A method for optimizing urban commuting patterns using artificial intelligence algorithms

Abstract

This invention relates to the field of urban governance technology, specifically to a method for optimizing urban commuting patterns using artificial intelligence algorithms. The method includes the following steps: identifying urban employment center data based on mobile signaling data and preprocessing the aggregated data; using a Gaussian mixture model to identify the spatial relative distribution characteristics and influence weights of employment locations and employment centers; using an XGBoost model to identify the spatial relative change patterns between residential and employment locations and analyze the job-residence relationship; using an XGBoost model to solve a gravity model to identify commuting patterns between residential and employment locations and obtain the commuting structure; using an intelligent optimization algorithm to simulate adjustments in employment centers and predict changes in employment locations, residential locations, and commuting flows to generate a commuting pattern; and evaluating the commuting pattern by collecting actual commuting data and user feedback. This invention can provide data support and scientific guidance for urban public facility layout, urban quality improvement, and refined urban governance.

Description

Technical Field

[0001] This invention relates to the fields of geospatial artificial intelligence and urban governance technology, and more specifically, to a method for optimizing urban commuting patterns using artificial intelligence algorithms. Background Technology

[0002] Carbon emissions from urban commuting are a significant component of urban carbon emissions. Optimizing urban commuting patterns plays a crucial role in reducing urban commuting carbon emissions and enhancing urban sustainable development capabilities and environmental friendliness. The urban job-housing spatial structure refers to the spatial distribution characteristics of the urban working population's residences and workplaces, as well as population mobility between residences and workplaces. The job-housing spatial pattern is a major determinant of commuting patterns. Optimizing urban commuting patterns by adjusting the job-housing spatial pattern is a core task in urban governance and one of the core means to reduce urban commuting carbon emissions.

[0003] In recent years, the rapidly emerging spatiotemporal big data and geographic artificial intelligence technologies have provided new ways to solve the above problems. New types of big data, mainly based on card swipe data, mobile phone signaling data, and LBS social data, are widely used in the research of work-life space.

[0004] Regarding the intrinsic relationship between urban spatial structure and urban commuting patterns, a considerable number of studies have measured the characteristics of both. The former explores the identification of diverse urban employment and residential spatial structures, while the latter focuses on assessing commuting distance, patterns, and demands, as well as calculating commuting efficiency. In terms of their relationship, by capturing the changing characteristics of urban residents' commuting, the correlation between urban commuting patterns and work-life space can be obtained, and changes can be effectively simulated using proxy models. Furthermore, some studies indicate that residents experiencing changes in their place of employment exhibit a commuting "self-balancing" characteristic, suggesting that optimizing local employment patterns or living units can promote a local "relative balance between work and residence," thereby gradually optimizing the overall spatial pattern.

[0005] On the other hand, existing research is still relatively weak in establishing the exact relationship between urban spatial structure and commuting patterns, resulting in a lack of solid and reliable empirical evidence and quantitative support for alleviating commuting problems by optimizing spatial structure.

[0006] There are currently no effective solutions to the problems in the relevant technologies. Summary of the Invention

[0007] To address the problems in related technologies, this invention proposes a method for optimizing urban commuting patterns using artificial intelligence algorithms, thereby overcoming the aforementioned technical issues in existing related technologies.

[0008] Therefore, the specific technical solution adopted by the present invention is as follows:

[0009] A method for optimizing urban commuting patterns using artificial intelligence algorithms, the optimization method comprising the following steps:

[0010] S1. Identify city employment center data based on mobile phone signaling data and preprocess the aggregated data;

[0011] S2. Use Gaussian mixture models to identify the spatial relative distribution characteristics of employment locations and employment centers and their influence weights.

[0012] S3. Analyze the relationship between work and residence by identifying the spatial relative change patterns of residence and workplace using the XGBoost model;

[0013] S4. Using the XGBoost model to solve the gravity model, we can identify the commuting patterns between residence and workplace and obtain the commuting structure.

[0014] S5. Simulates adjustments to employment centers through intelligent optimization algorithms, predicts changes in employment locations, residences, and commuter flows, and generates commuting patterns.

[0015] S6. Evaluate commuting patterns by collecting actual commuting data and user feedback.

