Study On BP Model For The Selection Of The Maximum Grade And Minimum Radius Of Curve Of Railway Line In Mountain Area

In the background that the main direction of China’s economic construction turns into the Midwest,the construction and development of mountain railway will become a hot issue.Mountain railway line is generally made up valley lines and uphill lines,There are many problems in the construction of railway line in mountain area,such as the different height,large landslide,mudslides and other natural disasters.However,the standards of the railway line is a bit higher,linear flat(maximum grade is generally not more than 20‰),which gives railway line in mountain area great difficulties to select line of work.The maximum grade and minimum radius of curve of railway line in mountain area influence not only on the scale of investment in railway construction,but also affect the transmission capacity of the line,and the impact of project operations is also more significant.During the standards selection of railway line in mountain area,the maximum grade and the minimum radius of curve is selected based on past experience in engineering practice,which lack a quantitative selection methods.The author try to establish a quantitative method of BP neural network model.There are great theoretical and practical significances.1.The neural network model is a highly nonlinear puzzle solving.It can fully approximate any complex nonlinear function,with a strong ability to deal with complex nonlinear problems.Through the analysis,BP neural network can be established highly nonlinear mapping of the maximum grade and the minimum radius of curve between their influence factors.Thus,it achieve the selection of maximum grade and minimum radius of curve to emulate.2.the author chooses the railroad grade,terrain conditions,transportation requirements and other factors as characterization.Then he chooses the earthwork,the length of the bridge and tunnel of engineering standards and economic indicators as the chosen measure of the maximum grade and minimum radius of curve.Therefore,they form the system of BP network model that selectes the maximum grade and the minimum radius of curve.3.BP neural network is introduced into railway line in mountain area.It can simulate Yiwan railway certain maximum grade and minimum radius of curve by the trained network model.The error between the simulation values and get the actual values are 7.5%in maximum grade and 0 in minimum radius of curve.Therefore,the established model of BP neural network is reasonable.The paper analyzes the impact of maximum grade and minimum radius of curve to engineering and operations.When the proposed railway line is in the uphill route selection,the maximum grade and minimum radius of curve may be appropriate to lower the standards,which makes the line not only reduce the project cost,but also makes it conform realistic situation.