On the cost estimation methods of college engineering based on BP neural network

The engineering cost estimation during the early stage of the project is not only the basis of feasibility study in the investment management of construction project, but is also the vital factor for the estimate of the bidding. What´s more, its accuracy directly erects influences on the decision of investment and construction cost control for the project. By collecting and clearing up the materials of the completed construction projects, this essay applies the BP neural network in the cost estimation of college engineering, defines the proper value which will influence the engineering cost and then builds up the model of neural network. For the purpose of realizing the model, it also applies the MATLAB in order to train, simulate and test the model. Furthermore, it uses the project case to test the model, the result of which proves that this kind of model is on the one hand of great generalization ability, but it can also precisely estimate the engineering cost.