装配式建筑项目风险动态测度及智能决策：基于复杂适应系统视角的研究

As a solution to a series of dilemmas and constraints witnessed in the urban and rural development and construction industry in China, prefabricated construction is envisaged to gain momentum owing to its potential benefits such as environment-friendly, better quality, cleaner and safer working environment. Potential benefits may not be fully exploited if its inherent weaknesses of fragmentation, discontinuity, and poor interoperability not being mitigated, which nurtures a variety of risks that impose significant adverse influence on the performance of prefabricated construction. As a result, delay, cost overrun, and other various problems frequently occurs in prefabricated construction projects. To help address these problems encountered in the prefabricated construction, many studies have investigated risk-related issues in the management of prefabricated construction. However, none of these studies developed an effective tool for managing construction risks of prefabricated construction by envisaging the key characteristics of risks in prefabricated construction, which include complexity, dynamic nature and adaptability. This study applies social network analysis and complex adaptive system approach to identify, analyze, evaluate and handle construction risks with consideration of various stakeholders involved in major production processes of prefabricated construction from complex adaptive system perspective, then an integrated and intelligent decision-making system is devised and developed to incorporate all the developed functions and models to facilitate the decision making by the involved stakeholders. employing multi-agent system method. This study contributes to current knowledge of the management of prefabrication construction by having developed models that offers an in-depth understanding of how construction performance of prefabricated construction is dynamically influenced by interrelationships and interactions underlying various risk variables. The developed integrated and intelligent decision-making system is of value in providing decision making services for stakeholders involved in prefabricated construction to minimize and mitigate the influence of corresponding construction risks prior to implementation.

装配式建筑作为转变城乡建设模式和促进建筑业可持续发展的一种重要手段，得到了国家层面的关注。但装配式建筑建设项目在实施过程中风险事件频发，导致建筑渗漏水、成本超支、工期延误和安全事故等工程问题，直接影响装配式建筑的推广。现有研究缺乏对装配式建筑项目风险所具有的复杂性、动态性、适应性特征的深入思考，无法有效应对装配式建筑项目所面临的风险。本课题拟从复杂适应系统的视角出发，借鉴利益相关者理论、社会网络理论、系统工程理论、决策支持系统理论、信息技术等多学科交叉研究成果，深入研究装配式建筑项目风险的形成机理及传导机制，构建符合复杂适应系统特征的装配式建筑项目风险动态测度模型，通过模拟仿真优化风险应对策略，并在此基础上设计开发集成化和智能化的风险决策支持系统。该课题将有助于装配式建筑项目利益相关者对项目风险进行前端决策、监控预警和风险应对，为提升装配式建筑项目风险管理水平提供理论依据及决策支持。

装配式建筑作为建筑工业化和可持续性发展的必要手段和途径之一，在我国得到大力推广和实施。然而发展过程中不可避免的出现各类问题，特别是涉及更多的新技术集成及众多利益相关者的协同，项目管理问题突出，所面临的风险因素和不确定性也更为复杂，导致装配式建筑项目实施效果不尽人意，阻碍其推广。因此围绕装配式建筑项目进行风险管理，从而规避和减少风险损失、提高项目绩效至关重要。已有研究虽然取得了一定成果，但仍欠缺对风险因素系统性、动态性和不确定性的分析，未能深入剖析风险对项目绩效的影响情况。.本项目首先通过文献分析、专家访谈和验证，构建了多利益相关方视角下的装配式建筑项目风险清单，共计识别了70个与利益相关方关联的风险；然后基于社会网络分析理论和建模，得到了12个关键风险因素和10个关键风险关系；其次将系统动力学和离散事件系统仿真结合，构建用于风险测度和仿真的混合模型，并对模型信度进行校验；模型信度通过检验后，将已识别的12个关键风险导入到模型进行蒙特卡洛模拟；最后基于已有研究成果提出风险应对措施和智能工作包模式的装配式建筑项目管控系统框架。研究成果有助于项目管理人员深入剖析风险不确定性以及相互作用的特征，为其预防和应对风险提供参考建议。