| With the increasing scale of urbanization,the rapid economic development has also brought huge challenges to urban management.The traditional management model has been unable to solve many of the current problems in urban management.The disposal and early warning of urban management cases has become an important part of fine urban refinement management.To achieve good urban management results,how to quickly find and efficiently handle urban management problems,how to promptly warn the escalation of cases,and how to accurately predict the spatiotemporal distribution and quantitative of cases changes,It is a problem that urban management workers need to solve urgently.This article belongs to the application research of smart urban management.Based on the real urban management business data and urgent application needs of Qingyuan city center in Guangdong Province,the spatial analysis technology and data prediction methods are used to study the spatiotemporal distribution characteristics of Qingyuan urban management cases and the changing trend of the number of cases The analysis result can provide scientific and predictive decision support for urban management departments in urban operation and management.The main research contents and features are as follows:(1)Propose the technical route of the research on the spatiotemporal distribution characteristics and quantity prediction of Qingyuan urban management cases.Including the macrotemporal distribution of urban management cases,spatiotemporal clustering model,spatiotemporal hotspots and other visual studies,the correlation analysis of influencing factors and the prediction analysis of the occurrence of cases.(2)Visualization study on spatiotemporal characteristics of urban management cases.First,the global spatial distribution characteristics of urban management cases based on the central urban area of Qingyuan are studied.The average center,standard distance and standard deviation ellipse are used to explore the density center and overall directional distribution of urban management cases;The average nearest neighbor and global spatial autocorrelation methods are used to analyze the spatial clustering of digital urban management cases,and a significance test is carried out.Finally,the core density estimation and hot spot analysis methods are used to study the continuous change and accurate case high concentration centers of each jurisdiction.The visualization resultsclearly expressed the global and local spatiotemporal distribution of Qingyuan urban management cases in the study area.(3)Research on the correlation between urban management cases and influencing factors.The factors affecting the occurrence of urban management cases are divided into two types,spatial factors and temporal factors.Based on the distribution of three types of urban interest points of Qingyuan city center,the number of urban management cases under different buffer radius is compared using buffer analysis tool determine the areas where the urban management department should focus on inspections;verify the correlation between the amount of urban management cases and time elements by comparing the impact of important holidays and major events on the amount of urban management cases.(4)Predictive analysis of time series of urban management cases.The gray prediction method,ARIMA prediction method and BP neural network prediction method are used to fit the monthly data of historical urban management cases to predict the future data;compare the relative error between the fitted value,predicted value and actual value of the above models,to get the fitting accuracy and prediction accuracy of different models.The result show that when the number of samples is sufficient,the BP neural network model has the best fitting effect.For the prediction of severe fluctuation data,the GM(1,n)model has the best prediction effect,which can be applied to the prediction of future data of urban management cases. |
PDF Full Text Request