Method for optimizing layout of deposit service based on multi-source data and GIS spatial analysis

Abstract

The application relates to a registration service layout optimization method based on multi-source data and GIS spatial analysis, which comprises the following steps: step S1, multi-source space-time data acquisition and preprocessing; step S2, space-time modeling based on passenger registration demand of multi-source space-time data; step S3, construction of a GIS network data set of a traffic station; step S4: establishment of a multi-objective location-allocation optimization model, comprising: step S4-1, setting of an objective function; step S4-2, setting of a constraint condition; step S4-3, solving of the multi-objective optimization model by using a genetic algorithm to obtain a set of Pareto optimal solutions; step S5: multi-scheme evaluation and three-dimensional visualization output. The demand modeling method of multi-source data fusion innovatively combines train schedules, passenger stay behaviors, individual attributes and building space data, constructs a more realistic "demand pulse" model and a demand intensity surface, and breaks through the limitation of relying only on passenger flow density.

Description

Technical Field

[0001] This invention belongs to the interdisciplinary field of geographic information technology and transportation hub public service facility planning, specifically involving a method for optimizing the layout of hosting services based on multi-source data and GIS spatial analysis. Background Technology

[0002] As major transportation hubs, the rationality of the layout of service facilities within high-speed railway stations directly impacts passenger travel efficiency and satisfaction. Luggage storage service is a crucial component of this. Currently, the layout planning of luggage storage points in high-speed railway stations mainly relies on traditional planning and design experience, which presents the following problems and shortcomings: Empirical vs. Static: Existing layouts are mostly based on macro-level passenger flow forecasts and designers' experience-based judgments, lacking quantitative analysis of the actual, dynamic, and fine-grained spatiotemporal distribution patterns of passenger flow. For example, luggage storage points at train stations may only be located at entrances / exits or waiting halls, failing to fully consider the actual needs of passengers on different trains at different times.

[0003] Supply and demand mismatch: Due to the lack of accurate demand forecasting, there is a serious shortage of service supply in some areas (such as transfer passages and near specific ticket gates), forming a "service desert"; while in other areas, service resources are idle, resulting in waste. This directly leads to problems such as long time for passengers to find storage points, long detours, and crowded queues.

[0004] Ignoring individual behavioral differences: Different passengers (such as business travelers and leisure travelers, transit passengers and originating and terminating passengers) have significant differences in their demand for luggage storage, acceptable walking distance, and service price sensitivity. Existing layout methods are unable to characterize this individual heterogeneity and cannot provide differentiated and refined layout solutions.

[0005] Lack of a scientific evaluation system: For existing or proposed layout schemes, there is a lack of a GIS-based spatial efficiency quantitative evaluation index system, such as spatial accessibility, service coverage, supply and demand matching degree, fairness, etc., which makes the scheme comparison and optimization lack an objective basis. Summary of the Invention

[0006] To address the aforementioned shortcomings in existing technologies, this invention provides a method for optimizing the layout of hosting services based on multi-source data and GIS spatial analysis, including: Step S1, Multi-source spatiotemporal data acquisition and preprocessing, includes collecting and fusing multi-source spatiotemporal data related to transportation hubs to construct a database; the multi-source spatiotemporal data includes: building space data, passenger flow spatiotemporal data, mobile phone signaling data, WiFi probe data, train timetable data and passenger attribute data; Step S2, based on the passenger storage demand of the multi-source spatiotemporal data, performs spatiotemporal modeling, including: dividing the internal space of the transportation station into regular grid units, calculating the weight of each factor for each grid for different factors, and generating a demand intensity surface; Step S3: Construct a GIS network dataset for transportation hubs, including: network element settings, impedance settings, and obstacle point settings; Step S4: Establish a multi-objective location-assignment optimization model, including: Step S4-1, setting the objective function; Step S4-2, setting constraints; Step S4-3, using a genetic algorithm to solve the multi-objective optimization model to obtain a set of Pareto optimal solutions. Step S5: Multi-option evaluation and 3D visualization output, including: the first option's storage points are located on the main passenger flow corridor; the second option's storage points are located at different floor heights; the third option's storage points are located next to the service counters at both ends of the connecting corridor.

