DECISION-MAKING SUPPORT SYSTEMS IN CONSTRUCTION PROJECTS BASED ON BAYES NETWORKS
Roman Osypenko, Оlena Lialiuk, Denys Melnyk · Modern technology materials and design in construction · 2024
The work presents support systems for decision-making in conditions of uncertainty or incomplete information inconstruction projects. With the help of integrated databases, the minimum level of information is determined, which can beused both for data extrapolation and for filling decision-making models. In order to estimate the probability of parameters, forexample, the cost or duration of their compilation when applying BIM modeling, a hypothesis calculation was carried out, whichis based on a probabilistic graph model of networks Bayes.On the one hand, BIM is a necessary technique both for the construction of new buildings, and on the other hand, itreceives particular attention and interest from owners of large construction funds who want to take advantage of buildinginformation modeling to have a coordinated system for joint use during construction, modernization and operation of buildingsand structures.Especially in a process related to the management and maintenance of large construction stocks, it involves the processingof uncertain information in BIM. When working with existing buildings, due to the absence and/or incomplete availability ofdocumentation, which entails significant investments in terms of time and additional costs.Therefore, to represent the reliability of existing data, it is worth introducing a tool based on a graphical probabilisticBayesian network model that offers valid decision support under uncertainty.