Dual Neural Network Error Classification for Electronic Apparatus
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing electronic apparatuses face challenges in accurately identifying and classifying errors, which can lead to inefficient operation and potential damage, due to the complexity of internal components and the variety of error types.
Innovation Solution
The electronic apparatus employs a dual neural network model approach, where a first neural network model is pre-trained using learning data for various error types, and a second neural network model is trained using processed data to classify error categories corresponding to these types. This allows for accurate identification of error categories based on probability values output by both models.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a single neural network model is used for error classification, then the device complexity is low, but the measurement precision of error types is insufficient
Solution Approach 1:
The patent divides the error classification task into two separate neural network models: a first model that identifies specific error types and a second model that categorizes errors by severity or category. This segmentation allows each model to specialize in one aspect, improving overall classification accuracy while managing complexity through functional division.
2Measurement precision
If multiple neural network models are used for error classification, then the measurement precision of error types is improved, but the device complexity increases
Solution Approach 1:
The patent combines two neural network models into a unified error classification system where the first model's output feeds into the second model. This merging approach leverages the strengths of both models - the first model's detailed error type identification and the second model's categorical classification - to achieve high precision without requiring each individual model to be overly complex.
3Reliability
If comprehensive learning data for all error types is used for training, then the reliability of error identification is improved, but the loss of time for data processing increases
Solution Approach 1:
The patent performs preliminary processing of learning data before training the neural network models. This includes organizing error data into structured formats, pre-labeling error types and categories, and preparing training datasets in advance. By performing these actions beforehand, the system reduces real-time data processing requirements while maintaining high training accuracy and model reliability.
Data Source
AI summary
An apparatus including: a sensor to detect information indicating a status of the electronic apparatus; a memory storing (i) a first neural network model and a second neural network model, pre-trained to classify an error type of the electronic apparatus, and (ii) one or more instructions; and a processor operatively coupled to the memory and configured to execute the one or more instructions stored in the memory that causes the electronic apparatus to acquire first error type information and second error type information by inputting the information indicating the status of the electronic apparatus into each of the first neural network model and the second neural network model, check an error category of the electronic apparatus based on the first error type information and second error type information, and control an operation of the electronic apparatus based on the error category.