[0007] Preferably, step S2 involves performing spatiotemporal modeling of passenger storage needs based on multi-source data, including: Step S2-1: The entropy weight method is used to assign weights to calculate the weights of factors such as dwell time, proximity, train frequency, and passenger type. Step S2-2: Using the inverse distance weighted interpolation method of GIS, weights are assigned to each grid, and the weights of all grids are used to generate an intensity surface representing passenger storage demand, including: Step S2-2-1: Calculate the distances from all unknown grids to known grids, as shown in the following formula: ; Step S2-2-2, the weight calculation is as follows: ; Step S2-2-3: Calculate the value of the mesh to be determined, as shown in the following formula: ; in, The value of the grid to be determined, The values ​​are for the known grid cells; x and y represent the coordinates of the known grid cells, and n represents the number of unknown grid cells. The weights of the different factors were obtained as follows: length of stay was 0.25; proximity was 0.35; train frequency was 0.3; and passenger type was 0.1.

[0008] Preferably, step S3 includes: Step S3-1, Network element settings: Set horizontal passages, stairs, and escalators as edges in graph theory; set walking speed as an attribute of the edges; set intersections, stairwells, and elevator entrances as nodes in graph theory. Step S3-2, Impedance setting: Use passage time as the main impedance; set different walking speeds for each side based on the type of different sides and the dwell time and passenger flow driving data. Step S3-3, Obstacle Point Setting: Set the ticket gates and security checkpoints as restriction points, allowing passage under restricted conditions to simulate the flow of traffic in a real transportation hub.

[0009] Preferably, step S4-1, setting the objective function, includes: Step S4-1-1: Set the decision variables as shown in the following formula: ; ; Step S4-1-2, set the objective function, including: Set the objective function to maximize total demand coverage: Obtain the sum of demand intensity of the serviced demand points to maximize the total passenger demand within the service area of ​​all check-in points, as shown in the following formula: ; Set the objective function to minimize the total passenger impedance: minimize the total travel time for all passengers to reach their nearest check-in point, as shown in the following formula: ; in, : Set of demand points, index ; Candidate point set, index ; Demand Points The intensity of demand; From the demand point to candidate point The passage time.

[0010] Preferably, step S4-2 involves setting constraints, including: Step S4-2-1: Set the number of register points The constraints are as follows: ; Step S4-2-2: Set a unique constraint for demand point allocation, as shown in the following formula: ; Step S4-2-3: Allocation depends on register point opening constraints, as shown in the following formula: ; Step S4-2-4: Set service threshold constraints, as shown in the following formula: ; in ; in, As an indicator variable, it represents the demand point. Is it at the storage point? Within the service scope; , indicating demand point to the storage point The passage time does not exceed the threshold ; This indicates that the threshold has been exceeded; Step S4-2-5: Set the service capacity constraints for the storage point, as shown in the following formula: ; Step S4-2-6: Set the minimum spacing constraint for register points, as shown in the following formula: ; The range of values ​​for the variables is as follows: ; This refers to the number of storage points that need to be set up, i.e., the set of all storage points. Two storage points and Spatial distance between them (usually straight-line distance or network distance); The minimum distance between any two registers j and l is the minimum distance threshold, which ensures that the two registers are not too close to each other, thus avoiding service overlap or self-competition. Decision variable, indicating whether or not a candidate point is present. or Set up lockers; Indicates settings, Indicates that no setting is required; For service threshold time; For storage points The upper limit of service capacity.

[0011] Preferably, step S4-3, solving the problem using a genetic algorithm, includes: inputting the objective function from step S4-1 into the genetic algorithm, outputting a Pareto optimal solution set, and providing multiple alternative solutions for decision-making.

[0012] Preferably, step S5 includes, The key evaluation metric for the first option, accessibility coverage, is calculated using the following function: Step S5-1-1, Service Area Generation: Use the service area analysis tool to set input parameters including: checkpoint location, time threshold, and walking speed; set output parameters including: polygonal service area for each checkpoint. Step S5-1-2, Spatial Overlay Analysis: Spatial connection is made between the service area polygons and the demand point grid to identify the demand points located in each service area; Step S5-1-3: Obtain the coverage calculation function, as shown in the following formula: ; in: Time threshold; The set of demand points that can be served by any storage point within t minutes; : Demand intensity at demand point i; : The set of all demand points.

