1,721,061 research outputs found
Stochastic Generalized Nash Equilibrium-Seeking in Merely Monotone Games
Full text linkWe solve the stochastic generalized Nash equilibrium (SGNE) problem in merely monotone games with expected value cost functions. Specifically, we present the first distributed SGNE-seeking algorithm for monotone games that require one proximal computation (e.g., one projection step) and one pseudogradient evaluation per iteration. Our main contribution is to extend the relaxed forward–backward operator splitting by the Malitsky (Mathematical Programming, 2019) to the stochastic case and in turn to show almost sure convergence to an SGNE when the expected value of the pseudogradient is approximated by the average over a number of random samples.Green Open Access added to TU Delft Institutional Repository 'You share, we take care!' - Taverne project https://www.openaccess.nl/en/you-share-we-take-care Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public.Team Sergio GrammaticoTeam Bart De Schutte
An incentive mechanism for agents playing competitive aggregative games
No full textWe propose an incentive mechanism for steering the strategies of noncooperative heterogeneous agents, each with strongly convex cost function depend- ing on the average among the agents’ strategies, and all sharing a convex constraint, toward a competitive aggregative equilibrium. We consider a co- ordinator agent having access to the average among the agents’ strategies and broadcasting incentive signals that affect the decentralized optimal responses of the agents. Our mechanism ensures, based on the Picard–Banach fixed point iteration, global convergence to an equilibrium
A distributed forward-backward algorithm for stochastic generalized Nash equilibrium seeking
No full textWe consider the stochastic generalized Nash equilibrium problem (SGNEP) with expected-value cost functions. Inspired by Yi and Pavel (2019), we propose a distributed generalized Nash equilibrium seeking algorithm based on the preconditioned forward-backward operator splitting for SGNEPs, where, at each iteration, the expected value of the pseudogradient is approximated via a number of random samples. Our main contribution is to show almost sure convergence of the proposed algorithm if the pseudogradient mapping is restricted (monotone and) cocoercive.Accepted Author ManuscriptTeam Bart De Schutte
Achieving a large domain of attraction with short-horizon linear MPC via polyhedral Lyapunov functions
Full text linkPolyhedral control Lyapunov functions (PCLFs) are exploited in this paper to propose a linear model predictive control (MPC) formulation that guarantees a “large” domain of attraction (DoA) even for short horizon. In particular, the terminal region of the proposed finite-horizon MPC formulation is chosen as a level set of an appropriate PCLF. For small dimensional systems, this terminal region can be explicitly computed as an arbitrarily close approximation to the entire (infinite-horizon) stabilizable set. Global stability of the origin is guaranteed by using an “inflated” PCLF as terminal cost. The proposed MPC scheme can be formulated as a (small dimensional) quadratic programming problem by introducing one additional scalar variable. Numerical examples show the main benefits and achievements of the proposed formulation in terms of trade-off between volume of the DoA, computational time and closed-loop performance
Training Generative Adversarial Networks via Stochastic Nash Games
No full textGenerative adversarial networks (GANs) are a class of generative models with two antagonistic neural networks: a generator and a discriminator. These two neural networks compete against each other through an adversarial process that can be modeled as a stochastic Nash equilibrium problem. Since the associated training process is challenging, it is fundamental to design reliable algorithms to compute an equilibrium. In this article, we propose a stochastic relaxed forward-backward (SRFB) algorithm for GANs, and we show convergence to an exact solution when an increasing number of data is available. We also show convergence of an averaged variant of the SRFB algorithm to a neighborhood of the solution when only a few samples are available. In both cases, convergence is guaranteed when the pseudogradient mapping of the game is monotone. This assumption is among the weakest known in the literature. Moreover, we apply our algorithm to the image generation problem.</p
Stochastic generalized Nash equilibrium seeking under partial-decision information
Full text linkWe consider for the first time a stochastic generalized Nash equilibrium problem, i.e., with expected-value cost functions and joint feasibility constraints, under partial-decision information, meaning that the agents communicate only with some trusted neighbors. We propose several distributed algorithms for network games and aggregative games that we show being special instances of a preconditioned forward–backward splitting method. We prove that the algorithms converge to a generalized Nash equilibrium when the forward operator is restricted cocoercive by using the stochastic approximation scheme with variance reduction to estimate the expected value of the pseudogradient.Team Sergio GrammaticoTeam Bart De Schutte
Guest Editorial: Introduction to IEEE Control Systems Letters Special Section on Multi-Agent Coordination for Energy Systems: From Model Based to Data-Driven Methods
Full text linkEditorialGreen Open Access added to TU Delft Institutional Repository 'You share, we take care!' - Taverne project https://www.openaccess.nl/en/you-share-we-take-care Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public.Team Sergio GrammaticoTeam Bart De Schutte
Mean Field Modeling of Large-Scale Energy Systems
No full textThis work proposes mean field game-type models for two instances of large- scale energy systems, namely plug-in electric vehicles and thermostatically controlled loads. Theoretical and numerical analysis show that both systems possess an equilibrium configuration which is optimal for the individuals and beneficial for the overall population
A new class of Lyapunov functions for the constrained stabilization of linear systems
No full textThe constrained stabilization of linear uncertain systems is investigated via the set-theoretic framework of control Lyapunov R-functions. A novel composition rule allows the design of a composite control Lyapunov function with external level set that exactly shapes the maximal controlled invariant set and inner sublevel sets arbitrarily close to any choice of smooth ones, generalizing both polyhedral and truncated ellipsoidal control Lyapunov functions. The feasibility test of the proposed smooth control Lyapunov functions can be cast into matrix inequalities conditions. The constrained linear quadratic control is addressed as an application
- …
