Research On Industrial Structure Transfering Prediction Model Of Cities Along Shanghai-nanjing Railway And Its Application

System’s structure and evolution direction will change under the influence of the impact. Thisarticle is trying to solve consequence of the stability which is led by the impact outside of thecomplicated big system. Research on the component element, and then divide the entire into differentpart in order to realize the structure. Construct different forecasting models for different subsystem.The new models appropriate for themselves contribute to forecasting evolution direction.The whole can be divided into three part through hierarchical division, model construction andresult forecasting so as to get hold of development tendency. The second charpter makes thedisorganized and complicated system into streaked part, thus get the first impression after separatingthe whole into different part with different levels which can be called regional central elements andnon-regional central elements. Charpter three and four build forecasting model to forecast the regionalcentral elements’ evolution direction. After building the interval GERT network model, charpter fiveaims at parameters between non-regional central elements to forecast transformation.Three models have been used in this article, hierarchical structure model, Markov predictionmodel and interval GERT network model. Based on former researchers’ work, the author try toimprove three models in tiny specific.(1) Hierarchical structure model. Improve the method of selecting turning points, and then getgrey fixed weight clustering analysis method which improving the choice of turning points. Select theobjects which can represent position and function of the factors in the system. Build the objects’system and then build the improving model with changing the turning points. Divide the factor’s level.Check the quality of the improving whitenization weight function. Design the solution for theimproving model so as to finish division for system structure.(2) Markov prediction model. Probability matrix is expanded from real probability matrix tointerval probability matrix with the improvement of traditional Markov prediction model. States ofdifferent interval range can be divided and states of objects in each time point can be determinedaccording to the error value of the original sequence and the prediction sequence. Get the objects’probability of interval numbers matrix. After normalization of the probability matrix, the improvedgrey Markov model can be got.(3) Interval GERT network model. Different from the traditional GERT, the information flowparameters is expanded from real number to interval number. Build network nodes’ sequences for different time sequences. Take into account the information arrows of the transformation betweenthemselves and different nodes. After getting different parameters’ interval sequences of informationarrows, the interval GERT network model can be built.Consider the encomic influence after the connection of Shanghai-Nanjing high railway. Throughhierarchical structure model built in charpter two, the cities along Shanghai-Nanjing high railway canbe separated into regional central cities and non-regional central cities. The Markov prediction modelwhich is stated in charpter three and charpter four can be used to forcast the evolution direction ofregional central cities along Shanghai-Nanjing high railway. The interval GERT nerwork model isused in charpter five which states the non-regional central cities’ parameters in information arrowsabout transformation of industry.