Anomaly Characterization Using Joint Historical and Time-Series Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current anomaly detection systems in manufacturing and energy-related fields struggle to distinguish between genuine opportunities for reducing energy or productivity losses and uncontrollable power surges or rapid transient events, leading to inefficient data mining and anomaly characterization.
Innovation Solution
A system that performs a joint historical and time-series analysis using binned inter-quartile range analysis to identify and classify anomalies, differentiating between surge anomalies and steady-state anomalies by calculating the derivative of recent time-series data and comparing it to predetermined thresholds.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional anomaly detection methods are used, then all anomalies are detected, but genuine opportunities for reducing energy losses cannot be distinguished from uncontrollable power surges
Solution Approach 1:
The patent segments anomalies into different categories (surge anomalies and steady-state anomalies) by analyzing the derivative magnitude of time-series data. This segmentation allows the system to distinguish between controllable and uncontrollable anomalies, enabling targeted remedial actions for genuine energy loss opportunities while ignoring transient power surges.
2Measurement precision
If detailed anomaly analysis is performed on all data points, then accurate anomaly detection is achieved, but system complexity and processing time increase
Solution Approach 1:
The patent applies local quality by performing detailed derivative analysis only on identified anomaly points rather than the entire dataset. The system first identifies anomalies using threshold-based detection, then applies localized derivative magnitude analysis only to these specific points to classify them as surge or steady-state anomalies, reducing overall computational complexity.
3Productivity
If comprehensive anomaly detection is applied, then all outliers are identified, but efficient remedial actions cannot be determined due to lack of anomaly classification
Solution Approach 1:
The patent converts the previously harmful or neutral information (derivative magnitude of time-series data) into a beneficial classification mechanism. By analyzing the magnitude of the derivative at anomaly points, the system transforms raw data into meaningful anomaly type classification, enabling users to take targeted remedial actions only on steady-state anomalies while ignoring surge anomalies.
Data Source
AI summary
One embodiment provides a system for facilitating anomaly detection and characterization. During operation, the system determines, by a computing device, a first set of testing data which includes a plurality of data points, wherein the first set includes a data series for a first variable and one or more second variables. The system identifies anomalies by dividing the first set into a number of groups and performing an inter-quartile range analysis on data in each respective group. The system obtains, from the first set, a second set of testing data which includes a data series from a recent time period occurring before a current time, and which further includes a first data point from the identified anomalies. The system classifies the first data point as a first type of anomaly based on whether a magnitude of a derivative of the second set is greater than a first predetermined threshold.