Scalar Vibration Data Thinning for Diagnostic Storage Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Continuous online machine vibration monitoring generates large volumes of scalar vibration data, making it unmanageable and requiring a method to thin the data without losing important diagnostic information.
Innovation Solution
A data thinning process that sets an amplitude range for scalar vibration measurement data, discarding values within that range and saving only values outside it, along with calculated average values and time gaps, to reduce storage requirements while preserving essential diagnostic information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If continuous online machine vibration monitoring is performed with high sampling rates, then measurement precision and diagnostic information quality are improved, but data storage requirements increase significantly
Solution Approach 1:
The patent extracts only the essential vibration information by comparing consecutive scalar values and storing only when changes exceed a threshold amplitude range. This selective extraction maintains diagnostic quality while dramatically reducing storage volume by discarding redundant data points that fall within normal variation ranges.
Solution Approach 2:
The patent changes the storage parameter from storing every scalar vibration value to storing only significant change events. By introducing an amplitude range threshold parameter, the system transforms continuous high-volume data storage into event-based sparse storage, reducing storage requirements while preserving diagnostic information.
2Loss of information
If all scalar vibration data values are stored in the database, then complete diagnostic information is preserved, but data management becomes unmanageable over long time frames
Solution Approach 1:
The patent extracts only meaningful vibration events by comparing each scalar value against the previous value and storing only when the change exceeds a predefined amplitude range. This extraction process eliminates redundant data while preserving diagnostic information, making long-term data management feasible.
Solution Approach 2:
The patent applies partial action by storing only a subset of vibration data points—specifically those representing significant changes—rather than storing all data points. This partial storage approach maintains diagnostic capability while reducing management complexity.
3Quantity of substance
If data thinning is applied to reduce storage requirements, then storage efficiency is improved, but there is risk of losing important vibration information
Solution Approach 1:
The patent uses feedback by continuously comparing each new scalar vibration value against the previous stored value and using the amplitude range threshold to determine whether to store the new value. This feedback mechanism ensures that only diagnostically significant changes are stored, preventing loss of important information while achieving data thinning.
Solution Approach 2:
The patent changes the storage decision parameter from storing all values to storing only when the absolute difference between consecutive values exceeds an amplitude range threshold. This parameter-based filtering reduces storage volume while preserving diagnostic information through intelligent selective storage.
Data Source
AI summary
A process for thinning scalar machine vibration data can reduce the amount of data storage required to store the data by several orders of magnitude, without losing any important machine diagnostic vibration information. The process assumes that because each scalar vibration measurement value has its own range of values, there is a unique delta change in value that does not significantly impact machine diagnostic information provided by the data. Some embodiments provide a method to automatically evaluate the delta change in value. The process can be used to thin data that have already been stored in a database, and also to thin the data in real-time during data collection. Data storage structures for storing the thinned scalar values and processes for displaying a trend plot to indicate where the scalar data have been thinned are also described.