[0016] Furthermore, the process of identifying urban employment center data based on mobile phone signaling data and preprocessing the aggregated data includes the following steps:

[0017] S11. Identify high-precision grid data in mobile phone signaling data, including the grid center coordinates of the commuting origin and destination, grid size, commuting scale corresponding to the grid, and resident population scale.

[0018] S12. Based on the commuting scale of the grid center coordinates, select the grid center with the largest commuting traffic as the initial point, and input the search radius required for the initial point center;

[0019] S13. The point with the largest cumulative number of commuter flows outside the search radius, centered on the initial search point, is designated as the employment center.

[0020] S14. Draw regions with the radius of the initial point centered on each employment center. Search for the next employment center outside the region and repeat this process until all employment centers that meet the conditions are identified, forming a set of employment center points.

[0021] Furthermore, the method of using Gaussian mixture models to identify the spatial relative distribution characteristics of employment locations and employment centers and their influence weights includes the following steps:

[0022] S21. Based on the cumulative number of commuter flows in each grid center, generate random commuter flow points within the grid center coordinates of the destination.

[0023] S22. Use Gaussian mixture models to identify the correlation between employment centers and the distribution of job positions and the distribution of the employed population;

[0024] S23. Use the head-and-tail splitting method to quantify the weights of each employment center.

[0025] Furthermore, the Gaussian mixture model used to identify the correlation between employment centers and the distribution of employment positions and the distribution of the employed population is as follows:

[0026] ;

[0027] In the formula, N represents the total working population of the city;

[0028] This indicates the influence of employment centers in the hybrid model;

[0029] This indicates the location of employment in the city. x The number of commuters employed;

[0030] k This represents the total number of elements in the set of employment centers in the city.

[0031] Represents a two-dimensional Gaussian density function;

[0032] The expression for the two-dimensional Gaussian density function is:

[0033] ;

[0034] In the formula, Indicates the first k Geographic coordinates of the employment center locations;

[0035] This indicates the rate at which the influence of the employment center is declining.

[0036] Furthermore, the quantification of the weights of each employment center using the head-and-tail segmentation method includes the following steps:

[0037] S231. Calculate the average of all values ​​obtained in the above steps. The part greater than or equal to the average is regarded as the head, and the part less than the average is regarded as the tail. The head-tail index of the tail is 1, and the head-tail index of the head is 2.

[0038] S232. Repeat the above process for the data located at the head after segmentation, and assign the head-tail index of 2 and 3 to the tail data and head data obtained by re-identification, respectively.

[0039] S233. Repeat step S232 until the required number of grades is reached. The head and tail index represents the weight of the employment center. The larger the head and tail index, the higher the grade.

[0040] Furthermore, the analysis of the work-residence relationship by identifying the spatial relative change patterns between residence and employment using the XGBoost model includes the following steps:

[0041] S31. Input the distribution data of the residence and employment of the commuting population in the two periods, and integrate the similarity weights of residence and employment to construct a homogeneous relationship network.

[0042] S32. Automatically extract the changing patterns and characteristics of residence and employment locations using homogeneous relationship networks;

[0043] S33. Use the extracted features as input points for the XGBoost model and iteratively train to obtain the optimal analysis model.

[0044] S34. By analyzing the feature importance of the XGBoost model, we can understand the spatial relative change patterns of residence and employment.

[0045] Furthermore, the step of using the XGBoost model to solve the gravity model to identify commuting patterns between residence and employment and obtaining the commuting structure includes the following steps:

[0046] S41. Collect the network coordinates of the place of residence and place of employment, commuting scale and population size, where commuting scale is the observed value of commuting flow and population size is the observed value of resident population.

[0047] S42. Establish a model of the relationship between population size and the distance between residence and employment location by solving the gravity model using XGBoost;

[0048] S43. Using the collected data, estimate the parameters in the gravitational model using the least squares method, and evaluate the gravitational model using cross-validation.

[0049] S44. Based on the established gravity model, predict the number of commuters and analyze the commuting patterns between residences and workplaces.

[0050] Furthermore, the step of generating commuting patterns by simulating adjustments to employment centers using intelligent optimization algorithms and predicting changes in employment locations, residences, and commuting flows includes the following steps:

[0051] S51. Establish an employment center based on the spatial relative distribution characteristics, work-residence relationship and commuting structure obtained from the above steps.

