Research On Risk Assessment Of Construction Project Based On BP Neural Network

With more and more sophisticated construction technology in China,people have more and more requirements for construction projects,not only to meet the needs of society,but also to beautify the environment,which greatly improves the difficulty of construction and increases the risk of the project.This also leads to the modern construction project in order to meet the needs of the society,the construction scale and investment amount are increasing,the risk of construction project becomes more complex,and the difficulty of risk evaluation also increases.It is very important to choose which evaluation method and build which model for modern construction project.This paper introduces an effective BP neural network model to evaluate the risk of construction projects.Through the characteristics of BP neural network model,the evaluation results are more accurate and more in line with the actual situation..In this paper,the construction project is divided into four stages: decision-making,preparation,implementation and operation.By using the methods of literature review,analytic hierarchy process,qualitative and quantitative analysis,the risk assessment is studied through BP neural network model.In this way,the risk assessment system of construction project can be more comprehensive and systematic.Based on this,the BP neural network model is designed.After training and testing,the risk of construction project is evaluated.Finally,the conclusion and countermeasures are given to control the risk and reduce the risk probability.The main research contents and results are as follows:(1)Through a large number of literature research.BP neural network,risk identification,evaluation and other related theories are reviewed.The successful and failed cases at home and abroad are compared,and the experience is drawn and summarized.(2)This paper analyzes and identifies the main risk factors in the four stages of decision-making,preparation,implementation and operation of construction projects,selects the corresponding risk factors and carries out quantitative treatment,and adopts the concept of analytic hierarchy process to build a reasonable risk evaluation index system for construction projects,which meets the principles of the system and conforms to the characteristics of actual construction projects.(3)Design the BP neural network mathematical model in line with the construction project,collect the relevant index data of the construction project,test and train the established model by using MATLAB software,and finally meet the evaluation requirements through continuous learning and adjustment,and verify the authenticity and effectiveness of the evaluation model through data testing.(4)Combined with the case,BP neural network risk evaluation model is used to evaluate the risk of “JHWZ construction project”.According to the results,corresponding countermeasures and suggestions are given for the four stages of construction project to reduce the probability and loss of risk.In a word,the paper takes the related theory as the guidance,verifies the effectiveness of BP neural network risk evaluation model with empirical evidence,and hopes to provide new ideas and methods for the research of construction project risk evaluation.