• Energy Research

Clustering strategies are becoming increasingly relevant to boost the scalability of distributed control methods by focusing the cooperation efforts on highly coupled agents. They are also relevant in systems where failing communication links and plug-and-play events are considered, which demand increased flexibility and modu larity. This article reviews commonalities and differences of those distributed strategies that exploit the degree of interaction between control agents to boost the mentioned properties, frequently leading to control structures where the communication network becomes a decision variable that may evolve dynamically. Taxonomies based on the control law employed, the criterion for selecting the network topology, its static/dynamical nature, the control architecture, and the provided theoretical properties, are given. Additionally, a review of applications in power networks, water systems, vehicle and traffic systems, renewable energy plants, and chemical processes is provided. Programa Español de Formación de Personal Académico FPU17/02653 Horizonte 2020 (Unión Europea) 789051 Ministerio de Economía y Competitividad ( España) DPI2017-86918-R Ministerio de Economía y Competitividad ( España) PID2020-119476RB-I00

Clustering strategies are becoming increasingly relevant to boost the scalability of distributed control methods by focusing the cooperation efforts on highly coupled agents. They are also relevant in systems where failing communication links and plug-and-play events are considered, which demand increased flexibility and modu larity. This article reviews commonalities and differences of those distributed strategies that exploit the degree of interaction between control agents to boost the mentioned properties, frequently leading to control structures where the communication network becomes a decision variable that may evolve dynamically. Taxonomies based on the control law employed, the criterion for selecting the network topology, its static/dynamical nature, the control architecture, and the provided theoretical properties, are given. Additionally, a review of applications in power networks, water systems, vehicle and traffic systems, renewable energy plants, and chemical processes is provided. Programa Español de Formación de Personal Académico FPU17/02653 Horizonte 2020 (Unión Europea) 789051 Ministerio de Economía y Competitividad ( España) DPI2017-86918-R Ministerio de Economía y Competitividad ( España) PID2020-119476RB-I00

This paper presents a novel clustering model predictive control technique where transitions to the best cooperation topology are planned over the prediction horizon. A new variable, the so-called transition horizon, is added to the optimization problem to calculate the optimal instant to introduce the next topology. Accordingly, agents can predict topology transitions to adapt their trajectories while optimizing their goals. Moreover, conditions to guarantee recursive feasibility and robust stability of the system are provided. Finally, the proposed control method is tested via a simulated eight-coupled tanks plant. Ministerio de Ciencia e Innovación FPU18{04476 Ministerio de Economía DPI2017-86918-R Ministerio de Economía DPI2015-67341-C02-01 Unión Europea No. 789051

This paper presents a novel clustering model predictive control technique where transitions to the best cooperation topology are planned over the prediction horizon. A new variable, the so-called transition horizon, is added to the optimization problem to calculate the optimal instant to introduce the next topology. Accordingly, agents can predict topology transitions to adapt their trajectories while optimizing their goals. Moreover, conditions to guarantee recursive feasibility and robust stability of the system are provided. Finally, the proposed control method is tested via a simulated eight-coupled tanks plant. Ministerio de Ciencia e Innovación FPU18{04476 Ministerio de Economía DPI2017-86918-R Ministerio de Economía DPI2015-67341-C02-01 Unión Europea No. 789051

Parabolic-trough solar collector fields are large-scale systems, so the application of centralized optimizationbased control methods to these systems is often not suitable for real-time control. As such, this paper formulates a novel coalitional control approach as an appropriate alternative to the centralized scheme. The key idea is to split the overall solar collector field into smaller subsystems, each of them governed by a local controller. Then, controllers are clustered into coalitions to solve a local optimization-based problem related to the corresponding subset of subsystems, so that an approximate solution of the original centralized problem can be obtained in a decentralized fashion. However, the operational constraints of the solar collector field couple the optimization problems of the multiple coalitions, thus limiting the ability to solve them in a fully decentralized manner. To overcome this issue, a novel population-dynamics-assisted resource allocation strategy is proposed as a mechanism to decouple the local optimization problems of the multiple coalitions. The proposed coalitional methodology allows to solve the multiple local subproblems in parallel, hence reducing the overall computational burden, while guaranteeing the satisfaction of the operational constraints and without significantly compromising the overall performance. The effectiveness of proposed approach is shown through numerical simulations of a 10- and 100-loop version of the ACUREX solar collector field of Plataforma Solar de Almería, Spain. Peer Reviewed

Parabolic-trough solar collector fields are large-scale systems, so the application of centralized optimizationbased control methods to these systems is often not suitable for real-time control. As such, this paper formulates a novel coalitional control approach as an appropriate alternative to the centralized scheme. The key idea is to split the overall solar collector field into smaller subsystems, each of them governed by a local controller. Then, controllers are clustered into coalitions to solve a local optimization-based problem related to the corresponding subset of subsystems, so that an approximate solution of the original centralized problem can be obtained in a decentralized fashion. However, the operational constraints of the solar collector field couple the optimization problems of the multiple coalitions, thus limiting the ability to solve them in a fully decentralized manner. To overcome this issue, a novel population-dynamics-assisted resource allocation strategy is proposed as a mechanism to decouple the local optimization problems of the multiple coalitions. The proposed coalitional methodology allows to solve the multiple local subproblems in parallel, hence reducing the overall computational burden, while guaranteeing the satisfaction of the operational constraints and without significantly compromising the overall performance. The effectiveness of proposed approach is shown through numerical simulations of a 10- and 100-loop version of the ACUREX solar collector field of Plataforma Solar de Almería, Spain. Peer Reviewed