WorksMathematics
Stability Filters for Receding-Horizon Games
Abstract
In receding-horizon games, self-interested players repeatedly solve a finite-horizon generalized Nash equilibrium problem (GNEP) and apply its first-stage actions. Ensuring stability of the resulting closed-loop system, however, is a challenging task that requires careful design of the players' strategies or additional constraints. We propose a stability filter that minimally modifies the variational equilibrium of the finite-horizon GNEP to enforce a prescribed decrease of a quadratic Lyapunov function. For receding-horizon games with linear dynamics, quadratic costs, and polyhedral coupling input constraints, whose associated GNEP is strongly monotone, we show that a filter always exists such that a closed-loop equilibrium exists, the filter is always feasible, does not intervene when the equilibrium already satisfies the decrease condition, and that renders the closed loop globally exponentially stable. Moreover, the filtered strategy is an -generalized Nash equilibrium, with vanishing exponentially along closed-loop trajectories.
Introduction
Many control problems involve several decision makers that are acting on the same dynamical system while pursuing individual objectives, for instance in demand-side management [1] or the charging of electric vehicles [2]. Coupled through the shared dynamics and the shared constraints, their interaction is naturally modeled as a generalized Nash equilibrium problem (GNEP) [3], [4], whose variational equilibrium (v-GNE) is characterized by a variational inequality with equal multipliers for all players [5], [6].
Receding-horizon games (RHGs) combine this model with the receding-horizon principle of model predictive control (MPC) [7], [8], [9]: at every time step, the players choose a sequence of future control actions by solving a finite-horizon GNEP, parameterized by the current state, and apply the first-stage actions of the resulting equilibrium. RHGs thus describe strategic agents that repeatedly re-plan their strategies in response to the evolving state.
Unlike in standard MPC, where the stage cost is chosen by the designer to optimize all control actions simultaneously, the costs in a RHG belong to the players and reflect their individual interests. Hence, the standard argument of using a single optimal cost as a Lyapunov function fails beyond potential games. Certifying closed-loop stability is of paramount importance for the applicability of RHGs in practice as, for example, agents can destabilize each other through an RHG even when their local dynamics are stable [10].
For strongly monotone games with stable linear dynamics, [10] derives, via dissipativity and monotone operator theory, a sufficient condition in the form of a linear matrix inequality (LMI) that certifies global asymptotic stability, together with guidelines for tuning the cost weights. Other works equip the finite-horizon game with terminal ingredients: [9] obtains an explicit stabilizing terminal cost for linear-quadratic games from an equivalent regulator in an augmented space, and [11], [12] derive conditions on the players’ objectives that guarantee asymptotic stability under terminal constraints or Lyapunov terminal costs via -quasi GNEPs. Closed-loop stability under input and state constraints is instead achieved in [13] by introducing in the GNEP a shared linear constraint on the first control moves that forces a certain polyhedral control Lyapunov function to decrease. These approaches either certify stability of a given RHG, with no remedy when the certificate fails, or modify the players’ costs or constraints, which may severely perturb the players’ private interests.
A different mechanism proposed in the MPC literature to ensure stability is the use of stability filters [14], [15], [16] and control barrier and control Lyapunov function based quadratic programs [17], [18], which minimally modify a nominal input to enforce safety or stability.
Contributions. For open-loop Schur stable linear dynamical systems, quadratic costs defining a game with strongly monotone pseudo-gradient, and compact polyhedral coupling constraints, we propose a stability filter that replaces the v-GNE at every time step by the closest feasible joint strategy that guarantees a prescribed decrease of a quadratic Lyapunov function. Since the filtered strategy is in general no longer an equilibrium, a central question is how much strategic optimality is sacrificed for stability. We establish the following properties:
The proposed convex filter is feasible for every state, leaves the v-GNE unchanged whenever it already satisfies the decrease condition, and intervenes by an amount bounded linearly in the distance to the equilibrium.
The filtered closed loop is globally exponentially stable, with a rate prescribed by the designer.
