Taxonomy-Based Diagnostic Libraries for Multi-Device Fault Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing diagnostic systems require time-consuming development and management of customized rule-based models for each device, making them inefficient for diagnosing devices with diverse characteristics in different environments.
Innovation Solution
A taxonomy-based diagnostic system that stores taxonomic information and diagnostic libraries, allowing for automatic diagnosis by matching device characteristics with pre-customized models, enabling efficient and customized device status determination.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If rule-based fault diagnosis is used for each device, then diagnosis accuracy is ensured, but time consumption for model development and management increases significantly
Solution Approach 1:
The patent creates a universal taxonomy-based diagnostic framework that can handle multiple device types and diagnostic scenarios through a single system architecture. The taxonomy structure allows one diagnostic system to serve multiple functions across different devices by organizing diagnostic knowledge hierarchically, eliminating the need to develop separate rule-based models for each device while maintaining specialized diagnostic accuracy.
Solution Approach 2:
The patent performs preliminary organization of diagnostic knowledge into a taxonomy structure during system setup, categorizing diagnostic rules, features, and procedures hierarchically before actual diagnosis needs to occur. This pre-structuring of diagnostic information allows rapid retrieval and application during runtime, significantly reducing the time required for model development and management while ensuring accurate diagnosis.
2Measurement precision
If customized diagnosis models are developed for each device and operating status, then diagnosis precision is improved, but device complexity of the diagnostic system increases
Solution Approach 1:
The patent segments the diagnostic system into a hierarchical taxonomy structure with distinct layers: device types, operational statuses, diagnostic procedures, and specific rules. This segmentation allows each component to be managed independently while maintaining overall system coherence, reducing complexity by breaking down the monolithic customized model approach into modular, reusable taxonomic elements that can be combined as needed.
Solution Approach 2:
The patent introduces a taxonomic dimension to organize diagnostic knowledge, transforming the flat, device-specific model approach into a multi-dimensional hierarchical structure. This additional organizational dimension allows the system to handle device complexity by mapping relationships across multiple levels (device type → operational status → diagnostic procedure → rule), enabling precise diagnosis without proportionally increasing system complexity.
3Manufacturing precision
If manual model development is performed for each device, then customization accuracy is achieved, but automation level of the diagnostic system remains low
Solution Approach 1:
The patent enables the diagnostic system to automatically generate and configure diagnostic models by querying the taxonomy structure based on device identification information. The system serves itself by autonomously retrieving appropriate diagnostic rules, features, and procedures from the pre-organized taxonomy, eliminating the need for manual model development while maintaining customization accuracy through the structured taxonomic framework.
Solution Approach 2:
The taxonomy structure acts as an intermediary between device identification and diagnostic model selection. Instead of direct manual model development, the system uses the taxonomy as a mediating layer that automatically maps device characteristics to appropriate diagnostic configurations, enabling automation while preserving the precision of customized diagnosis through the structured intermediate representation.
Data Source
AI summary
The present disclosure provides a taxonomy-based diagnostic system and method, and a readable storage medium. The disclosure includes a classification gateway configured to store taxonomic information including classification definition data for identifying a predetermined device and corresponding diagnostic data includes diagnoses made on the predetermined device, features that need to be extracted for diagnosis, and monitoring data required to extract each feature; a database configured to store a plurality of diagnostic libraries corresponding to the taxonomic information including diagnostic models for determining a status of a device based on features of the device; an application unit configured to receive first classification definition data that identifies the target device; a data acquisition unit that acquires monitoring data of the target device; a processing unit that determines status of the target device according to monitoring data of the target device and diagnostic models in diagnostic libraries.