Domain Knowledge-Informed Functional Outlier Detection for LQC
An ST-based method using failure-pattern knowledge to detect tiny anomalies in manufacturing time-series data.
2022.01 – 2022.12
·Motivation
- In manufacturing, time-series data are collected from multiple sensors for quality control.
- In line quality control (LQC), weak failures occur that are difficult to detect with conventional methods.
·Goal
Develop a method for detecting tiny anomaly patterns in manufacturing time-series data using Sequential Transformation (ST) and domain knowledge of failure patterns.
·Methodology
- ST maximizes the time-series pattern of a tiny anomaly sample through various computations.
- Domain knowledge of failure patterns defines new derivatives, combined with ST to improve performance.