The filtered strategy is an -generalized Nash equilibrium, with bounded by the v-GNE multipliers and the size of the intervention and vanishing exponentially
T. Tatarenko is with the Department of Control Theory and Intelligent Systems, TU Darmstadt, Germany. E-mail: [email protected].
A. Bemporad is with the IMT School for Advanced Studies, Piazza San Francesco 19, Lucca, Italy. E-mail: [email protected].
This work was funded by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation), project number 528033031. This work was also funded by the European Union (ERC Advanced Research Grant COMPACT, No. 101141351). Views and opinions expressed are however those of the authors only and do not necessarily reflect those of the European Union or the European Research Council. Neither the European Union nor the granting authority can be held responsible for them. along the closed loop. Hence strategic optimality is recovered asymptotically.
Organization. Section II introduces the RHG and the standing assumptions. Section III establishes existence of a closed-loop equilibrium, introduces the stability filter, proves global exponential stability of the filtered closed loop, and analyzes the resulting strategic loss; the main results are summarized in Theorem 4. A numerical example is provided in Section IV and the paper concludes with a discussion in Section V.
Problem Formulation
We consider the linear system
where is the system state at time and is the command input applied by all players. In game-theoretic MPC, is obtained from the first stage of the joint finite-horizon strategy, i.e., , where denotes the joint finite-horizon strategy generated at time by solving the receding-horizon game, and the matrix extracts the first-stage joint action from the finite-horizon strategy.
The strategy is partitioned into the decision blocks indexed by . Each player’s block is itself a finite-horizon sequence of length ,
where is player ’s action planned at time for prediction step , and collects the sequences of all other players. The cost of player is
with and where the predicted states are generated from the current state as
with . Since the game solved at time depends on only through , we omit the time index in and in all quantities derived from it below, which we treat as functions of a generic state .
Given and the sequences , each player solves
where the polyhedral constraint set
encodes shared linear inequality and, possibly, equality constraints the input variables of all players need to satisfy.
The players’ pseudo-gradient stacks the partial gradients . As from (4) is affine in , ’s are quadratic and the associated pseudogradient is affine, i.e.
with , , obtained from the standard condensed representation of (5), and , , denoting the corresponding -th block row/entry. Note that, for better generality, we retain the constant term in the affine representation of the pseudogradient, even though for each player when the game is generated from (3)–(6).
We impose the following standing assumptions.
Assumption 1: The following conditions hold:
is a nonempty and compact polyhedron;
is strongly monotone on , uniformly in , with constant , i.e., for all and all admissible .
for every player , the cost is -strongly convex in the player’s own decision variable, with ;
is Schur stable.
Remark 1 (Relation between Assumption 1.2 and 1.3): Since , the strong monotonicity inequality of Assumption 1.2 implies Assumption 1.3 with , as can be seen by testing the inequality on vectors such that . We retain it as a separate hypothesis only to allow, for each player, a constant tighter than . Such tighter constants directly sharpen the bounds of Theorems 2 and 3 below.
Under Assumption 1, for a given state , the variational generalized Nash equilibrium (v-GNE) of the game (5) is characterized by the variational inequality , which, by strong monotonicity of and Assumption 1.1, admits a unique solution for every . We denote the corresponding equilibrium solution mapping by
Computing at every time step is tractable: since is affine and strongly monotone and is a polyhedron, the v-GNE can be computed by centralized projection-type methods for strongly monotone variational inequalities, such as projected gradient [6]. Alternatively, it can be computed in a distributed manner, in which each player updates only its own decision variable, by gradient-play and learning algorithms with convergence rate guarantees for strongly monotone games [19], including settings with coupling constraints [20].
Since is affine in and strongly monotone in , the solution mapping is globally Lipschitz continuous [6 Thm. 2.3]. In particular,
Stability Filter Design and Guarantees
Closed-loop equilibrium
We first establish existence of an equilibrium of the nominal receding-horizon closed loop. Thus, existence of such an equilibrium does not need to be imposed separately.
Proposition 1: Suppose Assumption 1 holds. Then there exists satisfying
Proof: Since is Schur, is nonsingular. Define . The mapping is continuous because is (see (9)). Moreover, as any GNE solution is also feasible.
