-
Linear quadratic control of nonlinear systems with Koopman operator learning and the Nyström method
Authors:
Edoardo Caldarelli,
Antoine Chatalic,
Adrià Colomé,
Cesare Molinari,
Carlos Ocampo-Martinez,
Carme Torras,
Lorenzo Rosasco
Abstract:
In this paper, we study how the Koopman operator framework can be combined with kernel methods to effectively control nonlinear dynamical systems. While kernel methods have typically large computational requirements, we show how random subspaces (Nyström approximation) can be used to achieve huge computational savings while preserving accuracy. Our main technical contribution is deriving theoretic…
▽ More
In this paper, we study how the Koopman operator framework can be combined with kernel methods to effectively control nonlinear dynamical systems. While kernel methods have typically large computational requirements, we show how random subspaces (Nyström approximation) can be used to achieve huge computational savings while preserving accuracy. Our main technical contribution is deriving theoretical guarantees on the effect of the Nyström approximation. More precisely, we study the linear quadratic regulator problem, showing that the approximated Riccati operator converges at the rate $m^{-1/2}$, and the regulator objective, for the associated solution of the optimal control problem, converges at the rate $m^{-1}$, where $m$ is the random subspace size. Theoretical findings are complemented by numerical experiments corroborating our results.
△ Less
Submitted 3 March, 2025; v1 submitted 5 March, 2024;
originally announced March 2024.
-
Approximate Wasserstein Attraction Flows for Dynamic Mass Transport over Networks
Authors:
Ferran Arqué,
César A. Uribe,
Carlos Ocampo-Martinez
Abstract:
This paper presents a Wasserstein attraction approach for solving dynamic mass transport problems over networks. In the transport problem over networks, we start with a distribution over the set of nodes that needs to be "transported" to a target distribution accounting for the network topology. We exploit the specific structure of the problem, characterized by the computation of implicit gradient…
▽ More
This paper presents a Wasserstein attraction approach for solving dynamic mass transport problems over networks. In the transport problem over networks, we start with a distribution over the set of nodes that needs to be "transported" to a target distribution accounting for the network topology. We exploit the specific structure of the problem, characterized by the computation of implicit gradient steps, and formulate an approach based on discretized flows. As a result, our proposed algorithm relies on the iterative computation of constrained Wasserstein barycenters. We show how the proposed method finds approximate solutions to the network transport problem, taking into account the topology of the network, the capacity of the communication channels, and the capacity of the individual nodes. Finally, we show the performance of this approach applied to large-scale water transportation networks.
△ Less
Submitted 26 April, 2022; v1 submitted 19 September, 2021;
originally announced September 2021.
-
Operationally-Safe Peer-to-Peer Energy Trading in Distribution Grids: A Game-Theoretic Market-Clearing Mechanism
Authors:
Giuseppe Belgioioso,
Wicak Ananduta,
Sergio Grammatico,
Carlos Ocampo-Martinez
Abstract:
In future distribution grids, prosumers (i.e., energy consumers with storage and/or production capabilities) will trade energy with each other and with the main grid. To ensure an efficient and safe operation of energy trading, in this paper, we formulate a peer-to-peer energy market of prosumers as a generalized aggregative game, in which a network operator is only responsible for the operational…
▽ More
In future distribution grids, prosumers (i.e., energy consumers with storage and/or production capabilities) will trade energy with each other and with the main grid. To ensure an efficient and safe operation of energy trading, in this paper, we formulate a peer-to-peer energy market of prosumers as a generalized aggregative game, in which a network operator is only responsible for the operational constraints of the system. We design a distributed market-clearing mechanism with convergence guarantee to an economically-efficient and operationally-safe configuration (i.e., a variational generalized Nash equilibrium). Numerical studies on the IEEE 37-bus testcase show the scalability of the proposed approach and suggest that active participation in the market is beneficial for both prosumers and the network operator.
△ Less
Submitted 25 March, 2022; v1 submitted 28 July, 2021;
originally announced July 2021.
