Scenic spot flow prediction method, system, equipment terminal and storage medium thereof

Abstract

The present invention discloses a method, system, equipment terminal and storage medium for predicting the flow of tourists in a scenic area, and specifically relates to the field of scenic area management technology, and is used to solve the problem that the prior art lacks a comprehensive analysis and dynamic prediction method for the flow rules of tourists in scenic areas; by collecting multi-source tourist behavior data and performing preprocessing, a node relationship model of functional areas of scenic areas based on group behavior principles and geographic information systems is constructed; by calculating tourist congestion parameters and dynamic distribution parameters, a dynamic distribution state of functional areas is generated; based on historical passenger flow data and real-time monitoring data, a tourist flow propagation matrix is ​​constructed and migration weights are optimized; combining the propagation matrix and dynamic distribution parameters, the passenger flow of scenic areas in different time periods in the future is predicted, and time-based passenger flow distribution data is generated; finally, potential high-risk areas are marked by analyzing tourist density, and the accuracy and dynamic adaptability of passenger flow prediction in scenic areas are improved.

Description

Technical Field

[0001] The present invention relates to the field of scenic spot management technology, and more specifically, to a scenic spot passenger flow prediction method, system, equipment terminal and storage medium thereof. Background Art

[0002] During the peak season, tourists tend to flow in scenic spots like "fluids". When a functional area is crowded, tourists will naturally spread to surrounding areas to ease congestion. When a certain area is relatively idle, it will attract tourists to gather spontaneously. Existing crowd monitoring and prediction methods for scenic spots usually focus on a single data source or static measurement, which makes it difficult to capture the dynamic migration, congestion diffusion and re-condensation of tourists between different areas in a timely manner, and cannot provide scenic area management with an accurate portrayal of the laws of self-organized migration of tourists.

[0003] The existing technology lacks a comprehensive analysis of the spontaneous diffusion and cohesion patterns of tourists among different areas of the scenic area. When congestion occurs during peak hours, it is often impossible to effectively predict the flow trend of tourists, which will lead to a lag in the scenic area management's response to instantaneous changes in passenger flow. Summary of the invention

[0004] In order to overcome the above-mentioned defects of the prior art, the embodiments of the present invention provide a scenic spot passenger flow prediction method, system, device terminal and storage medium thereof to solve the problems raised in the above-mentioned background technology.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] The method for predicting the flow of people in a scenic spot includes the following steps:

[0007] S1: Collect multi-source tourist behavior data and pre-process the multi-source tourist behavior data;

[0008] S2: Based on the principles of group behavior and geographic information system data, a node relationship model of each functional area of ​​the scenic area is constructed;

[0009] S3: Calculate the tourist crowding parameters of each functional area based on the node relationship model and multi-source tourist behavior data, and generate dynamic distribution parameters of the functional area by combining the time dimension and space dimension;

[0010] S4: Compare and analyze historical passenger flow data with real-time monitoring data, construct a tourist flow propagation matrix between functional areas, and dynamically optimize and update the migration weights between functional areas through regression algorithms;

[0011] S5: Based on the updated tourist flow propagation matrix and the dynamic distribution parameters of the functional areas, the tourist flow of the scenic spot at different time periods in the future is predicted to generate the time-based tourist flow distribution data within the functional areas;

[0012] S6: Analyze the density of tourists in the functional areas based on the time-based crowd distribution data in the functional areas, and mark potential high-risk areas for crowd flow.

[0013] In a preferred embodiment, collecting multi-source tourist behavior data and preprocessing the multi-source tourist behavior data include:

[0014] Multi-source tourist behavior data includes image data collected by video surveillance equipment, location data from tourist smart terminals, and historical transaction data from the ticket sales system;

[0015] Preprocess multi-source tourist behavior data: identify and remove outliers in multi-source tourist behavior data; filter noise in multi-source tourist behavior data to eliminate irrelevant information; synchronize timestamps of multi-source tourist behavior data to ensure temporal consistency of data from different sources; standardize multi-source tourist behavior data to unify the format and scope of the data.

