Equipment State Estimation Using Distribution-Based Data Ranges
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing state estimation methods for equipment with high reliability, such as automatic ticket gates and power plants, face challenges in precision due to data deviations and variations, leading to lowered estimation accuracy when using average values or predefined value ranges.
Innovation Solution
An information processing apparatus that acquires monitoring data, calculates the distribution of monitoring values, and determines a specific 'interested range' for estimation, focusing on data within this range to enhance precision and account for variations, rather than relying solely on average values.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If average values or predefined value ranges are used for state estimation, then the estimation process is simple, but the estimation precision is lowered due to data deviations and variations
Solution Approach 1:
The patent segments the monitoring data by calculating distribution characteristics (mean, standard deviation) and dividing data into multiple ranges based on standard deviations from the mean. This segmentation allows selective use of data within specific ranges, improving estimation precision by excluding peculiar values while maintaining a structured, manageable process.
Solution Approach 2:
The patent changes the estimation approach from using fixed average values or predefined ranges to dynamically calculated distribution-based ranges. By computing mean and standard deviation from actual monitoring data and defining estimation ranges as mean ± n×standard deviation, the system adapts to data variations while maintaining process simplicity.
2Quantity of substance
If all monitoring data is used for estimation, then data utilization is maximized, but peculiar values and deviations reduce estimation accuracy
Solution Approach 1:
The patent extracts and excludes peculiar values by defining estimation ranges based on distribution characteristics (mean ± n×standard deviation). Data points falling outside these statistically determined ranges are automatically excluded from estimation calculations, effectively removing outliers while preserving the majority of valid data for accurate estimation.
Solution Approach 2:
The patent applies different treatment to different portions of the data distribution. Data within the calculated range (mean ± n×standard deviation) is used for estimation with high quality, while data outside this range is excluded. This local quality approach ensures that only statistically normal data contributes to the estimation, improving overall accuracy.
3Ease of operation
If fixed predefined ranges are used for estimation, then the estimation method is straightforward, but it cannot adapt to environmental changes and data variations
Solution Approach 1:
The patent transitions from static predefined ranges to dynamic ranges that automatically adjust based on monitoring data characteristics. By calculating mean and standard deviation from actual data and defining ranges as mean ± n×standard deviation, the system continuously adapts to environmental changes and data variations while maintaining a straightforward implementation through automated statistical calculations.
Solution Approach 2:
The patent incorporates feedback by using actual monitoring data to calculate distribution characteristics, which then define the estimation ranges. This closed-loop approach allows the system to automatically adapt to changing conditions: as data patterns change, the mean and standard deviation are recalculated, and the estimation ranges adjust accordingly, maintaining both simplicity and adaptability.
Data Source
AI summary
One embodiment of the present invention provides an information processing apparatus for precisely estimating a state of equipment. An information processing apparatus as one embodiment of the present invention includes: an acquirer; a calculator; and a determiner. The acquirer is configured to acquire first data about a predetermined target. The calculator is configured to calculate distribution of magnitude of a value included in the first data. The determiner is configured to determine a portion of a width of the distribution as a specific range used for estimating a state of the target based on the distribution.