The Research For Investment Predictive Model Based On The Control Technology

It is very important question in the economic field that the control predictive technique is used to analysis the relation of input and output, and it is very complex nonlinear problem too. In this paper, by studying the investment question deeply the predict models are pretended from three angles . They are: the output predictive model is set up by the relation of input and output, the optimal predictive model is set up according to the relation of input and output, and it can obtain the best optimal construction of the investment distribution on the premise of the most output, the quantitative prediction based on economic contrast is realized by using of the different data of input and output. For this, the investment predictive metheds are pretented based on the gray system theory, the artificial neural network theory and synergetic theory, and they can set up the three kinds of investment predictive model goodly.According to the gray system theory the new methed is predented to set up a investment predictive model. In this paper, the relation degrees between various investment are analyzed according to the relevant input and output data, and the prior sequence of investment construction is determined qualitatively. At the same time the concept of elasticity in the economics is lead into the gray prediction, and the calculation method of elasticity is given there. By making use of the gray model of GM(1,1) ,the predictive model is established about year production value . The output value is predicted accurately under the given various investment data. Especially, GM(2,4) model is studied to predict only by a few of history data . The experiment results indicate that this new method is simple and has a high accuracy.In accordance with the optimal question of investment construction, the improved BP algorithm is studied in order to build ANN model which can determine the nonlinear relations of input and output factors. Then the investmentinput is adjusted rationally in the ANN model in order that the best investment distributive proportion is predicted on the premise of non-change for investment total value. In this paper, the ANN model is builded which can reflect the relation of input and output,and the best distributive proportion is obtained. The experament result show that this method has ideal precision and good effect for prediction.In this paper, the synergetic theory is applicted succeessfully in the predictive model. The new method is presented to contrast economics based on the Haken pattern recognition. The synergetic network is builted and the prototype vector and adjoint vector of investment data are obtained. Finally,the economic contrast between different locality and different period is realized according to the synergetic pattern recognition of investment internal relation.