Simulation Method for Multi-crew Construction Processes Based on Large Language Model-powered Agent

doi:10.16182/j.issn1004731x.joss.25-0990

Abstract

Abstract:

Traditional construction simulation methods typically rely on predefined rules or static scheduling mechanisms, making it difficult to simulate interactive decision-making under complex constraints by dynamically adapting multi-crew construction scenarios. As a result, workforce idleness caused by process dependencies and spatial occupation fails to be solved. A simulation method for multi-crew construction processes based on a LLM-powered agent was proposed. The distributed crew agents endowed with construction scenario understanding and reasoning capabilities were constructed, as well as a centralized project manager agent. A “single-manager and multiple-crew” decision-making mechanism for multi-agent interaction in construction was designed. Autonomous decision-making and coordination mechanism under the constraints of process, space, and resource was simulated. The simulation results demonstrate that the proposed method can generate efficient construction workflows tailored to different scenarios and can give rise to macroscopic construction strategies adapted to scenarios. It also produces a multi-crew scheduling plan under zero-shot conditions, achieving advanced scheduling algorithm performance without iterative optimization or training and optimizing allocation of human resources for the case project.

Key words: construction of civil engineering, construction simulation, LLM, agent, multi-agent system

CLC Number:

TU722

Wang Yifan, Yang Bin, Wang Congjun. Simulation Method for Multi-crew Construction Processes Based on Large Language Model-powered Agent[J]. Journal of System Simulation, 2026, 38(2): 488-500.