Since is nonempty, compact, and convex, Brouwer’s fixed-point theorem implies that there exists such that . Define . Then and , or equivalently . This proves (10).
Remark 2: Proposition 1 establishes existence of , but Brouwer’s theorem does not rule out several fixed points of . None of the results that follow, namely global feasibility (Proposition 2), nonintervention (Proposition 3), the intervention bound (Proposition 4), closed-loop stability (Theorem 1), or the strategic-loss analysis in Section III-C, use uniqueness of . Each requires only that a single pair satisfying (10) to be fixed once and used consistently as the reference in the filter constraint defined below in (18). When (10) admits multiple solutions, should therefore be understood as one such pair, selected offline (for instance via Corollary 1 below, when applicable); every guarantee below then holds relative to that fixed choice, uniformly in the online state .
Corollary 1 (Sufficient condition for uniqueness): Let
with as in (9). If , then in Proposition 1 is the unique pair satisfying (10). Moreover, for any , the iterates satisfy
so , and hence , can be computed offline to any desired accuracy.
Proof: For , the Lipschitz bound (9) gives
If , is a contraction on , which is complete since it is a compact subset of . The Banach fixed-point theorem then gives a unique fixed point , together with the stated geometric convergence rate of the iterates . Since , it is likewise unique.
Stability filter
Since is Schur, for any there exists a unique satisfying the discrete-time Lyapunov equation
Define and
By (13), , hence
and
Choose a prescribed contraction factor
Throughout, denotes the first component of a fixed pair satisfying (10), in the sense of Remark 2. We define the stability-filtered strategy by
where .
The filter therefore selects the feasible strategy closest to the nominal v-GNE while enforcing a prescribed Lyapunov contraction defined by (17).
Remark 3: The constant depends on through and always satisfies , where is the spectral radius of . Conversely, for every there exists with , and then yields . Hence a suitable choice of brings arbitrarily close to , which enlarges the admissible range of contraction factors in (17).
We next establish the basic properties of the filter. We show that it is feasible for every state (Proposition 2), that it leaves the v-GNE unchanged whenever the latter already satisfies the contraction condition (Proposition 3), and that its deviation from the v-GNE is bounded linearly in (Proposition 4). Building on these results, Theorem 1 proves global exponential stability of the filtered closed loop, and Corollary 2 shows that the intervention vanishes exponentially along it.
Proposition 2: Under Assumption 1, the stability filter (18) is feasible for every .
Proof: Consider the constant candidate . By Proposition 1, . Moreover,
Consequently,
Thus is feasible for (18) for every .
Proposition 3: If the nominal equilibrium strategy satisfies
then .
Proof: If (19) holds, then is feasible for (18). The objective in (18) is nonnegative and equals zero at . Since , this point is the unique optimizer.
Proposition 4: Let
Then for every , where one may take
Proof: Since is feasible for (18), optimality of gives
Since , the Lipschitz property (9) yields
Using norm equivalence,
which proves the claim.
Theorem 1: Consider the filtered closed loop
initialized at . Then is globally exponentially stable. More precisely,
Proof: The contraction inequality follows directly from the constraint in (18). Hence, iteration yields . Finally,
gives (23).
Corollary 2: Along the filtered closed loop,
More precisely,
Proof: The result follows immediately from Proposition 4 and Theorem 1.
Remark 4: Corollary 2 measures the intervention in the space of strategies only. Closeness of to does not by itself imply that is close to an equilibrium in the strategic sense: equilibrium is characterized by the players’ unilateral optimality, i.e., by comparing with the best cost player can achieve against the filtered strategies of the others, rather than by the cost incurred at the v-GNE. Moreover, when coupling constraints are active at , a deviation may induce a regret of first order in , whose magnitude is governed by the associated multipliers, as opposed to the second-order effect one would expect at an unconstrained equilibrium. This loss of strategic optimality is quantified in the following subsection (Theorems 2 and 3).
