| Major construction projects are characterized by high complexity and numerous uncertainties,which can lead to a series of problems such as decision-making errors,cost overruns,and schedule delays,ultimately impacting the success of these projects.The key reason behind these issues lies in the insufficient attention given to the vulnerability of the construction systems.Existing research has predominantly focused on specific dimensions of engineering system vulnerability or has solely relied on a static perspective to assess the vulnerability of these systems,thereby overlooking the interrelationships and propagation effects among the influencing factors.In light of this,the present study adopts complex system theory and Bayesian network theory to systematically investigate the influencing factors and mechanisms underlying the vulnerability of major construction project systems.The specific research methodology,findings,and conclusions are outlined below:(1)The connotations of major construction project systems and the vulnerability of major construction project systems are defined,and the characteristics of major construction projects and the vulnerability of their systems are analyzed.The main contents of complex system theory and Bayesian network theory are explained,along with their relevance to this study.(2)Identification of factors influencing the vulnerability of major construction projects.Thirty-five vulnerability factors are initially identified through case analysis,literature review,and expert interviews.After designing,distributing,and collecting questionnaires,reliability analysis,exploratory factor analysis,and confirmatory factor analysis are comprehensively used to extract 30 vulnerability factors of major construction projects,which are divided into six dimensions: goal factors,environmental factors,technical factors,organizational factors,information factors,and management factors.(3)Analysis of the relationship and pathways of factors influencing the vulnerability of major construction projects is conducted.An explanatory structural model of the vulnerability of major construction projects is constructed using expert interviews and ISM,clarifying the mutual interaction and hierarchical transmission relationships among the influencing factors.The hierarchical structure of vulnerability factors of major construction projects is obtained.The parameter "λ = driving force/dependence" is introduced,and combined with the quadrant classification results of MICMAC based on "driving force-dependence," the vulnerability factors of major construction projects are categorized into the root layer,result layer,and transmission layer.The triggering modes of vulnerability in major construction projects are identified,including primary triggering,secondary triggering,and the coupling effect of both.There are a total of 57 transmission paths for the vulnerability of major construction projects,revealing the mechanisms of vulnerability factors and clarifying the process of their impact on major construction projects.(4)Sensitivity analysis and identification of critical transmission paths of factors influencing the vulnerability of major construction projects are conducted.The ISM structural hierarchy is transformed into a vulnerability BN topology structure,and the Leaky Noisy-Max model is introduced for modification,constructing a vulnerability inference model based on LNM-BN for major construction projects.Forward causal reasoning is used to predict the vulnerability status of major construction projects,while reverse causal reasoning is employed to analyze the probability of vulnerability factors occurring under causal effects,identifying 15 sensitive influencing factors and 9 key transmission paths.Based on the "posterior probability"-"λ = driving force/dependence"-"level of significance," a hierarchical classification of vulnerability influencing factors is proposed from the perspective of reducing the probability of factor occurrence and interrupting vulnerability transmission.A hierarchical division matrix of vulnerability influencing factors is constructed,where the three levels correspond to priority intervention,key control,and dynamic supervision,respectively.A vulnerability warning transmission network is depicted,confirming the mechanism of vulnerability factors influencing major construction projects.An empirical analysis is conducted using the ZW project as an example to validate the applicability of the LNM-BN inference method in the field of vulnerability of major construction projects.(5)Analysis of simulation and intervention strategies for the vulnerability of major construction projects’ systems is conducted.Based on the mechanism of vulnerability factors influencing major construction projects,a system dynamics model is established and validated to simulate the spatiotemporal dynamic evolution of vulnerability over a 60-month period.Using the framework of sensitive influencing factors,key transmission paths,and levels of significance(SCL),four critical primary influencing factors are selected: "qualifications,experience,and capabilities of project stakeholders"(ORV2),"attitudes and behaviors of government and the public"(ENV2),"intelligence level of information systems and platforms"(INV2),and "unreasonable project management decisions"(MAV1).Eight scenarios are designed,simulated,and compared for analysis.The results indicate that starting from the early stages of major construction projects,selecting project stakeholders with significant qualifications,experience,and capabilities,gaining support from the government and the public in terms of attitudes and behaviors,and reducing unreasonable project management decisions are effective intervention strategies for reducing the vulnerability of major construction project systems.Finally,intervention and control strategies are proposed,including "addressing the source and controlling key primary influencing factors","hierarchical monitoring of the occurrence probability of influencing factors" and "interruption of vulnerability transmission paths through reverse transmission."The innovation of this paper lies in:(1)The research is focused on the vulnerability of major construction project systems.30 influencing factors of system vulnerability of major construction projects are identified by considering six dimensions: objectives,environment,technology,organization,information,and management.A multi-level hierarchical structure of vulnerability in major construction project systems is constructed based on ISM.By combining MICMAC,the interrelationships among the vulnerability influencing factors are clarified,revealing the mechanisms of how these factors interact and contribute to the vulnerability of major construction project systems.(2)Bayesian network theory is introduced,and the dynamic transmission relationships of the influencing factors are considered.The Leaky Noisy-Max method is used to modify and optimize the BN model,significantly reducing the complexity of the BN model and the computational burden of its parameters.A vulnerability inference model for major construction project systems based on LNM-BN is constructed,thereby overcoming the limitations of static measurements in assessing the vulnerability of engineering systems.(3)Scenario design is carried out based on system dynamics(SD)and the sensitivity of influencing factors,key transmission paths,and their corresponding levels(SCL).Intervention and control strategies for the vulnerability of major construction project systems are proposed.The research findings can provide theoretical foundations and technical support for the management of vulnerability in major construction project systems. |