Figures/Tables 21

Fig. 1

Fig. 2

Fig. 3

Fig. 4

Fig. 5

Table 1

Table 2

Fig. 6

Fig. 7

Fig. 8

Fig. 9

Table 3

Table 4

Table 5

Fig. 10

Fig. 11

Fig. 12

Table 6

Fig. 13

Fig. 14

Fig. 15

References 31

[1]	Li Jue, Wang Hongwei. Modeling and Analyzing Multiteam Coordination Task Safety Risks in Socio-technical Systems Based on FRAM and Multiplex Network: Application in the Construction Industry[J]. Reliability Engineering & System Safety, 2023, 229: 108836.
[2]	Anderson N, Ones D S, Kepir Sinangil H, et al. Handbook of Industrial, Work & Organizational Psychology, Vol. 2: Organizational Psychology[M]. London: Sage Publications, 2001.
[3]	Teicholz P M. A Simulation Approach to the Selection of Construction Equipment: Stanford University[R]. CA: Stanford Construction Institute, 1963.
[4]	Halpin D W. An Investigation of the Use of Simulation Networks for Modeling Construction Operations[D]. Urbana-Champaign: University of Illinois at Urbana-Champaign, 1973.
[5]	Hu Di, Mohamed Yasser, Taghaddos Hosein, et al. A Simulation-based Method for Effective Workface Planning of Industrial Construction Projects[J]. Construction Management and Economics, 2018, 36(6): 328-347.
[6]	Zhang Hong, Li Heng. Simulation-based Optimization for Dynamic Resource Allocation[J]. Automation in Construction, 2004, 13(3): 409-420.
[7]	Mahdi Naeimi Rad, Maghrebi Mojtaba. Bi-level Integration of Discrete-event Simulation (DES) and Social Force Modeling (SFM) for Construction Labor Productivity[J]. Automation in Construction, 2024, 160: 105317.
[8]	Son JeongWook, Rojas E M, Shin Seung-Woo. Application of Agent-based Modeling and Simulation to Understanding Complex Management Problems in cem Research[J]. Journal of Civil Engineering and Management, 2015, 21(8): 998-1013.
[9]	Khodabandelu A, Park J. Agent-based Modeling and Simulation in Construction[J]. Automation in Construction, 2021, 131: 103882.
[10]	Ben-Alon Lola, Sacks Rafael. Simulating the Behavior of Trade Crews in Construction Using Agents and Building Information Modeling[J]. Automation in Construction, 2017, 74: 12-27.
[11]	钟登华, 吴康新, 练继亮. 基于多Agent的混凝土坝施工仿真与优化研究[J]. 系统仿真学报, 2008, 20(2): 485-489, 498.
	Zhong Denghua, Wu Kangxin, Lian Jiliang. Research on Construction Simulation and Optimization for Concrete Dam Based on Multi-agent[J]. Journal of System Simulation, 2008, 20(2): 485-489, 498.
[12]	Ahn S, Lee S. Methodology for Creating Empirically Supported Agent-based Simulation with Survey Data for Studying Group Behavior of Construction Workers[J]. Journal of Construction Engineering and Management, 2015, 141(1): 04014065.
[13]	Liu Boda, Yang Bin, Zhang Binghan, et al. Spatio-temporal Deduction of Floor Construction Based on the Agent Modeling of Construction Actors[J]. Automation in Construction, 2022, 142: 104487.
[14]	Mialon G, Dessi R, Lomeli M, et al. Augmented Language Models: A Survey[EB/OL]. (2023-02-15) [2025-09-07]. .
[15]	Park J S, O'Brien J, Cai Jun, et al. Generative Agents: Interactive Simulacra of Human Behavior[C]//Proceedings of the 36th Annual ACM Symposium on User Interface Software and Technology. New York: ACM, 2023: 2.
[16]	范文慧, 蒋沅. 人工智能时代的仿真科学与工程思考[J]. 系统仿真学报, 2025, 37(7): 1607-1623.
	Fan Wenhui, Jiang Yuan. Thinking on Simulation Science and Engineering in the Era of Artificial Intelligence[J]. Journal of System Simulation, 2025, 37(7): 1607-1623.
[17]	朴景华, 高宸, 张芳, 等. 大型社会模拟器:前沿与展望[J]. 系统仿真学报, 2025, 37(7): 1770-1790.
	Jinghua Piao, Gao Chen, Zhang Fang, et al. Large-scale Social Simulator: Frontiers and Perspectives[J]. Journal of System Simulation, 2025, 37(7): 1770-1790.
[18]	Gao Chen, Lan Xiaochong, Li Nian, et al. Large Language Models Empowered Agent-based Modeling and Simulation: A Survey and Perspectives[J]. Humanities and Social Sciences Communications, 2024, 11(1): 1259.
[19]	Zhang An, Chen Yuxin, Sheng Leheng, et al. On Generative Agents in Recommendation[C]//Proceedings of the 47th International ACM SIGIR Conference on Research and Development in Information Retrieval. New York: ACM, 2024: 1807-1817.
[20]	Hua Wenyue, Fan Lizhou, Li Lingyao, et al. War and Peace (WarAgent): Large Language Model-based Multi-agent Simulation of World Wars[EB/OL]. (2024-01-30) [2025-09-07]. .
[21]	Williams R, Hosseinichimeh N, Majumdar A, et al. Epidemic Modeling with Generative Agents[EB/OL]. (2023-07-11) [2025-09-07]. .
[22]	马辉, 张文静, 董美红. 装配式建筑吊装施工空间冲突分析与多目标优化[J]. 中国安全科学学报, 2020, 30(2): 28-34.
	Ma Hui, Zhang Wenjing, Dong Meihong. Spatial Conflict Analysis and Multi-objective Optimization for Prefabricated Building Hoisting Construction[J]. China Safety Science Journal, 2020, 30(2): 28-34.
[23]	Lazari Vasiliki, Chassiakos Athanasios, Karatzas Stylianos. Multi-objective Resource-constrained Scheduling in Large and Repetitive Construction Projects[J]. Algorithms, 2024, 17(8): 351.
[24]	吴杰, 王龙. 论多专业交叉施工项目管理[J]. 工业建筑, 2023, 53(增2): 831-832, 843.
[25]	Hart P E, Nilsson N J, Raphael B. A Formal Basis for the Heuristic Determination of Minimum Cost Paths[J]. IEEE Transactions on Systems Science and Cybernetics, 1968, 4(2): 100-107.
[26]	Brown T B, Mann B, Ryder N, et al. Language Models Are Few-shot Learners[C]//Proceedings of the 34th International Conference on Neural Information Processing Systems. Red Hook: Curran Associates Inc., 2020: 159.
[27]	陆淼嘉, 黄承媛, 滕靖. 基于多智能体的网购生鲜无人车配送调度仿真[J]. 系统仿真学报, 2022, 34(6): 1185-1195.
	Lu Miaojia, Huang Chengyuan, Teng Jing. Multi-agent Simulation for Online Fresh Food Autonomous Delivery[J]. Journal of System Simulation, 2022, 34(6): 1185-1195.
[28]	Han Qie, Shi Dianxi, Shen Tianlong, et al. Joint Optimization of Multi-UAV Target Assignment and Path Planning Based on Multi-agent Reinforcement Learning[J]. IEEE Access, 2019, 7: 146264-146272.
[29]	Taghaddos Maedeh, Taghaddos Hosein, Ulrich Hermann (Rick), et al. Hybrid Multi-mode Simulation and Optimization for Subarea Scheduling in Heavy Industrial Construction[J]. Automation in Construction, 2021, 125: 103616.
[30]	Yang Bin, Liu Boda, Han Yilong, et al. Multi-agent Reinforcement Learning-enhanced Decision-making of Crew Agents in Floor Construction Processes[J]. Advanced Engineering Informatics, 2026, 69, Part C: 103996.
[31]	马振鹏, 焦晗暘, 张哲, 等. 面向城市物流配送的车辆路径优化算法研究[J]. 系统仿真学报, 2025, 37(11): 2768-2777.
	Ma Zhenpeng, Jiao Hanyang, Zhang Zhe, et al. Research on Vehicle Path Optimization Algorithms for Urban Logistics and Distribution[J]. Journal of System Simulation, 2025, 37(11): 2768-2777.