[0013] Preferably, step S5 includes, The design of the calculation function for the key evaluation indicator of the second scheme, namely the supply-demand matching degree, includes: Step S5-2-1, Spatial Aggregation: Using each storage point as the center, generate Thiessen polygons to divide the service area or use network analysis to determine the nearest service point for each demand point; Step S5-2-2, Supply and demand data extraction: For each service area, calculate: Total demand intensity Service capabilities ; Step S5-2-3: Based on the Pearson correlation coefficient, calculate the matching degree using the function shown below: ; in: Number of storage locations; The average demand intensity across all service areas; : The average service capacity of each storage location.

[0014] Preferably, step S5 includes, For the third option, the design of the average walking distance / time calculation function includes: Step S5-3-1, Nearest Facility Analysis: Use the nearest facility point tool and set the input parameters to include: demand point as event point and storage point as facility point; Step S5-3-2 involves calculating the weighted average function, including obtaining the average walking time, as shown in the following formula: ; The average walking distance is obtained as shown in the following formula: ; in: Total number of demand points; Network travel time from demand point i to the nearest storage point; : The network path distance from demand point i to the nearest register point; Demand intensity at demand point i.

[0015] Preferably, step S5 further includes three-dimensional visualization output, which includes displaying the optimal layout scheme in the three-dimensional BIM model of the transportation station.

[0016] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. The demand modeling method involving multi-source data fusion in this application innovatively combines train timetables, passenger dwelling behavior, individual attributes and building space data to construct a more realistic "demand impulse" model and demand intensity surface, breaking through the limitations of relying solely on passenger flow density.

[0017] 2. This application focuses on the construction of a GIS network dataset for the micro-environment of transportation hubs, especially high-speed railway stations: It fully considers the impact of special facilities such as stairs, elevators, and ticket gates on traffic impedance, and establishes a highly realistic pedestrian network within the station, laying the foundation for accurate service area analysis.

[0018] 3. The location-allocation optimization model under multi-objective constraints involved in this application: For the special scenario of transportation hubs, an optimization model is constructed that integrates the three major objectives of "maximizing coverage, minimizing impedance, and maximizing fairness" and incorporates realistic constraints such as service capacity and minimum spacing, making the solution more practical and robust.

[0019] 4. This application relates to visualization-based decision support based on 3D BIM-GIS: it deeply integrates the optimization results of 2D with the 3D BIM model, providing an immersive and interactive environment for scheme display and comparison, which greatly improves the transparency and scientific nature of decision-making. Attached Figure Description

[0020] Figure 1 This is a flowchart of the hosting service layout optimization method based on multi-source data and GIS spatial analysis involved in this application. Detailed Implementation

[0021] To better understand this invention, the following description, in conjunction with the accompanying drawings and examples, will further illustrate the invention.

[0022] The improvement of the hosting service layout optimization method based on multi-source data and GIS spatial analysis lies in, for example, Figure 1 As shown, the method includes the following steps: Step S1, Multi-source spatiotemporal data acquisition and preprocessing, includes collecting and fusing multi-source spatiotemporal data related to transportation hubs to construct a basic database for analysis; the multi-source spatiotemporal data includes: building space data and passenger flow spatiotemporal data, including: mobile phone signaling data, WiFi probe data, train timetable data and passenger attribute data.

[0023] The collection of building space data includes: obtaining detailed 3D BIM models or 2D floor plans of transportation hubs, including the spatial location and attributes of elements such as entrances and exits, waiting areas, ticket gates, shops, restaurants, transfer passages (subway, bus), stairs, elevators, etc.

[0024] Passenger flow spatiotemporal data: This includes information on time, space, and passenger flow attributes, including: Mobile signaling data includes: anonymized passenger location data, used to analyze the macroscopic movement trajectories and gathering hotspots of people within the station.

[0025] WiFi probe data: Deploy probe devices at key nodes within the station to obtain the time and strength of received mobile phone signals, in order to capture more detailed passenger dwell and movement behaviors.

[0026] Train timetable data: Obtaining the time, train number, and gate location of all arriving and departing trains is key to predicting the pulse distribution of demand.

[0027] Passenger attribute data: Through questionnaires or APP data, obtain social attributes such as passengers' travel purpose (business / tourism), amount of luggage, transfer time, acceptable walking distance and service price.

