Study Of Economic Forecasting In Harbour Engineering Project

The core of management is decision-making, and scientific forecasts provide assurance for right decision. Economic forecasting is widely applied in every area of society and economy, it is important to grasp accurately the situation of future development for planning the future development stratagem or investing an engineering project. In the harbor project, forecasting the port throughput precisely is the key to deciding the scale of harbor construction.On the basis of some basic notions of economic forecasting, three forecasting models and one combined forecasting model are proposed in this paper. Basic notions, operating(modeling) process and indexes of checkout of artificial neural network (ANN), genetic algorithm(GA), grey theory and linear regression model are analyzed. By improving BP neural network and combining genetic algorithm with BP neural network, the performance of model is advanced. Some limitations in the grey model (GM) are ameliorated. The checkout and amending method of commonly-used linear regression are studied.Above three models are brought forward from different aspects, and they are all unilateral. So a combined forecasting model is introduced, and a quadratic programming algorithm for the optimal weight coefficient of combined forecasting is presented in this paper.As an example the throughput of Tianjin port is studied from different aspects. It is shown that the precision of GA&BP model is higher than GM and linear regression, the precision of GM model is the lowest, and the precision of combined forecasting is the highest. The combined forecasting is close to practical case due to considering variouse facters.