CN-121997283-A - Urban crowd activity and space structure dynamic deduction method, system, terminal and storage medium based on single traffic flow

CN121997283ACN 121997283 ACN121997283 ACN 121997283ACN-121997283-A

Abstract

The invention belongs to the technical field of traffic geographic information analysis, and discloses a method, a system, a terminal and a storage medium for dynamically deducting urban crowd activities and space structures based on single traffic flow, wherein the method comprises the steps of constructing a riding flow network, and identifying a space distribution mode by combining a hierarchical clustering algorithm through a flow similarity measurement method based on shared single traffic order data in an urban range; the method comprises the steps of carrying out quantity aggregation on the identified space distribution patterns according to grid units to construct a space grade distribution model, combining urban built-up environment characteristics with space grade distribution in the space grade distribution model to construct a comprehensive data set suitable for machine learning regression modeling, quantitatively analyzing nonlinear influence mechanisms of the urban built-up environment characteristics on different space distribution patterns based on the comprehensive data set and a machine learning model interpretation method of game theory, and deducing the internal relevance between urban crowd activities and space structures. The invention comprehensively realizes the dynamic evolution process of the urban space structure.

Inventors

MA DING
LIU SHIGANG
Gao Qinxin
LI XIAOMING
CHEN YEBIN
GUO RENZHONG
Zou Shanlai
WU JIAQIN

Assignees

深圳大学

Dates

Publication Date: 20260508
Application Date: 20251205

Claims (10)

1. A city crowd activity and space structure dynamic deduction method based on single traffic flow is characterized by comprising the following steps: Constructing a riding stream network, and identifying a space distribution mode by combining a hierarchical clustering algorithm through a flow similarity measurement method based on shared bicycle order data in a city range; the identified spatial distribution patterns are aggregated in number according to grid cells, and a spatial grade distribution model is constructed; Combining the urban built environment characteristics with the spatial grade distribution in the spatial grade distribution model to construct a comprehensive data set suitable for machine learning regression modeling; based on the comprehensive data set, and combined with a machine learning model interpretation method of game theory, nonlinear influence mechanisms of the urban built-up environment features on different spatial distribution modes are quantitatively analyzed, and internal relevance between urban crowd activities and spatial structures is deduced.
2. The method for dynamically deducting the activity and the spatial structure of the urban population based on the single traffic as claimed in claim 1, wherein the step of constructing the riding stream network, based on the shared single vehicle order data in the urban range, identifies the spatial distribution pattern by combining the flow similarity measurement method with the hierarchical clustering algorithm comprises the following steps: acquiring shared bicycle order data in the city range; Acquiring a travel flow movement track according to the shared bicycle order data; Constructing the riding stream network based on the travel stream movement track; Based on the riding stream network, riding stream similarity evaluation is carried out by using the flow similarity measurement method, and the spatial distribution mode is identified by using the hierarchical clustering algorithm.
3. The method for dynamically deducting urban crowd activity and spatial structures based on single traffic according to claim 2, wherein said using the flow similarity measurement method for evaluating the similarity of riding streams comprises: For any two riding flow paths in the riding flow network, a dynamically adjusted space buffer zone is built by taking the starting point and the end point of each riding flow path as the center; When the start point and the end point of the two riding flow paths simultaneously fall into the space buffer zone, the two riding flow paths are judged to be similar flow paths in space.
4. The method for dynamically deducting urban crowd activities and spatial structures based on single traffic according to claim 2, wherein the spatial distribution mode comprises an aggregation mode, a divergent mode and a convergent mode; the identifying the spatial distribution pattern by using the hierarchical clustering algorithm comprises the following steps: Initializing each riding flow path as an independent flow cluster; Calculating the midpoint position of each riding flow path, constructing a KD tree space index structure based on all midpoint positions, combining the current clusters from bottom to top based on the KD tree space index structure, and identifying to obtain the aggregation mode; Selecting a riding flow path set with adjacent starting points and random directions based on a starting point adjacency judging method, and identifying to obtain the divergence mode; and selecting a riding flow path set with adjacent end points and convergent directions based on an end point proximity judging method, and identifying to obtain the convergence mode.
5. The method for dynamically deducting urban crowd activities and spatial structures based on single traffic according to claim 1, wherein the step of aggregating the identified spatial distribution patterns according to the number of grid cells to construct a spatial rank distribution model comprises the steps of: Counting the total number of starting points and ending points of riding streams in a grid in an aggregation mode, counting the number of starting points in a divergent mode and counting the number of ending points in an aggregation mode by adopting a grid space counting method to obtain scale grades in each mode; And constructing the space grade distribution model according to the counted scale grade.
6. The method for dynamically deriving motion and spatial structure of urban population based on single-vehicle flow according to claim 1, wherein said combining urban as-built environmental features with spatial level distribution in the spatial level distribution model constructs a comprehensive dataset suitable for machine learning regression modeling, comprising: And constructing a data set containing urban built environment characteristics, and carrying out associated modeling with the space grades in the space grade distribution model to obtain the comprehensive data set, wherein the urban built environment characteristics comprise urban density characteristics, urban design characteristics, urban diversity characteristics, public transportation accessibility characteristics and destination accessibility characteristics.
7. The method for dynamically deducting urban crowd activities and spatial structures based on single traffic according to claim 1, wherein the method for quantitatively analyzing the nonlinear influence mechanism of the urban built-up environmental features on different spatial distribution modes based on the comprehensive data set and combined with a machine learning model interpretation method of game theory to deduct the inherent relevance between the urban crowd activities and the spatial structures comprises the following steps: Based on the comprehensive data set, a regression model is established by adopting a machine learning algorithm with nonlinear fitting capability; quantitatively analyzing the prediction contribution degree of the urban built environment features to different spatial distribution modes by using a machine learning model interpretation method of game theory based on the regression model; And outputting global feature importance ranking and local feature dependency relationship maps of the urban built-up environment features according to the prediction contribution degree to obtain a deduction result of the inherent relevance between the urban crowd activities and the spatial structure.
8. A city crowd activity and space structure dynamic deduction system based on single traffic flow is characterized by comprising: the spatial distribution pattern recognition module is used for constructing a riding stream network, and recognizing a spatial distribution pattern by combining a hierarchical clustering algorithm through a flow similarity measurement method based on shared bicycle order data in a city range; the space grade distribution model construction module is used for carrying out quantity aggregation on the identified space distribution modes according to grid cells to construct a space grade distribution model; The comprehensive data set construction module is used for combining the urban built environment characteristics with the spatial grade distribution in the spatial grade distribution model to construct a comprehensive data set suitable for machine learning regression modeling; The riding flow mode interpretation module is used for quantitatively analyzing nonlinear influence mechanisms of the urban built-up environment characteristics on different spatial distribution modes based on the comprehensive data set and a machine learning model interpretation method of a game theory, and deducing the internal relevance between urban crowd activities and spatial structures.
9. A terminal comprising a processor and a memory, wherein the memory stores a single-traffic-based urban population activity and spatial structure dynamic deduction program, and the single-traffic-based urban population activity and spatial structure dynamic deduction program, when executed by the processor, is configured to implement the operations of the single-traffic-based urban population activity and spatial structure dynamic deduction method according to any one of claims 1-7.
10. A computer readable storage medium, wherein the computer readable storage medium stores a single-traffic-based urban population activity and spatial structure dynamic deduction program, which when executed by a processor, is configured to implement the operations of the single-traffic-based urban population activity and spatial structure dynamic deduction method according to any one of claims 1 to 7.

