Digital Twin Synchronization Using Surrogate Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional methods for monitoring and updating the state or condition of complex vehicle subsystems are time-consuming and labor-intensive, especially when dealing with large and integrated systems, as they require extensive data processing and model recalibration with changes or upgrades, making them impractical in terms of breadth of coverage and labor costs.
Innovation Solution
The implementation of a digital twin system that uses a surrogate model trained with simulated data to generate estimates of parameters or variables based on operational data, allowing the digital system model to be executed more quickly and efficiently to update the digital twin, thereby synchronizing it with the current state of the subsystem.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If physics-based models or model-based methods are used to generate state or condition of operative subsystems, then measurement precision is improved, but loss of time increases due to time-consuming execution and model recalibration
Solution Approach 1:
The patent creates a digital twin as a virtual copy of the physical subsystem that replicates its state and behavior. This digital twin is updated using surrogate models that approximate the complex physics-based models, allowing rapid state generation without executing the full physical model simulations each time, thus reducing time loss while maintaining measurement precision.
Solution Approach 2:
The system performs preliminary actions by pre-computing and storing surrogate models that approximate the complex physics-based models. These surrogate models are trained in advance on historical data and simulation results, so that when actual state generation is needed, the pre-trained surrogate models can quickly provide accurate estimates without re-executing the time-consuming physics-based models.
2Measurement precision
If traditional analysis methods are used to monitor complex integrated systems, then measurement precision is improved, but device complexity increases and labor costs increase
Solution Approach 1:
The digital twin serves as a simplified virtual replica that captures the essential state and behavior of the complex physical subsystem without requiring complex real-time analysis methods. The surrogate models provide simplified mathematical representations that can be executed efficiently, reducing device complexity while maintaining the ability to monitor system state with high precision.
Solution Approach 2:
The system enables self-service monitoring where the digital twin automatically updates itself using surrogate models without requiring complex manual analysis procedures. The surrogate models automatically process operational data and update the digital twin state, eliminating the need for labor-intensive traditional analysis methods while maintaining measurement precision.
3Measurement precision
If pattern recognition methods are used to monitor subsystem state, then measurement precision is improved, but loss of time increases due to requirement for large training data and model re-training
Solution Approach 1:
The system performs preliminary training of surrogate models using historical operational data and simulation results before actual monitoring begins. These pre-trained surrogate models are then used to quickly generate system state without requiring continuous re-training, significantly reducing the time loss associated with pattern recognition methods while maintaining measurement precision through accurate state generation.
Data Source
AI summary
The present application relates to a system comprising a processor configured to execute a digital system model based on simulated conditions to generate simulated data. The processor may also be configured to train a surrogate model using at least the simulated data to approximate the digital system model of the system and generate, using the trained surrogate model, estimated values for conditions or parameters of the systems based on operational data, wherein the operational data includes sensor data or in-service data from the system. Further, the processor may be configured to execute the digital system model of the system to generate simulation data based on the operational data and the estimated values or parameters generated by the surrogate model and synchronize or update a digital twin of the system based on the simulation data, wherein the digital twin represents a state or condition of the system.