tensorkelechi/arxiv-cstext · Datasets at Hugging Face

text stringlengths 0 89.3k
10125 memory slot M3is initialized with tubes com
puted by 42 for initial condition x z 00 and
memory slot M2is left empty The memory is then up
dated by Algorithm 2 in every timestep and by Algo
rithm 1 every ten timesteps where the new tubes are
stored in M1and the previous ones are copied to M2 In
Figure 6 these updates are clearly visible by the spikes
in the convex combination variables The spikes occur
because we copy the previous tubes to M2 ie when the
tubes in M1blue lines are updated it is beneficial to fuse
them with the previous ones in M2green lines How
ever after a few timesteps it is preferrable to mainly use
the new tubes which is evident by increasing λ1blue
and decreasing λ2green towards the next tube update
Fork50 the trajectories are approaching the origin
and the tubes in M3yellow lines are increasingly used
which is expected since they are optimized for a region
around the origin Finally we discourage the use of the
fallback tubes in M0red lines with a regularization cost
onλ0 which is evident by their low usage ie these
tubes are only used if necessary to retain feasibility
7 Conclusions
This paper introduced filterbased system level tube
MPC SLTMPC for constrained discretetime linear
systems with additive disturbances and model uncer
tainties We detailed a generalization of the filterbased
SLTMPC method introduced in 25 and proposed a
modified version which admits closedloop guarantees
in receding horizon By introducing a new terminal
controller combined with an online optimized terminal
set we proved recursive feasibility and inputtostate
stability of the proposed method Additionally we in
troduced a new asynchronous computation scheme for
filterbased SLTMPC which significantly reduces its
computational demand by separating the tube and nom
inal trajectory optimizations into different processes
Finally we showed the benefits and effectiveness of the
proposed methods on two numerical examples
15105 0 5 10 15
u1D46500255075100125150u1D467Primary process 46
0 50 100 150
Timestepu1D458000204060810u1D706u1D45Au1D7060u1D7061u1D7062u1D7063Fig 6 Closedloop trajectories in the xzplane of the
VTOL vehicle with the primary process 46 starting
in x z 10 125top and the corresponding convex
combination parameters λm m 0 M1 bottom
Table 2
Computation times for the VTOL vehicle example with hori
zonN 10
Comp Time ms
Min Median
SLTMPC 32 49075 71976
Asynchronous computation
Primary Process 46 779 1659
Secondary Process 42 4735 6853
References
1 M V Kothare V Balakrishnan and M Morari Robust
constrained model predictive control using linear matrix
inequalities Automatica vol 32 no 10 pp 13611379
1996
2 B Kouvaritakis J Rossiter and J Schuurmans Efficient
robust predictive control IEEE Transactions on Automatic
Control vol 45 no 8 pp 15451549 2000
3 J B Rawlings D Q Mayne and M Diehl Model Predictive
Control Theory Computation and Design Nob Hill
Publishing 2 ed 2019
174 B Kouvaritakis and M Cannon Model Predictive Control
Classical Robust and Stochastic Springer 1 ed 2016
5 L Chisci J A Rossiter and G Zappa Systems with
persistent disturbances predictive control with restricted
constraints Automatica vol 37 no 7 pp 10191028 2001
6 D Q Mayne M M Seron and S Rakovi c Robust model
predictive control of constrained linear systems with bounded
disturbances Automatica vol 41 no 2 pp 219224 2005
7 M Zanon and S Gros On the Similarity Between Two
Popular Tube MPC Formulations in 2021 European Control
Conference ECC pp 651656 2021
8 S V Rakovi c B Kouvaritakis R Findeisen and M Cannon
Homothetic tube model predictive control Automatica
vol 48 no 8 pp 16311638 2012
9 S V Rakovi c W S Levine and B A cikmese Elastic
tube model predictive control in Proc of American Control
Conference ACC pp 35943599 2016
10 J Sieber A Zanelli S Bennani and M N Zeilinger
System level disturbance reachable sets and their
application to tubebased MPC European Journal of
Control p 100680 2022
11 J L ofberg Approximations of closedloop minimax MPC
inProc 42nd IEEE Conf Decis Control vol 2 pp 1438
1442 2003
12 P J Goulart E C Kerrigan and J M MacIejowski
Optimization over state feedback policies for robust control
with constraints Automatica vol 42 no 4 pp 523533
2006
13 J Sieber S Bennani and M N Zeilinger A System Level
Approach to TubeBased Model Predictive Control IEEE
Control Systems Letters vol 6 pp 776781 2022
14 J Anderson J C Doyle S H Low and N Matni System
level synthesis Annu Rev Control vol 47 pp 364393
2019
15 M Bujarbaruah U Rosolia Y R St urz and F Borrelli A

