Pith. sign in

REVIEW 2 minor 20 references

An explicit controller from the unconstrained problem yields performance bounds and minimum horizon for MPC on constrained positive graph systems.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

T0 review · grok-4.3

2026-06-29 10:42 UTC pith:R6JQVNDV

load-bearing objection The paper gives a structure-based way to build an explicit feasible controller for capacity-constrained positive incidence systems, from which graph-computable bounds and a shortest stabilizing horizon follow for terminal-free MPC.

arxiv 2605.28550 v1 pith:R6JQVNDV submitted 2026-05-27 math.OC cs.SYeess.SY

Model Predictive Control for Constrained Linear Positive Systems on Graphs

Roland Schurig , David Ohlin , Anders Rantzer , Emma Tegling , Rolf Findeisen This is my paper
classification math.OC cs.SYeess.SY
keywords model predictive controlpositive systemsgraphscapacity constraintsincidence formperformance boundsstabilizing horizonconvex optimization
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper establishes that for positive linear systems on graphs with capacity constraints, the structure of the unconstrained optimal control problem allows construction of a suboptimal but admissible controller. This controller provides explicit, graph-computable bounds on the performance of a model predictive controller without terminal conditions. It also determines the minimum horizon length needed for stability. A convex optimization problem computes the tightest such bounds and horizon. This approach exploits the positivity and incidence structure to achieve guarantees that are usually unavailable for constrained MPC.

Core claim

By leveraging the analytic structure of the unconstrained problem, an explicit suboptimal admissible controller is constructed for positive linear systems in incidence form with linear cost and capacity constraints. This yields graph-computable performance bounds and a minimum stabilising horizon length for model predictive control without terminal conditions, with the optimal bound and horizon computed via a convex program.

What carries the argument

The explicit suboptimal admissible controller constructed by leveraging the analytic structure of the unconstrained problem, which remains feasible under capacity constraints and enables computation of performance bounds.

Load-bearing premise

The systems are positive linear systems in incidence form with linear cost, and the analytic structure of the unconstrained problem directly enables construction of an admissible controller that remains feasible under capacity constraints.

What would settle it

Finding a positive system instance where the constructed controller becomes infeasible when capacity constraints are active, or where the predicted minimum horizon fails to stabilize the closed loop.

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

  • Graph-computable performance bounds are available for the MPC.
  • A minimum stabilising horizon length can be determined.
  • Efficient computation of the optimal bound and horizon is possible via convex program.
  • System structure enables explicit MPC guarantees typically not available.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • This method could extend to other classes of positive systems beyond incidence form.
  • The approach might improve scalability of control design for large logistics or compartmental networks.
  • It suggests that positivity can be exploited more broadly to simplify constrained optimal control problems.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

0 major / 2 minor

Summary. The paper considers model predictive control for constrained linear positive systems on graphs, modeled in incidence form with linear costs and capacity constraints on states and inputs. Leveraging the analytic structure of the unconstrained problem, an explicit suboptimal admissible controller is constructed. This yields graph-computable performance bounds and a minimum stabilising horizon length for an MPC scheme without terminal conditions. A convex program is given to compute the optimal bound and horizon efficiently.

Significance. If the derivations hold, the work offers a meaningful advance in structure-exploiting MPC for positive systems common in routing, logistics, and compartmental models. The explicit controller construction, graph-computable bounds, and convex program for horizon selection provide practical guarantees that are often unavailable without terminal costs or heavy computation, strengthening applicability to network problems.

minor comments (2)
  1. [Abstract] Abstract: the phrase 'graph-computable performance bounds' is introduced without a brief definition or reference to the incidence matrix properties that enable it; adding one sentence would improve immediate clarity for readers.
  2. The manuscript would benefit from a small illustrative graph example (e.g., 3-4 nodes) showing the explicit controller, bound computation, and convex program output to make the claims more concrete.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for their supportive summary and recommendation of minor revision. No major comments were provided in the report, so there are no specific points requiring rebuttal or revision at this stage. We will address any minor issues identified during the revision process.

