15]
N
vo i = ab / a ij (xt i - xt j), (S21a) Let G be a strongly connected and weight-balanced digraph.
.i
j =1 Assume the reference signals satisfy sup t ! [0, 3) u (t) = l i 1 3,
N .
xo = u -a (x i - u i) - b / a ij (xt i - xt j) - v i, (S21b)
i .i for i ! {1, f, N}, and sup t ! [0, 3) P N u (t) = c 1 3. For any
j =1 a, b ! R 2 0, (S21) over G starting from x i (0) ! R and v i (0) ! R
with R Nj = 1 v j (0) = 0, where each agent i ! {1, f, N} commu-
for each i ! {1, f, N}, where xt i (t) = x i (t ik) for t ! [t ik, t ik +1), with
nicates with its neighbors at times {t ik} k ! N 1 R $ 0, starting at
{t ik} 1 R $ 0 denoting the sequence of times at which agent i
t i1 = 0, determined by (S22) with e ! R N2 0, satisfies
communicates with its in-neighbors. The basic idea is that
agents share their information with neighbors when the uncer- c +b L e
tainty in the outdated information is such that the monotonic limsup x i (t) - u avg (t) # , (S24)
t"3 bmt 2
convergent behavior of the overall network can no longer be
guaranteed. The design of such triggers is challenging because for i ! {1, f, N} with an exponential rate of convergence of
of the following requirements: 1) triggers need to be distributed min {a, bmt 2}. Furthermore, the interexecution times of agent
so that agents can check them with the information available i ! {1, f, N} are lower bounded by (S23).
to them from their out-neighbors, 2) they must guarantee the The expected tradeoff between the desire for longer
absence of Zeno behavior (the undesirable situation where an interevent time and the adverse effect on systems convergence
rj
j=1
bounds on the maximum number of communication rounds
N
(and associated energy spent) by each agent i ! {1, f, N}
0
1
(and hence the network) during any given time interval. It is
N
xi −
interesting to analyze how this lower bound depends on the
−1
various problem ingredients: x i is an increasing function of 0 5 10 15 20
e i and a decreasing function of a and c i . Through the lat- t
ter variable, the bound also depends on the graph topology (a)
and the design parameter b. Given the definition of c i, we 5
can deduce that the faster an input of an agent is changing 4
Agents
(larger l i) or the farther the agent initially starts from the av- 3
erage of the inputs, the more often that agent would need to 2
trigger communication. The connection between the network 1
performance and the communication overhead can also be ob- 0 5 10 15 20
served here. Increasing b or decreasing e i to improve the ul- t
timate tracking error bound (S24) results in smaller inter-event (b)
times. Given that the rate of convergence of (S21) under (S22)
is min {a, bmt 2}, decreasing a to increase the inter-event times FIGURE S2 A comparison between the event-triggered algo-
slows down the convergence. rithm (S21) employing the event-triggered communication law
(S25) and the Euler discretized implementation of (25), as
When the interaction topology is a connected graph, the
described in (S15) with fixed step size [S15]. Both of these
properties of the Laplacian enable the identification of an al- algorithms use a = 1 and b = 4. The network is a weight-bal-
ternative event-triggered communication law that, compared to anced digraph of five agents with unit weights. The inputs are
(S22), has a longer inter-event time but similar dynamic aver- r 1 (t) = 0.5 sin (0.8t), r 2 (t) = 0.5 sin (0.7t) + 0.5 cos (0.6t), r 3 (t) =
age tracking performance. Consider the sequence of commu- sin (0.2t) + 1, r 4 (t) = atan (0.5t), and r 5 (t) = 0.1 cos (2t). In plot
(a), the black (respectively, gray) lines correspond to the track-
nication times {t ik} k ! N determined by
ing error of the event-triggered algorithm (S21) employing
event-triggered law (S25) with e i / ^2 d iout h = 0.1 [respectively,
t ik + 1 = argmax " t ! 6t ik, 3 h xt i ^ t h - x i ^ t h
2
the Euler discretized algorithm (S15) with fixed step size
N d = 0.12]. Recall from “Euler Discretizations of Continuous-
# 1
4d iout
/ a ij ^ t h xt i ^ t h - xt j ^ t h +
2 1 e 2}.
4d iout
i (S25) Time Dynamic Average Consensus Algorithms” that conver-
j=1
gence for (S15) is guaranteed if d ! (0, min {a -1, b -1 (d max out -1
) }),
which, for this example, results in d ! (0, 0.125). The horizontal
Compared to (S22), the extra term 1/ (4d iout) R Nj = 1 a ij (t) ; xt i (t) -
blue lines show the ! 0.05 error bound for reference. Part (a)
xt j (t) ; 2 in the communication law (S25) allows agents to have shows that both algorithms exhibit comparable tracking per-
longer inter-event times. Formally, the interexecution times formance. Part (b) shows the communication times of each
of agent i ! {1, f, N} implementing (S25) are lower bound- agent using the event-triggered strategy. The number of times
ed by that agents {1, f, 5} communicate in the time interval [0, 20]
is (39, 40, 42, 40, 39), respectively, when implementing event-
1 ln 1 + triggered communication (S25). These numbers are sig-
c m, (S26)
ae i
xi =
a 2cr i d iout nificantly smaller than the communication rounds required
by each agent in the Euler discretized algorithm (S15)
for positive constants cr i; see [S15] for explicit expressions. (20/0.12 - 166 rounds).
Numerical examples in [S15] show that the implementation
of (S25) for connected graphs results in inter-event times lon- [S23] W. P. M. H. Heemels, K. H. Johansson, and P. Tabuada, “An in-
troduction to event-triggered and self-triggered control,” in Proc. IEEE
ger than the ones of the event-triggered law (S22). Figure S2 Conf. Decision and Control, 2012, pp. 3270–3285.
shows one of those examples. Similar results can also be de- [S24] L. Hetel et al., “Recent developments on the stability of systems
rived for time-varying, jointly connected graphs (see [S15] for a with aperiodic sampling: An overview,” Automatica, vol. 76, pp. 309–
335, Feb. 2017.
complete exposition). [S25] C. Nowzari, E. García, and J. Cortés. Event-triggered control and
communication of network systems for multi-agent consensus. 2017.
[Online]. Available: arXiv: 1712.00429
REFERENCES [S26] L. Ding, Q. L. Han, X. Ge, and X. M. Zhang, “An overview of recent
[S22] H. Karl and A. Willig, Protocols and Architectures for Wireless advances in event-triggered consensus of multiagent systems,” IEEE
Sensor Networks. Hoboken, NJ: Wiley, 2005. Trans. Cybern., vol. 48, no. 4, pp. 1110–1123, 2018.