Strategic loss induced by stability filtering
By the construction and results above, the filtered strategy is feasible for the original game and renders the closed loop globally exponentially stable (Theorem 1). However, whenever the filter intervenes, need not itself be a GNE. What remains is therefore to characterize its optimality in terms of the players’ cost functions , i.e., to quantify the loss of strategic optimality incurred for the sake of stability.
For a feasible strategy , define the maximum unilateral regret
Thus is an -GNE whenever its maximum unilateral regret is bounded by .
For , the v-GNE optimality conditions imply the existence of common multipliers and satisfying (see, e.g., [6, Thm. 12.1)
We first establish that the multipliers required in the subsequent analysis can always be selected locally bounded.
Lemma 1: Under Assumption 1, for every compact set there exists a constant such that, for every , there exists a KKT multiplier pair associated with satisfying
In particular, the inequality multipliers admit a locally bounded selection in a neighborhood of .
The proof, which relies on Hoffman's error bound for linear systems [21], is given in Appendix I.
Theorem 2: Let , , and . Under Assumption 1, let be any KKT multiplier pair associated with . Then
Proof: Fix a player and any feasible unilateral deviation , i.e., . By strong convexity of with respect to its own decision variable,
Since is affine, . Thus, by the v-GNE stationarity condition (see (27)),
Let . Since and differ only in player ’s decision block, . Furthermore, both and satisfy the equality constraints, and hence . Therefore, by (31)
since and . Substitution of (32) into (30) gives
where we used . Since the inequality (33) holds for every feasible unilateral deviation, taking the supremum over and then the maximum over players gives
Finally, implies
which proves (29). ■
Remark 5: The bound (29) separates the strategic loss into two qualitatively different contributions:
The second term is quadratic in the stability intervention . In contrast, moving away from an inequality that is active at the nominal v-GNE and has a positive equilibrium multiplier can generate a first-order loss. Thus, the potentially first-order strategic effect of stability filtering is entirely associated with releasing positively priced active constraints.
Theorem 3. Under Assumption 1, for every initial condition there exists a constant such that, along the filtered closed loop (22) initialized at ,
Consequently, , for every . In particular, regardless of the initial condition.
Proof: Fix and define the sublevel set
Since is a positive-definite quadratic, is a compact ellipsoid, and every satisfies
By the contraction inequality, along the filtered closed loop; since trivially, induction on gives for every , so contains the entire trajectory.
By Lemma 1 applied to the compact set , there is a valid KKT multiplier selection with
where is a finite constant depending on . By Proposition 4, for every . Using Theorem 2,
where . Therefore, for every ,
By (37), for every . Thus,
Substituting into (39) gives, for every and in particular for every ,
which proves (35).
Finally, by Theorem 1,
so combining with (35),
and therefore , for every . ■
Main result
We now collect the results of this section into a single statement, which constitutes the main result of the paper: the stability filter renders the receding-horizon game globally exponentially stable while preserving strategic optimality asymptotically.
Theorem 4. Suppose Assumption 1 holds. Then the following statements are true.
Equilibrium: there exists with and .
Global feasibility: the filtered strategy defined in (18) exists for every .
Exact nonintervention: if satisfies (19), then .
Global exponential stability: for every , the trajectory of the filtered closed loop , , satisfies
Strategic recovery: for every , the maximum unilateral regret along the filtered closed loop satisfies ; in particular, .
Numerical example
We consider a random system (1) with and , whose matrix has eigenvalues and . Each of the players controls one input over the horizon . After reordering the stage vector in (3) as , player ’s weight is , where has random Gaussian entries and is the th unit vector. The low-rank term couples each player’s input to the state in a selfish way. The set contains the shared constraints and and the local bounds at all prediction steps. The game is strongly monotone with and . Since and , satisfies (10). We set in (13), which gives , and take and . At each , is computed with the QP-based method of [22] in NashOpt [23]; when (19) fails, (18) is solved as a second-order cone program by Clarabel [24].