Circularity Check

0 steps flagged

No significant circularity detected

full rationale

The abstract and available description contain no equations, self-referential definitions, or fitted parameters presented as predictions. The claimed construction leverages the analytic structure of the unconstrained problem to build an admissible controller, yielding graph-computable bounds, without any visible reduction of outputs to inputs by construction. No self-citation load-bearing steps, uniqueness theorems, or ansatzes are referenced in the provided text. The derivation chain appears self-contained against external structure-exploiting MPC results for positive systems.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract-only review; no equations or detailed assumptions are provided, so the ledger is empty.

pith-pipeline@v0.9.1-grok · 5647 in / 1081 out tokens · 27801 ms · 2026-06-29T10:42:18.542072+00:00 · methodology

0 comments
read the original abstract

Positive systems describing networks with inherently non-negative states and inputs arise naturally in routing, logistics, and compartmental modelling. We consider problems modelled as positive linear systems in incidence form with linear cost. The addition of capacity constraints on states (storage) and inputs (flows between nodes) significantly increases the problem complexity. Leveraging the analytic structure of the unconstrained problem, an explicit suboptimal admissible controller is constructed. This yields graph-computable performance bounds and a minimum stabilising horizon length for a model predictive controller without terminal conditions. A convex program enables efficient computation of the optimal bound and horizon. These results highlight how system structure enables explicit MPC guarantees that are typically not available.

Figures

Figures reproduced from arXiv: 2605.28550 by Anders Rantzer, David Ohlin, Emma Tegling, Roland Schurig, Rolf Findeisen.

Figure 1
Figure 1. Figure 1: Graph with n = 5 vertices and goal vertex g = 6. Coloured edges indicate inputs used in the optimal unconstrained solution. Exam￾ple 1 uses costs s⊤ = (10 5 1 3 2), r⊤ = (1 5 5 1 1 1 1 1 1) and capacity constraints x¯ = 1, u¯⊤ = ( 1 4 1 4 1 4 1 4 1 4 1 4 1 1 1). remains nonnegative and prevents more commodities being taken out of a vertex than are currently present. Example 1. Consider the graph in [PITH_… view at source ↗
Figure 2
Figure 2. Figure 2: Evolution of the states x(·) := xµ¯(·, x¯) in the example of Section V using MPC (µ¯ = µN , left) and the suboptimal feedback law (µ¯ = ˆµ, right). t u1,k(t) 0 1 2 3 4 .25 t u2,k(t) 0 1 2 3 4 .25 t ui(t) 0 1 2 3 4 1 t u1,k(t) 0 1 2 3 4 .25 t u2,k(t) 0 1 2 3 4 .25 t ui(t) 0 1 2 3 4 1 k = 1 k = 2 k = 3 k = 1 k = 2 k = 3 i = 3 i = 4 i = 5 [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Controls u(·) := ¯µ(xµ¯(·, x¯)), using MPC (µ¯ = µN , left) and the admissible feedback law (µ¯ = ˆµ, right) in the example of Section V. The top plot shows the controls associated with vertex 1, the middle plot the controls associated with vertex 2, and the bottom plot the controls associated with the remaining vertices. In both cases, u1,2(·) = u1,3(·) and u2,2(·) = u2,3(·). V. NUMERICAL EXAMPLE We demon… view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

20 extracted references · 1 canonical work pages

  1. [1]

    Optimal control of compartmental models: The exact solution,

    F. Blanchini, P. Bolzern, P. Colaneri, G. De Nicolao, and G. Giordano, “Optimal control of compartmental models: The exact solution,”Auto- matica, vol. 147, no. C, jan 2023

    2023

  2. [2]

    Explicit solution to Bellman equation for positive systems with linear cost,

    A. Rantzer, “Explicit solution to Bellman equation for positive systems with linear cost,” in2022 IEEE 61st Conf. Dec. Cont. (CDC), 2022, pp. 6154–6155

    2022

  3. [3]

    Optimal control of linear cost networks,

    D. Ohlin, E. Tegling, and A. Rantzer, “Optimal control of linear cost networks,”Europ. J. Cont., vol. 80, p. 101068, 2024

    2024

  4. [4]

    Heuristic search for linear positive systems,

    D. Ohlin, A. Rantzer, and E. Tegling, “Heuristic search for linear positive systems,”arXiv: 2410.17220 [math.OC], 2024

    work page arXiv 2024

  5. [5]

