@article{Lu2026, 
author = {Yan Lu and Rui Lu and Peng Qi and Jian Lu and Yang Gao and Le Yang},
title = {Intelligent Generation Method for Annular Triangular Pipe Truss Segmentation Based on NSGA-Ⅱ Algorithm},
year = {2026},
journal = {Journal of Tianjin University (Science and Technology)},
volume = {59},
number = {8},
pages = {815-826},
keywords = {multi-objective optimization, non-dominated sorting genetic algorithmⅡ(NSGA-Ⅱ) algorithm, triangular pipe truss, segmented hoisting},
url = {https://www.sciopen.com/article/10.11784/tdxbz202510027},
doi = {10.11784/tdxbz202510027},
abstract = {In this paper, the non-dominated sorting genetic algorithmⅡ(NSGA-Ⅱ)is applied to the segmental construction process of circular triangular pipe trusses, realizing the multi-objective optimization of construction cost, construction period and structural safety. The expression method for the segmentation of triangular pipe trusses is simplified, a basic unit division method for triangular pipe trusses is proposed, and the concept of basic unit matrix is put forward. The physical model is converted into a mathematical model, thus transforming the segmentation problem of the integral truss structure into a permutation and combination problem of basic units. The generation of different segmentation schemes is achieved by generating different arrays and conducting permutation operations on them. A multi-objective optimization model for cost, construction period and safety is established, which enables the optimization selection of hoisting machinery and the calculation of jig frame layout cost varying with segmentation conditions. A construction period optimization model with the hoisting time of the inner ring truss as the critical path is determined, and a safety optimization model with the configuration degree as an evaluation index for structural safety is also established. By comparing the optimization results of the NSGA-Ⅱ algorithm with the segmentation results of a case project, it can be concluded that the results calculated by the algorithm achieve certain improvements in terms of cost, construction period and safety, which verifies the feasibility of the algorithm. The weights of cost, construction period and safety are determined as 0.0438, 0.4247 and 0.5315 respectively by a method of weighting with the entropy weight method and correction with the analytic hierarchy process. Finally, Segmentation Scheme 1 has the highest relative closeness degree, and its cost, construction period and safety coefficient are optimized by 5.2%, 24.0% and 116.0% respectively compared with those under the original scheme.}
}