CN-121880840-B - Online data analysis method and system for fusion radiation calorimetric diagnosis

CN121880840BCN 121880840 BCN121880840 BCN 121880840BCN-121880840-B

Abstract

The invention discloses an online data analysis method and an online data analysis system for fusion radiation calorimetric diagnosis, and belongs to the technical field of nuclear fusion experiments. The method comprises the steps of firstly establishing a fusion radiation calorimetric diagnosis expert rule base based on multi-dimensional parameter data fusion, creating a paging man-machine interface and a unified canvas frame nested on an interface front page, then obtaining current discharge cannon numbers on line, obtaining multi-signal source time sequence data from a current pulse tree of a remote pulse database, storing the multi-signal source time sequence data into a local pulse base, then processing and analyzing the data according to the diagnosis expert rule, realizing automatic identification and compensation of abnormal channels of diagnosis, on-line calculation of radiation intensity and total radiation power, and finally integrating and visualizing various signal data under the unified canvas frame.

Inventors

DUAN YANMIN
QIAN JING
Mao Songtao
LIN SHIYAO
XU LIQING
ZANG QING
LIU HAIQING

Assignees

中国科学院合肥物质科学研究院

Dates

Publication Date: 20260512
Application Date: 20260320

Claims (10)

1. An on-line data analysis method for fusion radiation calorimetric diagnosis, comprising the steps of: s1, establishing a fusion radiation thermal diagnosis expert rule base based on multi-dimensional parameter data fusion; s2, creating a paging man-machine interface, and nesting a first page of the paging man-machine interface to create a unified canvas frame and an internal subgraph thereof; s3, acquiring a current discharge gun number of a fusion plasma discharge experiment on line; S4, acquiring time sequence data of multiple signal sources in a fusion experiment remote pulse database based on the current discharge gun number, wherein the multiple signal sources comprise a multi-channel original voltage signal of a radiation calorimetric diagnosis detector array, a multi-channel input power signal of a plasma heating system and a plasma discharge current signal, and the multiple signals of the multiple signal sources have different sampling rates, data starting time, data ending time and data duration; s5, storing the acquired time sequence data of the multiple signal sources into a local pulse database; S6, processing time sequence data of multiple signal sources according to fusion radiation thermal diagnosis expert rules, identifying and compensating radiation thermal diagnosis abnormal channels, calculating multi-channel radiation intensity by using the compensated radiation thermal diagnosis data, and finally calculating the total radiation power of the current discharge gun number according to the multi-channel radiation intensity data; s7, integrating various signal data of the visualized current discharge gun number based on a unified canvas frame; S8, returning to S3, and continuing to wait for obtaining the next current discharge gun number until all the discharge gun numbers are completed.
2. The online data analysis method for fusion radiation calorimetric diagnosis of claim 1, wherein fusion radiation calorimetric diagnosis expert rules based on multi-dimensional parameter data fusion are used for plasma discharge current type division, data segmentation, radiation calorimetric diagnosis original data preprocessing, abnormal channel identification, abnormal channel compensation, multi-channel radiation intensity online calculation, radiation total power online calculation and elimination of temperature drift influence possibly existing under long pulse discharge, wherein the abnormal channel identification is performed in a plurality of times, and comprises abnormal channel pre-identification, abnormal channel secondary identification based on plasma discharge current type and discharge duration and correlation analysis, abnormal channel tertiary identification based on plasma discharge current type and discharge duration and multi-channel radiation intensity trend data, and the multi-dimensional parameters comprise multi-signal source time sequence data obtained online, a plurality of signal time sequence statistical characteristic values, time sequence curve waveform change types and trend data thereof, radiation calorimetric diagnosis detector characteristics and correlation analysis data among a plurality of signals.
3. The online data analysis method for fusion radiation calorimetric diagnosis according to claim 2, wherein the data segments are divided into three types according to different purposes, namely, a data segment for calculating correlation analysis data, a data segment for calculating time sequence statistical characteristic values, a data segment for calculating radiation intensity and total radiation power of radiation calorimetric diagnosis, wherein the data segment for calculating correlation analysis data comprises a data segment for correlation analysis between plasma discharge current and multi-channel radiation calorimetric diagnosis signals and a data segment for correlation analysis between radiation calorimetric diagnosis multi-channels, and the data segment for calculating time sequence statistical characteristic values comprises a data segment for calculating plasma discharge current characteristic values and a data segment for calculating multi-channel radiation calorimetric diagnosis characteristic values.
4. The method for analyzing the online data for fusion radiation thermal diagnosis according to claim 2, wherein the preprocessing of the radiation thermal diagnosis raw data comprises the steps of removing baseline drift and zero-phase low-pass filtering of the multichannel radiation thermal diagnosis raw data online, segmenting data according to purposes, calculating time domain statistical characteristic values of the radiation thermal diagnosis data, calculating characteristic values of a detector used for radiation thermal diagnosis, detecting and counting waveform change types of a multichannel radiation thermal diagnosis time sequence curve by adopting an overlapped sliding window, wherein the time domain statistical characteristic values comprise maximum value, minimum value, mean value, median value, variance, maximum first-order difference, minimum first-order difference, maximum second-order difference, minimum second-order difference, skewness and kurtosis, and the characteristic values of the detector used for radiation thermal diagnosis comprise thermal time constant, baseline drift rate, high-low frequency noise ratio, environmental temperature fluctuation amplitude and signal-to-noise ratio.
5. The online data analysis method for fusion radiation calorimetric diagnosis according to claim 2, wherein the abnormal channel pre-identification specifically comprises identifying an abnormal channel which does not reflect any physical information and has obvious abnormal characteristics according to the time domain statistical characteristic value of the multi-channel radiation calorimetric diagnosis data and the statistical analysis of the waveform change type of the time sequence curve, and identifying an abnormal channel with unreliable data according to the characteristic value of a detector used for the multi-channel radiation calorimetric diagnosis.