[0016] In a preferred embodiment, based on the group behavior principle and geographic information system data, a node relationship model of each functional area of ​​the scenic area is constructed, including:

[0017] Obtaining geographic information system data of the scenic area, which includes geographic boundary information of functional areas, traffic path information and location data of important landmarks in the scenic area;

[0018] Based on the GIS data, the scenic area is divided into multiple functional areas, which are connected by access paths, and the adjacency relationship and path attributes between the functional areas are recorded;

[0019] A node relationship model of functional areas is constructed based on the principle of group behavior. The node relationship model includes the following contents: each functional area is defined as a node in the node relationship model; the traffic paths between functional areas are defined as edges connecting nodes, and the attributes of the edges include path length, traffic capacity and congestion coefficient; the geographical boundaries, entrance and exit locations and accessible directions of each functional area are recorded.

[0020] In a preferred embodiment, the tourist crowding parameters of each functional area are calculated based on the node relationship model and multi-source tourist behavior data, and the dynamic distribution parameters of the functional area are generated by combining the time dimension and the space dimension, including:

[0021] Calculate the real-time tourist density of the functional area, which is the ratio of the current number of tourists in the functional area to the area of ​​the functional area;

[0022] By analyzing the trajectory data of tourists in the functional areas, the average length of stay of tourists in the functional areas is calculated;

[0023] Divide the tourist data by time period to generate the distribution of the number of tourists in each functional area at different time periods;

[0024] Based on the adjacency relationship and path attributes between functional areas, the tourist flow ratio between functional areas and adjacent areas is calculated;

[0025] Based on the real-time tourist density, tourist flow ratio and average stay time of the functional areas, the dynamic distribution parameters are obtained by comprehensive calculation: the real-time tourist density, tourist flow ratio and average stay time are dimensionlessly processed, the real-time tourist density, tourist flow ratio and average stay time are assigned weight coefficients respectively, the real-time tourist density, the inverse of the tourist flow ratio and the average stay time are multiplied by their corresponding weight coefficients and then added to obtain the dynamic distribution parameters.

[0026] In a preferred embodiment, historical passenger flow data is compared and analyzed with real-time monitoring data to construct a tourist flow propagation matrix between functional areas, and the migration weights between functional areas are dynamically optimized and updated through a regression algorithm, including:

[0027] Obtain historical passenger flow data and real-time monitoring data for functional areas, and compare the number of tourists and flow changes by time period and functional area;

[0028] The number of tourists flowing between functional areas is counted based on real-time monitoring data, and the proportion of tourists flowing is calculated in combination with historical passenger flow data;

[0029] A tourist flow propagation matrix is ​​constructed based on the adjacency relationship between functional areas and the tourist flow ratio. The elements in the tourist flow propagation matrix represent the tourist flow relationship between adjacent functional areas.

[0030] Based on the tourist flow propagation matrix, the regression algorithm is used to calculate the migration weights between functional areas, and the weights are dynamically optimized according to real-time monitoring data. The optimization goal is to minimize the prediction error of the migration weights.

[0031] In a preferred embodiment, based on the updated tourist flow propagation matrix combined with the dynamic distribution parameters of the functional area, the flow of tourists in the scenic area at different time periods in the future is predicted to generate the time-based distribution data of the flow of tourists in the functional area, including:

[0032] A prediction model is constructed based on the updated tourist flow propagation matrix. The prediction model takes the tourist flow relationship between functional areas as the input variable and the number of tourists in the functional areas in the future as the output variable.

[0033] The historical passenger flow data and real-time monitoring data of the functional area are input into the prediction model as time series data, and the dynamic distribution parameters of the functional area and the tourist flow propagation matrix are also input;

[0034] The prediction model is used to calculate the passenger flow data of functional areas at different time periods in the future, including the number of future tourists in each functional area and their flow trend;

[0035] The predicted functional area flow data is recorded as periodic flow distribution data according to time periods.

