Predicting particle accelerator failures using binary classifiers

Abstract

Particle accelerator failures lead to unscheduled downtime and lower reliability. Although simple to mitigate while they are actually happening such failures are difficult to predict or identify beforehand. In this work we propose using machine learning approaches to predict machine failures via beam current measurements before they actual occur. To demonstrate this technique in this paper we examine beam pulses from the Oakridge Spallation Neutron Source (SNS). By evaluating a pulse against a set of common classification techniques we show that accelerator failure can be identified prior to actually failing with almost 80% accuracy. We also show that tuning classifier parameters and using pulse properties for refining datasets can further lead to almost 92% accuracy in classification of bad pulses. Most importantly, in the paper we establish there is information about the failure encoded in the pulses prior to it, so we also present a list of feasible next steps for increasing pulse classification accuracy.

Uncertainty aware anomaly detection to predict errant beam pulses in the SNS accelerator

“…The goal in this paper is to predict the onset of conditions leading to errant beam pulses at least one pulse in advance from beam measurements. Similar questions are studied in [11] and [12]. Rescic et al [11] demonstrate that a Random Forest (RF) is capable of identifying a beam loss event at SNS one pulse in advance using data from one beam current monitor.…”

“…Similar questions are studied in [11] and [12]. Rescic et al [11] demonstrate that a Random Forest (RF) is capable of identifying a beam loss event at SNS one pulse in advance using data from one beam current monitor. In [12], the authors examine the potential to predict interlocks (reflecting beam shutoff) using multiple measurements along the accelerator.…”

“…We utilize a Differential Current Monitor (DCM) that uses data from two beam current sensors located upstream and downstream of the source. The sensors used in this study are the same as those in [11]. Our approach specifically addresses the use of a limited amount of monitoring devices to identify beam loss events, avoids extensive computational loads and the use of data from multiple measurements, can adjust the threshold to change the FP and TP rates, ca determine is the trained model is outdated, and is not affected by anomalies the model is not trained on.…”

A Novel Approach for Classification and Forecasting of Time Series in Particle Accelerators

“…There also lacks a quantified validation of the model performance among all the jitters. Rescic et al applied various binary classification algorithms on beam pulses from the SNS to identify the last pulse ahead of the failure from normal pulses [15], in which predictive power is embedded, but their approach depends highly on the discrete pulse structure and cannot be directly adapted to continuous cases. It is also worth noting that both studies deal with univariate time series, i.e., RF power for Donon et al and beam current for Rescic et al Our work presents a novel and more ambitious approach to classify a continuous stream of multivariate time series data by cutting out windows from different regions, and trigger an alarm for a potential beam interruption before it happens.…”

A Novel Approach for Classification and Forecasting of Time Series in Particle Accelerators

“…There also lacks a quantified validation of the model performance among all the jitters. Rescic et al applied various binary classification algorithms on beam pulses from the SNS to identify the last pulse ahead of the failure from normal pulses [15], in which predictive power is embedded, but their approach depends highly on the discrete pulse structure and cannot be directly adapted to continuous cases. It is also worth noting that both studies deal with univariate time series, i.e.…”