Management And Prediction Of Construction Cost Based On BIM5D

The capacity of cost management which serves as a key part of construction project management,has a direct impact on the profitability of enterprises.The problems of traditional cost management modes,such as extensive operation,insufficient technical competence,poor communication,incapability of realizing dynamic management,etc.The first step of cost management is effective cost prediction.Accurate and dynamic cost prediction is propitious for enterprises to make correct decisions,improve cash flow and raise profits.However,the cost predication has been not put enough importance to,and it is still difficult for the existing methods to satisfy the requirements of timeliness and accuracy.In the process of project construction,there is a lack of effective methods and complete systems for dynamic prediction of its direct costs.With the development of computer technologies and the popularization of intelligent concepts,BIM technology and BP neural network technology emerge as the times require;and BIM5 D platform,which integrates the schedule and cost information,has included the life-cycle data of relevant projects and simulates the whole process of projects through virtual construction.And the improved BP neural network model has strong fitting ability for complex non-linear systems.First of all,this thesis analyzes the existing problems in current cost management,and introduces the concept and advantages of BIM5 D technology and its applications in cost management.Then,the dynamic prediction of direct costs is chosen as the core in cost management;and by referring to the documents,the demands and shortages of the existing cost prediction methods are studies.Finally,the BIM5 D technology and BP neural network model are introduced into the cost prediction of project construction,and they are organically combined,thus a BP neural network prediction model that is optimized by BIM5D-based genetic algorithm is constructed.First,Through obtain engineering quantities through dynamic information of BIM5 D platform.On the basis of a thorough understanding of principles,characteristics,industrial standards and modeling process of BIM technology,the BIM5 D platform of engineering projects is constructed using Revit,Guanglian Da BIM5 D and other software,and the dynamic data of engineering quantities and material usage is obtained.After that,Through predict the unit price of materials according to the BP neural network model optimized by genetic algorithm.Whiling understanding its working principles deeply,phase space reconstruction and genetic algorithm are used to optimize BP neural network model based on its limitations;the optimized BP neural network model is realized at the software level by programming;the data collection of unit price and the training of model have led to an effective prediction of raw materials' unit price.Finally,Through establish a cost predication system relying on actual cases.Through the introduction of a residential building project of an education bureau,the specific application of cost management based on BIM5 D platform is explained in detail.Combining with BIM5 D and BP neural network technology,the dynamic prediction of direct costs is realized,which embodies the advantages of the optimized prediction model in accuracy and stability.In terms of theoretical construction and practical application,this thesis expounds the feasibility and superiority of BIM5 D technology in cost management and cost prediction,striving to provide new ideas for future researches.