Digital Twin Model Derivation Using Weighted Hierarchical Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current digital twin technologies face challenges in effectively generating and utilizing hierarchical data sets from multiple sensors to predict and manage events in target areas, such as underground utility tunnels, leading to difficulties in determining the occurrence of management events and visualizing potential hazards.
Innovation Solution
An apparatus and method that collect sensing data from multiple sensors, preprocess it to create weighted hierarchical data sets, and use pre-trained classification models to derive information about specific digital twin models, enabling the determination of event occurrences and visualization of management events through dynamic visual information generation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If sensing data from multiple sensors is collected and processed to create hierarchical data sets for digital twin models, then the accuracy of event prediction and management is improved, but the system complexity and data processing requirements increase
Solution Approach 1:
The system segments the complex monitoring task by creating multiple digital twin models, each representing specific event types (fire, flood, earthquake, etc.). Each digital twin model processes specific hierarchical data sets from relevant sensors, dividing the overall complex system into manageable specialized components that can be independently trained and executed.
Solution Approach 2:
The patent introduces digital twin models as intermediary components between raw sensor data and event prediction outcomes. These digital twins act as mediators that receive hierarchical data sets, apply pre-trained classification models, and generate probability values for specific events, simplifying the overall system architecture while improving prediction accuracy.
2Productivity
If pre-trained classification models are used to derive information about digital twin models, then the processing speed and efficiency are improved, but the initial training time and computational resources required increase
Solution Approach 1:
The system performs preliminary action by pre-training classification models offline before deployment. The digital twin models are trained in advance using historical data and event patterns, storing learned knowledge in the model structure. During actual operation, these pre-trained models rapidly process new hierarchical data sets without requiring retraining, achieving fast real-time prediction while the initial training time is amortized over extended operational use.
3Reliability
If hierarchical data sets with weights are generated from sensor data, then the relevance and accuracy of digital twin model derivation are improved, but the data preprocessing complexity increases
Solution Approach 1:
The patent applies local quality by assigning different weights to different sensors and data sources within the hierarchical data sets. Each sensor's contribution is locally optimized based on its relevance to specific event types and digital twin models. This weighted approach enhances the reliability of model derivation by emphasizing critical data sources while downplaying less relevant ones, without requiring complete redesign of the entire preprocessing system.
Data Source
AI summary
An apparatus configured to drive a digital twin model includes a data collection unit configured to collect sensing data from a plurality of sensors mapped to a plurality of digital twin models for a target area to be managed, a preprocessing unit configured to generate a hierarchical data set assigned with a weight by performing a preprocessing process on the collected sensing data, and a derivation unit configured to derive information about at least one digital twin model corresponding to the hierarchical data set among the plurality of digital twin models by using a pre-trained classification model.