[0052] S52. Adjust the employment pattern according to the optimization objective and constraints, and calculate the commuting cost function;

[0053] S53. Use the adaptive differential evolution optimization algorithm to continue to adjust the employment pattern and perform iterative optimization.

[0054] S54. Repeat the above steps until the optimization goal of generating the employment pattern is achieved.

[0055] Furthermore, the step of adjusting the employment pattern according to the optimization objective and constraints, and calculating the commuting cost function, includes the following steps:

[0056] S521. Adjust the weight of the employment center based on the adjustment objectives and data analysis;

[0057] S522. Based on the spatial relative distribution characteristics and influence weights of employment locations and employment centers, recalculate the distribution of employment locations;

[0058] S523. Based on the spatial relative change pattern of residence and employment, recalculate the distribution of residence;

[0059] S524. Recalculate the commuter flow distribution based on the commuting patterns between residence and workplace;

[0060] S525. Recalculate the average commuting distance based on the calculation results obtained from the above steps;

[0061] S526. Evaluate the degree of optimization of the adjusted commuting distance.

[0062] Furthermore, the iterative optimization of the employment pattern using the adaptive differential evolutionary optimization algorithm includes the following steps:

[0063] S531. Select two individuals from the parent individuals and generate a difference vector by subtracting their vectors.

[0064] S532. Select another individual and sum it with the difference vector to generate an experimental individual;

[0065] S533. Perform crossover operations between parent individuals and corresponding experimental individuals to generate new offspring individuals;

[0066] S534. Perform a selection operation between parent and offspring individuals, and save the qualified individuals to the next generation group.

[0067] The beneficial effects of this invention are as follows:

[0068] 1. This invention provides a method for optimizing the urban work-residence spatial structure by considering the characteristics of current work-residence centers and commuting patterns. It explores the derivation technology of the optimal urban employment spatial pattern based on intelligent optimization algorithms, which can provide data support and scientific guidance for the urban public facility pattern, urban quality improvement, and refined urban governance.

[0069] 2. This invention identifies the spatial relative change patterns between residence and employment through the XGBoost model. The XGBoost model can identify the spatial relative change patterns between residence and employment by analyzing a large amount of data, which helps urban planning and transportation planning departments to better understand population flow patterns. At the same time, by analyzing the spatial changes in the relationship between work and residence, it can help urban planning and transportation planning departments to better understand people's travel needs and commuting patterns.

[0070] 3. This invention uses the XGBoost model to solve a gravity model to identify commuting patterns between residences and workplaces. By analyzing these patterns, we can understand people's commuting methods, distances, and times, which helps reveal commuting patterns between different areas of a city. Furthermore, the analysis results can help transportation planning departments better understand people's commuting needs and travel patterns, optimize the layout of urban transportation networks, public transportation lines, and transportation facilities, improve transportation efficiency and convenience, reduce traffic congestion, alleviate commuting pressure, and improve urban traffic conditions. Attached Figure Description

[0071] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0072] Figure 1 This is a flowchart of a method for optimizing urban commuting patterns using artificial intelligence algorithms according to an embodiment of the present invention;

[0073] Figure 2 This is a general technical roadmap for a method of optimizing urban commuting patterns using artificial intelligence algorithms according to an embodiment of the present invention;

[0074] Figure 3 This is a schematic diagram illustrating the working principle of the adaptive differential evolution algorithm in a method for optimizing urban commuting patterns using artificial intelligence algorithms according to an embodiment of the present invention. Detailed Implementation

[0075] To further illustrate the various embodiments, the present invention provides accompanying drawings, which are part of the disclosure of the present invention. These drawings are mainly used to illustrate the embodiments and can be used in conjunction with the relevant descriptions in the specification to explain the operating principles of the embodiments. With reference to these drawings, those skilled in the art should be able to understand other possible implementation methods and the advantages of the present invention.

[0076] According to an embodiment of the present invention, a method for optimizing urban commuting patterns using artificial intelligence algorithms is provided.

[0077] The present invention will now be further described in conjunction with the accompanying drawings and specific embodiments, such as... Figures 1 - 3 As shown, according to an embodiment of the present invention, a method for optimizing urban commuting patterns using artificial intelligence algorithms includes the following steps:

[0078] S1. Identify city employment center data based on mobile phone signaling data and preprocess the aggregated data.