6. The online data analysis method for fusion radiation calorimetric diagnosis according to claim 2, wherein the specific step of performing abnormal channel secondary identification based on plasma discharge current type and discharge duration and correlation analysis comprises the following steps: A1. According to the plasma discharge current Ip data, calculating the real discharge duration of the current discharge gun number; A2. carrying out data segmentation processing on the plasma discharge current and the radiation calorimetric diagnosis raw data according to the real discharge time length and the diagnosis expert rules, and carrying out data alignment on the processed plasma discharge current and radiation calorimetric diagnosis data; A3. Calculating a time domain statistical characteristic value of the plasma discharge current Ip; A4. Dividing the type of the Ip into six types including a class I test gun, a class II non-breakdown current waveform, a class III breakdown but fast broken current waveform, a class IV breakdown but incomplete current waveform, a class V breakdown and complete current waveform and a class VI long pulse discharge; A5. removing an abnormal channel obtained by pre-identifying the abnormal channel, thereby obtaining a normal channel data set after pre-identifying the abnormal channel by the radiation calorimetric diagnosis under the current discharge gun number; A6. Correlation analysis data are obtained, and correlation coefficient matrixes and p-value matrix correlation analysis data among a plurality of normal channels of the radiation thermal diagnosis are obtained through calculation according to the pre-identified normal channel data set and plasma discharge current data, and correlation coefficient matrixes and p-value matrix correlation analysis data among the normal channels of the radiation thermal diagnosis and the plasma discharge current are calculated; A7. And combining the plasma discharge current type and the discharge duration, and carrying out anomaly detection on the correlation analysis data, thereby obtaining an anomaly channel after the current discharge gun number is secondarily identified in the radiation calorimetric diagnosis.
7. The online data analysis method for fusion radiation calorimetric diagnosis according to claim 2, wherein the abnormal channel compensation means compensation by replacing all data of the automatically identified radiation calorimetric diagnosis abnormal channel with data of a nearest neighbor channel or by taking an average value of data of an adjacent normal channel according to an abnormal channel compensation strategy in a diagnosis expert rule base.
8. The online data analysis method for fusion radiation calorimetric diagnosis of claim 1, wherein the method is characterized in that the method comprises the steps of creating a unified canvas frame and sub-graphs and molecular graphs in the unified canvas frame to automatically draw various signal time sequence curves and automatically play the curves according to the sub-graphs, wherein the various signals drawn by the molecular graphs comprise multichannel original data of a radiation calorimetric diagnosis detector array and multichannel data after baseline drift is removed, online calculated multichannel radiation intensity, online calculated total radiation power, input power of different plasma heating systems and plasma discharge current, and the input power signals of the different plasma heating systems have different sampling rates, data starting moments, data ending moments and data duration.
9. An on-line data analysis system for fusion radiation calorimetric diagnostics, comprising: The pulse database operation module reads, writes and creates multi-signal source data in a current pulse tree named by a current discharge gun number; The plasma discharge current data processing module is used for calculating a time domain statistical characteristic value and a real discharge time length of a plasma discharge current corresponding to the current discharge gun number, dividing the plasma discharge current type and judging the plasma discharge current type of the current discharge gun number; The diagnosis original data preprocessing module is used for preprocessing the radiation calorimetric diagnosis original data, segmenting the data according to the real discharge time length, the data time length and the data purpose, calculating time domain statistics characteristic values of all channels, detecting and counting time sequence curve waveform change types of all channels, and calculating detector characteristic values; the correlation analysis data module is used for calculating a correlation coefficient matrix and a p-value matrix among multiple channels; The abnormal channel identification and compensation module is used for identifying the radiation quantity thermal diagnosis abnormal channel of the current discharge gun number for a plurality of times according to the rule of a diagnosis expert; The radiation intensity and total radiation power online calculation module is used for calculating the multi-channel radiation intensity according to the abnormal channel compensated radiation calorimetric diagnosis data, the multi-detector array characteristic parameters and the spatial position parameters thereof, and carrying out surface integral and volume integral online calculation of the total radiation power by a weighted addition method; And the visualization module is used for creating a unified canvas frame and subgraphs thereof for integrally displaying various signal time sequence curves, and the molecular diagram draws the various signal time sequence curves and automatically plays the various signal time sequence curves according to the subgraph sequence.
10. The online data analysis system for fusion radiation thermal diagnosis according to claim 9, wherein the abnormal channel identification module comprises a pre-identification module, a secondary identification module and a tertiary identification module, wherein the pre-identification module pre-identifies an abnormal channel by applying the diagnostic expert rule according to all channel time domain statistical characteristic values, time sequence curve waveform change types and detector characteristics detected and counted by the diagnostic raw data preprocessing module so as to obtain a pre-identified abnormal channel combination BadChan _gp1, the secondary identification module secondarily identifies the abnormal channel by adopting a robust probability analysis and an IRQ four-way bit method according to relevance analysis data obtained by the relevance analysis data module, a plasma discharge current type and a discharge duration of a current discharge gun number, so as to obtain a secondarily identified abnormal channel combination BadChan _gp2, the tertiary identification module firstly obtains multi-channel radiation intensity trend data by adopting a linear regression method based on signal amplitude, then obtains the abnormal channel combination BadChan _gp3 by combining with the plasma discharge current type and the discharge duration, and finally obtains a final combination 354 of abnormal channels BadChan _gp3_gp3 and BadChan.

