Sensor Data Correction for Maintenance-Shifted Anomaly Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data processing systems fail to accurately correct for deviations in sensor readings due to equipment maintenance, leading to incorrect machine learning models and recognition of apparatus states.
Innovation Solution
A data processing apparatus that includes a data acquisition unit, storage unit, detection unit, calculation unit, correction unit, learning processing unit, and output unit, which corrects sensor data for deviations caused by equipment maintenance by calculating and applying reference values to smooth and correct data variations, generating a machine learning model for anomaly detection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If sensor data is used directly without correction for equipment maintenance deviations, then data processing is simple, but measurement precision deteriorates leading to incorrect machine learning models
Solution Approach 1:
The system performs preliminary correction of sensor data by calculating reference values before machine learning processing. The correction unit pre-processes sensor readings by subtracting calculated reference values that represent expected deviations during maintenance periods, thereby improving measurement precision before the data is used for anomaly detection.
Solution Approach 2:
The invention introduces an intermediary correction mechanism that acts between raw sensor data and machine learning processing. The correction unit serves as a mediator that adjusts sensor readings by reference values derived from maintenance schedules and historical data, eliminating the need for complex post-processing while improving accuracy.
2Measurement precision
If sensor data is corrected using reference values for maintenance deviations, then measurement precision improves, but device complexity increases
Solution Approach 1:
The correction unit performs multiple functions: it calculates reference values, corrects sensor data, and integrates with the machine learning model. By making the correction mechanism multi-functional, the system avoids adding separate complex subsystems while achieving improved measurement precision through a unified processing approach.
Solution Approach 2:
The system changes the parameter of sensor readings by subtracting reference values that represent maintenance-induced deviations. This parameter transformation converts raw sensor data into corrected data that accurately reflects actual equipment states, improving measurement precision through mathematical adjustment rather than physical modification.
3Reliability
If uncorrected sensor data is used for machine learning, then processing time is reduced, but reliability deteriorates due to incorrect anomaly detection
Solution Approach 1:
The correction of sensor data is performed preliminarily before machine learning processing, ensuring that the input data to the anomaly detection model is already corrected. This preliminary correction improves reliability by eliminating maintenance-induced deviations from the training and detection data, without requiring additional processing time during actual anomaly detection operations.
Data Source
AI summary
(SOLUTION) Provided is a data processing apparatus including: a data acquisition unit for acquiring time-series data regarding running of an apparatus; a stop acquisition unit for acquiring a stop of the running of the apparatus; and a correction unit for correcting, after changing a reference value according to an amount of variation in the data between before and after a stop period of the apparatus, the data using the value. Provided is a data processing apparatus including: a data acquisition unit for acquiring time-series data regarding running of an apparatus; a stop acquisition unit for acquiring a stop of the running of the apparatus; an output unit for outputting an amount of variation in the data between before and after a stop period of the apparatus; and a correction unit for correcting, after changing a reference value upon receiving an instruction corresponding to the output, the data using the value.