Motor Anomaly Diagnosis Using Control-Bandwidth Data Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing anomaly diagnosis methods for drive machines using motor systems suffer from reduced accuracy due to the analysis of unselected data, including redundant information irrelevant to anomalies.
Innovation Solution
An anomaly diagnosis apparatus that generates a command value to control the motor or drive machine, selects relevant time-series data based on a comparison between the control bandwidth and a threshold determined from the drive machine's resonance frequency, and uses this data for anomaly determination.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If all collected time-series data is analyzed for anomaly diagnosis, then comprehensive monitoring is achieved, but analysis accuracy decreases due to redundant data
Solution Approach 1:
The patent extracts and analyzes only the relevant time-series data corresponding to the control bandwidth range, separating useful anomaly-diagnostic information from redundant data. This extraction approach maintains comprehensive monitoring while eliminating data that does not contribute to anomaly detection accuracy.
Solution Approach 2:
The patent applies local quality by focusing analysis resources on the specific frequency range (control bandwidth) where anomalies manifest, rather than uniformly analyzing all data. This localized approach concentrates computational effort where it is most needed for accurate anomaly detection.
2Reliability
If comprehensive data analysis is performed, then all potential anomaly indicators are captured, but computational resources are wasted on redundant data
Solution Approach 1:
The patent extracts only the time-series data within the control bandwidth range for analysis, removing redundant frequency components. This extraction maintains reliable anomaly detection by focusing on the relevant frequency range while reducing computational resource consumption proportionally to the amount of excluded redundant data.
3Measurement precision
If no data selection is applied, then data processing is simple, but anomaly diagnosis accuracy is reduced
Solution Approach 1:
The patent performs preliminary data selection by filtering time-series data according to the control bandwidth range before anomaly analysis. This preliminary action prepares the data in advance, ensuring that only relevant data enters the anomaly detection process, thereby improving accuracy without requiring complex processing during the actual diagnosis phase.
Data Source
AI summary
An anomaly diagnosis apparatus includes a command generation unit that generates a command value to specify the operation of a motor or a drive machine, a drive control unit that performs feedback control on the motor based on a control gain so that the operation of the motor or the drive machine follows the command value, a data switching unit that switches selected time-series data by selecting data from time-series data indicating the state of the motor or the drive machine, based on the result of a comparison between a control bandwidth determined from the control gain and a threshold determined from the drive machine, and an anomaly determination unit that determines an anomalous state of the motor or the drive machine, based on the selected time-series data.