[0079] Specifically, the mobile phone signaling data is high-precision grid data, which includes the grid center coordinates of the commuting origin and destination. The grid size is The commuting scale corresponding to the grid is Flow The size of the permanent resident population is Pop。

[0080] In one embodiment, the step of identifying urban employment center data based on mobile phone signaling data and preprocessing the aggregated data includes the following steps:

[0081] S11. Identify high-precision grid data in mobile phone signaling data, including the grid center coordinates of the commuting origin and destination, grid size, commuting scale corresponding to the grid, and resident population scale.

[0082] S12. Based on the grid center coordinates commuting scale Flow The grid center with the highest commuter traffic was selected as the initial point. LMP 0 Input the search radius required for the initial point center. R 0 ;

[0083] S13, Search Initial Point LMP 0 Center, search radius R 0 The point with the largest cumulative number of commuters outside the city is designated as the employment center. LMP 1 ;

[0084] S14, respectively by employment center LMP 1 Draw the radius of the initial point centered on the center. R 0 Within that area, search for the next employment center outside that area. LMP 2 This process is repeated until all eligible employment centers are identified, forming a set of employment center points.

[0085] Specifically, the employment center point set is represented as .

[0086] S2. Use Gaussian mixture models to identify the spatial relative distribution characteristics of employment locations and employment centers and their influence weights.

[0087] Specifically, the employment center extraction results from step S1 are input, and the distribution characteristics of the number of employees in each employment location are identified through a Gaussian mixture model, and the influence weight of the employment center is quantified.

[0088] In one embodiment, identifying the spatial relative distribution characteristics of employment locations and employment centers and their influence weights using a Gaussian mixture model includes the following steps:

[0089] S21. Based on the center of each grid group Cumulative number of commuter flows Flow Random commuter traffic points are generated within the grid center coordinates of the destination.

[0090] Specifically, random commuter traffic points are generated within the grid center coordinates of the destination. arrive Generate one within the range Random point.

[0091] S22. Use Gaussian mixture models to identify the correlation between employment centers and the distribution of job positions and the distribution of the employed population.

[0092] Specifically, the Gaussian mixture model used to identify the correlation between employment centers and the distribution of employment positions and the distribution of the employed population is:

[0093] ;

[0094] In the formula, N represents the total working population of the city;

[0095] The weights, also known as the mixing coefficients, represent the influence of employment centers in the mixed model.

[0096] This indicates the location of employment in the city. x The number of commuters employed;

[0097] k This represents the total number of elements in the set of employment centers in the city.

[0098] Represents a two-dimensional Gaussian density function;

[0099] The expression for the two-dimensional Gaussian density function is:

[0100] ;

[0101] In the formula, Indicates the first k Geographic coordinates of the employment center locations;

[0102] This indicates the rate at which the influence of the employment center is declining.

[0103] Considering the universality and stability of the decay characteristics of urban center influence, all Gaussian mixture models are set to share a common... .

[0104] Its expression is:

[0105] ;

[0106] and The solution is usually obtained by using expectation maximization (...) EM )algorithm.

[0107] in, Covariance represents the direction of the linear correlation between variables.

[0108] Its expression is:

[0109] ;

[0110] S23. Use the head-and-tail splitting method to quantify the weights of each employment center.

[0111] In one embodiment, quantifying the weights of each employment center using the head-and-tail segmentation method includes the following steps:

[0112] S231. Calculate the average of all values ​​obtained in the above steps. Consider the portion greater than or equal to the average as the head, and the portion below the average as the tail. The head-tail index of the tail is... HT i The head-to-tail index is 1. HT i It is 2;

[0113] S232. Repeat the above process for the segmented head data, and assign head-tail indices to the re-identified tail data and head data respectively. HT i For 2 and 3;

[0114] S233. Repeat step S232 until the required number of levels is reached or the head data cannot be further subdivided, based on the head-tail index. HT i Representing the weight of employment centers, head and tail index HT i The larger the size, the higher the level.

[0115] S3. Analyze the relationship between work and residence by identifying the spatial relative change patterns of residence and employment location using the XGBoost model.

