Investigation of the J-TEXT plasma events by k-means clustering algorithm

Various types of plasma events emerge in specific parameter ranges and exhibit similar characteristics in diagnostic signals,which can be applied to identify these events.A semisupervised machine learning algorithm,the k-means clustering algorithm,is utilized to investigate and identify plasma events in the J-TEXT plasma.This method can cluster diverse plasma events with homogeneous features,and then these events can be identified if given few manually labeled examples based on physical understanding.A survey of clustered events reveals that the k-means algorithm can make plasma events(rotating tearing mode,sawtooth oscillations,and locked mode)gathering in Euclidean space composed of multi-dimensional diagnostic data,like soft x-ray emission intensity,edge toroidal rotation velocity,the Mirnov signal amplitude and so on.Based on the cluster analysis results,an approximate analytical model is proposed to rapidly identify plasma events in the J-TEXT plasma.The cluster analysis method is conducive to data markers of massive diagnostic data.

Overview of machine learning applications in fusion plasma experiments on J-TEXT tokamak

Machine learning research and applications in fusion plasma experiments are one of the main subjects on J-TEXT.Since 2013,various kinds of traditional machine learning,as well as deep learning methods have been applied to fusion plasma experiments.Further applications in the real-time experimental environment have proved the feasibility and effectiveness of the methods.For disruption prediction,we started by predicting disruptions of limited classes with a short warning time that could not meet the requirements of the mitigation system.After years of study,nowadays disruption prediction methods on J-TEXT are able to predict all kinds of disruptions with a high success rate and long enough warning time.Furthermore,cross-device disruption prediction methods have obtained promising results.Interpretable analysis of the models are studied.For diagnostics data processing,efforts have been made to reduce manual work in processing and to increase the robustness of the diagnostic system.Models based on both traditional machine learning and deep learning have been applied to real-time experimental environments.The models have been cooperating with the plasma control system and other systems,to make joint decisions to further support the experiments.

Machine learning application to predict the electron temperature on the J-TEXT tokamak

The reliability of diagnostic systems in tokamak plasma is of great significance for physics researches or fusion reactor.When some diagnostics fail to detect information about the plasma status,such as electron temperature,they can also be obtained by another method:fitted by other diagnostic signals through machine learning.The paper herein is based on a machine learning method to predict electron temperature,in case the diagnostic systems fail to detect plasma temperature.The fully-connected neural network,utilizing back propagation with two hidden layers,is utilized to estimate plasma electron temperature approximately on the J-TEXT.The input parameters consist of soft x-ray emission intensity,electron density,plasma current,loop voltage,and toroidal magnetic field,while the targets are signals of electron temperature from electron cyclotron emission and x-ray imaging crystal spectrometer.Therefore,the temperature profile is reconstructed by other diagnostic signals,and the average errors are within 5%.In addition,generalized regression neural network can also achieve this function to estimate the temperature profile with similar accuracy.Predicting electron temperature by neural network reveals that machine learning can be used as backup means for plasma information so as to enhance the reliability of diagnostics.