Missing Data Compensation Using GAN Historical Signal Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Missing data in sensor signals due to external factors like power supply abnormalities and electromagnetic interference can lead to misjudgment of equipment status, resulting in decreased production line yield or forced stops.
Innovation Solution
A missing data compensation method that inputs sensing signals, searches for similar historical data sections, calculates data relation diagrams, and uses a feature recognition model to select a candidate data section with maximum similarity to compensate for missing values, thereby generating compensated data sections.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If sensors are used to detect equipment data, then equipment status evaluation is enabled, but missing values occur due to external factors such as power supply abnormality, electromagnetic interference, and sensor overheating
Solution Approach 1:
The patent creates a virtual copy of the missing sensor data by training a generative adversarial network (GAN) model with historical complete sensor data. The trained model generates compensated data that replicates the statistical characteristics and temporal patterns of the original sensor data, effectively creating a digital twin of the missing information without requiring physical sensor redundancy.
Solution Approach 2:
The patent performs preliminary training of the GAN model using historical sensor data before actual compensation is needed. The model learns the statistical properties and temporal correlations of sensor data in advance, so when missing values occur, the pre-trained model can immediately generate compensated data without requiring real-time complex calculations or additional sensor installations.
2Measurement precision
If missing data is not compensated, then data processing is simple, but misjudgment occurs when determining equipment status
Solution Approach 1:
The patent replaces traditional mechanical data processing methods (interpolation, deletion, or imputation) with an intelligent neural network-based GAN model. This substitution enables the system to automatically learn complex temporal patterns and statistical relationships in sensor data, providing more accurate compensation while maintaining computational efficiency through the pre-trained model's ability to generate data rapidly.
3Productivity
If traditional methods are used to handle missing data, then implementation is simple, but production line yield decreases or forced stops occur
Solution Approach 1:
The patent implements a feedback mechanism where the GAN model is continuously trained and refined using historical sensor data that includes both complete and previously compensated data. The model learns from past compensation accuracy and adjusts its parameters to improve future data generation, creating a self-improving system that enhances both reliability and productivity over time through iterative optimization.
Data Source
AI summary
A missing data compensation method, missing data compensation system and non-transitory computer-readable medium are provided in this disclosure. The method includes the following operations: inputting a sensing signal by a sensor; searching for a historical data sections similar to a first data section from the plurality of historical data sections to generate a plurality of candidate data sections; calculating a plurality of data relation diagrams according to the first data section and the candidate data sections, respectively; utilizing a feature recognition model to calculate a plurality of similarity values according to the data relation diagrams; selecting a candidate data section corresponding to the maximum similarity value as a sample data section; and utilizing the data in the sample data section to compensate the data in the first data section to generate compensated data section.