Figure 1 shows the closed loop from . During the transient , the v-GNE alone would increase ; the filter intervenes at these steps, with decreasing from to , in agreement with Corollary 2. For , satisfies (19) and is applied unchanged (Proposition 3). As guaranteed by Theorem 1, at every step, and (Figure 2(a)). During the transient, the v-GNE lies on the edge of , while the filter moves the applied input away from it and onto the edge for (Figure 2(b)).
Figure 1. Closed-loop states (top) and inputs (bottom). Solid: applied input . Dotted: first move of the v-GNE. Dashed: bounds. The red ellipse marks the steps where the filter intervenes.
Figure 2. (a) Closed-loop trajectory with level sets of through every 4 steps. (b) versus for the v-GNE first move (crosses) and the applied input (circles), with the bounds of the shared constraints (dashed). Red arrows: corrections applied by the filter.
Conclusion
In this paper, we have proposed the use of a control-input filtering mechanism to ensure desired behavior and stability of a given constrained closed-loop game-theoretic MPC system. We have illustrated via a simple numerical example the effectiveness of the approach in guiding the system towards the desired equilibrium while satisfying all constraints. Future work will focus on extending the methodology to control unstable open-loop systems, handling state-constraints, and solving set-point tracking problems.
Proof of Lemma 1
The proof relies on the following classical error bound for linear systems.
Lemma 2 (Hoffman [21]). Let and . There exists , depending only on and , such that for all and for which the polyhedron
is nonempty, and for all ,
where denotes the componentwise positive part.
Proof of Lemma 1: Fix and write and for the active set. Consider
Every is a KKT multiplier pair associated with : (27) and hold by definition, (28) holds since , and (29) holds since whenever . Conversely, since is polyhedral, the KKT conditions (27) are necessary for to solve [6], so a KKT pair exists, and by (29) it lies in . Hence .
With , the set has the form of Lemma 2 with
where collects the rows of the identity indexed by . The matrices and depend only on , , and ; the state enters only through the right-hand side . Let be the corresponding Hoffman constant.
At the inequality constraints are satisfied, , and the equality residual is . Lemma 2 therefore gives
Since is nonempty and closed, it contains an element of minimal norm, which is a KKT multiplier pair by Step 1 and satisfies
The active set varies with , but it is a subset of the finitely many constraint indices. Hence is independent of , and
Finally, is continuous with respect to (see (9)), hence bounded on the compact set , which gives (30) with .
References
References
- [1]H. Chen, Y. Li, R. H. Louie, and B. Vucetic, “Autonomous demand side management based on energy consumption scheduling and instantaneous load billing: An aggregative game approach,” IEEE Transactions on Smart Grid, vol. 5, no. 4, pp. 1744–1754, 2014.DOI
- [2]Z. Ma, D. S. Callaway, and I. A. Hiskens, “Decentralized charging control of large populations of plug-in electric vehicles,” IEEE Transactions on control systems technology, vol. 21, no. 1, pp. 67–78, 2011.
- [3]F. Facchinei and C. Kanzow, “Generalized Nash equilibrium problems,” 4OR, vol. 5, no. 3, pp. 173–210, 2007.
- [4]A. Fischer, M. Herrich, and K. Schönfeld, “Generalized Nash equilibrium problems-recent advances and challenges,” Pesquisa Operacional, vol. 34, no. 3, pp. 521–558, 2014.DOI
- [5]A. A. Kulkarni and U. V. Shanbhag, “On the variational equilibrium as a refinement of the generalized Nash equilibrium,” Automatica, vol. 48, no. 1, pp. 45–55, 2012.DOI
- [6]J.-S. Pang and F. Facchinei, Finite-dimensional variational inequalities and complementarity problems : vol. 1, ser. Springer series in operations research. New York, Berlin, Heidelberg: Springer, 2003.
- [7]W. van den Broek, “Moving horizon control in dynamic games,” Journal of Economic Dynamics and Control, vol. 26, no. 6, pp. 937–961, 2002.
