Distributed Model Predictive Control For Demand Response On Thermal Devices in Building Blocks

Menon, Ramanunni Parakkal; Amblard, Frédéric; Page, Jessen

doi:10.1109/isgteurope.2019.8905624

Cited by 1 publication

(1 citation statement)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Braun et al [12] and Worthmann et al [16] use hierarchical model predictive control to coordinate the DO of different agents, and Menon et al [17] use distributed model predictive control for demand response. The authors of [8,9,18,19] use approaches where no central control unit is present.…”

Section: Related Workmentioning

confidence: 99%

Demand response through decentralized optimization in residential areas with wind and photovoltaics

Jochem

Fïchtner

2021

Energy

View full text Add to dashboard Cite

A paradigm shift has to be realized in future energy systems with high shares of renewable energy sources. The electrical demand has to react to the fluctuating electricity generation of renewable energy sources. To this end, flexible electrical loads like electric heating devices coupled with thermal storage or electric vehicles are necessary in combination with optimization approaches. In this paper, we develop a novel privacy-preserving approach for decentralized optimization to exploit load flexibility. This approach, which is based on a set of schedules, is referred to as SEPACO-IDA. The results show that our developed algorithm outperforms the other approaches for scheduling based decentralized optimization found in the literature. Furthermore, this paper clearly illustrates the suboptimal results for uncoordinated decentralized optimization and thus the strong need for coordination approaches. Another contribution of this paper is the development and evaluation of two methods for distributing a central wind power profile to the local optimization problem of distributed agents (Equal Distribution and Score-Rank-Proportional Distribution). These wind profile assignment methods are combined with different decentralized optimization approaches. The results reveal the dependency of the best wind profile assignment method on the used decentralized optimization approach.

show abstract