Sensor-Based Device Diagnosis Across Variable Operating States
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing diagnostic systems for industrial devices face challenges in detecting abnormalities when sound samples corresponding to specific operation states are not pre-stored, making it difficult to compare and diagnose device conditions effectively.
Innovation Solution
A diagnostic apparatus and method that creates and compares feature amount vectors from sensor values, including normalized and spectral intensity components, to determine the normal or abnormal state of a device, even when samples for specific operation states are not available, by using a computer-readable medium to store and process sensor data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If sound samples are pre-stored for all operation states, then diagnostic accuracy is improved, but device complexity and data storage requirements increase
Solution Approach 1:
The system pre-stores reference sensor values for multiple predetermined operation states (temperature, humidity, load, speed) before actual diagnosis. These reference values serve as baseline data for comparison, enabling accurate diagnostic decisions without requiring complete sound sample libraries for every possible operation state.
Solution Approach 2:
The system changes the approach from storing complete sound samples to storing extracted feature amounts (parameters) such as spectral centroid, spectral rolloff, and zero-crossing rate. This parameter extraction and normalization process reduces data storage requirements while maintaining diagnostic accuracy through comparative analysis of normalized feature vectors.
2Reliability
If sound samples are pre-stored for specific operation states, then diagnostic capability for those states is improved, but adaptability to unknown or varying operation states deteriorates
Solution Approach 1:
The system creates a universal diagnostic framework that works across multiple operation states by storing reference sensor values for various predetermined states (different temperatures, humidities, loads, speeds). The normalization process enables the same diagnostic algorithm to adapt to any operation state by comparing against appropriate reference data, making the system versatile without requiring state-specific diagnostic models.
Solution Approach 2:
The system creates normalized feature amount vectors that copy the essential characteristics of sound signals across different operation states. By normalizing feature amounts based on reference values for each operation state, the system preserves the relative characteristics of abnormal sounds while removing variations due to operating conditions, enabling consistent diagnostic capability across all states.
3Speed
If feature amount comparison is performed without normalization, then processing speed is improved, but measurement precision deteriorates
Solution Approach 1:
The system applies normalization as a parameter transformation process that converts raw feature amounts into standardized values based on reference data for each operation state. This parameter change (normalization) preserves processing efficiency while significantly improving measurement precision by enabling accurate comparisons across different operating conditions through consistent scaling and reference-based adjustment.
Data Source
Figure 1
Figure 2
Figure 3(a)~3(b)
AI summary
A diagnostic apparatus (1) of the present invention includes: an operation state determination unit (21) that determines, at a reference time at which whether a device is normal or abnormal is known, which operation state of the device at the time among a plurality of operation states; a reference data creation unit (22) that repeats a process of storing sensor values acquired from the device while changing the operation state at the reference time in a storage unit in association with each of the determined operation states until there is no non-corresponding operation state with which the acquired sensor value is not yet associated; a diagnostic data creation unit (23) that acquires, at a diagnosis time at which it is not known whether the device is normal or abnormal, an operation state and a sensor value of the device at the time; and a diagnosis unit (24) that reads the sensor value associated with the acquired operation state from the storage unit and compares the sensor value acquired at the diagnosis time with the read sensor value to display a result of determination on whether the device is normal or abnormal.