Application of evolutionary algorithm (EA) in automated clash-free rebar placement

Abstract

Rebar placement is a time-consuming and labor-intensive task. Especially in cases where the position and shape need to be adjusted to ensure the rebar does not conflict with each other. Currently, the issue of rebar positioning to avoid clashes is often overlooked during the design phase, leading to repeated rework or correction. This study presents an approach to optimize the position of rebar by applying an evolutionary algorithm in parametric design. First, the concept of rebar collision is defined, and a mathematical formula for collision is developed to measure the severity of collision. A formula to control the deviation of the new layout compared to the reference layout is also constructed. Second, an optimization model based on the NSGA_II multi-objective evolutionary algorithm is built, controlled by national standards code for reinforced concrete design. Applying this process to a case study saves more than 85% of the time in finding optimal solutions compared to the current manual method.