Research And Design Of Land Economic Compensation Based On Neural Networs

With the continuous development of China's economy and the continuous construction of major projects,there are more and more tasks for land expropriation and housing demolition.In the process of land expropriation and housing demolition,although there are the policy support from the state province city and district,but in the real work,there are still many problems.In order to solve the price problem in the process of land expropriation preferably,this thesis studies a compensation algorithm for land expropriation and housing demolition,and gives a specific land expropriation economic compensation model an d verification system.First of all,the present national land expropriation standards and basic land expropriation methods are introduced,with the emphasis on taking Lanzhou city as an example,and the current land expropriation status and land expropriatio n methods of Lanzhou city are introduced in detail.Secondly,this thesis summarizes and analyzes the neural network algorithm,describes the steps and process of establishing BP neural network,and discusses its advantages and disadvantages.In combination with the methods for implementing land requisition of the Lanzhou city,I establish the model of economic compensation for land expropriation,considering the factors that affect the price of land requisition compensation through specific data sample,via screening,normalized processing,training by function of the neural network,I draw the conclusion:adopt the traditional BP algorithm for training data are not suitable for requisition of land economic compensation model.Using the improved LMBP algorithm to model again and learn network,after 12 iteration,the algorithm expected simulation test requirements,and reflect well the real situation about studying affected factors for interrelated constrains on the price of land expropriation compensation.the network convergence speed is fast and fitting effect is good,results compared with the actual price,error values are in the controllable range,which is consistent with the actual data in our daily practical work,so it illustrates using LMBP is correct.On this premise,I explore establishing land expropriation compensation system by using VC++ in accord with Lanzhou basic situation.