[0036] In a preferred embodiment, the density of tourists in the functional area is analyzed based on the time-based crowd flow distribution data in the functional area, and potential high-risk areas for crowd flow are marked, including:

[0037] Calculate the tourist density in the functional area based on the area of ​​the functional area and the distribution data of the flow of people in different time periods;

[0038] The high-risk tourist density threshold of the functional area is set according to the management needs of the scenic spot, the calculated tourist density of the functional area is compared with the high-risk density threshold, and the functional area where the tourist density is greater than the high-risk density threshold is marked as a potential high-risk area for human flow.

[0039] On the other hand, the present invention provides a scenic spot flow prediction system, comprising:

[0040] Multi-source data processing module: collects multi-source tourist behavior data and pre-processes the multi-source tourist behavior data;

[0041] Functional area modeling module: Based on the principles of group behavior and geographic information system data, the node relationship model of each functional area of ​​the scenic area is constructed;

[0042] Dynamic parameter generation module: calculates the tourist crowding parameters of each functional area based on the node relationship model and multi-source tourist behavior data, and generates dynamic distribution parameters of the functional area by combining the time dimension and space dimension;

[0043] Propagation matrix optimization module: compares and analyzes historical passenger flow data with real-time monitoring data, constructs a tourist flow propagation matrix between functional areas, and dynamically optimizes and updates the migration weights between functional areas through regression algorithms;

[0044] Future passenger flow prediction module: Based on the updated tourist flow propagation matrix and the dynamic distribution parameters of the functional areas, the passenger flow of the scenic area at different time periods in the future is predicted, and the time period passenger flow distribution data within the functional area is generated;

[0045] Risk area marking module: Analyze the density of tourists in functional areas based on the time-based passenger flow distribution data in the functional areas, and mark potential high-risk passenger flow areas.

[0046] On the other hand, the present invention provides a device terminal, which includes: a processor, a memory, and a program or instruction stored in the memory and executable on the processor, and a method for predicting the flow of people in a scenic spot is implemented when the program or instruction is executed by the processor.

[0047] On the other hand, the present invention provides a computer-readable storage medium, on which a program or instruction is stored, and when the program or instruction is executed by a processor, a method for predicting the flow of people in a scenic spot is implemented.

[0048] The technical effects and advantages of the scenic spot flow prediction method, system, device terminal and storage medium thereof of the present invention are as follows:

[0049] 1. By collecting and preprocessing multi-source tourist behavior data, based on the principles of group behavior and geographic information system data, a node relationship model of each functional area of ​​the scenic area is constructed, so as to accurately characterize the geographical boundaries and traffic attributes between functional areas; by calculating the tourist congestion parameters and dynamic distribution parameters, and comprehensively considering the dynamic changes in the time dimension and space dimension, it can accurately reflect the distribution status and flow pattern of tourists in the scenic area, especially when the tourist flow shows "spontaneous diffusion" and "spontaneous condensation" phenomena during peak periods, it can timely capture and analyze this dynamic migration process, providing basic data support for subsequent crowd flow prediction and risk warning.

[0050] 2. Based on the updated tourist flow propagation matrix and the dynamic distribution parameters of the functional areas, the migration weights between functional areas are optimized through the regression algorithm, which realizes the high-precision prediction of the scenic spot flow at different time periods in the future. By generating the time-based passenger flow distribution data within the functional areas, potential high-density areas can be identified in advance, and high-risk areas can be marked based on the calculated results of tourist density, thereby providing accurate decision-making basis for scenic spot management. Compared with the existing technology, it not only has significant improvements in data comprehensiveness and prediction accuracy, but also can dynamically adapt to changes in tourist flow characteristics, solving the problem of lagging response to peak-period congestion risks in traditional methods. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 It is a schematic diagram of the method for predicting the flow of people in a scenic spot according to the present invention;

[0052] Figure 2 It is a structural schematic diagram of the scenic spot passenger flow prediction system of the present invention. DETAILED DESCRIPTION

[0053] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0054] Embodiment 1: Figure 1 A method for predicting the flow of people in a scenic spot is provided, which comprises the following steps:

[0055] S1: Collect multi-source tourist behavior data and pre-process the multi-source tourist behavior data.

[0056] S2: Based on the principles of group behavior and geographic information system data, a node relationship model of each functional area of ​​the scenic area is constructed.

