Context-Aware Diagnostic Modeling for Abnormal Sound Source Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing diagnostic techniques for image forming apparatuses fail to accurately determine the source of abnormal sounds due to insufficient operation information, leading to incorrect diagnoses.
Innovation Solution
A diagnostic device that receives context information and sensing information, using a model to determine the state of a target item based on the sensing information and context information, allowing for more accurate abnormality detection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If operation timing chart is used to estimate abnormal sound source, then diagnosis can be performed, but diagnosis accuracy is insufficient
Solution Approach 1:
The patent segments the diagnosis process by dividing operation information into multiple types (operation timing chart, operation log, maintenance log) and segments abnormal sounds into individual sound components. Each segment is analyzed separately and integrated to achieve comprehensive and accurate diagnosis of the abnormal sound source.
Solution Approach 2:
The patent adds new dimensions to the diagnosis by incorporating multiple types of operation information (timing, logs, maintenance records) and analyzing sound components in the frequency domain. This multi-dimensional approach transforms the insufficient 1D timing chart into a comprehensive multi-faceted analysis system.
2Measurement precision
If simple operation timing chart is used, then device complexity is low, but diagnosis accuracy deteriorates
Solution Approach 1:
The patent creates a universal diagnosis system that handles multiple types of operation information (timing charts, operation logs, maintenance logs) and multiple abnormal sound types through a unified processing framework. This multi-functional approach improves accuracy without proportionally increasing complexity.
Solution Approach 2:
The patent introduces an intermediary information processing layer that integrates diverse operation information types and sound component analyses. This intermediary layer coordinates the complex information flow and produces accurate diagnosis results without requiring direct complex interactions between all information elements.
3Reliability
If operation information is insufficient, then information processing is simple, but sound generation source estimation becomes incorrect
Solution Approach 1:
The patent performs preliminary collection and organization of multiple types of operation information (timing charts, operation logs, maintenance logs) before conducting abnormal sound analysis. This preliminary preparation ensures that comprehensive information is available for reliable sound source estimation without processing complexity during the actual diagnosis.
Solution Approach 2:
The patent merges multiple information sources (operation timing chart, operation log, maintenance log) and sound component analyses into a unified diagnosis result. This combination approach ensures reliable sound source estimation by integrating complementary information from different sources.
Data Source
AI summary
A diagnostic device includes a reception unit and a determination unit. The reception unit is configured to receive context information and sensing information. The context information corresponds to a certain operation of a target item that constitutes a target device. The context information is a piece of a plurality of pieces of context information each describing an operation of the target item determined depending on a type of operation of the target device. The sensing information is on a physical quantity that varies in accordance with the operation of the target item. The determination unit is configured to determine a state of the target item based on the sensing information detected while the target item is performing the certain operation, and based on a model corresponding to the received context information. The model is a model of one or more models respectively defined for one or more pieces of the context information.