Use of efficient task allocation algorithm for parallel real-time EMT simulation

• Published in: Electric power systems research Vol. 189; p. 106604
• Main Authors: Bruned, Boris, Rault, Pierre, Dennetière, Sébastien, Martins, Ian Menezes
• Format: Journal Article
• Language: English
• Published: Amsterdam Elsevier B.V 01.12.2020; Elsevier Science Ltd
• Subjects: Algorithms; Computer peripherals; Electricity distribution; Electromagnetism; Graph partitioning; Hardware-in-the-loop; Hardware-in-the-loop simulation; Microprocessors; Operations research; Optimization; Parallel processing; Parallel simulation; Real time; Real-time simulation; Simulation; Task allocation problem; Graph partitioning; Parallel simulation; Task allocation problem; Hardware-in-the-loop; Real-time simulation; Optimization
• ISSN: 0378-7796; 1873-2046
• DOI: 10.1016/j.epsr.2020.106604

•Description of efficient task allocation algorithm from operational research techniques for parallel real-time EMT simulation.•Performances asserted on very large networks and quality of solution verified with exact resolution methods.•Validation of each mapping strategy through a real EMT case study with real-time Hardware-in-the-Loop simulation. Real-time EMT (Electromagnetic Transients) simulation relies on multi-cores computers to accelerate the simulation through parallelization. It also increases simulation accuracy by allowing a lower time step. First, the network has to be split into several tasks using a separation technique. Then, each task has to be allocated/mapped to a processor. This paper focuses on this problem which can be formulated as a TAP (Task Allocation Problem). To find optimal task allocation, operational research techniques can be used. Heuristics such as graph partitioning allow getting fast solutions. Their performances are asserted with very large networks and real-time simulator architectures, both from TSO (Transmission System Operator) grids. Exact resolution methods are used to verify solution quality. The validation of each task mapping strategy is done through a real EMT case study which involves real-time Hardware-in-the-Loop simulation.