[0116] Specifically, XGBoost (eXtreme Gradient Boosting) is a machine learning model based on gradient boosting decision trees, and it is a powerful model that is widely used in data science competitions and practical applications.

[0117] XGBoost employs a gradient boosting algorithm, combining multiple weak learners (typically decision trees) into a single strong learner, thereby improving the model's predictive power and generalization performance. Simultaneously, XGBoost incorporates a regularization term into its objective function to control model complexity and prevent overfitting. Adjusting the regularization parameter controls both model complexity and generalization ability. Furthermore, XGBoost helps understand and interpret data by analyzing the importance of each feature within the model. It evaluates feature importance by calculating metrics such as feature coverage, gain, and weight, thus aiding in feature selection and feature engineering.

[0118] Input the residential and employment distribution data of the commuter population in two periods, and fit the impact of changes in employment positions on residential patterns using the XGBosst model.

[0119] In one embodiment, the analysis of the work-residence relationship by identifying the spatial relative change patterns between residence and workplace using the XGBoost model includes the following steps:

[0120] S31. Input the distribution data of the residence and employment of the commuting population in the two periods, and integrate the similarity weights of residence and employment to construct a homogeneous relationship network.

[0121] S32. Automatically extract the changing patterns and characteristics of residence and employment locations using homogeneous relationship networks;

[0122] S33. Use the extracted features as input points for the XGBoost model and iteratively train to obtain the optimal analysis model.

[0123] S34. By analyzing the feature importance of the XGBoost model, we can understand the spatial relative change patterns of residence and employment.

[0124] The assumed XGBoost model has a total of T A decision tree, whose model equation is:

[0125] ;

[0126] in Indicates the first jA regression tree function.

[0127] In the model, the first t The objective function for the trees is:

[0128] ;

[0129] Where Ω represents the complexity of a regression tree, which requires... t The cumulative complexity of each regression tree is:

[0130] ;

[0131] Among them, hyperparameters and We can control the severity of our punishments;

[0132] T This indicates the number of nodes in the current regression tree leaf;

[0133] The sum of the squares of the values ​​of each leaf node is called the sum of squares. L2 Regular terms.

[0134] When used to determine the correspondence between residence and employment, XGBoost is used to calculate the time at which a person resides in the grid. i The equation for the working population is:

[0135] ;

[0136] in, and They represent t and t-1 Time Grid i The number of jobs available;

[0137] and They represent Correspondingly and These are the local coefficients of the geographic coordinates;

[0138] express t-1 Time Grid i The number of working people residing in the area;

[0139] Indicates the corresponding local coefficients;

[0140] Indicates the intercept;

[0141] This represents the residual term of the model.

[0142] This invention takes into account that the distribution of residential and workplace locations each have spatial autocorrelation, and there is also spatial autocorrelation between the two. Therefore, it introduces geographic coordinate variables. This will be reflected in the data.

[0143] S4. Using the XGBoost model to solve the gravity model, the commuting pattern between residence and workplace is identified, and the commuting structure is obtained.

[0144] Input the network center coordinates of the origin and destination, commuting scale, and population size, and use XGBoost as described in step S3 to solve the gravity model, where the commuting scale is the observed value of the commuting flow and the population size is the observed value of the resident population.

[0145] In one embodiment, the step of using the XGBoost model to solve the gravity model to identify commuting patterns between residence and workplace to obtain the commuting structure includes the following steps:

[0146] S41. Collect the network coordinates of the place of residence and place of employment, commuting scale and population size, where commuting scale is the observed value of commuting flow and population size is the observed value of resident population.

[0147] S42. Establish a model of the relationship between population size and the distance between residence and employment location by solving the gravity model using XGBoost;

[0148] S43. Using the collected data, estimate the parameters in the gravitational model using the least squares method, and evaluate the gravitational model using cross-validation.

[0149] S44. Based on the established gravity model, predict the number of commuters and analyze the commuting patterns between residences and workplaces.

[0150] Specifically, the expression for the gravity model is:

[0151] ;

[0152] in, This represents the size of the commuter population from the origin grid to the destination grid;

[0153] k Indicates constant coefficients;

[0154] and They represent i and j The number of employed people residing and working;

[0155] This indicates the distance between the number of employed people residing in a given area and the number of people working in that area.