Description

Urban crowd activity and space structure dynamic deduction method, system, terminal and storage medium based on single traffic flow Technical Field The invention relates to the technical field of traffic geographic information analysis, in particular to a method, a system, a terminal and a storage medium for dynamically deducting urban crowd activities and space structures based on single traffic flow. Background With the acceleration of the urban process and the continuous rise of population density, urban traffic systems face structural problems such as increased commute pressure, serious separation of jobs and liveness, unbalanced layout of functional areas and the like. In order to effectively cope with the challenges, realize efficient organization and sustainable development of urban space, deep analysis on urban structures and dynamic evolution processes thereof is required to be carried out by means of novel technical means. The urban traveling behavior is understood from the geographic flow perspective, so that the interaction rules among different areas can be revealed, and the urban complex dynamic evolution mechanism can be reflected more comprehensively by combining multiple flow modes for analysis. The development of big data and sensing technology makes the high space-time resolution data of urban traffic flow increasingly abundant. The shared bicycle travel data has the characteristics of definite starting and ending point (OD), time stamp, wide coverage and the like, and becomes an important data source for researching urban microscopic travel behaviors. In the existing research, urban travel hot spot areas are identified through clustering and statistical analysis of shared bicycle OD data, spatial analysis is further carried out by combining population distribution, land utilization data, built environment data and the like, the urban morphology is also researched based on space-time evolution characteristics analysis of riding flow, and main driving factors are identified by utilizing traditional statistical models such as a factor analysis method, a principal component analysis method, a maximum likelihood method and the like. The urban microscopic trip behaviors have diversity, heterogeneity, selectivity and dynamics, namely subjective intention of human activity behaviors is difficult to predict or describe by using a simple physical model. The traditional model has the defects in revealing the complex nonlinear relation behind the travel behaviors and the threshold effect, and the variable change does not have obvious influence on the use mode after the condition reaches a certain critical point. Meanwhile, the traditional research is insufficient in systematic analysis of various flow modes, the dynamic evolution process of the urban space structure is difficult to comprehensively reflect, and the deep research and scientific intervention on the complex urban dynamic structure are limited. In summary, the prior art has a problem that it is difficult to fully reflect the dynamic evolution process of the urban spatial structure, and therefore, the prior art needs to be improved. Disclosure of Invention Aiming at the defects of the prior art, the invention provides a method, a system, a terminal and a storage medium for dynamically deducting the activities and space structures of urban crowd based on single traffic flow, which are used for solving the problems that the prior art is insufficient in systematic analysis of multiple flow modes and is difficult to comprehensively reflect the dynamic evolution process of the urban space structure. The technical scheme adopted for solving the technical problems is as follows: in a first aspect, the present invention provides a method for dynamically deducting urban crowd activities and spatial structures based on single traffic, comprising: Constructing a riding stream network, and identifying a space distribution mode by combining a hierarchical clustering algorithm through a flow similarity measurement method based on shared bicycle order data in a city range; the identified spatial distribution patterns are aggregated in number according to grid cells, and a spatial grade distribution model is constructed; Combining the urban built environment characteristics with the spatial grade distribution in the spatial grade distribution model to construct a comprehensive data set suitable for machine learning regression modeling; based on the comprehensive data set, and combined with a machine learning model interpretation method of game theory, nonlinear influence mechanisms of the urban built-up environment features on different spatial distribution modes are quantitatively analyzed, and internal relevance between urban crowd activities and spatial structures is deduced. In one implementation, the construction of the riding stream network, based on the shared bicycle order data in the city range, identifies the spatial distribution pattern by com