BP Neural Network Based Cost Prediction Model for Transmission Projects

A BP neural network based method for predicting the cost of transmission projects is proposed to quickly and accurately compare the costs and assess the reasonableness of different plans with limited information during the early stage of the decision-making process.The factors influencing the project cost are also analyzed.Furthermore,by putting these factors into BP neural network as input variables and the project cost as an output variable,a BP neural network model of three-tier structure is constructed using actual project data for data training and prediction model verification.The example using this method shows it can forecast the project cost quickly and accurately,which is suitable for comparison of project cost and assessing the reasonableness of different plans.