Unit Space Updating for Rare Plant State Retention
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing unit space update methods may mistakenly determine a plant operation state as abnormal during startup or load fluctuations due to the deletion of low-frequency occurrence bundles of detected values over time.
Innovation Solution
A unit space update device and method that includes a detected value acquiring unit and a deletion processing unit, which randomly deletes or selectively removes bundles with high frequency of appearance or smallest difference in detected values, to maintain a balanced unit space for accurate plant operation state monitoring.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If the oldest bundle of detected values is deleted from the unit space to update it for each predetermined cycle, then the unit space can be updated to respond to change in tendencies of detected values caused by aging deterioration and seasonal changes, but bundles of detected values acquired in operation states with low frequency of occurrence (such as plant startup and load fluctuations) may be deleted, leading to mistaken determination of abnormal operation states
Solution Approach 1:
The patent changes the deletion criterion from time-based (oldest bundle) to frequency-based (bundles with highest frequency of appearance). This parameter change ensures that operation states with low frequency of occurrence are preserved in the unit space, preventing mistaken determination of abnormal states during startup or load fluctuations, while still maintaining adaptability to aging and seasonal changes through regular updates.
2Productivity
If bundles of detected values are updated continuously to reflect current plant conditions, then the monitoring system can respond to seasonal changes and aging deterioration, but low-frequency operation states may be lost due to continuous overwriting of the unit space
Solution Approach 1:
The patent applies the principle of discarding and recovering by selectively removing bundles with highest frequency of appearance while preserving bundles representing low-frequency operation states. This ensures that information about rare but important operation states (startup, load fluctuations) is recovered and retained in the unit space, preventing information loss while maintaining continuous update capability for current plant conditions.
Data Source
AI summary
A unit space update device, which generates a unit space, includes: a detection value acquisition unit which acquires a bundle of detection values detected at a constant cycle; and a deletion processing unit which deletes one bundle from the bundles of the detection values constituting the unit space, when adding the bundle of the detection values to the unit space. The deletion processing unit includes at least one among: a first processing unit which randomly deletes one bundle from the bundles of the detection values forming the unit space; and a second processing unit which, with respect to the detection value of the predetermined evaluation item among the plurality of evaluation items, deletes any one bundle of the bundles including the detection value with the highest appearance frequency in the unit space, or one of two bundles having the smallest difference between the detected values in the unit space.