[0028] Step S2 involves performing spatiotemporal modeling of passenger storage demand based on multi-source data, including: dividing the internal space of the transportation station into regular grid cells, calculating the demand weight of each factor for each grid cell based on different factors, and generating a demand intensity surface; the different factors include: dwell time, proximity, train frequency, and passenger type; Using multi-source spatiotemporal data, a spatiotemporal distribution model of passenger storage demand is established in a GIS platform, including step S2-1, dividing the internal space of the transportation station into regular grid units (such as 10m×10m). For each grid, the entropy weight method is used to assign weights to calculate the weights of factors such as dwell time, proximity, train number drive and passenger type.

[0029] The entropy weight method assigns weights based on the dispersion (variability) of the data for each factor; the greater the data difference, the higher the weight. The weights for stay time, proximity, train frequency, and passenger type are 0.25, 0.35, 0.3, and 0.1, respectively.

[0030] The calculation of dwell time weighting includes: calculating the average dwell time of passengers within the grid based on mobile signaling / WiFi data. The longer the dwell time, the higher the demand for luggage storage may be (e.g., in waiting areas).

[0031] The calculation of proximity weights includes: calculating the GIS spatial distance from the grid to key facilities (such as ticket gates, shops, restaurants, and restrooms). The closer to shops and restaurants, the higher the demand weight; if the distance to the ticket gate is too close (<50 meters), the weight decreases, because passengers may board directly.

[0032] The calculation of train number driving weights includes: dynamically and significantly increasing the grid demand weight of the corresponding ticket gate area within a specific time window before and after the train's arrival, simulating a "demand pulse".

[0033] The calculation of passenger type weights includes: through data fusion, identifying areas where tour groups or families who may carry large luggage congregate, and further increasing the demand weight of those areas.

[0034] Step S2-2: Based on the above factors, use the inverse distance weight interpolation method of GIS to set weights for each grid (network unit) and generate a continuous "passenger storage demand intensity surface" by combining the weights of all grids.

[0035] The steps of the inverse distance weighted interpolation method are as follows: Step S2-2-1: Calculate the distances from all unknown grids to known grids (locations that have been measured or whose data are available), as shown in the following formula: ; Step S2-2-2, the weight calculation is as follows: ; Step S2-2-3: Calculate the values ​​of the mesh to be interpolated, i.e., the unknown mesh, as shown in the following formula: ; in, The value of the grid to be determined, The values ​​are for the known grid cells. x and y represent the coordinates of the known grid cells, and n represents the number of unknown grid cells.

[0036] Step S3: Construct a high-precision GIS network dataset within the station and model existing facilities. Building a network dataset in GIS that reflects actual traffic conditions is fundamental for accurate service area analysis, including: Step S3-1, Network element settings: Set the walking speed of horizontal passages, stairs and escalators as edges, and set intersections, stairwells and elevator entrances as nodes.

[0037] Step S3-2, Impedance Setting: Use travel time as the primary impedance. Based on the type of side (horizontal passageway, stairs, escalator) and congestion levels (calculated from passenger flow data: passenger dwell time, passenger flow-driven data), set different walking speeds for each side. For example, set the walking speed for horizontal passageways to 1.2 m / s and for upward stairs to 0.7 m / s.

[0038] Step S3-3, Obstacle Point Setting: Set the ticket gates and security checkpoints as restricted points, allowing passage only under specific conditions (such as after ticket inspection) to simulate real station flow.

[0039] Step S4: Establish a multi-target location-allocation optimization model. This is the core algorithm of this invention. The model aims to select the optimal K locations for setting up the storage lockers from all candidate locations (determined through spatial analysis in Step 1, such as vacant corners, walls, etc.), including: Step S4-1, set the objective function, including: Step S4-1-1: Set the decision variables as shown in the following formula: ; ; Step S4-1-2, set the objective function, including: Set the objective function to maximize total demand coverage: obtain the sum of demand intensity for the serviced demand points. , is used to represent maximizing the sum of demand intensity of all assigned demand points, thereby maximizing the total passenger demand within the service area of ​​all check-in points, as shown in the following formula: ; Set a target function to minimize total passenger impedance, which means minimizing the sum of the products of demand intensity and travel time at each service point: minimizing the total travel time for all passengers to reach their nearest checkpoint. Minimize total passenger impedance (minimize the total weighted travel time of all passengers to their assigned checkpoints): ; in, : Set of demand points (grid), indexed as ; : Candidate point set, index is ; Demand Points The intensity of demand (obtained from step S2); From the demand point to candidate point The passage time.

