Research On GNGO-ELM Assembled Construction Cost Prediction Model Based On EPC Mode

Compared with traditional buildings,prefabricated building have the construction characteristics of prefabrication of components and assembly construction.This construction method greatly reduces energy consumption,improves production efficiency,shortens construction period,and effectively improves project quality and construction environment.However,the high construction cost has seriously hindered the development of prefabricated building.Therefore,effective control of the construction cost of prefabricated building is crucial to its development.Aiming at the problem that the cost of prefabricated building is higher than that of traditional buildings and difficult to predict,this thesis takes the cost analysis of prefabricated buildings under Engineering Procurement Construction(EPC)mode as the starting point,and proposes to optimize the initial parameters in the extreme learning machine(ELM)model based on the improved Northern Goshawk Optimization Algorithm(GNGO),so as to predict the cost of prefabricated building.This thesis first analyzes the research status of prefabricated building and construction cost prediction,and expounds the relevant theoretical knowledge of EPC mode,prefabricated building and cost prediction.Secondly,based on the analysis and comparison of the cost increment of prefabricated building on the basis of traditional buildings based on the EPC general contracting mode,the cost influencing factors in the whole life cycle of prefabricated buildings under the EPC mode are summarized,and the cost influencing index system of prefabricated building under the EPC mode is constructed.The weight of each index is calculated through the subjective and objective combination weighting method,and the fuzzy comprehensive evaluation method is used to evaluate the index system.Then,multiple strategy fusion was used to improve the Northern Goshawk Algorithm(NGO),and the improved Northern Goshawk Algorithm(GNGO)was verified to be superior to the NGO algorithm in terms of optimization accuracy and convergence speed through testing functions.Finally,the GNGO algorithm is used to optimize the input weights and bias values in the ELM model,and a GNGO-ELM prefabricated building cost prediction model is constructed.Combined with actual engineering project cases,the effectiveness of the GNGO-ELM prediction model is verified through the Matlab R2020 a platform simulation sample data,and compared with the results of other prediction models.This thesis constructs the GNGO-ELM prefabricated building cost prediction model based on EPC mode,and verifies the feasibility of GNGO-ELM model in prefabricated building cost prediction combined with actual engineering projects,which provides a reference for future research on prefabricated building cost prediction methods.