10125 memory slot M3is initialized with tubes com

puted by 42 for initial condition x z 00 and

memory slot M2is left empty The memory is then up

dated by Algorithm 2 in every timestep and by Algo

rithm 1 every ten timesteps where the new tubes are

stored in M1and the previous ones are copied to M2 In

Figure 6 these updates are clearly visible by the spikes

in the convex combination variables The spikes occur

because we copy the previous tubes to M2 ie when the

tubes in M1blue lines are updated it is beneficial to fuse

them with the previous ones in M2green lines How

ever after a few timesteps it is preferrable to mainly use

the new tubes which is evident by increasing λ1blue

and decreasing λ2green towards the next tube update

Fork50 the trajectories are approaching the origin

and the tubes in M3yellow lines are increasingly used

which is expected since they are optimized for a region

around the origin Finally we discourage the use of the

fallback tubes in M0red lines with a regularization cost

onλ0 which is evident by their low usage ie these

tubes are only used if necessary to retain feasibility

7 Conclusions

This paper introduced filterbased system level tube

MPC SLTMPC for constrained discretetime linear

systems with additive disturbances and model uncer

tainties We detailed a generalization of the filterbased

SLTMPC method introduced in 25 and proposed a

modified version which admits closedloop guarantees

in receding horizon By introducing a new terminal

controller combined with an online optimized terminal

set we proved recursive feasibility and inputtostate

stability of the proposed method Additionally we in

troduced a new asynchronous computation scheme for

filterbased SLTMPC which significantly reduces its

computational demand by separating the tube and nom

inal trajectory optimizations into different processes

Finally we showed the benefits and effectiveness of the

proposed methods on two numerical examples

15105 0 5 10 15

u1D46500255075100125150u1D467Primary process 46

0 50 100 150

Timestepu1D458000204060810u1D706u1D45Au1D7060u1D7061u1D7062u1D7063Fig 6 Closedloop trajectories in the xzplane of the

VTOL vehicle with the primary process 46 starting

in x z 10 125top and the corresponding convex

combination parameters λm m 0 M1 bottom

Table 2

Computation times for the VTOL vehicle example with hori

zonN 10

Comp Time ms

Min Median

SLTMPC 32 49075 71976

Asynchronous computation

Primary Process 46 779 1659

Secondary Process 42 4735 6853

References

1 M V Kothare V Balakrishnan and M Morari Robust

constrained model predictive control using linear matrix

inequalities Automatica vol 32 no 10 pp 13611379

1996

2 B Kouvaritakis J Rossiter and J Schuurmans Efficient

robust predictive control IEEE Transactions on Automatic

Control vol 45 no 8 pp 15451549 2000

3 J B Rawlings D Q Mayne and M Diehl Model Predictive

Control Theory Computation and Design Nob Hill

Publishing 2 ed 2019

174 B Kouvaritakis and M Cannon Model Predictive Control

Classical Robust and Stochastic Springer 1 ed 2016

5 L Chisci J A Rossiter and G Zappa Systems with

persistent disturbances predictive control with restricted

constraints Automatica vol 37 no 7 pp 10191028 2001

6 D Q Mayne M M Seron and S Rakovi c Robust model

predictive control of constrained linear systems with bounded

disturbances Automatica vol 41 no 2 pp 219224 2005

7 M Zanon and S Gros On the Similarity Between Two

Popular Tube MPC Formulations in 2021 European Control

Conference ECC pp 651656 2021

8 S V Rakovi c B Kouvaritakis R Findeisen and M Cannon

Homothetic tube model predictive control Automatica

vol 48 no 8 pp 16311638 2012

9 S V Rakovi c W S Levine and B A cikmese Elastic

tube model predictive control in Proc of American Control

Conference ACC pp 35943599 2016

10 J Sieber A Zanelli S Bennani and M N Zeilinger

System level disturbance reachable sets and their

application to tubebased MPC European Journal of

Control p 100680 2022

11 J L ofberg Approximations of closedloop minimax MPC

inProc 42nd IEEE Conf Decis Control vol 2 pp 1438

1442 2003

12 P J Goulart E C Kerrigan and J M MacIejowski

Optimization over state feedback policies for robust control

with constraints Automatica vol 42 no 4 pp 523533

2006

13 J Sieber S Bennani and M N Zeilinger A System Level

Approach to TubeBased Model Predictive Control IEEE

Control Systems Letters vol 6 pp 776781 2022

14 J Anderson J C Doyle S H Low and N Matni System

level synthesis Annu Rev Control vol 47 pp 364393

2019

15 M Bujarbaruah U Rosolia Y R St urz and F Borrelli A