Author: Paskvan, D.R.
Paper Title Page
WEPAB323 High Performance DAQ Infrastructure to Enable Machine Learning for the Advanced Photon Source Upgrade 3434
 
  • G. Shen, N.D. Arnold, T.G. Berenc, J. Carwardine, E. Chandler, T. Fors, T.J. Madden, D.R. Paskvan, C. Roehrig, S.E. Shoaf, S. Veseli
    ANL, Lemont, Illinois, USA
  
  Funding: Argonne National Laboratory’s work was supported by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences, under contract DE-AC02-06CH11357.
It is well known that the ef­fi­ciency of an ad­vanced con­trol al­go­rithm like ma­chine learn­ing is as good as its data qual­ity. Much re­cent progress in tech­nol­ogy en­ables the mas­sive data ac­qui­si­tion from a con­trol sys­tem of mod­ern par­ti­cle ac­cel­er­a­tor, and the wide use of em­bed­ded con­trollers, like field-pro­gram­ma­ble gate ar­rays (FPGA), pro­vides an op­por­tu­nity to col­lect fast data from tech­ni­cal sub­sys­tems for mon­i­tor­ing, sta­tis­tics, di­ag­nos­tics or fault record­ing. To im­prove the data qual­ity, at the APS Up­grade pro­ject, a gen­eral-pur­pose data ac­qui­si­tion (DAQ) sys­tem is under ac­tive de­vel­op­ment. The APS-U DAQ sys­tem col­lects high-qual­ity fast data from un­der­neath em­bed­ded con­trollers, es­pe­cially the FPGAs, with the man­ner of time-cor­re­la­tion and syn­chro­nously sam­pling, which could be used for com­mis­sion­ing, per­for­mance mon­i­tor­ing, trou­bleshoot­ing, and early fault de­tec­tion, etc. This paper pre­sents the de­sign and lat­est progress of APS-U high-per­for­mance DAQ in­fra­struc­ture, as well as its prepa­ra­tion to en­able the use of ma­chine learn­ing tech­nol­ogy for APS-U, and its use cases at APS. 		 
DOI • reference for this paper ※ https://doi.org/10.18429/JACoW-IPAC2021-WEPAB323  
About • paper received ※ 19 May 2021       paper accepted ※ 24 June 2021       issue date ※ 29 August 2021  
Export • reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)