[0040] Step S4-2, set constraints, including: number of registers. Constraints (set by the administrator); service capacity limit constraint for each storage point; minimum distance constraint between any two storage points (to avoid self-competition). Furthermore, each demand point (grid) must be within a specific time threshold (e.g., 5 minutes) of its nearest storage point.

[0041] Specifically, set constraints, including: Step S4-2-1: Set the number of register points The constraints are as follows: ; Step S4-2-2: Set a unique constraint for the assignment of requirement points (allowing them to not be overridden), as shown in the following formula: ; Step S4-2-3: Allocation depends on register point opening constraints, as shown in the following formula: ; Step S4-2-4: Set service threshold constraints (allocation is only allowed when the passage time does not exceed the threshold), as shown in the following formula: ; in ; in, As an indicator variable, it represents the demand point. Is it at the storage point? Within the service scope; , indicating demand point to the storage point The passage time does not exceed the threshold ; This indicates that the threshold has been exceeded; Step S4-2-5: Set the service capacity constraints for the storage point, as shown in the following formula: ; Step S4-2-6: Set the minimum spacing constraint for register points (to avoid self-competition), as shown in the following formula: ; The range of values ​​for the variables is as follows: ; Service threshold time (e.g., 5 minutes); : The number of storage points to be set, i.e., the set of all storage points (set by the administrator); Storage point The upper limit of service capacity; Two storage points and Spatial distance between them (usually straight-line distance or network distance); The minimum distance between any two registers j and l is the minimum distance threshold, which ensures that the two registers are not too close to each other, thus avoiding service overlap or self-competition. Decision variable, indicating whether or not a candidate point is present. or Set up lockers; Indicates settings, This indicates that no settings are required.

[0042] Step S4-3, Solving Algorithm: The genetic algorithm is used to solve the multi-objective optimization model in the GIS modeling toolbox to obtain a set of Pareto optimal solutions.

[0043] The multi-objective optimization model function is a bi-objective integer programming model. Based on the objective function described in step S4-1 above, a genetic algorithm is used to solve for the Pareto optimal solution set of this model. The genetic algorithm solution includes: Step S4-3-1, set the encoding, including: Binary encoding is used, and the chromosome length is equal to the number of candidate points. gene locus Indicates whether it is at the candidate point Set up lockers.

[0044] The constraint must be satisfied: the number of 1s in the chromosome is equal to... And the distance between any two candidate points corresponding to 1 is not less than .

[0045] Step S4-3-2, initialize the population, including: Randomly generated Each individual (chromosome) is selected randomly. Candidate points that satisfy the minimum distance constraint are constructed. If the constraint is not satisfied, they are regenerated or repaired.

[0046] Step S4-3-3, Fitness Assessment (Decoding and Target Value Calculation): For each individual (i.e., a set of selected storage points) The following heuristic allocation algorithm is used to calculate... And two target values: a. For each demand point Find all selected registers that meet the following conditions. set b. If Not empty, will Assigned to middle Minimum register However, capability constraints need to be checked: if after allocation Total demand If the allocation is successful, then the allocation is valid; otherwise, no allocation is made. (Right now (Not covered). c. If If empty, no allocation will be performed. It will not be covered.

[0047] Calculate based on the allocation results: ; If an individual violates the minimum distance constraint, a maximum penalty value is imposed (e.g., , ).

[0048] Step S4-3-4, Non-dominated sorting and crowding calculation (NSGA-II mechanism): Perform a non-dominated ranking of all individuals in the population: compare the target values ​​of two individuals. If individual A is no worse than individual B in both objectives and is better in at least one objective, then A dominates B. Individuals are stratified (frontier hierarchy) based on the dominance relationship.

[0049] Calculate the crowding degree of individuals within the same non-dominated layer: sort by each objective function value, calculate the sum of objective value differences between adjacent solutions of each individual as the crowding degree (reflecting the sparsity of solutions in the objective space).

[0050] Step S4-3-5, select operation: Using a binary tournament selection: randomly select two individuals, prioritizing the individual with the lower non-dominant ranking; if the rankings are the same, select the individual with the higher crowding.

[0051] Step S4-3-6, cross operation: Single-point crossover is used: two parent individuals are randomly selected and a portion of their genes are exchanged at a random location.

