Prefabricated Building Component Recommendation System Based On Dynamic Hybrid Forecasting Model

Prefabricated construction,owing to its advantage of less construction waste,less cost,financial savings,standard quality,short construction period and safety,has become an important issue in the construction industrialization and city sustainability.Prefabricated buildings are multi-stage,multi-disciplinary and multi-participant building types with concentrated information of the life cycle.Building Information Modeling(BIM)technology can further promote the development of prefabricated construction.However,there are many limits in the integration of BIM technology and prefabricated building design methods including dilemma in reusing existing BIM components and knowledge,less support of intelligent prefabricated building management platform.Therefore,in the transforming and upgrading process of the traditional construction industry into modern informatization,it should emphasize the informatization degree of prefabricated construction and combine information management technology.Based on BIM library,developing BIM component recommendation system and using intelligent algorithm to recommend BIM components to enhance the acquisition and sharing of components,which can greatly improve the design efficiency.In order to further explore the intelligent application of BIM technology on prefabricated construction,strengthen information transmission and sharing of prefabricated building and improve design efficiency and quality,this study integrates artificial neural network(ANN)with the gray forecasting model,proposes prefabricated building component recommendation system based on dynamic hybrid forecasting model.According to requirement,a prefabricated building knowledge sharing platform is established.The functions of BIM component search and recommendation are realized on the platform.This research proposes a novel algorithm recommendation forecasting system framework,and completes the following main tasks:1.By summarizing domestic and foreign research on BIM component library and artificial intelligence recommendation system,application of BIM component library in various fields and possibility of integrating artificial intelligence technology with BIM are analyzed.A sharing mechanism of prefabricated building BIM components with artificial intelligence based on recommendation system is proposed.This research builds a prefabricated building knowledge sharing platform with a BIM component library and a knowledge library as data support.2.In order to predict the dynamic change of user preferences and knowledge attributes,this research based on GM(1,1),MFGMn(1,1),artificial neural network,proposes a dynamic weighting adjustment algorithm for hybrid grey model to provide a flexible combination to adapt unstable time series and algorithmic support for complete recommendation prediction function for prefabricated building knowledge sharing platform.3.The recommendation system based on probabilistic matrix factorization and optimized hybrid grey model is established to achieve an information filtering and sharing.An illustrative case is presented to verify the effectiveness of recommendation system and the quality of recommendation.The results have shown that this method can accurately provide effective BIM component recommendations and help users find the needed knowledge and components easily and effectively.