Distribution Network Construction Cost Estimation Model Based On RBF Neural Network

Cost estimation is the main basis for investment plan and fund raising, is a more important data which affects directly the investment control. According to statistics, the weight of the cost estimation affectting investment control is up to 55%-65% during pre-project stage. Due to the cost estimation often occurs in the early stages, many unpredictable things that affect the project investment will appear in the process of projects, so the preparation of the cost estimation is very difficult. The phenomenon of high surpluses and conservative estimation are widespread, which directly increase the pressure of fund raising and lead the extensive enterprise management. With the State Grid proposes efforts to solve the "two weak" of the power grid of UHV and distribution network, the investment of distribution network engineering projects increases year by year. But the study of distribution network engineering cost estimation is rare currently. So how to find a feasible method of cost estimation for distribution network engineering has a very far-reaching significance.In this thesis, based on extensive research of the domestic and foreign government engineering cost estimation method and model, through the "significant cost(CS)" models and engineering consulting historical data to identify the estimated impact of a major distribution network engineering investment factors, and then the model of distribution network engineering cost estimation is established on the base of rbf neural network theory. The example shows that investment of the distribution network construction project can be estimated fast and accurately by the model, and then the capital error can be controlled within 10%, which reduces the pressure of fund raising and improves the level of whole process cost management.