[0057] S3: Based on the node relationship model and multi-source tourist behavior data, the tourist congestion parameters of each functional area are calculated, and the dynamic distribution parameters of the functional area are generated by combining the time dimension and the space dimension.

[0058] S4: Compare and analyze historical passenger flow data with real-time monitoring data, construct a tourist flow propagation matrix between functional areas, and dynamically optimize and update the migration weights between functional areas through regression algorithms.

[0059] S5: Based on the updated tourist flow propagation matrix and the dynamic distribution parameters of the functional areas, the tourist flow of the scenic area at different time periods in the future is predicted to generate the time-based tourist flow distribution data within the functional areas.

[0060] S6: Analyze the density of tourists in the functional areas based on the time-based crowd distribution data in the functional areas, and mark potential high-risk areas for crowd flow.

[0061] Collect multi-source tourist behavior data and pre-process the multi-source tourist behavior data, including:

[0062] Multi-source tourist behavior data includes image data collected by video surveillance equipment, positioning data of tourist smart terminals, and historical transaction data of the ticket sales system:

[0063] Multiple video surveillance devices are deployed in the scenic area, and each device is set with different shooting angles and area coverage. The video surveillance equipment continuously records the dynamic behavior image data of tourists in the scenic area, and transmits the collected image data to the data processing server through a wired or wireless network. The image data contains the real-time location, dynamic trajectory and visual information of the tourists' crowding level.

[0064] Tourist smart terminals include smartphones or portable devices with GPS positioning functions. Through the scenic area navigation application installed by tourists, the geographical location data of tourists is collected in real time. The collected positioning data is recorded in the form of longitude and latitude, with timestamp information, and uploaded to the central data processing platform of the scenic area.

[0065] The scenic spot's electronic ticket system records each visitor's entry time, ticket type, and entrance location. Historical transaction data is regularly exported through the data interface and stored in a structured database for subsequent analysis and processing.

[0066] Preprocess multi-source tourist behavior data:

[0067] Identify and remove outliers in multi-source tourist behavior data:

[0068] In the preprocessing stage, the collected image data, positioning data and historical transaction data are comprehensively checked to identify abnormal values ​​in the data. For example, for image data, the image blur or loss caused by monitoring equipment obstruction or weather conditions (such as rain and fog) is identified; for positioning data, the data points whose longitude and latitude deviate from the reasonable range (such as beyond the geographical boundaries of the scenic area) are identified; for historical transaction data, the integrity of the ticket purchase information is checked, such as whether there are duplicate records or invalid records.

[0069] Outliers are removed through automated algorithms, such as using an outlier detection model based on historical distribution, or by defining the anomaly range through rules and then removing them in batches.

[0070] Filter the noise in multi-source tourist behavior data to eliminate irrelevant information:

[0071] Filter the noise in multi-source data; for example: for image data, use the convolutional neural network (CNN) algorithm to separate and filter non-tourist targets (such as animals, scenic area equipment) in the picture, and only retain key image information related to tourist behavior; for positioning data, use the Kalman filter algorithm to correct the jitter phenomenon in the positioning trajectory and eliminate errors caused by weak signals or interference; for historical transaction data, clean irrelevant fields (such as payment methods, marketing information, etc.) and only retain data related to time, location and ticket purchase category.

[0072] Synchronize timestamps of multi-source visitor behavior data to ensure temporal consistency of data from different sources:

[0073] To ensure the consistency of timestamps of video surveillance data, positioning data, and transaction data, a global time synchronization server is used to uniformly time-calibrate the data:

[0074] The frame timestamps of video surveillance data are aligned to the second level, and the timestamps of positioning data are calibrated with millisecond accuracy to ensure continuity. For historical transaction data, the missing time information is completed based on the ticket purchase and entry time as the reference time points.

[0075] Standardize multi-source tourist behavior data to unify the format and scope of the data:

[0076] The data from the three sources are standardized, including but not limited to: image data is standardized to a resolution of 1080p and a frame rate of 30fps; positioning data is represented by a unified WGS84 geographic coordinate system, and the timestamp is converted to the UTC standard time format; historical transaction data is converted to a structured data table format, and the fields are uniformly named "entry time", "ticket purchase category", "entrance location", etc. In addition, the numerical data is normalized, and the latitude and longitude values ​​and timestamp values ​​in the positioning data are adjusted to the range of [0, 1] to facilitate the input of subsequent analysis algorithms.

