Vibration Sample Classification With Periodic Model Retraining
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional supervised learning clustering models are not configured for retraining and have long retraining periods, making them ineffective in environments with changing conditions, leading to misidentification of input data as anomalous.
Innovation Solution
A system that periodically retrains machine learning clustering and anomaly detection models using short retraining periods, allowing for adaptability in dynamic environments and effective use in applications of short duration.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If conventional supervised learning clustering models are used, then the system can categorize input data into clusters, but the model cannot be retrained or has long retraining periods, making it unable to adapt to changing environmental conditions
Solution Approach 1:
The patent implements dynamic retraining by periodically updating the clustering model with new data batches. The system transitions from a static conventional model to a dynamic adaptive model that continuously learns from incoming data, adjusting cluster assignments and model parameters based on changing environmental conditions while maintaining operational continuity.
Solution Approach 2:
The system employs periodic retraining cycles where the model is retrained at predetermined intervals using accumulated data batches. This periodic action allows the model to adapt to environmental changes systematically, balancing the need for adaptability with computational resource constraints and operational requirements.
2Reliability
If conventional supervised clustering models with long retraining periods are used, then the model structure can be simple, but the model cannot provide accurate predictions in dynamic environments, leading to hyper-classification of environmental data as anomalous
Solution Approach 1:
The patent ensures continuous model improvement by continuously accumulating data batches and performing incremental retraining operations. The useful action of model adaptation continues without interruption, maintaining prediction accuracy in dynamic environments while managing computational complexity through efficient data batch processing and incremental learning strategies.
3Productivity
If conventional models with long retraining periods are used, then computing resource consumption during operation can be lower, but the models are not economically effective in applications of short duration
Solution Approach 1:
The system applies partial retraining by processing data in batches and performing incremental model updates rather than complete retraining cycles. This partial action approach provides sufficient adaptability for short-duration applications while minimizing computing resource consumption, achieving economic effectiveness by matching the retraining intensity to the application duration requirements.
Data Source
Figure 1
Figure 2
Figure 3~7
AI summary
A sampling device receives, from a transducer computing device located within a predefined proximity to an equipment in an operating environment, a vibration sample from the operating environment and increments a retrain counter. In response to determining that the incremented retrain counter does not meet or exceed a retrain threshold, the sampling device predicts, using a model, an anomalous or non-anomalous designation for the vibration sample and a cluster assignment, to a particular cluster of a set of clusters, for the vibration sample when the model predicts the non-anomalous designation for the vibration sample. The sampling device receives a subsequent vibration sample and further increments the retrain counter. In response to determining that the further incremented retrain counter exceeds a retrain threshold, the sampling device receives a subsequent set of vibration samples and retrains, using the subsequent vibration sample and the subsequent set of vibration samples, the model.