A distributed predictive control approach for periodic flow-based networks: application to drinking water systems

Abstract

This paper proposes a distributed model predictive control (MPC) approach designed to work in a cooperative manner for controlling flow-based networks showing periodic behaviours. Under this distributed approach, local controllers cooperate in order to enhance the performance of the whole flow network avoiding the use of a coordination layer. Alternatively, controllers use both the monolithic model of the network and the given global cost function to optimize the control inputs of the local controllers but taking into account the effect of their decisions over the remainder subsystems conforming the entire network. In this sense, a global (all-to-all) communication strategy is considered. Although the Pareto optimality cannot be reached due to the existence of non-sparse coupling constraints, the asymptotic convergence to a Nash equilibrium is guaranteed. The resultant strategy is tested and its effectiveness is shown when applied to a large-scale complex flow-based network: the Barcelona drinking water supply system.

A framework for receding-horizon control in infinite-horizon aggregative games

“…For this reason -especially when addressing large-population noncooperative settings-applications of the game-theoretical framework on (model predictive) control schemes have considered noncooperative mechanisms as a means to devise distributed control laws (Scattolini, 2009;Li and Marden, 2013;Christofides et al, 2013). Starting from the work of van den Broek (2002) on receding horizon solutions for a linear quadratic game, several distributed MPC applications of Nash games have been proposed, amongst others, for robotic formation control (Gu, 2008), water distribution networks (Ramirez-Jaime et al, 2016;Grosso et al, 2017), freeway traffic control (Pisarski and de Wit, 2016), and economic process optimization (Lee and Angeli, 2014).…”

A non-centralized predictive control strategy for wind farm active power control: A wake-based partitioning approach

“…where l = l + 1 and f q (q = 1, 2, 3) are given in (6), 7and (9). As a consequence of using the auxiliary Boolean variable δ i jl , constraints (10d)-(10f) must be added to the original problem (5), [27] and the original no-empty and exclusive constraints in (5b) and (5c) must be rewritten as (10b) and (10c), respectively.…”

“…In the current work, a similar approach is presented assuming that the air flow within a wind farm can be modeled as a simplified flow-based distribution network. Many engineering systems have been modeled as flow-based distribution systems [9,26], which consist of several elements of diverse nature, e.g., storage, actuator, joint, sink, source and flow. Unlike other energy sources, wind cannot be stored, and hence the wind flow in a farm can be obtained identifying only the following elements:…”

“…This idea has been applied to several large-scale systems, e.g., traffic systems, energy systems, smart grids and water systems [7,8,9]. The application of partitioning techniques for controlling wind farms in a decentralized manner has been started only recently [10,11,12].…”

Accelerated Multi-Agent Optimization Method over Stochastic Networks

“…The advancement on information, computation and communication technologies promotes the deployment of distributed approaches to solve complex large-scale problems, e.g., in power networks [1], [2] and water networks [3]. On one hand, such approaches offer flexibility and scalability.…”