[0077] Based on the principles of group behavior and geographic information system data, a node relationship model of each functional area of ​​the scenic area is constructed, including:

[0078] Obtain the GIS data of the scenic area, which includes the geographic boundary information of the functional area, the access path information and the location data of important landmarks in the scenic area:

[0079] Through the scenic area management system interface, call and download geographic information system data, including the following three categories:

[0080] Geographic boundary information of functional areas: coordinate data composed of polygon vertices. The vertex coordinates are expressed in the WGS84 geographic coordinate system. The boundary of each functional area is defined by a set of closed polygon vertex coordinates. For example: the boundary of a functional area in a scenic area is composed of (longitude 1, latitude 1), (longitude 2, latitude 2), (longitude 3, latitude 3)...

[0081] Passage path information: The passage path within the scenic area is represented by the coordinates of the path starting point and the end point, including the path length and path direction. For example: the starting point of a path is (longitude A, latitude A), the end point is (longitude B, latitude B), the path length is 500 meters, and the direction is one-way.

[0082] Important landmark location data: including the latitude and longitude coordinates of the scenic area entrance, exit, main attractions and service facilities. Each landmark records its functional type (such as entrance, restroom, core attraction, etc.).

[0083] The acquired GIS data are processed as follows: the integrity of each set of coordinate data is verified one by one, for example, ensuring that the boundary polygon data is closed, the access path data has the starting and ending coordinates, and the landmark data has a valid functional identifier; the coordinate points beyond the scenic area boundary are detected and eliminated, for example, the boundary points do not conform to the polygon data of the scenic area's geographic scope; all geographic data are unified into the WGS84 coordinate system, the path length is unified into a value in meters, and the time information is converted into the UTC time format.

[0084] Based on the GIS data, the scenic area is divided into multiple functional areas, which are connected by access paths, and the adjacency relationship and path attributes between the functional areas are recorded:

[0085] According to the geographical boundary information, the scenic area is divided into multiple functional areas. The scope of each functional area is defined by the coordinates of the closed polygon vertices. According to the access path information, it is determined whether there is an adjacency relationship between the functional areas, and the adjacent areas are connected by the access path. The entrance and exit positions of each functional area are marked, and its category information (such as main entrance, secondary entrance, scenic spot entrance, etc.) is recorded.

[0086] Path length: The path length is obtained by calculating the straight-line distance between the start and end points, and taking terrain factors (such as slope) into account.

[0087] Traffic capacity: Based on the path width and terrain conditions, the maximum number of tourists passing through each path per unit time is set, in persons per minute.

[0088] Direction of travel: According to the scenic area management rules, the marked paths are one-way or two-way.

[0089] Special areas such as core scenic spots and transportation hubs are marked and their category information is recorded; the area of ​​each functional area is calculated and the area data is stored as one of the node attributes.

[0090] The node relationship model of functional areas is constructed based on the principle of group behavior. The node relationship model includes the following contents: each functional area is defined as a node in the node relationship model; the passage paths between functional areas are defined as edges connecting nodes, and the attributes of the edges include path length, traffic capacity and congestion coefficient; the geographical boundaries, entrance and exit locations and passable directions of each functional area are recorded:

[0091] Each functional area is defined as a node, and the attributes of the node include the following: geographic boundary: a closed boundary expressed in polygonal coordinates; area: the area of ​​the area obtained by the formula for calculating the area of ​​the geographic boundary polygon; entrance and exit locations: recording the entrance and exit coordinates on the boundary of the functional area, classified by functional category (such as main entrance and exit); direction of travel: the direction attribute of the entrance and exit of the area, including one-way or two-way.

