Time-Series Anomaly Classification for Launch Vehicle Propulsion Systems: Fast Statistical Detectors Enhancing LSTM Accuracy and Data Quality

TL;DR

This paper introduces a statistical relabeling method to improve LSTM-based anomaly classification in launch vehicle propulsion systems, enhancing detection accuracy and data quality using real-time telemetry analysis.

Contribution

It proposes a novel Mahalanobis distance-based statistical detector for relabeling training data, significantly improving LSTM classifier performance in simulation-based anomaly detection.

Findings

Relabeling improved precision by 7%.

Relabeling improved recall by 22%.

Method demonstrated on digital twin simulations.

Abstract

Supporting Go/No-Go decisions prior to launch requires assessing real-time telemetry data against redline limits established during the design qualification phase. Family data from ground testing or previous flights is commonly used to detect initiating failure modes and their timing; however, this approach relies heavily on engineering judgment and is more error-prone for new launch vehicles. To address these limitations, we utilize Long-Term Short-Term Memory (LSTM) networks for supervised classification of time-series anomalies. Although, initial training labels derived from simulated anomaly data may be suboptimal due to variations in anomaly strength, anomaly settling times, and other factors. In this work, we propose a novel statistical detector based on the Mahalanobis distance and forward-backward detection fractions to adjust the supervised training labels. We demonstrate our method on digital twin simulations of a ground-stage propulsion system with 20.8 minutes of operation per trial and O(10^8) training timesteps. The statistical data relabeling improved precision and recall of the LSTM classifier by 7% and 22% respectively.