[0052] Repair offspring: Ensure that the number of 1s in the offspring is 1. If the condition is not met, add or delete 1 randomly; at the same time, check the minimum distance constraint, and if it is not met, fix it by local adjustment (such as replacing conflict points).

[0053] Step S4-3-7, Mutation operation: Each gene locus is flipped (0 becomes 1 or 1 becomes 0) with a small probability, and the same repair is performed to meet the requirements. And minimum distance constraint.

[0054] Step S4-3-8, Iterative operation: Repeat steps S4-3-3 to S4-3-7 to generate a new generation of population until the maximum number of iterations is reached.

[0055] Step S4-3-9, output the Pareto optimal solution set: The set of individuals in the final population that belong to the first non-dominated layer (i.e., not dominated by any other solution) is the Pareto optimal solution set, providing multiple alternatives for decision-making.

[0056] Step S5: Multi-scheme evaluation and 3D visualization output. Scheme evaluation: For the multiple layout schemes obtained from the solution, use the service area analysis tool in network analysis within GIS to calculate the key evaluation indicators for each scheme, including: Suppose that the genetic algorithm in step S4 yields three Pareto optimal solutions: Step S5-1, Obtain the first solution (coverage priority): The luggage storage points are located at: the entrance to the waiting hall, next to the central ticket gate, and in the catering area; J=3 luggage storage points. Features of the luggage storage points: distributed along main passenger flow routes, providing wide service coverage.

[0057] For the first option, accessibility coverage will be used as a key evaluation indicator to obtain the proportion of total demand within a 3-minute or 5-minute walking range. Specifically, accessibility coverage is the proportion of the total demand intensity of all points reachable from the check-in point within a specific walking time to the total demand intensity of the entire station.

[0058] The specific calculation function for the key evaluation indicator, accessibility coverage, in the first option is as follows: S5-1-1, Service Area Generation: Using a GIS network analysis tool like "Service Area Analysis" (such as ArcGIS's Service Area Analysis), input: location of the storage point, time threshold (3 minutes, 5 minutes), and walking speed (e.g., 1.2 m / s); output: polygonal service areas (isochronous circles) for each storage point. S5-1-2, Spatial Overlay Analysis: Spatial Join is performed between the service area polygons and the demand point grid to identify the demand points located within each service area; S5-1-3, Obtain the coverage calculation function, as shown in the following formula: ; in: Time threshold (3 or 5 minutes); The set of demand points that can be served by any storage point within t minutes; : Demand intensity at demand point i; : The set of all demand points.

[0059] Step S5-2, obtain the second solution (balanced type): Storage points are distributed at different floor levels: such as the arrival hall on the basement level, the middle of the waiting hall on the second floor, and the elevator entrance in the commercial area on the third floor. J=3 storage points. Features: Three-dimensional distribution, accommodating both arriving and waiting passengers.

[0060] For the second option, the supply-demand matching degree is used as a key evaluation indicator, and the spatial correlation coefficient between demand intensity and service capacity is calculated. Specifically, the supply-demand matching degree measures the degree of matching between the service capacity of storage points and demand intensity in spatial distribution, and is calculated using the spatial correlation coefficient.

[0061] The design of the calculation function for the key evaluation indicator of the second scheme, namely the supply-demand matching degree, includes: S5-2-1, Spatial Aggregation: For each storage point, generate a Voronoi diagram to divide the service area or use network analysis to determine the nearest service point for each demand point; S5-2-2, Supply and Demand Data Extraction: For each service area, calculate: Total demand intensity Service capabilities (Determined by the locker capacity).

[0062] S5-2-3, Calculation of the matching degree function based on Pearson correlation coefficient: ; in: Number of storage locations; The average demand intensity across all service areas; : The average service capacity of each storage location.

[0063] Step S5-3, obtain the third option (efficiency-first type). The storage points are located next to the service counters at both ends of the connecting corridor. J = 2 storage points. Features: maximized spacing, reduced self-competition.

[0064] For the third option, the average walking distance / time is used as the key evaluation indicator, and the average distance / time from all demand points to the nearest storage point is calculated. Specifically, the average walking distance / time is the weighted average distance or time from all demand points to the nearest storage point.

