Distributed Optimal Heating Control of a Residential Building Resilient to Cybersecurity Issues

Abstract

Predictive control and optimization in buildings proved to be a promising approach in increasing energy efficiency of the sector as one of the largest energy consumer. Zone digitalization is still one of the ongoing issues in older buildings, and has only recently started to be interesting in the residential sector due to high prices of required expert knowledge and automation equipment. However, buildings systems digitalization and networking also brought to fore the system security issues. Distributed approach to zone digitalization and predictive control, enabled by recent advances in embedded technology, implies both hardware topology and control algorithm structure. The paper focuses on a case where each zone holds a separate controller with tailored temperature setpoint prediction and model predictive control algorithm, which independently calculate the optimal heating control laws of the corresponding zones. Furthermore, the controllers are mutually and iteratively bidding toward the joint energy efficiency goal of the whole building. Such control structure enables fast digitalization and optimal joint operation of the building while keeping the independency of the users and retaining the data privacy. Only essential data is transmitted to the central coordinator in form of a summed information, which cannot extrapolate particular user data. Additionally, single zone controller security breach does not inflict damage to the whole system. System resiliency to security issues is therefore strongly increased.