Analyzing the relationship between urban macroeconomic development and transport infrastructure system based on neural network

Abstract

Urban planners have always concerned about the relationship between macroeconomic trend of development and transport infrastructure planning. In the past, urban planners presented qualitative analysis for the construction of transport infrastructure depended on experience. Based on the acquired data of transport infrastructure system and urban macroeconomic, we take the urban development indicators as independent variable and transport infrastructure system indicators as dependent variable. Then, we extract factors and reduce the interference of weak correlation by the method of principal component analysis (PCA). In the aspect of establishing linkage model, BP neural network simplifies topology structures well by adjusting the discrete input, and the minimum error between actual value and predicted value was used as training objectives. Finally, we get a quantitative analysis of the relationship between urban macroeconomic and transport infrastructure. We take Changsha city in China as a case, reducing dimension by PCA and extracting 8 main indicators from original 14 indicators of urban macroeconomic. The precision reaches to 0.99992 and error value is less than 5%. We work out the best weight matrix W1 and the offset value matrix B1, B2 for urban development and transport system in Changsha city. It provides quantitative methodology for policymakers in transport system planning.