Research On The Relationship Between Block Spatial Form And Crowd Temporal Spatial Behavior In The Main Urban Area Of Lhasa Based On Multi-source Big Data

With the peaceful liberation of Tibet,economic development has been strongly supported by the state,autonomous regions,aid provinces and cities.With the rapid development of urbanisation and the rapid increase in economic levels,the people of Lhasa have greater demand for comfort,convenience and diversity of urban living facilities,and the focus of urban development has gradually shifted from the incremental planning period to the stock planning period.In recent years,with the continuous development of data analysis technology and information and communication technology,the methods and data available for studying the spatial behaviour of residents are becoming more and more abundant,and urban planners are paying more and more attention to using the laws of human behaviour to analyse the laws and characteristics of urban spatial form.This paper takes the main urban area of Lhasa city as the research object,based on the spatio-temporal behaviour perspective of the population,and takes macro,meso and micro scales as the starting point respectively,based on spatio-temporal behaviour,spatial morphology and environmental behaviour as the theoretical support,using multi-source big data including: Baidu map heat map,POI data of the main urban area of Lhasa city,OSM road network data,building vector data,bus and rail traffic The data is quantified by GIS spatial analysis,and the spatial morphology indicators are quantified by machine learning semantic segmentation algorithm in python language,and the corresponding indicator elements are selected from three scales to build a quantified system of spatial morphology and spatial behaviour of crowds at different scales.The spatial and temporal characteristics of the perceived behaviour of the crowd in the main urban area of Lhasa are analyzed and summarized.Firstly,at the macroscopic scale,the overall quantitative analysis is carried out from the external representation of the spatio-temporal behaviour of the urban population and the elements that carry the spatial pattern of the urban area.Using big data as the basic data,combined with GIS spatial analysis,spatial autocorrelation and other methods,the spatial dimension is taken as all administrative streets in the main urban area,and the temporal dimension is combined with different time sections of different seasons,weekdays and rest days,and the whole day,to analyse the spatio-temporal characteristics of crowd behaviour and the functional differences of each spatial morphological element.Based on the quantitative results of each indicator,the correlation between the two and the degree of influence and differences are further investigated through the construction of OLS and GWR models.Secondly,based on the typical neighbourhoods in the main urban area of Lhasa,the results of the study were summarised,summarised and analysed in four types of neighbourhoods,based on the morphological elements of the physical environment of the neighbourhoods that affect the spatio-temporal activities of the crowd.At the same time,the spatial morphological elements of each neighbourhood are characterised by a combination of quantitative and qualitative methods,using multi-source big data as the base data,and by constructing a spatial morphological radar image of each neighbourhood.On the other hand,the vectorised Baidu heat map is also used as the base data for the external representation of the spatio-temporal behaviour of the urban population,and the spatio-temporal behavioural activity distribution characteristics of the population in each typical neighbourhood are summarised.At the same time,a mesoscopic index system for spatial patterns and spatio-temporal behaviour of the population is constructed,and the correlation between them and the degree of influence and differences are further investigated through person correlation analysis based on the quantitative results of each index.Then,at the microscopic scale,based on the various behavioural activities and perceptions and feelings generated by people in the neighbourhoods,a specific empirical analysis is carried out on the perception of spatio-temporal behaviour of people in four types of typical neighbourhood spaces selected from the main city of Lhasa,a total of 13 neighbourhoods.Based on the basic information of the research questionnaire on the perception of spatial behaviour of the crowd,the analysis is based on the three aspects of comfort,safety and diversity.The results are analysed in terms of comfort,safety and diversity.Then,through the semantic segmentation algorithm of machine learning and GIS spatial analysis,the characteristic system of morphological elements indicators under microscopic scale is constructed.Based on the quantitative results of each indicator,the correlation,degree of influence and differences between them are further investigated by conducting person correlation analysis.Finally,suggestions are made to improve the vitality of urban space,optimise the quality of urban space and guide the behavioural patterns of residents: 1.improve the layout of functional services and facilities to activate the vitality of neighbourhood space;2.improve the quality of neighbourhood space based on the spatial use needs of people;3.optimise the construction of urban space based on the spatial behavioural patterns of people.