Research On The Soil And Water Conservation And The Land Reclamation For The Land Remporary In Railway Constructionbased On BP Neural Network Model

The course of railway construction projects in addition to the main square outside the permanent, temporary works covering a larger number. But, the existing land reclamation and soil-water conservation of railway construction temporary works are often focused on the theoretical design of the project but far away the regional eco-environmental characteristics design. This often makes the soil-water conservation and land reclamation works not good ecological restoration effects. There are not yet to find a better method of discriminated analysis to the rational design of water-soil conservation and the suitability of land reclamation.Though on analyzing China's soil-water conservation and land reclamation in the status quo, the thesis combined with the concept, characteristics, application and theoretical basis of artificial neural network theory and point out the method that analysis the soil-water conservation and the suitability of land reclamation by use of BP neural network model. Artificial Neural Network (ANN) is a structural mimic biological neural connection type system. It can be used to design for pattern analysis, signal processing and so on. BP (Back Propagation) neural network is a kind of artificial neural networks, which can achieve arbitrary nonlinear mapping optimization between the input and output.The BP (Back Propagation) neural network have adaptability, fault-tolerant and the advantages of self-organization which is the traditional statistical methods can not to do. Therefore, there are broad application prospects by using of BP neural network to analysis the soil and water conservation and land reclamation in railway temporary works.Based on the investigation of existed railway projects, we can establish the the BP neural network analysis model about soil-water conservation design and land reclamation. The model's establishment is though to use 6 technical parameters of soil-water conservation and 7 potential factors of land reclamation, then to use the platform tangent function (tansig), the logarithmic function (logsig), the training function (trainlm), the performance function (mse) and the simulation function (sim), such as MATLAB function of the five kinds of input data samples for analysis and nonlinear optimization.Make the temporary occupies of Guiyang - Guangzhou railway (part of Guizhou) to be analyzed by the established BP neural network model. Through input 3 group of the typical wall residue field of soil-water conservation design and 2 typical residue field of land reclamation to analyze by BP neural network analysis model, we can drown an appropriate and suitable conclusion. So, we can see that the 3 groups to determine the typical slag block wall design is reasonable and the 2 residue field of land reclamation were appropriate and sub-suitable for reclamation.The samples were used form railway construction projects is both well data of soil-water conservation and land reclamation. The result can provide guidance and reference to Research soil-water conservation and rational design of new rail construction.