Sheet Conveying Sound Analysis for Abnormal Feed Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing sheet conveying devices struggle to accurately detect and prevent abnormal conveyance conditions, such as document slippage or deformation, using conventional machine learning methods that rely on support vector machines and quantitative feature classification.
Innovation Solution
A novel sheet conveying device that employs a sound collector to gather operating sounds, extracts features, and uses circuitry to select appropriate index data sets for determining abnormal conveyance based on specific conveyance conditions, enhancing the accuracy of abnormal conveyance detection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional machine learning methods using support vector machines are used to classify operating sounds into normal conveyance, document deformation, and document slippage, then the system can detect abnormal conveyance conditions, but the detection accuracy is insufficient and false classifications occur
Solution Approach 1:
The patent segments the index data into multiple sets corresponding to different conveyance conditions (e.g., single document conveyance, multiple document conveyance, different paper types). Instead of using a single unified classification model, the system divides the classification task into multiple specialized models, each optimized for specific conveyance scenarios. This segmentation allows the system to achieve higher detection accuracy by selecting the appropriate index data set based on the current conveyance condition, thereby reducing false classifications.
2Measurement precision
If a single set of index data is used for all conveyance conditions, then the system structure remains simple, but the detection accuracy deteriorates under varying conveyance conditions
Solution Approach 1:
The patent implements a dynamic selection mechanism that automatically chooses the appropriate set of index data based on the current conveyance condition. The system monitors conveyance parameters (such as conveyance speed, document thickness, and stack height) and dynamically switches between different index data sets. This dynamic approach allows the system to maintain high detection accuracy across varying conditions without requiring a completely complex reconfiguration, as the selection logic is integrated into the existing control flow.
Solution Approach 2:
The patent changes the parameters of the index data based on conveyance conditions by preparing multiple sets of index data with different characteristics optimized for specific conditions. Each set contains parameters tuned for particular scenarios (e.g., different threshold values for slippage detection, different frequency ranges for deformation detection). The system selects and applies the appropriate parameter set based on real-time conveyance conditions, thereby achieving high detection accuracy without permanently increasing physical system complexity.
3Measurement precision
If multiple sets of index data are prepared for different conveyance conditions, then the detection accuracy improves, but the data management and selection process becomes more complex
Solution Approach 1:
The patent incorporates feedback mechanisms that monitor the effectiveness of the selected index data set and adjust selections based on detection results. The system evaluates whether the current index data set is producing accurate classifications and uses this feedback to refine future selections. This feedback loop simplifies data management by automatically learning which index data sets work best under specific conditions, reducing the need for manual configuration and selection complexity while maintaining high detection accuracy.
Data Source
AI summary
A sheet conveying device includes a conveyor, a sound collector, and circuitry. The conveyor conveys a sheet. The sound collector collects an operating sound when the sheet is conveyed. The circuitry is to extract a feature amount of the operating sound collected by the sound collector, select, among multiple sets of index data serving as index to determine whether an abnormal conveyance of the sheet is to be occurred, a set of index data corresponding to a conveyance condition of the sheet, and determine whether the abnormal conveyance of the sheet is to be occurred based on the feature amount and the set of index data selected from the multiple sets of index data.