Conditional Evaluation Models for Device Abnormality Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for detecting abnormalities in devices fail to accurately consider operating conditions, leading to incomplete detection of device failures.
Innovation Solution
An evaluation device that generates characteristic data by combining operation data and state data, defines classes based on these data, divides the data into groups, and uses conditional models to evaluate the operating state of the device, considering both time and operational conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a conventional evaluation model is used to detect device abnormalities, then the detection process is simple, but the detection accuracy is insufficient because operating conditions are not considered
Solution Approach 1:
The patent segments the evaluation model into multiple conditional models, each corresponding to different operating conditions. The system divides the overall evaluation process into condition identification and condition-specific evaluation stages, allowing each model to focus on specific operating scenarios rather than attempting to handle all conditions universally. This segmentation improves detection accuracy for each condition while managing overall system complexity through modular organization.
Solution Approach 2:
The patent implements a dynamic evaluation system that adapts to changing operating conditions. The system dynamically selects and switches between different conditional models based on the current operating state of the device. This dynamic approach allows the evaluation process to respond to varying conditions in real-time, improving detection accuracy without requiring a single overly complex static model.
2Reliability
If operating conditions are considered in the evaluation model, then detection accuracy improves, but the model complexity increases
Solution Approach 1:
The patent segments the complex evaluation task into manageable conditional models, where each model handles a specific operating condition. This segmentation allows the system to consider multiple operating conditions and their interactions without creating an unmanageably complex monolithic model. Each conditional model can be developed, validated, and maintained independently, improving reliability while controlling complexity.
Solution Approach 2:
The patent introduces condition identification as an intermediary component that bridges the gap between raw operating data and the conditional models. This intermediary layer processes and categorizes operating conditions, then routes them to the appropriate conditional models. This mediation simplifies the overall system architecture by decoupling condition analysis from condition-specific evaluation, making the system more reliable and easier to maintain.
3Measurement precision
If multiple conditional models are used to evaluate different operating conditions, then evaluation precision improves, but the computational burden increases
Solution Approach 1:
The patent performs preliminary action by pre-identifying and categorizing operating conditions before applying the conditional models. The condition identification component prepares and structures the operating condition data in advance, so that when evaluation is needed, the system can quickly select and apply the appropriate conditional model without excessive computational overhead. This preliminary preparation reduces processing time while maintaining high evaluation precision.
Solution Approach 2:
The patent segments the evaluation process into distinct stages: condition identification and condition-specific evaluation. This segmentation allows the system to process only the relevant conditional model for the current operating state rather than evaluating all possible conditions. By dividing the computational task into focused segments, the system achieves high precision evaluation while minimizing unnecessary computational burden and processing time.
Data Source
AI summary
An evaluation device of an embodiment includes a storage, a data generator, a class definer, a characteristic data divider, and an evaluator. The data generator is configured to generate a set of characteristic data from both a set of first data and at least a set of second data, the at least set of second data being associated in time information with the set of first data, the set of characteristic data representing a plurality of characteristics. The characteristic data divider is configured to divide the plurality of sets of characteristic data into a plurality of groups on the basis of the plurality of classes defined by the class definer and condition of operations included in the set of first data. The evaluator is configured to evaluate a operating state using a first model defined for each of the plurality of groups divided by the characteristic data divider.