Time-Series Event Segment Detection Using Protruding Waveform Patterns
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data analysis devices cannot extract time-series data of event sections when there is no event information indicating the occurrence timing of events.
Innovation Solution
A time-series data processing device that extracts protruding data with ascending and descending legs from time-series data and defines an occurrence pattern to detect matching segments, allowing for the extraction of event sections without event information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If event information indicating occurrence timings of events is used to extract time-series data of event sections, then the extraction accuracy of event sections is improved, but the device cannot extract time-series data of event sections in the case where there is no event information
Solution Approach 1:
The system performs self-service by automatically defining occurrence patterns of protruding data without requiring external event information. The occurrence pattern defining unit creates patterns based on the intrinsic characteristics of the time-series data itself, allowing the system to autonomously identify event sections even when no event information is available from the production line.
Solution Approach 2:
The system changes the approach from relying on event timing parameters to using waveform shape parameters. By defining occurrence patterns based on the characteristics of protruding data (ascending and descending legs), the system transforms the extraction method to work with intrinsic data features rather than external event information, enabling extraction in both scenarios.
2Ease of operation
If event information from production line is required for data extraction, then the extraction process is simplified with clear event boundaries, but the data analysis device becomes dependent on external information systems
Solution Approach 1:
The system achieves self-service by internally defining occurrence patterns based on protruding data characteristics. This eliminates the need for external event information systems, making the data analysis device independent while maintaining the ability to extract event sections effectively.
Solution Approach 2:
The occurrence pattern defining unit provides universal functionality that works both with and without event information. The system can operate in two modes: using predefined event information when available, or automatically defining patterns from data characteristics when event information is unavailable, making the device adaptable to various operational contexts.
Data Source
AI summary
A time-series data processing device (10) includes: a protruding data extracting unit (2) for extracting, from time-series data (1) which is a sequence of values obtained from sequential observation with the elapse of time, protruding data including an ascending leg a value of which continuously rises with respect to time and a descending leg a value of which continuously drops with respect to time; an occurrence pattern defining unit (3) for defining an occurrence pattern of protruding data in the time-series data (1); and an occurrence pattern detecting unit (4) for detecting, as a segment (5), one or more pieces of protruding data matching the occurrence pattern defined by the occurrence pattern defining unit (3), from among a set of pieces of protruding data extracted by the protruding data extracting unit (2).