Development of a real-time framework for parallel data stream processing⋆

Abstract

Abstract This paper presents the Korean Superconducting Tokamak Advanced Research (KSTAR) real-time framework for the parallel data stream processing framework (RT-ParaPro). RT-ParaPro is a framework used to develop a program that simultaneously processes data streamed over a real-time network, sends the data over a network, and archives the data in real-time. In most fusion experimental devices, each device processes the data needed for real-time control and transmits them to the other real-time systems in real-time via the network. By using RT-ParaPro, it is possible to simplify the configuration of a program that performs a series of processes and shorten the development time. Unlike other real-time frameworks that focus on real-time control, RT-ParaPro is specialized for the parallel data stream processing and transmission of data over a real-time network. By using this framework, the L-H transition detection system using machine learning (LHML), which determines whether plasma is in low-confinement mode (L-mode) or high-confinement mode (H-mode) in real-time by using machine learning, and the reflective memory (RFM) archiving system, which stores various RFM channel data to MDSplus, has been developed and operated in KSTAR. To evaluate the real-time performance of this framework, we tested the consistency of the thread period by varying the period of the thread. The test results show that the thread control period is consistent. The period of the thread has a jitter of about 8 μ s not only at a low control cycle rate (1 kHz) but also at a control cycle rate of 100 kHz.