Shear Wall Structural Layout Using Physics-Enhanced GAN Design
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing intelligent structural design methods for shear wall buildings struggle to learn both physical mechanisms and data simultaneously, leading to low practical value and inefficiency in design outcomes.
Innovation Solution
A physics-enhanced data-driven method using a generative adversarial network to integrate architectural design images and text data, performing training to generate structural design images that meet physical performance criteria, including vectorization and multi-degree-of-freedom mechanics calculations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional artificial design methods are used, then design results can be obtained, but the process is time-consuming and labor-intensive with low efficiency
Solution Approach 1:
The patent replaces the mechanical manual design process with an intelligent system that uses generative adversarial networks (GANs) and physics-enhanced data-driven algorithms to automatically generate structural design images, eliminating the need for manual drawing and reducing design time significantly
Solution Approach 2:
The system creates synthetic training data by copying and generating structural design images through the GAN model, allowing the system to learn from and reproduce design patterns without requiring manual creation of each design image, thereby improving efficiency
2Productivity
If data-driven intelligent algorithms are used, then design speed increases, but the algorithms cannot learn physical mechanisms and empirical knowledge, resulting in low practical value
Solution Approach 1:
The patent merges data-driven GAN algorithms with physics-based verification methods by integrating a physics engine that calculates structural performance (strength, stiffness, stability) and combines this with the generative design process, allowing the system to both generate designs quickly and verify their physical validity
Solution Approach 2:
The system implements feedback mechanisms where physics-based verification results are fed back into the design process, allowing the GAN model to learn from physical performance data and adjust future generations to better satisfy structural constraints, thereby improving reliability while maintaining speed
3Measurement precision
If neural networks are trained only on image data, then image characteristics can be learned effectively, but physical mechanisms of structural design cannot be learned directly
Solution Approach 1:
The patent introduces a physics engine as an intermediary between the neural network and physical reality, where the physics engine calculates structural performance based on design images and feeds this information back to the GAN model, enabling the network to learn physical mechanisms indirectly through this mediator
Solution Approach 2:
The system adds a new dimension to the training process by incorporating physics-based verification results as an additional training signal, transforming the neural network from learning only image characteristics to learning both image patterns and physical constraints through the extended training framework
4Reliability
If traditional design methods are used, then physical constraints can be satisfied, but the process lacks intelligence and automation
Solution Approach 1:
The system enables self-service automation where the GAN model automatically generates design images, the physics engine automatically verifies structural constraints, and the system automatically iterates on designs without requiring manual intervention, achieving high automation while maintaining constraint satisfaction through the integrated automated workflow
Data Source
AI summary
The present application provides a physics-enhanced data-driven method and device for intelligent structural design of shear wall building, and the method includes: obtaining an architectural design image and a basic design condition text to be processed; inputting the architectural design image and the basic design condition text into a structural design model, and obtaining a structural design image; the structural design model being obtained by performing a structural design image generation capability training and a physical performance optimization training for a physics-enhanced data-driven generative adversarial network; and vectorizing the structural design image and the architectural design image to obtain a structural design result of shear wall building. The physics-enhanced data-driven method and device for intelligent structural design of shear wall building provided by the present application improve the efficiency and the reliability of structural design.