[0156] and These represent the elasticity of population size at the origin and destination, respectively.

[0157] This indicates a measure of distance decay effect;

[0158] This indicates the error term.

[0159] S5. Through intelligent optimization algorithms, simulate the adjustment of employment centers and predict changes in employment location, residence location, and commuting flow to generate commuting patterns.

[0160] The simulation study adjusts employment centers to predict changes in employment locations, residences, and commuting flows. An intelligent optimization algorithm is used to fine-tune commuting costs. The results of steps S2, S3, and S4 are then substituted to find the most suitable adjustment scheme for the work-residence spatial structure, serving as the basis for optimizing the urban commuting pattern.

[0161] In one embodiment, the step of generating a commuting pattern by simulating adjustments to employment centers using intelligent optimization algorithms and predicting changes in employment locations, residences, and commuting flows includes the following steps:

[0162] S51. Establish an employment center based on the spatial relative distribution characteristics, work-residence relationship and commuting structure obtained from the above steps.

[0163] S52. Adjust the employment pattern according to the optimization objective and constraints, and calculate the commuting cost function.

[0164] Specifically, the process of adjusting the employment pattern based on the optimization objective and constraints, and calculating the commuting cost function, includes the following steps:

[0165] S521. Adjust the weight of the employment center based on the adjustment objectives and data analysis;

[0166] S522. Based on the spatial relative distribution characteristics and influence weights of employment locations and employment centers, recalculate the distribution of employment locations;

[0167] S523. Based on the spatial relative change pattern of residence and employment, recalculate the distribution of residence;

[0168] S524. Recalculate the commuter flow distribution based on the commuting patterns between residence and workplace;

[0169] S525. Recalculate the average commuting distance based on the calculation results obtained from the above steps;

[0170] S526. Evaluate the degree of optimization of the adjusted commuting distance.

[0171] Specifically, the commuting cost function for evaluating the degree of optimization is:

[0172] ;

[0173] in, This indicates the adjusted average commuting distance;

[0174] C Simulated commuting distance representing the current situation;

[0175] This indicates the benefit of the optimization strategy, namely the reduction in average commuting distance; the smaller the value, the greater the reduction.

[0176] The value represents the regularization term and the cost of the optimization strategy, namely the magnitude of change in the urban employment and housing patterns. The smaller the value, the closer the optimization result is to the current situation, and the smaller the economic and social cost required to achieve structural optimization.

[0177] in The calculation formula is:

[0178] ;

[0179] in, This represents the regularization coefficient, which is adjusted to balance the benefits of commuting distance reduction with the costs of structural changes.

[0180] and Representing the center before and after optimization respectively j The weight.

[0181] S53. Use the adaptive differential evolution optimization algorithm to continue to adjust the employment pattern and perform iterative optimization.

[0182] Specifically, the use of the adaptive differential evolutionary optimization algorithm to iteratively optimize the employment pattern includes the following steps:

[0183] S531. Select two individuals from the parent individuals and generate a difference vector by subtracting their vectors.

[0184] S532. Select another individual and sum it with the difference vector to generate an experimental individual;

[0185] S533. Perform crossover operations between parent individuals and corresponding experimental individuals to generate new offspring individuals;

[0186] S534. Perform a selection operation between parent and offspring individuals, and save the qualified individuals to the next generation group.

[0187] S54. Repeat the above steps until the optimization goal of generating the employment pattern is achieved.

[0188] S6. Evaluate commuting patterns by collecting actual commuting data and user feedback.

[0189] In one embodiment, assessing commuting patterns by collecting actual commuting data and user feedback includes the following steps:

[0190] Collect relevant commuting data, including commuting time, commuting mode, commuting distance, and traffic congestion, through methods such as questionnaires, traffic surveys, and traffic monitoring systems.

[0191] The collected commuting data is analyzed using statistical methods and data visualization to examine indicators such as commuting time, commuting distance, and traffic congestion, identifying problems and areas for improvement.

[0192] We collected user feedback on commuting patterns through questionnaires, interviews, and other methods to understand users' satisfaction and dissatisfaction with commuting time, traffic congestion, and transportation choices, as well as their suggestions and opinions.

