Research On Risk Evaluation Of Project Program Based On Neural Network

Since risks exist in the entire period of project, effective management of risks is of greatimportance; while risk assessment is a vital link in risk management, therefore, the riskassessment directly determines the success or failure of an entire project. Establishing whatkinds of assessment system and using which assessment method to assess the risks ofconstruction project are issues with increasing significance.Starting from the concepts of risk management for construction project and riskassessment for project, this thesis firstly describes relevant knowledge such as thecharacteristics, connotation, technologies, content and procedures of risk assessment forconstruction project in detail. Secondly, the thesis introduces and analyzes such problems asthe principles and train of thought to establish risk assessment index of construction projectas well as how to quantize assessment index; in accordance with the analyzed principles andtrain of thought to establish risk assessment index of construction project, reasonable riskassessment index system for construction project is established and assessment index systemis successfully quantized. Thirdly, according to the established index system and thecharacteristics of neural network, the thesis designs BP neural network model and the modelhas been studied and tested; meanwhile, the established BP neural network is emulatedthrough using MATLAB software. Lastly, risk assessment is conducted to Penang’s theSecond Bridge in Malaysia by making use of the established index system and BP neuralnetwork model; the assessed result indicates that the major risks of Penang’s the SecondBridge project lie in time limit risk, quality risk and fund risk. In allusion to time limit risk,the thesis provides relevant control measures; in allusion to quality risk, the thesisestablishes quality management control system; in allusion to fund risk, the thesis putforwards relevant suggestions.