Semi-Supervised Fault Diagnosis for Machine Condition Monitoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing machine condition monitoring algorithms rely heavily on supervised approaches that require annotated training data, which can be difficult to obtain and prone to errors, and unsupervised techniques often fail to capture meaningful patterns in sensor signals without proper supervision.
Innovation Solution
A semi-supervised framework incorporating time series segmentation, prototype selection, and supervised learning to generate a discriminative model for machine condition monitoring, allowing for the classification of machine conditions without extensive user annotation, using multivariate sensor data to identify normal and failure behaviors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If supervised approaches are used for machine condition monitoring, then classification accuracy can be improved, but the requirement for annotated training data increases significantly
Solution Approach 1:
The system performs self-training by automatically generating pseudo-labels for unlabeled data and iteratively improving the model without requiring extensive expert annotation. The algorithm monitors its own performance and adapts to new failure patterns autonomously
Solution Approach 2:
The approach changes the parameter of data annotation from manual expert labeling to automatic pseudo-labeling, transforming the nature of training data preparation while maintaining classification accuracy through iterative refinement
2Ease of operation
If unsupervised techniques are used for machine condition monitoring, then ease of operation is improved, but the ability to capture meaningful patterns in sensor signals deteriorates
Solution Approach 1:
The system segments the learning process into distinct phases: initial unsupervised clustering to discover patterns, followed by supervised fine-tuning to refine classification. This segmentation allows the system to benefit from both unsupervised ease of operation and supervised accuracy
Solution Approach 2:
The system performs preliminary unsupervised clustering to identify potential failure patterns before applying supervised classification. This preliminary action prepares the data structure to enable accurate pattern recognition without requiring manual annotation
3Measurement precision
If supervised approaches are used for fault diagnosis, then diagnostic accuracy is improved, but device complexity increases due to requirement for expert knowledge
Solution Approach 1:
The system automatically adapts to new failure modes and refines its diagnostic capabilities without requiring external expert intervention. It performs self-updating by learning from newly labeled data, reducing the complexity burden of expert knowledge management
Data Source
AI summary
A computer-implemented method for performing machine condition monitoring for fault diagnosis includes collecting multivariate time series data from a plurality of sensors in a machine and partitioning the multivariate time series data into a plurality of segment clusters. Each segment cluster corresponds to one of a plurality of class labels related to machine condition monitoring. Next, the segment clusters are clustered into segment cluster prototypes. The segment clusters and the segment cluster prototypes are used to learn a discriminative model that predicts a class label. Then, as new multivariate time series data is collected from the sensors in the machine, the discriminative model may be used to predict a new class label corresponding to segments included in the new multivariate time series data. If the new class label indicates a potential fault in operation of the machine, a notification may be provided to one or more users.