[0092] The traffic paths between functional areas are defined as the edges connecting the nodes. The attributes of the edges include: path length: in meters; traffic capacity: the maximum number of tourists passing through per unit time (persons / minute); congestion coefficient: the current congestion status value of the path calculated based on historical data or real-time monitoring data (such as a floating point number between 0 and 1, 0 means unobstructed and 1 means complete congestion).

[0093] Establish a graph structure containing nodes and edges, and record the adjacent edge information of each node; establish an association table for each edge, associating the starting node and the end node, as well as the attribute data of the edge (such as path length, traffic capacity, congestion coefficient, etc.); verify the connectivity of the graph structure through a topological algorithm to ensure that the traffic paths between all functional areas are fully represented in the model.

[0094] Based on the node relationship model and multi-source tourist behavior data, the tourist crowding parameters of each functional area are calculated, and the dynamic distribution parameters of the functional area are generated by combining the time dimension and the space dimension, including:

[0095] Calculate the real-time tourist density of the functional area. The real-time tourist density is the ratio of the current number of tourists in the functional area to the area of ​​the functional area:

[0096] The geographical boundaries and area data of the functional area are extracted from the node relationship model; the current number of tourists in the functional area is collected from the real-time monitoring equipment, and the tourist distribution information is updated in real time.

[0097] The real-time tourist density is calculated based on the current number of tourists in the functional area and the area of ​​the functional area. The real-time tourist density is the ratio of the current number of tourists in the functional area to the area of ​​the functional area.

[0098] The real-time tourist density is measured in people per square meter (person / ㎡), which serves as the basic indicator of the degree of crowding in functional areas.

[0099] By analyzing the trajectory data of tourists in the functional areas, the average length of stay of tourists in the functional areas is calculated:

[0100] Tourist trajectory data is extracted from multi-source tourist behavior data, including the timestamps of tourists entering and leaving the functional areas.

[0101] For each tourist, the time difference between entry and exit time is calculated as the individual stay duration; the individual stay durations of all tourists in the functional area are averaged to obtain the average stay duration of the functional area.

[0102] Divide the tourist data by time period and generate the distribution of the number of tourists in each functional area at different time periods:

[0103] The number of tourists per minute in the functional area is extracted from the real-time tourist data and analyzed in combination with historical data.

[0104] Group the tourist data by time periods (such as minutes, hours), count the changes in the number of tourists in each time period, and generate time dimension distribution parameters. For example, time dimension distribution parameters = {time period 1: number of tourists, time period 2: number of tourists, ...}.

[0105] Based on the adjacency relationship and path attributes between functional areas, the tourist flow ratio between functional areas and adjacent areas is calculated:

[0106] Extract the adjacency relationship and path attributes of functional areas from the node relationship model, including path length, traffic capacity and traffic direction;

[0107] The number of tourists flowing from one functional area to an adjacent functional area is counted from the tourist trajectory data.

[0108] For each functional area, the tourist flow ratio is calculated, which is the ratio of the number of tourists flowing out to the adjacent area to the total number of tourists in the current functional area.

[0109] According to the adjacency and directionality of the passage paths, the flow ratios of all adjacent areas are normalized to ensure that the total ratio is 1.

[0110] Based on the real-time tourist density, tourist flow ratio and average stay time of the functional area, the dynamic distribution parameters are calculated comprehensively:

[0111] The real-time tourist density, tourist flow ratio and average stay time are dimensionless, and weight coefficients are assigned to them respectively. The real-time tourist density, the inverse of the tourist flow ratio and the average stay time are multiplied by their corresponding weight coefficients and then added to obtain the dynamic distribution parameters.

[0112] Among them, the weight coefficients of real-time tourist density, tourist flow ratio and average length of stay are all greater than 0, and their settings are based on the actual management needs of the functional area and the importance of crowding risk assessment. In specific settings, the degree of influence of each parameter on crowding and safety risks is usually determined based on historical data and expert experience. For example, the real-time tourist density directly reflects the crowding state in the area, so the weight coefficient is usually high; the tourist flow ratio represents the mobility between areas, and its inverse is positively correlated with the crowding risk, so the weight is second; the average length of stay affects the length of time tourists stay in the area, and plays an important role in assessing long-term crowding risks, so the weight is moderate. The specific value of the weight coefficient can be tuned by historical data fitting or multi-objective optimization algorithm to meet the needs of actual scenarios.

