Abnormality Detection Control with Adaptive Algorithm Switching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current algorithms for abnormality detection in machine learning lack flexibility to adapt to varying production site needs, such as processing time and accuracy requirements, and do not provide an environment for switching between multiple algorithms effectively.
Innovation Solution
A control device that includes a feature extraction part, a processing part, a determination part, and a switching part, which calculates feature quantities, refers to a learning model, generates determination results, and switches algorithms based on predetermined conditions such as processing time, determination results, and production process changes, allowing for real-time control and adaptive abnormality detection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a single algorithm is used for abnormality detection, then the system is simple to operate, but it cannot adapt to varying processing time and accuracy requirements at different production sites
Solution Approach 1:
The system dynamically switches between multiple algorithms based on predetermined conditions such as processing time requirements and accuracy needs. The switching part selects appropriate algorithms from multiple candidates, allowing the system to adapt its behavior to varying production site requirements without requiring manual reconfiguration.
Solution Approach 2:
The control device incorporates multiple algorithms with different characteristics (high accuracy/long processing time and low accuracy/short processing time) within a single system. This multi-functional approach allows the same device to serve different needs - strict abnormality detection when time permits, and rapid detection when time is critical - without requiring separate systems.
2Adaptability or versatility
If multiple algorithms are provided to meet various needs, then adaptability improves, but the system becomes more complex and difficult to manage
Solution Approach 1:
The system automatically selects the appropriate algorithm based on predetermined conditions without requiring manual intervention. The switching part monitors conditions such as processing time requirements and automatically switches between algorithms, making the system self-managing and eliminating the need for operators to understand or configure multiple algorithms manually.
Solution Approach 2:
The system uses feedback from condition monitoring to automatically adjust algorithm selection. By continuously evaluating predetermined conditions such as processing time constraints and abnormality severity, the switching part provides feedback-driven algorithm selection that maintains operational simplicity while achieving adaptability.
3Measurement precision
If a high-accuracy algorithm is used, then abnormality detection precision improves, but processing time increases
Solution Approach 1:
The system changes the parameter of algorithm selection based on processing time requirements. When sufficient time is available, high-accuracy algorithms are selected; when time is constrained, low-accuracy algorithms with shorter processing times are selected. This parameter-based switching resolves the contradiction by making accuracy and time requirements explicit selection criteria rather than fixed trade-offs.
Solution Approach 2:
The system dynamically adjusts the detection accuracy level based on real-time conditions. Rather than using a fixed algorithm, the system transitions between different accuracy levels by selecting appropriate algorithms according to predetermined conditions, allowing optimal balance between accuracy and processing time for each specific situation.
Data Source
AI summary
The disclosure provides an environment in which it is possible to switch an algorithm involved in an abnormality detection process. A control device: calculates a feature quantity from a state value acquired from a monitored object; uses a learning model on the basis of the calculated feature quantity to execute one of a plurality of types of algorithms for calculating a value indicating the probability that an abnormality is occurring in the monitored object; determines, on the basis of the calculated value, whether the abnormality is occurring; and switches, in accordance with a condition defined in advance, the one algorithm that is executed.