Electric Machine Behavior Prediction Using Multi-Physics Surrogate Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Predicting the behavior of electric machines during the initial design stage is challenging due to the complexity of multiple interacting components and physical domains, leading to increased time-to-market and potential design errors when noise and vibration analysis is performed later in the design lifecycle.
Innovation Solution
A computer-implemented method and system that generates a simulated dataset for electromagnetic, structural, and acoustic properties of electric machines using parametric models, trains artificial neural network models, and predicts behavior by orchestrating the execution of these models for custom design parameters, enabling fast and accurate early-design stage behavior prediction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If noise and vibration behavior analysis is performed using traditional multi-physical models during early design stage, then prediction accuracy is improved, but computational time and complexity increase significantly
Solution Approach 1:
The system performs preliminary actions by pre-computing and storing the relationship between design parameters and noise/vibration behavior in a dataset during the training phase. This pre-computed knowledge is then rapidly applied during early design stage predictions, eliminating the need for time-consuming multi-physical simulations at that stage while maintaining high accuracy.
Solution Approach 2:
The system creates a simplified copy or surrogate model (the trained machine learning model) that replicates the complex multi-physical behavior of electric machines. This copy can be executed rapidly during early design stage without requiring the full computational resources of the original multi-physical models, thus reducing computational time while preserving prediction accuracy.
2Measurement precision
If traditional multi-physical models are used for noise and vibration analysis, then prediction accuracy is improved, but device complexity increases
Solution Approach 1:
The system replaces the complex mechanical multi-physical simulation models with a data-driven machine learning model. The trained model captures the essential physics-based relationships between design parameters and noise/vibration behavior without requiring explicit representation of electromagnetic, structural, and acoustic physics domains, thus reducing model complexity while maintaining prediction accuracy.
3Use of energy by stationary object
If noise and vibration analysis is delayed to later design stage, then computational resources are reduced, but time-to-market increases due to redesign requirements
Solution Approach 1:
The system performs noise and vibration behavior prediction in advance during the early design stage using the trained model. This preliminary assessment allows designers to identify and correct potential issues before committing to detailed design and manufacturing, thereby avoiding costly redesigns and reducing time-to-market despite the additional computational investment in model training.
Data Source
AI summary
A system and method of predicting behavior of at least one electric machine is provided, wherein the method includes: generating a simulated-dataset including simulated design results, (e.g., individually), for electromagnetic properties, structural properties, and acoustic properties of the electric machine, wherein the simulated-dataset is generated by simulating at least one operating condition of the electric machine on parametric models generated from design parameters of the electric machine; training artificial neural network models using the design parameters and the simulated design results output from the parametric models in response to at least one operating condition of the electric machine; and predicting behavior of the electric machine by orchestrating execution of the artificial neural network models for custom design parameters.