Abstract:Structural optimization can lead to efficient designs for high-rise concrete buildings. However, the high computational cost of the structural analysis step hinders its application in practical cases. With regards to this problem, a novel hybrid optimization framework is developed in this study. Discrete variables, code-stipulated constraints and other construction requirements under multiple loading scenarios are considered. The hybrid optimization framework contains two steps: in the first step, a computationally efficient approximation of the optimization formulation is constructed using a combined response surface methodology (RSM). Combining all the design constraints, three sub-optimization problems are proposed based on the relaxation model of the original problem to update the response surface model in the near-optimum region. When the response surface model has sufficient accuracy in the near-optimum region, in the second step, a discrete Particle Swarm Optimization (PSO) technique is applied to search for the optimal solution utilizing the response surface model. By integrating the commercially available ETABS, a dedicated optimization software with an independent interface is developed and details for practical software development are included in this paper. Three examples of high-rise concrete building design are presented to demonstrate the effectiveness of the proposed optimization framework, and the results are compared to those of a general PSO algorithm. It is shown that, compared to the conventional design approach, a 14.3% reduction in material usage is achieved by the hybrid optimization framework. Additionally, the proposed framework significantly improves the computational efficiency compared to PSO. This methodology can therefore be used by engineers to optimize member sizes for high-rise buildings so that stable and sustainable designs can be achieved.

Learning to simulate and design for structural engineering

Size Optimization Design of Members for Shear Wall High-Rise Buildings

Discrete Sizing Optimization Method Based on Dividing Rectangles Algorithm and Local Response Surface for Steel Frame Structures

A Practical Discrete Sizing Optimization Methodology for the Design of High-Rise Concrete Buildings

Differentiable automatic structural optimization using graph deep learning

Structural Design Recommendations in the Early Design Phase using Machine Learning

Computer Simulation for intuitive StruCturing

Integrated design of control/structural systems based on genetic algorithm

Design-Informed Generative Modelling using Structural Optimization

Automatic structural optimization design using sensitivity-based method

A differentiable structural analysis framework for high-performance design optimization

Structural Optimization, the Past, Present and Future

Accelerating Optimization Design of Bio-inspired Interlocking Structures with Machine Learning

Digital twin-based online structural optimization? Yes, it's possible!

Integrated method for intelligent structural design of steel frames based on optimization and machine learning algorithm

Simple and effective strategies for achieving diverse and competitive structural designs

Two-stage automatic structural design of steel frames based on parametric modeling and multi-objective optimization

Physical Design using Differentiable Learned Simulators

Semi-supervised Learning Method Incorporating Structural Optimization for Shear-Wall Structure Design Using Small and Long-Tailed Datasets

Structural optimization based on CAD–CAE integration and metamodeling techniques

Real-time design of architectural structures with differentiable mechanics and neural networks