CAD Structure Generation for Target Physical Parameters
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The current reliance on manual design processes for creating structures with specific physical parameter values is time-consuming and inefficient, especially in industries like automotive and aerospace where precision is critical.
Innovation Solution
A method utilizing a processor-implemented model trained with machine learning on datasets of CAD structures, allowing for the generation of structures that meet target physical parameter values, thereby reducing the need for manual design input.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If manual design processes are used to create structures with specific physical parameter values, then design flexibility and control are maintained, but the time required to create new structures increases significantly
Solution Approach 1:
The patent replaces manual mechanical design processes with an automated machine learning system. The ML model trained on CAD structures and physical parameter data automatically generates structure designs, substituting human manual operations with an automated computational system that processes design parameters and generates optimized structures.
Solution Approach 2:
The system enables self-service design generation where the machine learning model autonomously creates structure designs based on input physical parameter requirements. The model independently processes design specifications, generates candidate structures, and produces designs without requiring continuous manual intervention or expert guidance at each design step.
2Productivity
If manual design processes are used, then design expertise can be applied, but design efficiency and productivity decrease
Solution Approach 1:
The system performs preliminary action by pre-training the machine learning model on extensive datasets of CAD structures and their physical parameters before actual design generation. This preliminary training phase enables the model to quickly generate designs during operation, significantly improving design efficiency without requiring complex real-time computations or manual expertise during the design generation process.
3Manufacturing precision
If more manual refinement is performed to reduce flaws, then structure quality improves, but the time and resources required increase
Solution Approach 1:
The system incorporates feedback mechanisms where the machine learning model is trained on datasets containing physical parameter measurements from actual structures. This feedback from real-world performance data enables the model to generate designs with fewer flaws initially, reducing the need for time-consuming manual refinement cycles while maintaining or improving structure quality.
Data Source
AI summary
The present disclosure provides a method for generating a structure of an object with a CAD system that meets a target physical parameter value. First a target physical parameter value is set. A model trained using machine learning on a dataset comprising data corresponding to structures of objects having known structural and physical parameter values is obtained, the model establishing relations between the structural parameter values and the physical parameter values. Based on that model and the target parameter value, a generated structure of the object is constructed.