[0193] By comprehensively analyzing commuting data and user feedback, we can evaluate commuting patterns, compare commuting conditions in different regions and at different times, identify problems and key areas for improvement, and take into account user needs and traffic planning requirements.

[0194] Based on the assessment results, suggestions are made to improve commuting patterns, propose traffic improvement plans for traffic congestion points, optimize the traffic network layout, and improve public transportation services.

[0195] Based on the proposed improvement suggestions, an implementation plan was developed, and improvement measures were implemented step by step.

[0196] After the improvement measures are implemented, the effectiveness is monitored and evaluated. By collecting commuting data and user feedback again, the differences before and after the improvement are compared, and the effectiveness and feasibility of the improvement measures are assessed.

[0197] In summary, by utilizing the above-mentioned technical solutions of this invention, this invention provides a method for optimizing the urban work-residence spatial structure by considering the characteristics of current work-residence centers and commuting patterns. It explores a technology for deriving the optimal urban employment spatial pattern based on intelligent optimization algorithms, which can provide data support and scientific guidance for urban public facility layout, urban quality improvement, and refined urban governance. This invention identifies the spatial relative change patterns between residences and employment locations using the XGBoost model. The XGBoost model can identify these patterns by analyzing large amounts of data, helping urban planning and transportation planning departments better understand population flow patterns. Furthermore, by analyzing the spatial changes in the work-residence relationship, it can help urban planning and transportation planning departments better understand people's travel needs and commuting patterns. This invention utilizes the XGBoost model to solve a gravity model to identify commuting patterns between residences and workplaces. By analyzing these patterns, information such as commuting methods, distances, and times can be obtained, helping to reveal commuting patterns between different areas of a city. Furthermore, the analysis results can help transportation planning departments better understand people's commuting needs and travel patterns, optimize the layout of urban transportation networks, public transportation lines, and transportation facilities, improve transportation efficiency and convenience, reduce traffic congestion, alleviate commuting pressure, and improve urban traffic conditions.

[0198] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for optimizing urban commuting patterns using artificial intelligence algorithms, characterized in that, The optimization method includes the following steps: S1. Identify city employment center data based on mobile phone signaling data and preprocess the aggregated data; S2. Use Gaussian mixture models to identify the spatial relative distribution characteristics of employment locations and employment centers and their influence weights. S3. Analyze the relationship between work and residence by identifying the spatial relative change patterns of residence and workplace using the XGBoost model; S4. Using the XGBoost model to solve the gravity model, we can identify the commuting patterns between residence and workplace and obtain the commuting structure. S5. Simulates adjustments to employment centers through intelligent optimization algorithms, predicts changes in employment locations, residences, and commuter flows, and generates commuting patterns. S6. Evaluate commuting patterns by collecting actual commuting data and user feedback; The method of identifying the spatial relative distribution characteristics of employment locations and employment centers and their influence weights using Gaussian mixture models includes the following steps: S21. Based on the cumulative number of commuter flows in each grid center, generate random commuter flow points within the grid center coordinates of the destination. S22. Use Gaussian mixture models to identify the correlation between employment centers and the distribution of job positions and the distribution of the employed population; S23. Use the head-and-tail splitting method to quantify the weight of each employment center; Analyzing the relationship between work and residence by identifying the spatial relative changes in residence and workplace using the XGBoost model includes the following steps: S31. Input the distribution data of the residence and employment of the commuting population in the two periods, and integrate the similarity weights of residence and employment to construct a homogeneous relationship network. S32. Automatically extract the changing patterns and characteristics of residence and employment locations using homogeneous relationship networks; S33. Use the extracted features as input points for the XGBoost model and iteratively train to obtain the optimal analysis model. S34. By analyzing the feature importance of the XGBoost model, we can understand the spatial relative change patterns between residence and employment. The process of using the XGBoost model to solve the gravity model to identify commuting patterns between residence and workplace and obtain commuting structure includes the following steps: S41. Collect the network coordinates of the place of residence and place of employment, commuting scale and population size, where commuting scale is the observed value of commuting flow and population size is the observed value of resident population. S42. Establish a model of the relationship between population size and the distance between residence and employment location by solving the gravity model using XGBoost; S43. Using the collected data, estimate the parameters in the gravitational model using the least squares method, and evaluate the gravitational model using cross-validation. S44. Based on the established gravity model, predict the number of commuters and analyze the commuting patterns between residences and workplaces.

