Research On Rebar Purchasing Price Forecasting Of Construction Project Based On Fusion Model

At present,the construction of China’s engineering projects is in a stage of vigorous development,and the procurement of materials in the construction of engineering projects is an indispensable link.How to effectively control the procurement cost of materials is the key content of project management.In this context,this paper studies how to accurately predict the purchase price of the rebar which is widely used in construction projects.Based on the results of the forecast,it is expected to provide decision support for the procurement personnel and project managers who can implement a better procurement strategy,thereby reducing the procurement cost of the project and optimizing the operation quality of the project.In the existing related research,most scholars are concerned about the forecast of rebar market price or the forecast of related good’s market prices,such as the price of electricity and agricultural products.However,there are significant differences between the organization’s internal procurement of rebar and above problems.For example,the forecast of market rebar prices does not need to consider the impact of the supplier’s supply cycle.Based on the above analysis,in order to better solve the problem of the rebar’s purchase price forecast from the internal perspective of the organization,this paper first conducted field investigation and related literature review and analyzed the solutions to the related problem’s(single model prediction)advantages and disadvantages,This paper introduces the concept of ensemble learning,and uses the idea of ensemble learning to improve the accuracy of the rebar purchase price prediction through the integration of multiple models.The main research work of this paper is as follows:(1)Establishing a predictive index system for the purchase price of rebar in construction projects.Through the field investigation and other related methods,the characteristics and modes of project procurement and the problems in the procurement process were analyzed.Combined with the characteristics of project procurement and the principle of index system designed,the price forecasting index of rebar in construction project was constructed.It is verified by relevant characteristic analysis.(2)Simulation of the price prediction model of rebar purchase based on machine learning.Based on grid search and five-fold cross-validation,the task of random forest and XGBoost model is completed,and the structure of BP neural network is designed.The experimental results show that the prediction performance of the model has achieved relatively good results.(3)Simulation of the price prediction model of rebar purchase based on Multi-model fusion.Random forest,XGBoost and BP neural network are the primary learners of the fusion model.The fusion of the three models is completed based on the Stacking method.Experiments show the accuracy of fusion model is the most excellent when BP neural network is the secondary learner in the fusion model...