Data Processing Device for Electrical Apparatus Inference
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for inferring the operational status of electrical apparatuses, such as those with inverters, face accuracy issues due to fluctuations in current and power consumption, requiring extensive training data to cover these fluctuations effectively.
Innovation Solution
A data processing device and method that acquire measurement data, extract fluctuation components, and derive feature values from current and power consumption patterns to infer operational statuses, distinguishing between fluctuation and non-fluctuation states for improved accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If measurement data is processed to cover fluctuations in current and power consumption, then the accuracy of inference is improved, but the amount of training data required increases
Solution Approach 1:
The patent extracts fluctuation components from measurement data by separating them from reference components. This extraction allows the system to focus only on the relevant fluctuation patterns rather than processing entire datasets, thereby improving inference accuracy while reducing the amount of training data needed.
Solution Approach 2:
The patent transforms measurement data by changing its representation parameters - extracting fluctuation components and representing them as feature values. This parameter transformation enables more efficient use of training data while maintaining or improving inference accuracy for electrical apparatuses with fluctuating power consumption.
2Productivity
If fluctuation components are extracted and feature values are acquired, then processing load is reduced, but the complexity of data processing increases
Solution Approach 1:
The patent segments measurement data into reference components and fluctuation components. This segmentation simplifies the processing load by allowing separate handling of stable reference patterns and variable fluctuation patterns, reducing the overall computational burden despite the added extraction step.
Solution Approach 2:
The patent performs preliminary extraction of fluctuation components and acquisition of feature values before the main inference process. This preliminary action prepares the data in advance, reducing the processing load during actual inference operations even though it adds initial processing complexity.
Data Source
AI summary
In order to solve the above-described problem, there is provided a data processing device (1) including a measurement data acquisition unit (10) that acquires measurement data indicating a temporal change in at least one of current consumption and power consumption of an electrical apparatus, a fluctuation component extraction unit (20) that extracts a fluctuation component related to fluctuation in the current consumption and a fluctuation component related to fluctuation in the power consumption from the measurement data, and a feature value acquisition unit (30) that acquires a feature value of the fluctuation component.