Distributed transactive control in distribution systems with microgrids

Microgrids are considered as a cornerstone in the evolution to a smarter grid. However, this evolution brings some critical challenges for the control in a real-time implementation. We present two control algorithms to operate a power system with microgrids and other two to operate microgrids in ord...

Full description

Autores:
Baron Prada, Eder David
Tipo de recurso:
Fecha de publicación:
2018
Institución:
Universidad Nacional de Colombia
Repositorio:
Universidad Nacional de Colombia
Idioma:
spa
OAI Identifier:
oai:repositorio.unal.edu.co:unal/68807
Acceso en línea:
https://repositorio.unal.edu.co/handle/unal/68807
http://bdigital.unal.edu.co/70045/
Palabra clave:
0 Generalidades / Computer science, information and general works
62 Ingeniería y operaciones afines / Engineering
Distributed Control
Distributed Optimization
Transactive Control
Projection Algorithms
Asynchronous Algorithms
Game Theory
Microgrids
Control distribuido
Optimización distribuida
Control transaccional
Algoritmos de proyección
Algoritmos asíncronos
Teoría de juegos
Rights
openAccess
License
Atribución-NoComercial 4.0 Internacional
Description
Summary:Microgrids are considered as a cornerstone in the evolution to a smarter grid. However, this evolution brings some critical challenges for the control in a real-time implementation. We present two control algorithms to operate a power system with microgrids and other two to operate microgrids in order to reach the optimal social welfare. We consider three types of agents: photovoltaic generators, conventional generators and smart loads. These agents can be aggregated into a microgrid or interact directly in the power system depending on their power. To optimize the microgrids, we use two strategies. First one is based on projected consensus algorithm, where each agent iteratively optimizes its local utility function based on local information obtained from its neighbors and global information obtained through a distributed finite-time average algorithm. The second one is based on populations game theory; specifically we use a centralized replicator dynamics where a central agent iteratively optimizes the system status. To optimize the whole power system we use two strategies, first an asynchronous algorithm based on primal-dual optimization is proposed, where we consider that agents update the primal variables and a "virtual agent" updates the dual variables. Our last algorithm is a distributed transactive control algorithm based on populations games to dynamically manage the distributed generators and smart loads in the system to reach the optimum social welfare. Agents are considered non-cooperative, and they are individually incentive-driven. The proposed algorithm preserve stability while guarantee optimality conditions considering several constraints in the system on the real-time operation. We show numerical results of the proposed control strategies.