Research On Spatial And Temporal Characteristics Of Taxi Carbon Emissions Based On Nonnegative Matrix Factorization

The travel habits and travel modes of urban residents vary according to travel needs.Due to their flexibility,convenience,comfort,and all-weather operation,taxis can meet the individual needs of residents for "point" to "point" travel.Therefore,it has become an important supplementary part of the urban public transport system.The taxi GPS trajectory data combined with the urban POI data can more accurately discover the activity rules of urban residents’ taxi travel.Therefore,based on the above background and under the development goal of "Double Carbon",this paper takes the GPS trajectory data,POI data,and bus network data of taxis in Xi’an as examples,and uses POI data as an intermediary to explore the spatiotemporal characteristics of taxi carbon emissions.Based on the interactive relationship with the types of residents’ travel activities,and taking the carbon emission reduction of taxis as the starting point,optimization suggestions are put forward for the status quo of the bus network layout.The specific research contents are as follows:In this paper,firstly,the dimensionality reduction of the space-time matrix of taxi carbon emissions is carried out based on the non-negative matrix factorization algorithm(NMF),and the space-time characteristics of several types of taxi carbon emissions are extracted;The driving factors are analyzed to reveal the types of residents’ travel activities corresponding to the spatio-temporal characteristics of different types of taxi carbon emissions and analyze their similarities and differences;finally,the two typical residents’ travel activities corresponding to workday commuting and entertainment travel Taking the spatio-temporal characteristics of taxi carbon emissions as an example,explore its spatial autocorrelation,coupling coordination relationship and relative development degree with the bus network density,and aim at reducing taxi carbon emissions,from 24-hour time periods and specific spatial units From the perspective of the grid,more realistic and targeted,more refined suggestions for the layout of the bus network and the optimization of the frequency of departures are proposed.The research found that:(1)The spatio-temporal distribution characteristics of taxi carbon emissions on working days and non-working days are significantly different.Compared with non-working days,the travel destinations of residents corresponding to the spatio-temporal characteristics of taxi carbon emissions on working days are more concentrated,while those on non-working days are more scattered;the carbon emissions of non-working days peak early The occurrence time of the evening peak is 1-2 hours later than that of working days,reflecting the slower rhythm of residents’ travel activities on nonworking days.(2)The driving force analysis results of different POI types on the spatio-temporal characteristics of each taxi’s carbon emissions show that the explanatory power of all detection factors interacted on working days and non-working days is greater than that of any single factor,indicating that taxi carbon emissions Quantitative changes are the result of multiple factors acting together.At the same time,the explanatory power of the interaction of accommodation factors(residential districts,accommodation services)on working days and non-working days is higher than the explanatory power of interaction with other influencing factors,indicating that the impact of accommodation factors on taxi travel Changes in carbon emissions have the strongest driving effect.(3)The high-high agglomeration areas of taxi carbon emissions and bus network density corresponding to the two types of residents’ travel activities on weekday commuting and life and entertainment are mainly concentrated in the central area of the main urban area of Xi’an,and the low-high cluster areas are mainly concentrated in the central area of the main urban area of Xi’an The high agglomeration area is distributed in the outer edge of the high-high agglomeration area,and the low-low agglomeration area is distributed in the fringe area of the main urban area of Xi’an;the coupling and coordination state of taxi carbon emissions and bus network density of two types of residents’ travel activities Both are poor,and in terms of relative development level,the carbon emissions of taxis "lag" behind the density of public transport network.