Description

Online data analysis method and system for fusion radiation calorimetric diagnosis Technical Field The invention belongs to the technical field of nuclear fusion experiments, and particularly relates to an online data analysis method and system for fusion radiation calorimetric diagnosis. Background A plurality of nuclear fusion experimental devices at home and abroad comprise EAST, JET, DIII-D, ASDEX-U, JT-60U, alcator C-Mod, BEST, ITER and the like, and a radiation calorimetric diagnosis system is established or is to be established. On a full superconducting tokamak EAST device, as one of key diagnoses of plasma discharge parameters, a bolometric diagnosis mainly measures total power of radiation and distribution thereof during a plasma discharge experiment, and measures radiation of a wide energy range using a metal film resistance bolometric detector. Under the electromagnetic environment of the EAST device with complex site, the radiation calorimetric diagnosis of the multi-detector array with spatial distribution can generate abnormal channel combination which changes due to the change of the plasma discharge condition corresponding to the current discharge gun number, thereby influencing the radiation calorimetric diagnosis data quality and the application effect. And, during the operation of the large tokamak device, the plasma discharge frequency is high and the number of daily discharge times is large. Therefore, it is important for a multi-channel bolometric diagnostic system to achieve efficient and reliable online data analysis. At present, an online data analysis system for radiation calorimetric diagnosis meets the basic requirements of data analysis, and the main defects are that: (1) Abnormal channels of multi-channel radiation calorimetric diagnosis cannot be automatically identified and compensated based on the current discharge gun number; (2) The unmanned aerial vehicle interaction interface can not automatically load, set and store various parameters required by on-line calculation of radiation intensity and total radiation power, wherein the various parameters required by calculation of the total radiation power comprise physical parameters and control parameters such as fusion experiment devices, plasma discharge experiment related parameters, radiation calorimetric diagnosis detector characteristic parameters, detector related installation position information and the like; (3) The visual multi-channel radiation calorimetric diagnosis original signals, the multi-channel radiation intensity signals obtained through on-line calculation and the experimental data of the radiation total power signals obtained through on-line calculation are not integrated under a unified framework. Disclosure of Invention The invention aims to solve the defects of the prior art, and provides an online data analysis method and an online data analysis system for fusion radiation thermal diagnosis, which aim to realize efficient online data analysis of fusion radiation thermal diagnosis by applying diagnostic expert rules in a fusion radiation thermal diagnosis expert rule base based on multidimensional parameter data fusion, lay a foundation for providing accurate absolute radiation power measurement data in a nuclear fusion plasma discharge experiment, and provide reliable data support for researching a plasma energy loss mechanism, developing related control technologies and the like. In order to achieve the aim of the invention, the invention adopts the following technical scheme: An on-line data analysis method for fusion radiation calorimetric diagnosis, comprising the steps of: s1, establishing a fusion radiation thermal diagnosis expert rule base based on multi-dimensional parameter data fusion; s2, creating a paging man-machine interface, and nesting a first page of the paging man-machine interface to create a unified canvas frame and an internal subgraph thereof; s3, acquiring a current discharge gun number of a fusion plasma discharge experiment on line; S4, acquiring time sequence data of multiple signal sources in a fusion experiment remote pulse database based on the current discharge gun number, wherein the multiple signal sources comprise a multi-channel original voltage signal of a radiation calorimetric diagnosis detector array, a multi-channel input power signal of a plasma heating system and a plasma discharge current signal, and the multiple signals of the multiple signal sources have different sampling rates, data starting time, data ending time and data duration; s5, storing the acquired time sequence data of the multiple signal sources into a local pulse database; S6, processing time sequence data of multiple signal sources according to fusion radiation thermal diagnosis expert rules, identifying and compensating radiation thermal diagnosis abnormal channels, calculating multi-channel radiation intensity by using the compensated radiation thermal diagnosis data, and finally calculat