    On the infinite horizon performance of receding horizon controllers,

    L. Grüne and A. Rantzer, “On the infinite horizon performance of receding horizon controllers,”IEEE Trans. Aut. Cont., vol. 53, no. 9, pp. 2100–2111, 2008

    2008

  6. [6]

    S. P. Boyd and L. Vandenberghe,Convex optimization. Cambridge University Press, 2004

    2004

  7. [7]

    Mathematical methods of organizing and planning production,

    L. V . Kantorovich, “Mathematical methods of organizing and planning production,”Management Science, vol. 6, no. 4, pp. 366–422, July 1960

    1960

  8. [8]

    A note on two problems in connexion with graphs,

    E. W. Dijkstra, “A note on two problems in connexion with graphs,” Numer. Math. 1, p. 269–271, 1959

    1959

  9. [9]

    Learning to act using real- time dynamic programming,

    A. G. Barto, S. J. Bradtke, and S. P. Singh, “Learning to act using real- time dynamic programming,”Artificial Intelligence, vol. 72, no. 1, pp. 81–138, 1995

    1995

  10. [10]

    An optimal control approach to dynamic routing in networks,

    F. Moss and A. Segall, “An optimal control approach to dynamic routing in networks,”IEEE Trans. Aut. Cont., vol. 27, no. 2, pp. 329–339, 1982

    1982

  11. [11]

    Grüne and J

    L. Grüne and J. Pannek,Nonlinear Model Predictive Control: Theory and Algorithms, 2nd ed., ser. Comm. Cont. Eng. Springer, 2017

    2017

  12. [12]

    Borrelli, A

    F. Borrelli, A. Bemporad, and M. Morari,Predictive Control for Linear and Hybrid Systems. Cambridge University Press, 2017

    2017

  13. [13]

    A quasi-infinite horizon nonlinear model predictive control scheme with guaranteed stability,

    H. Chen and F. Allgöwer, “A quasi-infinite horizon nonlinear model predictive control scheme with guaranteed stability,”Automatica, vol. 34, no. 10, pp. 1205–1217, 1998

    1998

  14. [14]

    Stability and feasibility of state constrained mpc without stabilizing terminal constraints,

    A. Boccia, L. Grüne, and K. Worthmann, “Stability and feasibility of state constrained mpc without stabilizing terminal constraints,”Syst. & Cont. Let., vol. 72, pp. 14–21, 2014

    2014

  15. [15]

    Linear programming based routing design for a class of positive systems with integral and capacity constraints,

    H. Arneson and C. Langbort, “Linear programming based routing design for a class of positive systems with integral and capacity constraints,” IFAC Proc. Vol., vol. 42, no. 20, pp. 352–357, 2009

    2009

  16. [16]

    A model predictive control framework for robust management of multi-product, multi-echelon demand networks,

    M. Braun, D. Rivera, M. Flores, W. Carlyle, and K. Kempf, “A model predictive control framework for robust management of multi-product, multi-echelon demand networks,”An. Rev. Cont., vol. 27, no. 2, pp. 229–245, 2003

    2003

  17. [17]

    A model predictive control framework for constrained uncertain positive systems,

    J. Zhang, X. Jia, R. Zhang, and Y . Zuo, “A model predictive control framework for constrained uncertain positive systems,”I. J. Syst. Sci., vol. 49, no. 4, pp. 884–896, 2018

    2018

  18. [18]

    Economic model predictive control for robust optimal operation of sparse storage networks,

    F. Lejarza and M. Baldea, “Economic model predictive control for robust optimal operation of sparse storage networks,”Automatica, vol. 125, p. 109346, 2021

    2021

  19. [19]

    Constrained model predictive control for positive systems,

    H. Mehrivash and M. H. Shafiei, “Constrained model predictive control for positive systems,”IET Cont. Theo. & Appl., vol. 13, no. 10, pp. 1491–1499, 2019

    2019

  20. [20]

    R. A. Horn and C. R. Johnson,Matrix Analysis, 2nd ed., ser. Comm. Cont. Eng. Cambridge University Press, 2013

    2013