Neural network prediction of some classes of tokamak disruptions

Abstract

The use of neural network algorithms for predicting minor and major disruptions in tokamaks is explored by analysing disruption data from the TEXT tokamak with two network architectures. Future values of the fluctuating magnetic signal are predicted based on L past values of the magnetic fluctuation signal measured by a single Mirnov coil. The time step used (=0.04 ms) corresponds to the experimental data sampling rate. Two kinds of approach are adopted for the network: the contiguous future prediction and the multi-time-scale prediction. Both networks are trained through the back-propagation algorithm with inertial terms and the strengths of the results are compared. The use of additional diamagnetic signals as a method of increasing the performance is suggested. The degree of success indicates that the magnetic fluctuations associated with the TEXT disruption data may be characterized by a low dimensional dynamical system

Forecasting disruptions in the ADITYA tokamak using neural networks

“…(a) Unlike the previous articles [4] and [6] where a time lag was reported for the predicted instant of Figure 3. The first disruptive discharge, shot 6690, was used for forecasting.…”

“…To begin with, the ANN was trained with only one diagnostic signal. This was to test the performance of the network with similar input information as that already used in Refs [4] and [6]. First, one Mirnov probe was used as input, followed by the SXR signal.…”

Application of Neural Networks for the Measurement of Electronic Temperature in Nuclear Fusion Experiments

“…MLPs have been successfully used in several nuclear research applications, such as plasma control [6][7][8], plasma parameter extraction [9][10][11], Gaussian fitting [12] and online detection of disruption precursors [13,14]. The MLP architecture with BP error learning is also used in the application described in this paper.…”

Detection of Disruptions in the High-β Spherical Torus NSTX

“…Neural networks are typically trained, in the sense that a predetermined sample of input and output data are used to determine the optimal values of the coefficients of the network. Neural networks have been used to predict various forms of disruptions on ASDEX Upgrade [84][85][86], DIII-D [87], ADITYA [88,89], TEXT [90,91], JET [85,92,93], and JT-60 [94,95].…”

“…Typical input data to the network include a measure of the plasma β (the normalized β (β N ) [96,97] or poloidal β (β P )), edge safety factor, plasma density (or Greenwald fraction [98,99]), locked mode amplitude, input power, radiated power, shape parameters, internal inductance, confinement time, and neutron emission. Some early work also used soft X-ray emission [87][88][89]91], high(er) frequency magnetic probes [87][88][89][90], or Dα monitors [87][88][89] though these diagnostics have typically not been used in more recent studies [85,86,92,93,95].…”