- [8]S. Hall, L. Guerrini, F. Dörfler, and D. Liao-McPherson, “Receding horizon games for modeling competitive supply chains,” IFAC-PapersOnLine, vol. 58, no. 18, pp. 8–14, 2024.arxiv.org/abs/2401.09853
- [9]E. Benenati and S. Grammatico, “Kinematic, “Linear-quadratic dynamic games as receding-horizon variational inequalities,” IEEE Transactions on Automatic Control, vol. 71, no. 4, pp. 2404–2417, 2026.arxiv.org/abs/2408.15703
- [10]S. Hall, G. Belgioioso, F. Dörfler, and D. Liao-McPherson, “Stability certificates for receding-horizon games,” IEEE Transactions on Automatic Control, vol. 71, no. 5, pp. 3511–3518, 2026.arxiv.org/abs/2404.12165
- [11]F. M. Hante and A. Topalovic, “Stabilizing GNEP-based model predictive control: Quasi-GNEPs and end constraints,” Optimization Online preprint, 2024.
- [12]——, “Stabilizing model predictive control for generalized Nash equilibrium problems using approximation by α-quasi-GNEPs and Lyapunov end cost,” in 2025 European Control Conference (ECC), 2025, pp. 2427–2432.DOI
- [13]E. Dimou, F. Fabiani, and A. Bemporad, “Receding horizon games with stability guarantees,” 65th IEEE Conference on Decision and Control, 2026, available on line at https://cse.lab.imtlucca.it/~bemporad/publications/papers/cdc26_gtmpc.pdf.
- [14]K. P. Wabersich and M. N. Zeilinger, “A predictive safety filter for learning-based control of constrained nonlinear dynamical systems,” Automatica, vol. 129, p. 109597, 2021.
- [15]K. P. Wabersich, A. J. Taylor, J. J. Choi, K. Sreenath, C. J. Tomlin, A. D. Ames, and M. N. Zeilinger, “Data-driven safety filters: Hamilton-Jacobi reachability, control barrier functions, and predictive methods for uncertain systems,” IEEE Control Systems Magazine, vol. 43, no. 5, pp. 137–177, 2023.DOI
- [16]K.-C. Hsu, H. Hu, and J. F. Fisac, “The safety filter: A unified view of safety-critical control in autonomous systems,” Annual Review of Control, Robotics, and Autonomous Systems, vol. 7, pp. 47–72, 2024.arxiv.org/abs/2309.05837
- [17]A. D. Ames, X. Xu, J. W. Grizzle, and P. Tabuada, “Control barrier function based quadratic programs for safety critical systems,” IEEE Transactions on Automatic Control, vol. 62, no. 8, pp. 3861–3876, 2017.DOI
- [18]A. D. Ames, K. Galloway, K. Sreenath, and J. W. Grizzle, “Rapidly exponentially stabilizing control Lyapunov functions and hybrid zero dynamics,” IEEE Transactions on Automatic Control, vol. 59, no. 4, pp. 876–891, 2014.DOI
- [19]T. Tataraneko, W. Shi, and A. Nedić, “Geometric convergence of gradient play algorithms for distributed Nash equilibrium seeking,” IEEE Transactions on Automatic Control, vol. 66, no. 11, pp. 5342–5353, 2021.
- [20]T. Tataraneko and M. Kangkampour, “Convergence rate of payoff-based generalized Nash equilibrium learning,” European Journal of Control, open access, 2025.
- [21]A. J. Hoffman, “On approximate solutions of systems of linear inequalities,” Journal of Research of the National Bureau of Standards, vol. 49, no. 4, pp. 263–265, 1952.
- [22]A. Bemporad and T. Tateraneko, “Solving monotone linear-quadratic generalized Nash equilibrium problems via quadratic programming,” arXiv preprint 2608.07336, 2026.arxiv.org/abs/2608.07336
- [23]A. Bemporad, “NashOpt: A Python library for computing generalized Nash equilibria and game design,” arXiv preprint 2512.23636, 2025, https://github.com/bemporad/nashopt.
- [24]P. J. Goulart and Y. Chen, “Clarabel: An interior-point solver for conic programs with quadratic objectives,” Mathematical Programming Computation, pp. 1–83, 2026.arxiv.org/abs/2405.12762


