Demand Analysis Of Urban Underground Space Resources Development And Utilization Based On Grey Neural Network

With the increase in population and urban expansion,the role of underground space in alleviating traffic congestion,environmental pollution,flood drainage and flood prevention has been increasingly recognized.The three-dimensional development from above ground to underground has become a new trend in the development of major cities in the world.For China,where land resources are scarce and in a stage of rapid urbanization,the development of underground space is particularly important and has a "late-mover advantage".Underground space development has the characteristics of irreversibility and long construction period,based on the analysis of the main factors affecting the urban underground space demand,scientific predictions are made on the rational development and utilization of urban underground space,this is very important for the development of the city and the use of underground space resources.However,the current analysis of the underground space demand of a city is mostly based on the perspectives of engineering geology,geotechnical exploration and other development feasibility,lack of analysis based on economics,and focus on theoretical argumentation.The mathematical models used are relatively one-sided,which is not conducive to analyze and forecast the demand for urban underground space.Taking into account the fact that urban land in China is state-owned and urban underground space is developed by the government,and so far,most of the urban underground space is still public goods and quasi-public goods.This paper takes the government as the main body,constructs a theoretical model of urban underground space development,and derives the main factors that affect the rational use of urban underground space from the theoretical model.Then,based on the main factors derived from the theoretical model,collect data on the development and utilization of underground space in 11 first-and second-tier cities in China over the years,and use Lasso regression to further filter and find that the most important variable that affects the amount of underground space development is:urban residents The number of cars owned by one hundred people,the average price of land sold in cities,the per capita GDP,and the tertiary industry’s regional GDP.Based on the selected variables,a gray neural network model and a widely used multiple linear regression model were established to fit and predict the development area of the underground space in Nanjing and Hangzhou,which are representative of the development trend.After comparison and optimization,the series type The gray neural network model has the best prediction effect.The prediction method of this model is simple and the process is efficient,it can provide theoretical support for the development and construction of urban underground space.The model is further used to predict the underground space demand of Nanjing and Hangzhou from 2021 to 2030,and it is found that the current underground space development plans of the two cities are unreasonable and urgently need to be adjusted.