Data stream processing meets the Advanced Metering Infrastructure: possibilities, challenges and applications
Licentiate thesis, 2020
Energy production is increasingly distributed, weather dependent and located in the distribution network, close to consumers.
Energy consumption is increasing throughout society and the electrification of transportation is driving distribution networks closer to the limits.
Operating the networks closer to their limits also increases the risk for faults.
Continuous monitoring of the distribution network closest to the customers is needed in order to mitigate this risk.
The Advanced Metering Infrastructure introduced smart meters throughout the distribution network.
Data stream processing is a computing paradigm that offers low latency results from analysis on large volumes of the data.
This thesis investigates the possibilities and challenges for continuous monitoring that are created when the Advanced Metering Infrastructure and data stream processing meet.
The challenges that are addressed in the thesis are efficient processing of unordered (also called out-of-order) data and efficient usage of the computational resources present in the Advanced Metering Infrastructure.
Contributions towards more efficient processing of out-of-order data are made with eChIDNA and TinTiN. Both are systems that utilize knowledge about smart meter data to directly produce results where possible and storing only data that is relevant for late data in order to produce updated results when such late data arrives. eChIDNA is integrated in the streaming query itself, while TinTiN is a streaming middleware that can be applied to streaming queries in order to make them resilient against out-of-order data.
Eventual determinism is defined in order to formally investigate the deterministic properties of output produced by such systems.
Contributions towards efficient usage of the computational resources of the Advanced Metering Infrastructure are made with the application LoCoVolt.
LoCoVolt implements a monitoring algorithm that can run on equipment that is localized in the communication infrastructure of the Advanced Metering Infrastructure and can take advantage of the overlap between the communication and distribution networks.
All contributions are evaluated on hardware that is available in current AMI systems, using large scale data obtained from a real production AMI.
Data stream processing
Advanced Metering Infrastructure
Joris Van Rooij
Chalmers, Computer Science and Engineering (Chalmers), Networks and Systems (Chalmers)
eChIDNA: Continuous Data Validation in Advanced Metering Infrastructures
2018 IEEE International Energy Conference (ENERGYCON),; (2018)p. 1-6
Paper in proceeding
TinTiN: Travelling in time (if necessary) to deal with out-of-order data in streaming aggregation
DEBS 2020 - Proceedings of the 14th ACM International Conference on Distributed and Event-Based Systems,; (2020)p. 141-152
Paper in proceeding
LoCoVolt: Distributed Detection of Broken Meters in Smart Grids through Stream Processing
DEBS '18 Proceedings of the 12th ACM International Conference on Distributed and Event-based Systems,; (2018)p. 171-182
Paper in proceeding
HARE: Self-deploying and Adaptive Data Streaming Analytics in Fog Architectures
Swedish Research Council (VR) (2016-03800), 2017-01-01 -- 2020-12-31.
STAMINA - WASP
Wallenberg AI, Autonomous Systems and Software Program, 2016-04-04 -- 2020-04-06.
INDEED: Information and Data-processing in Focus for Energy Efficiency
Chalmers, 2020-01-01 -- .
Future factories in the Cloud (FiC)
Swedish Foundation for Strategic Research (SSF) (GMT14-0032), 2016-01-01 -- 2020-12-31.
Integrated cyber-physical solutions for intelligent distribution grid with high penetration of renewables (UNITED-GRID)
European Commission (EC) (EC/H2020/773717), 2017-11-01 -- 2020-04-30.
STAMINA - GE
Göteborg Energi, Foundation for Research and Developmen, 2017-01-01 -- 2021-12-31.
Areas of Advance
Opponent: David Eyers, University of Ontago, Dunedin, New Zealand