Description du projet :
Over the last decade, the electric grid changed dramatically with the introduction of distributed and volatile power production means, such as private solar panels. Thus,Distribution System Operators (DSOs) are facing daily challenges to handle frequent grid operational limit violations (GOLV). Additionally, installed grid measurement devices, though they provide huge amount of data, give only a partial monitoring of Low Voltage (LV) grids.
Consequently, interpretation of these data is hard, and decision even harder.
Innovative part of this project is to valorize available data to DSOs through a data-driven, and automated process. It aims at providing necessary information and analysis for secure and optimal operation of distribution grids using descriptive, diagnostic, predictive, and
preventive data analytics applications.
Key needed researches:
Development of data-driven algorithms based on machine learning techniques aiming
at 1) identification of main trends, correlations and irregularities on voltage/current/power profiles of LV grids, 2) complex event identification regarding several correlated GOLVs/events, 3) prediction of GOLVs few hours ahead of real-time, and 4) proposition of control actions for flexible resources to prevent GOLVs.
Study and modeling of data orchestration for harmonizing different input data from various sources (e.g., measurement devices, smart meters, etc.) to be used in the abovementioned algorithms and to provide a fast and efficient environment for running the algorithms.
To develop a big-data platform with above data-driven applications, called 'Grid Data Digger', various competencies from different engineering fields, including computer science, machine learning, embedded software, and power systems are brought together.
DEPsys, as the main implementation partner, will commercialize the 'Grid Data Digger' platform and sell it to its new and existing DSO customers. The income increase is estimated as 8 MCHF by 2022.
Forschungsteam innerhalb von HES-SO:
Dassatti Alberto, Fatemi Nastaran, Hochet Guillaume, Gavin Serge, Pena Carlos Andrés, Burella Pérez Julián Mariano, Kissling Simon, Meier Christopher, Bürer Mary Jean, Bozorg Mokhtar, Fesefeldt Marten, Rodriguez Zaloña Oscar, Paruta Paola, Ataee Shabnam, Carpita Mauro
Partenaires académiques: IICT; IESE; Rodriguez Zalona Oscar, ReDS
Durée du projet:
08.05.2019 - 31.05.2021
Montant global du projet: 225'379 CHF
Statut: Laufend
SWITCH edu-ID
Verwaltung
