Neural Network Condition Quantification Using Hypertubes
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Complex and time-consuming analyses are required to determine the condition of machines equipped with sensors, which can be costly and inefficient for large-scale distributed organizations.
Innovation Solution
A neural network is used to quantify the condition of an item by defining 'good' and 'bad' hypertubes in a neural state space based on sensed attribute values, with the condition being determined by the location of a current item state hyperpoint relative to these hypertubes, and visual indicators are provided to represent the condition.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If human analysts manually examine sensor data to determine machine condition, then analysis accuracy can be maintained, but analysis time and cost increase significantly
Solution Approach 1:
The patent replaces the mechanical human analysis process with an automated neural network system. The neural network receives sensor data and automatically determines machine condition, eliminating manual intervention while maintaining assessment accuracy through learned patterns from training data.
Solution Approach 2:
The patent introduces a neural network as an intermediary between sensor data and condition assessment. This intermediary processes the raw sensor inputs through learned transformations to produce condition determinations, acting as a bridge that automates the analysis while preserving accuracy.
2Reliability
If comprehensive sensor data analysis is performed to accurately determine machine condition, then condition assessment quality improves, but system complexity and cost increase
Solution Approach 1:
The patent transforms the condition determination problem into a parameter-based classification task. The neural network learns to map sensor parameter combinations to condition states through training, changing the approach from complex qualitative analysis to quantitative parameter evaluation that is more systematic and less complex.
Solution Approach 2:
The patent segments the condition determination process into distinct phases: training phase where the neural network learns from labeled data, and operation phase where it applies learned patterns to new sensor data. This segmentation allows the system to achieve high reliability through proper training while keeping the operational complexity manageable.
Data Source
AI summary
Quantification of a condition of a selected item using a neural network is disclosed. A device defines a good hypertube in a neural state space based on good item state points obtained from one or more items that exhibit desired operating characteristics, and a bad hypertube in the neural state space based on bad item state points obtained from one or more items that exhibit undesirable operating characteristics. A current item state hyperpoint is determined in the neural state space based on a current item state point of the selected item. A condition of the selected item is quantified as a function of a location of the current item state hyperpoint with respect to at least a portion of the good hypertube and with respect to at least a portion of the bad hypertube.