-
Distributed Augmented Lagrangian Method for Link-Based Resource Sharing Problems of Multi-Agent Systems
Authors:
Wicak Ananduta,
Angelia Nedić,
Carlos Ocampo-Martinez
Abstract:
A multi-agent optimization problem motivated by the management of energy systems is discussed. The associated cost function is separable and convex although not necessarily strongly convex and there exist edge-based coupling equality constraints. In this regard, we propose a distributed algorithm based on solving the dual of the augmented problem. Furthermore, we consider that the communication ne…
▽ More
A multi-agent optimization problem motivated by the management of energy systems is discussed. The associated cost function is separable and convex although not necessarily strongly convex and there exist edge-based coupling equality constraints. In this regard, we propose a distributed algorithm based on solving the dual of the augmented problem. Furthermore, we consider that the communication network might be time-varying and the algorithm might be carried out asynchronously. The time-varying nature and the asynchronicity are modeled as random processes. Then, we show the convergence and the convergence rate of the proposed algorithm under the aforementioned conditions.
△ Less
Submitted 28 February, 2021;
originally announced March 2021.
-
Accelerated Multi-Agent Optimization Method over Stochastic Networks
Authors:
Wicak Ananduta,
Carlos Ocampo-Martinez,
Angelia Nedić
Abstract:
We propose a distributed method to solve a multi-agent optimization problem with strongly convex cost function and equality coupling constraints. The method is based on Nesterov's accelerated gradient approach and works over stochastically time-varying communication networks. We consider the standard assumptions of Nesterov's method and show that the sequence of the expected dual values converge t…
▽ More
We propose a distributed method to solve a multi-agent optimization problem with strongly convex cost function and equality coupling constraints. The method is based on Nesterov's accelerated gradient approach and works over stochastically time-varying communication networks. We consider the standard assumptions of Nesterov's method and show that the sequence of the expected dual values converge toward the optimal value with the rate of $\mathcal{O}(1/k^2)$. Furthermore, we provide a simulation study of solving an optimal power flow problem with a well-known benchmark case.
△ Less
Submitted 16 December, 2020; v1 submitted 8 September, 2020;
originally announced September 2020.
-
Primal-dual evolutionary dynamics for constrained population games
Authors:
Juan Martinez-Piazuelo,
Nicanor Quijano,
Carlos Ocampo-Martinez
Abstract:
Population games can be regarded as a tool to study the strategic interaction of a population of players. Although several attention has been given to such field, most of the available works have focused only on the unconstrained case. That is, the allowed equilibrium of the game is not constrained. To further extend the capabilities of population games, in this paper we propose a novel class of p…
▽ More
Population games can be regarded as a tool to study the strategic interaction of a population of players. Although several attention has been given to such field, most of the available works have focused only on the unconstrained case. That is, the allowed equilibrium of the game is not constrained. To further extend the capabilities of population games, in this paper we propose a novel class of primal-dual evolutionary dynamics that allow the consideration of constraints that must be satisfied at the equilibrium of the game. Using duality theory and Lyapunov stability theory, we provide sufficient conditions to guarantee the asymptotic stability and feasibility of the equilibria set of the game under the considered constraints. Furthermore, we illustrate the application of the developed theory to some classical population games with the addition of constraints.
△ Less
Submitted 20 August, 2020;
originally announced August 2020.
-
Event-triggered Partitioning for Non-centralized Predictive-Control-based Economic Dispatch of Interconnected Microgrids: Technical Report
Authors:
Wicak Ananduta,
Carlos Ocampo-Martinez
Abstract:
A non-centralized model predictive control (MPC) scheme for solving an economic dispatch problem of electrical networks is proposed in this paper. The scheme consists of two parts. The first part is an event-triggered repartitioning method that splits the network into a fixed number of non-overlapping sub-systems {(microgrids)}. The objective of the repartitioning procedure is to obtain self-suffi…
▽ More
A non-centralized model predictive control (MPC) scheme for solving an economic dispatch problem of electrical networks is proposed in this paper. The scheme consists of two parts. The first part is an event-triggered repartitioning method that splits the network into a fixed number of non-overlapping sub-systems {(microgrids)}. The objective of the repartitioning procedure is to obtain self-sufficient microgrids, i.e., those that can meet their local loads using their own generation units. However, since the algorithm does not guarantee that all the resulting microgrids are self-sufficient, the microgrids that are not self-sufficient must then form a coalition with some of their neighboring microgrids. This process becomes the second part of the scheme. By performing the coalition formation, we can decompose the economic dispatch problem of the network into coalition-based sub-problems such that each subproblem is feasible. Furthermore, we also show that the solution obtained by solving the coalition-based sub-problems is a feasible but sub-optimal solution to the centralized problem. Additionally, some numerical simulations are also carried out to show the effectiveness of the proposed method.
△ Less
Submitted 15 July, 2020;
originally announced July 2020.
