| Prefabricated buildings have been widely promoted in China in recent years because of their high efficiency,low pollution and fast construction progress.Compared with traditional cast-in-place buildings,prefabricated buildings include three stages: production,transportation and construction,and each link is interrelated and affects each other,which increases the risk of construction progress.At the same time,the risk management method of construction progress is not perfect,which restricts the advantages of prefabricated building construction progress.Robust project schedule planning theory uses time buffer to deal with project schedule risks and improve project schedule performance.At present,there is a lack of research on the robust construction schedule of prefabricated buildings.In this paper,the formulation of robust construction schedule based on decentralized buffering is studied.By studying the risk factors and risk size of start-up delay of prefabricated building construction activities,the construction schedule is established by using Bayesian network to solve the activity delay probability and improve the insertion time buffering method.The main research work is as follows:First,the key activities of prefabricated building projects and the risk factors of delay in starting key activities are identified.The construction process of prefabricated buildings is divided into stacking,hoisting and connection stages,and the key construction activities in each stage are identified by field investigation.Using the method of literature analysis,the risk factors affecting the start-up time of key construction activities are initially identified.According to the expert investigation and the entropy weight TOPSIS method,the risk factors are sorted and screened according to their importance,and the final list of risk factors is formed.At the same time,the probability of risk factors is obtained by expert investigation,which lays the foundation for the later Bayesian network calculation.Secondly,based on Bayesian network,the risk assessment model of start-up delay of prefabricated building construction activities is constructed.The risk factors of start-up delay and key construction activities are input into Bayesian network as network nodes,and using GTT structure learning algorithm of Ge NIe software to determine Bayesian network structure.Then,the structure is optimized according to causal correlation and expert investigation.The triangular fuzzy number representation is introduced as the prior probability of the parent node,and the conditional probability algorithm of the child node is adjusted according to the number of the parent nodes,so as to construct the risk assessment model of the construction delay of the assembled building.The model is used to calculate the probability of each key construction activity starting delay event,that is,the risk of construction activity starting delay.Then,a robust construction schedule based on the improved STC(Starting Time Criticality)decentralized buffering method is established.Aiming at the defects of the original STC distributed buffer method in solving the activity delay probability and calculating the penalty cost,Bayesian network is used to solve the activity delay probability,and the direct and indirect costs caused by the activity delay in the assembled building construction site are analyzed,so as to improve the calculation method of the penalty cost.On the basis of the construction schedule based on the improved STC,this paper considers the preference of project managers in the robustness of completion and planning.The construction method of robust schedule based on the improved STC is researched.Finally,the robust construction schedule based on the improved STC method is made by using a prefabricated residential project in Daxing District,Beijing.The simulation analysis of the improved STC method and the equal time buffer method is carried out.Through the analysis of robust schedule based on improved STC method,the difference of insertion time buffer size under different managers’ robust preferences is obtained.At the same time,Monte Carlo is used to simulate the total project duration of the schedule made by different time buffering methods.The improved STC method has a good comprehensive performance in terms of on-time completion rate and shortest completion period.In this paper,the theory of robust project scheduling is introduced into the preparation of assembly building construction schedule,which enriches the existing preparation methods of assembly building schedule,improves the index system of assembly building activity start-up delay risk,and establishes an activity start-up delay risk assessment model based on Bayesian network,which provides a more scientific and effective method for assembly building construction schedule management.At the same time,the existing STC decentralized buffer method is improved to improve the risk probability of activity delay,which also enriches the robust project scheduling theory. |
