Automatic and optimal rebar layout in reinforced concrete structure by decomposed optimization algorithms

In structural designs, the rebar layout often relies on the empirical knowledge from engineers, resulting in higher labor cost, low efficiency, and low accuracy. This paper focuses on automatically designing collision-free rebar layout in Reinforced Concrete (RC) structures. Due to a great number of rebars with various sizes and shapes, the rebar layout task easily suffers from congestions and collisions. An improved method by decomposing the original layout task into simple subtasks is proposed, and each rebar layout subtask is modeled as an optimal trajectory planning problem with collision-free constraints. The Particle Swarm Optimization (PSO), Differential Evolution (DE), and Neighborhood Field Optimization (NFO) algorithms are respectively utilized in the decomposed method to complete subtasks. The PSO, DE, and NFO algorithms are verified on beam-column joints in a RC structure; and the performance with respect to computation time and path length is evaluated. The experimental results show that the decomposed optimization is effective for automatic rebar layout and the PSO is the best algorithm.

Automated summary

This paper focuses on automatically designing collision-free rebar layout in Reinforced Concrete (RC) structures, utilizing Particle Swarm Optimization (PSO), Differential Evolution (DE), and Neighborhood Field Optimization (NFO) algorithms to complete subtasks. Experimental results show that the decomposed optimization is effective for automatic rebar layout, with the PSO algorithm performing the best.