Risk Evaluation And Risk Early Warning Study Of Chemical New Material Product R＆D Projects

The frequency of new material product iterations is gradually increasing,driving the technological development of enterprises,and R＆D activities occupy the core productivity of enterprises.The chemical new material product R＆D projects are systematic and complex,with high initial investment and great uncertainty of output,and often accompanied by great risks while possibly generating great benefits.Therefore,whether we can effectively evaluate and warn the risks of chemical new material product R＆D projects is the key to the successful completion of the projects.Firstly,by combing the relevant literature on R＆D project risk management,this thesis uses the fishbone analysis method to effectively identify the risk shadow factors of chemical new material product R＆D projects,and uses the four quadrant diagram of risk factors to screen the preferred risk influencing factors of chemical new material product R＆D projects and analyze the risk influencing factors at the same time.Secondly,the hierarchical analysis method is selected to calculate the weights of the index system,after which the research method of vague synthetical assess is used to assess the R＆D projects of new chemical material products.Then,the weights of each risk factor are calculated by PCA,to obtain the values of the input neurons of the BP neural network model,by using a BPNN model is constructed to monitor and analyze the risks of chemical new material product R＆D projects.Finally,by introducing a example to proving the availability on the BPNN model,and a risk response strategy is proposed for the degree of early warning of chemical new material product R＆D projects.The results of the study show that the potential risks of chemical new material product R＆D projects in each stage（four stages of project,R＆D,acceptance,and delivery）are identified using fishbone analysis,with a total of 30 risk factors,and 18 risk factors are filtered out through a fourquadrant diagram of risk factors and classified into five categories.Secondly,from the hierarchical analysis,the most important risk factor index weight of the product R＆D project level 1 is management risk,followed by technical risk,the third and fourth are financial risk and environmental risk respectively,and the least important risk factor is market risk,from the fuzzy comprehensive evaluation,financial risk and technical risk are evaluated as the second highest risk,and management risk,environmental risk and market risk are evaluated as medium risk.Thirdly,the principal components derived from dimensionality reduction using principal component analysis are 5,totaling 18 attribute indicators,which are used as input neurons for the BP neural network model.The BP neural network risk warning model can reach 100% accuracy in predicting mediumrisk items,and the accuracy decreases in predicting low-risk and high-risk items,and overall,the BP neural network warning model works well.