[0065] For the third option, the design of the average walking distance / time calculation function includes: S5-3-1, Analysis of Recent Facilities: Using the "Closest Facility" tool in GIS network analysis, input the following: demand points as event points and storage points as facility points. S5-3-2, Weighted average calculation function, includes: Average walking time: ; Average walking distance: ; in: Total number of demand points; Network travel time from demand point i to the nearest storage point; : The network path distance from demand point i to the nearest register point; Demand intensity at demand point i (as weight) Step S5, 3D visualization output, involves making 3D visualization decisions by visualizing the optimal layout schemes in the 3D BIM model of the transportation hub. Managers can intuitively view the location of the lockers and their 3D service range (such as a 3D "bubble" generated with a radius of 5 minutes' walking time) from any angle, and view various evaluation indicators in conjunction with these parameters, achieving a "what you see is what you get" scientific decision-making process.

[0066] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0067] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0068] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0069] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0070] The above are merely embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention are included within the scope of the claims of the present invention pending approval.

Claims

1. A method for optimizing the layout of hosting services based on multi-source data and GIS spatial analysis, characterized in that, The method includes the following steps: Step S1, Multi-source spatiotemporal data acquisition and preprocessing, includes collecting and fusing multi-source spatiotemporal data related to transportation hubs to construct a database; the multi-source spatiotemporal data includes: building space data, passenger flow spatiotemporal data, mobile phone signaling data, WiFi probe data, train timetable data and passenger attribute data; Step S2, based on the passenger storage demand of the multi-source spatiotemporal data, performs spatiotemporal modeling, including: dividing the internal space of the transportation station into regular grid units, calculating the weight of each factor for each grid for different factors, and generating a demand intensity surface; Step S3: Construct a GIS network dataset for transportation hubs, including: network element settings, impedance settings, and obstacle point settings; Step S4: Establish a multi-objective location-assignment optimization model, including: Step S4-1, setting the objective function; Step S4-2, setting constraints; Step S4-3, using a genetic algorithm to solve the multi-objective optimization model to obtain a set of Pareto optimal solutions. Step S5: Multi-option evaluation and 3D visualization output, including: the first option's storage points are located on the main passenger flow corridor; the second option's storage points are located at different floor heights; the third option's storage points are located next to the service counters at both ends of the connecting corridor.

2. The hosting service layout optimization method based on multi-source data and GIS spatial analysis as described in claim 1, characterized in that, Step S2 involves performing spatiotemporal modeling of passenger storage needs based on multi-source data, including: Step S2-1: The entropy weight method is used to assign weights to calculate the weights of factors such as dwell time, proximity, train frequency, and passenger type. Step S2-2: Using the inverse distance weighted interpolation method of GIS, weights are assigned to each grid, and the weights of all grids are used to generate an intensity surface representing passenger storage demand, including: Step S2-2-1: Calculate the distances from all unknown grids to known grids, as shown in the following formula: ； Step S2-2-2, the weight calculation is as follows: ； Step S2-2-3: Calculate the value of the mesh to be determined, as shown in the following formula: ； in, The value of the grid to be determined, The values ​​are for the known grid cells; x and y represent the coordinates of the known grid cells, and n represents the number of unknown grid cells. The weights of the different factors were obtained as follows: length of stay was 0.25; proximity was 0.35; train frequency was 0.3; and passenger type was 0.

1.

3. The hosting service layout optimization method based on multi-source data and GIS spatial analysis as described in claim 1, characterized in that, Step S3 includes: Step S3-1, Network element settings: Set horizontal passages, stairs, and escalators as edges in graph theory; set walking speed as an attribute of the edges; set intersections, stairwells, and elevator entrances as nodes in graph theory. Step S3-2, Impedance setting: Use passage time as the main impedance; set different walking speeds for each side based on the type of different sides and the dwell time and passenger flow driving data. Step S3-3, Obstacle Point Setting: Set the ticket gates and security checkpoints as restriction points, allowing passage under restricted conditions to simulate the flow of traffic in a real transportation hub.

4. The hosting service layout optimization method based on multi-source data and GIS spatial analysis as described in claim 1, characterized in that, Step S4-1, set the objective function, including: Step S4-1-1, set the decision variables as shown in the following formula: ； ； Step S4-1-2, set the objective function, including: Set the objective function to maximize total demand coverage: Obtain the sum of demand intensity of the serviced demand points to maximize the total passenger demand within the service area of ​​all check-in points, as shown in the following formula: ； Set the objective function to minimize the total passenger impedance: minimize the total travel time for all passengers to reach their nearest check-in point, as shown in the following equation: ； in, : Set of demand points, index ; Candidate point set, index ; Demand Points The intensity of demand; From the demand point to candidate point The passage time.

