This work proposes a novel technique to conduct back‐analysis of lateral displacement of deep cement mixing (DCM) columns in deep excavation construction. For the first time, we propose a process to investigate both soil and underground structure end‐to‐end automatically. The novel technique is a complex combination of three crucial factors: (1) a nature‐inspired optimization algorithm (O), (2) a three‐dimensional PLAXIS Geotechnical Engineering software (P) and (3) a modern Python language programming (P), hereinafter called the a nature‐inspired optimization algorithm (O), a three‐dimensional PLAXIS Geotechnical Engineering software (P) and a modern Python language programming (P) (OPP) technique. The novel meta‐heuristic algorithm simulated the co‐evolved partnership behavior of shrimps and goby fishes, termed Shrimp and Goby Association (SGA), which plays an important role in complex analyses. A series of exams to determine the SGA's performance is conducted on 38 benchmark test functions (IEEE Congress on Evolutionary Computation 2017 and 2019) and three real‐world engineering design problems to showcase its applicability. The metaheuristic and PLAXIS 3D analysis work well together, which makes the back‐analysis technique powerfully to determinate stiffness parameters instead of the traditional approaches. Based on the optimized parameters, the lateral deflection of DCM and soil are well predicted for excavation. This study proposes a technique to estimate efficiently the stiffness parameter for very soft soil. As a consequence of the optimization process, an equation to determine the stiffness parameter of DCM columns from laboratory test is also proposed. Based on the obtained results, this research provides a comprehensive methodology for predicting risk, enhancing safety, saving time and money, and effectively designing and constructing underground structures.

中文翻译：

---

软土深基坑大规模深层水泥搅拌校准的非线性优化方法

这项工作提出了一种对深基坑施工中深层水泥搅拌（DCM）柱横向位移进行反分析的新技术。我们首次提出了一种自动端到端研究土壤和地下结构的过程。这项新技术是三个关键因素的复杂组合：(1) 受自然启发的优化算法 (O)、(2) 三维 PLAXIS 岩土工程软件 (P) 和 (3) 现代 Python 语言编程 (P) ），以下称为自然启发优化算法（O）、三维 PLAXIS 岩土工程软件（P）和现代 Python 语言编程（P）（OPP）技术。这种新颖的元启发式算法模拟了虾和虾虎鱼共同进化的伙伴关系行为，称为虾和虾虎鱼协会（SGA），在复杂分析中发挥着重要作用。针对 38 个基准测试函数（2017 年和 2019 年 IEEE 进化计算大会）和三个实际工程设计问题进行了一系列测试，以确定 SGA 的性能，以展示其适用性。元启发式和 PLAXIS 3D 分析可以很好地协同工作，这使得反分析技术能够取代传统方法来确定刚度参数。基于优化参数，可以很好地预测开挖时 DCM 和土体的横向挠度。本研究提出了一种有效估计非常软土的刚度参数的技术。作为优化过程的结果，还提出了根据实验室测试确定 DCM 柱刚度参数的方程。基于所获得的结果，本研究提供了预测风险、增强安全性、节省时间和金钱以及有效设计和施工地下结构的综合方法。