Flow-Path Sensor Data Reduction Using Time-Series Similarity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current data management systems in industrial plants face challenges in efficiently reducing data volume for target sensors without compromising data integrity, especially when measurement data from upstream or downstream sensors shows similarity, leading to unnecessary data storage and transmission.
Innovation Solution
A data management system that acquires measurement data from multiple sensors along a flow path, reduces data volume by comparing time-series changes between sensors, and selectively deletes or reduces data samples based on similarity, ensuring efficient storage and transmission while maintaining critical data points.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If data volume reduction is performed by deleting measurement data from target sensors, then storage and transmission burdens are reduced, but data integrity may be compromised
Solution Approach 1:
The patent creates virtual copies of measurement data from upstream or downstream sensors when the target sensor's data can be reproduced based on similarity of time-series changes. This allows the system to reduce actual stored data volume while maintaining data integrity through reconstruction capabilities, effectively copying essential information from related sensors.
Solution Approach 2:
The patent changes the parameter of data representation by storing compressed or reduced-resolution data for target sensors when similarity conditions are met, while maintaining full-resolution data for reference sensors. This parameter transformation allows volume reduction without complete loss of information, as the full data can be reconstructed when needed.
2Reliability
If all measurement data from multiple sensors is recorded, then complete data availability is maintained, but storage and transmission resources are wasted
Solution Approach 1:
The patent applies local quality by differentiating the treatment of data from different sensors based on their position in the flow path and similarity relationships. Reference sensors (upstream or downstream) maintain full data quality and availability, while target sensors undergo selective data reduction. This localized differentiation optimizes the balance between data availability and volume reduction on a sensor-by-sensor basis.
Solution Approach 2:
The patent performs preliminary analysis of time-series similarity between sensors before deciding on data reduction strategies. By pre-identifying which sensors have highly correlated measurements, the system can proactively reduce data volume for those sensors while ensuring reference data is available for reconstruction, preventing the need for full data storage.
3Device complexity
If data volume reduction is performed without considering time-series similarity, then processing is simplified, but unnecessary data is deleted reducing accuracy
Solution Approach 1:
The patent implements feedback by continuously monitoring the similarity between time-series data from different sensors and using this information to dynamically adjust data reduction decisions. The system feeds back similarity metrics to determine whether target sensor data can be safely reduced or if full data must be retained, ensuring that accuracy requirements are met while maximizing volume reduction where appropriate.
Data Source
AI summary
Provided a data management system which includes a data acquisition unit that acquires measurement data obtained by measuring a fluid flowing in a flow path from each of a plurality of sensors, a data recording unit that records the acquired measurement data, and a data volume reduction unit that reduces a data volume to be recorded for a target sensor based on the measurement data acquired from another sensor installed in either an upstream or a downstream from itself in the flow path among the plurality of sensors.