5. The hosting service layout optimization method based on multi-source data and GIS spatial analysis as described in claim 1, characterized in that, Step S4-2, set constraints, including: Step S4-2-1: Set the number of register points The constraints are as follows: ； Step S4-2-2: Set a unique constraint for demand point allocation, as shown in the following formula: ； Step S4-2-3: Allocation depends on register point opening constraints, as shown in the following formula: ； Step S4-2-4: Set service threshold constraints, as shown in the following formula: ； in ; in, As an indicator variable, it represents the demand point. Is it at the storage point? Within the service scope; , indicating demand point to the storage point The passage time does not exceed the threshold ; This indicates that the threshold has been exceeded; Step S4-2-5: Set the service capacity constraints for the storage point, as shown in the following formula: ； Step S4-2-6: Set the minimum spacing constraint for register points, as shown in the following formula: ； The range of values ​​for the variables is as follows: ; The number of storage points to be set; For any two registers and Spatial distance between them; The minimum distance between any two register points j and l; Decision variable, indicating whether or not a candidate point is present. or Set up lockers; Indicates settings, Indicates that no setting is required; For service threshold time; For storage points The upper limit of service capacity; The range of values ​​for the variables is as follows: ; This refers to the number of storage points that need to be set up, i.e., the set of all storage points. Let j be the minimum distance between any two register points j and l. For service threshold time; For storage points The upper limit of service capacity.

6. The hosting service layout optimization method based on multi-source data and GIS spatial analysis as described in claim 1, characterized in that, Step S4-3, solving the problem using a genetic algorithm, includes: inputting the objective function from step S4-1 into the genetic algorithm, outputting a Pareto optimal solution set, and providing multiple alternative solutions for decision-making.

7. The hosting service layout optimization method based on multi-source data and GIS spatial analysis as described in claim 1, characterized in that, Step S5 includes, The key evaluation metric for the first option, accessibility coverage, is calculated using the following function: Step S5-1-1, Service Area Generation: Use the service area analysis tool to set input parameters including: checkpoint location, time threshold, and walking speed; set output parameters including: polygonal service area for each checkpoint. Step S5-1-2, Spatial Overlay Analysis: Spatial connection is made between the service area polygons and the demand point grid to identify the demand points located in each service area; Step S5-1-3: Obtain the coverage calculation function, as shown in the following formula: ； in: Time threshold; The set of demand points that can be served by any storage point within t minutes; : Demand intensity at demand point i; : The set of all demand points.

8. The hosting service layout optimization method based on multi-source data and GIS spatial analysis as described in claim 1, characterized in that, Step S5 include, The design of the calculation function for the key evaluation indicator of the second scheme, namely the supply-demand matching degree, includes: Step S5-2-1, Spatial Aggregation: Using each storage point as the center, generate Thiessen polygons to divide the service area or use network analysis to determine the nearest service point for each demand point; Step S5-2-2, Supply and demand data extraction: For each service area, calculate: Total demand intensity Service capabilities ; Step S5-2-3: Based on the Pearson correlation coefficient, calculate the matching degree using the function shown below: ； in: Number of storage locations; The average demand intensity across all service areas; : The average service capacity of each storage location.

9. The hosting service layout optimization method based on multi-source data and GIS spatial analysis as described in claim 1, characterized in that, Step S5 includes, For the third option, the design of the average walking distance / time calculation function includes: Step S5-3-1, Nearest Facility Analysis: Use the nearest facility point tool and set the input parameters to include: demand point as event point and storage point as facility point; Step S5-3-2 involves calculating the weighted average function, including obtaining the average walking time, as shown in the following formula: ； The average walking distance is obtained as shown in the following formula: ； in: Total number of demand points; Network travel time from demand point i to the nearest storage point; : The network path distance from demand point i to the nearest register point; Demand intensity at demand point i.

10. The hosting service layout optimization method based on multi-source data and GIS spatial analysis as described in claim 1, characterized in that, Step S5 also includes 3D visualization output, which includes displaying the optimal layout scheme in the 3D BIM model of the transportation station.

Images