2. The method for optimizing urban commuting patterns using artificial intelligence algorithms according to claim 1, characterized in that, The process of identifying urban employment center data based on mobile phone signaling data and preprocessing the aggregated data includes the following steps: S11. Identify high-precision grid data in mobile phone signaling data, including the grid center coordinates of the commuting origin and destination, grid size, commuting scale corresponding to the grid, and resident population scale. S12. Based on the commuting scale of the grid center coordinates, select the grid center with the largest commuting traffic as the initial point, and input the search radius required for the initial point center; S13. The point with the largest cumulative number of commuter flows outside the search radius, centered on the initial search point, is designated as the employment center. S14. Draw regions with the radius of the initial point centered on each employment center. Search for the next employment center outside the region and repeat this process until all employment centers that meet the conditions are identified, forming a set of employment center points.

3. The method for optimizing urban commuting patterns using artificial intelligence algorithms according to claim 1, characterized in that, The Gaussian mixture model used in identifying the correlation between employment centers and the distribution of employment positions and the employment population is as follows: ； In the formula, N represents the total working population of the city; This indicates the influence of employment centers in the hybrid model; This indicates the location of employment in the city. x The number of commuters employed; k This represents the total number of elements in the set of employment centers in the city. Represents a two-dimensional Gaussian density function; The expression for the two-dimensional Gaussian density function is: ； In the formula, Indicates the first k Geographic coordinates of the employment center locations; This indicates the rate at which the influence of the employment center is declining.

4. The method for optimizing urban commuting patterns using artificial intelligence algorithms according to claim 3, characterized in that, The method of quantifying the weights of each employment center using the head-and-tail segmentation method includes the following steps: S231. Calculate the average of all values ​​obtained in the above steps. The part greater than or equal to the average is regarded as the head, and the part less than the average is regarded as the tail. The head-tail index of the tail is 1, and the head-tail index of the head is 2. S232. Repeat the above process for the data located at the head after segmentation, and assign the head-tail index of 2 and 3 to the tail data and head data obtained by re-identification, respectively. S233. Repeat step S232 until the required number of grades is reached. The head and tail index represents the weight of the employment center. The larger the head and tail index, the higher the grade.

5. The method for optimizing urban commuting patterns using artificial intelligence algorithms according to claim 1, characterized in that, The process of generating commuting patterns by simulating adjustments to employment centers and predicting changes in employment locations, residences, and commuter flows using intelligent optimization algorithms includes the following steps: S51. Establish an employment center based on the spatial relative distribution characteristics, work-residence relationship and commuting structure obtained from the above steps. S52. Adjust the employment pattern according to the optimization objective and constraints, and calculate the commuting cost function; S53. Use the adaptive differential evolution optimization algorithm to continue to adjust the employment pattern and perform iterative optimization. S54. Repeat the above steps until the optimization goal of generating the employment pattern is achieved.

6. A method for optimizing urban commuting patterns using artificial intelligence algorithms according to claim 5, characterized in that, The process of adjusting the employment pattern based on optimization objectives and constraints, and calculating the commuting cost function, includes the following steps: S521. Adjust the weight of the employment center based on the adjustment objectives and data analysis; S522. Based on the spatial relative distribution characteristics and influence weights of employment locations and employment centers, recalculate the distribution of employment locations; S523. Based on the spatial relative change pattern of residence and employment, recalculate the distribution of residence; S524. Recalculate the commuter flow distribution based on the commuting patterns between residence and workplace; S525. Recalculate the average commuting distance based on the calculation results obtained from the above steps; S526. Evaluate the degree of optimization of the adjusted commuting distance.

7. A method for optimizing urban commuting patterns using artificial intelligence algorithms according to claim 6, characterized in that, The use of the adaptive differential evolutionary optimization algorithm to iteratively optimize the employment pattern includes the following steps: S531. Select two individuals from the parent individuals and generate a difference vector by subtracting their vectors. S532. Select another individual and sum it with the difference vector to generate an experimental individual; S533. Perform crossover operations between parent individuals and corresponding experimental individuals to generate new offspring individuals; S534. Perform a selection operation between parent and offspring individuals, and save the qualified individuals to the next generation group.

Images