-
Quality-oriented monitoring task management
The SmartC2Net monitoring system takes grid controller requirements and actual communication network Quality-of-Service (QoS) metrics (delay, loss, etc.) into account to deliver the best monitoring data quality possible. To do so, it can dynamically change monitoring configurations (access techniques, update frequencies, protocols, etc.) during runtime to react to uncontrollable QoS changes (e.g. congestion, link overload, link failure, etc.). In addition, the adaptive reconfigurations consider priorities of monitoring variables as well as the overall data quality of the full set of monitoring tasks. In case a monitoring task cannot be executed to deliver a sufficient data quality, the issue is reliably reported to the requesting controller.
-
Fault detection and prediction algorithms
The SmartC2Net fault management system has the aim to detect (possibly predict) and localize faults that can happen in the smart grid ICT domain. Since it exploits a highly cross-technology based approach on the events and monitoring traces that can be collected in the system where it is installed, it can also also be adapted to detect grid domain fault scenarios. The framework reveals anomalies to support online diagnosis activities of complex systems. This is a promising approach w.r.t. traditional detection mechanisms (e.g., based on heartbeats) that can be inadequate or even cannot be applied. The fault management implements a configurable detection framework to reveal faults based on online statistical analysis techniques, and it is designed for systems that operate under variable and non-stationary conditions. This allows to continuously track and predict system behaviours, thus reducing the probability of false alarms and improving the detection accuracy.
-
Grid demand inference from cellular network monitoring
The solution is able to forecast the short time frame energy demand fluctuation with a precision higher than current methods. Currently the available energy providers’ infrastructures are not able to estimate and predict real-time fluctuation of the energy demand and are not scalable enough to integrate, with low cost and effort, hardware elements able to estimate energy demand in real-time. Moreover new kinds of energy demand profiles, such as the ones coming from EVs, nomadic users in high density urban areas or diffusion of conditioning appliances due to climate change, are also affecting validity of historical energy demand statistics used by DSOs to predict at real time the short time frame energy demand. Instead of using statistical data coming from DSO, the solution is based on innovative proxies of the energy demand such as the real time people distribution and the ability to measure new type of consumption, such as EVs. Distributor System Operators can use these estimation to self-manage the energy demand, distribution and storage at real-time, without any user intervention.
For more details, please see public deliverables: D2.1, D2.2 and D2.3.