任务名称	数量	安全距离/m	所需材料
接头清理	14	1.2
预制件吊装	14	2.5	预制件
灌浆	14	1.1
钢筋绑扎	22	1.5	钢筋
模板搭建	22	2.0	模板

班组名称	工作效率或任务时长期望	工人数目	工作区域占地面积/m²	移动速度/(m/s)
接头清理班组	11.8 min	3	4	0.4
预制件吊装班组	11.5 min	5	5	0.3
灌浆班组	5 min	1	1.5	0.6
钢筋绑扎班组	0.021 m³/min	2	2.4	0.7
模板搭建班组	0.495 m²/min	2	3.2	0.5

班组配置	指标	随机	最近距离	优先级规则	GA	RL	大模型智能体
5-Agent	平均值	24.20	24.19	23.93	24.02	24.02	23.87
5-Agent	最小值	23.90	23.91	23.37	23.07	23.22	23.57 (-0.5)
6-Agent	平均值	18.13	16.96	16.95	16.77	16.41	16.62 (-0.21)
6-Agent	最小值	17.39	16.79	16.73	16.55	15.27	15.94 (-0.67)
7-Agent	平均值	18.09	16.63	17.47	16.80	17.47	16.17
7-Agent	最小值	17.06	16.25	16.48	16.66	16.48	14.51

班组配置	指标	随机	最近距离	优先级规则	GA	RL	大模型智能体
5-Agent	平均值	3.81	3.79	3.96	3.55	3.63	3.54
5-Agent	最小值	3.51	3.71	3.62	3.40	3.45	3.40
6-Agent	平均值	3.97	3.43	3.28	3.26	2.48	2.69 (-0.21)
6-Agent	最小值	3.54	3.01	2.94	3.07	2.28	2.11
7-Agent	平均值	5.78	5.26	5.57	5.01	4.24	4.43 (-0.19)
7-Agent	最小值	5.26	4.89	4.74	4.92	3.83	3.74

班组配置	GA	RL
5-Agent	21.17	9.65
6-Agent	12.82	10.57
7-Agent	10.64	10.11