[0113] The larger the dynamic distribution parameter, the higher the real-time tourist density in the functional area, the lower the proportion of tourist flow (higher after taking the inverse), or the longer the average length of stay. This comprehensively reflects that the degree of crowding and detention in the functional area is serious, which may lead to an increase in crowding risks and safety hazards in the area, such as poor traffic and difficulty in evacuation. In this case, intervention measures need to be taken, such as guiding tourists to adjacent areas or limiting the number of people entering. The smaller the dynamic distribution parameter, the lower the real-time tourist density in the functional area, the higher the proportion of tourist flow (lower after taking the inverse), or the shorter the average length of stay, indicating that the distribution of tourists in the area is relatively uniform, the mobility is good, and the crowding and safety risks are low. At this time, the area is in a relatively safe and comfortable state, and the existing management strategy can be maintained without additional intervention.

[0114] Compare and analyze historical passenger flow data with real-time monitoring data, build a tourist flow propagation matrix between functional areas, and dynamically optimize and update the migration weights between functional areas through regression algorithms, including:

[0115] Obtain historical traffic data and real-time monitoring data for functional areas, and compare the number of tourists and flow changes by time period and functional area:

[0116] The historical passenger flow data is extracted through the scenic area management database: the historical number of tourists in the functional areas is recorded by time period; the historical number of tourist flows between functional areas, and the flow paths and number of tourists are recorded by time period.

[0117] Real-time monitoring data is obtained through the scenic area’s video surveillance system and visitor positioning equipment: the number of tourists in the current functional area; the paths and number of tourists entering or leaving the functional area within a unit of time.

[0118] Compare and analyze historical passenger flow data with real-time monitoring data, calculate the number of tourists and flow change trends by time period and functional area. Analyze whether the flow of tourists conforms to historical trends and record the deviation.

[0119] The number of tourists flowing between functional areas is counted based on real-time monitoring data, and the proportion of tourists flowing is calculated based on historical passenger flow data:

[0120] Based on real-time monitoring data, the number of tourists entering, leaving and flowing to other functional areas in each functional area per unit time is counted.

[0121] The tourist flow ratio between functional areas is calculated by historical passenger flow data and real-time monitoring data. The calculation formula is: ;in, Indicates from the functional area Flow to functional area proportion of tourist flows; Indicates from the functional area Flow to functional area the number of tourists; Indicates functional area Total number of tourists, including stay and outgoing tourists; Indicates the starting functional area number of the tourist flow, specifically which functional area the tourist departs from; Indicates the target functional area number of tourist flow, specifically refers to the destination functional area of ​​tourist flow.

[0122] The calculated tourist flow ratio is stored in the database according to functional areas and time periods for the subsequent construction of the diffusion matrix.

[0123] The tourist flow propagation matrix is ​​constructed based on the adjacency relationship between functional areas and the tourist flow ratio. The elements in the tourist flow propagation matrix represent the tourist flow relationship between adjacent functional areas:

[0124] The adjacency relationship between functional areas, including the travel direction and capacity information of the paths, is extracted as the topological basis of the tourist flow propagation matrix.

[0125] According to the adjacency relationship and the proportion of tourist flow, the tourist flow propagation matrix is ​​constructed. The calculation formula of the tourist flow propagation matrix is: ;in, represents the tourist flow propagation matrix, is the total number of functional areas, Indicates from the functional area Flow to functional area proportion of tourist flows.

[0126] The constructed tourist flow propagation matrix is ​​stored in the database, and each element of the tourist flow propagation matrix represents the tourist flow relationship between functional areas.

[0127] Based on the tourist flow propagation matrix, the regression algorithm is used to calculate the migration weights between functional areas, and the weights are dynamically optimized according to real-time monitoring data. The optimization goal is to minimize the prediction error of the migration weights:

[0128] Based on the tourist flow propagation matrix and tourist flow ratio, the migration weights between functional areas are preliminarily calculated. The weight calculation formula is: ;in, Indicates from the functional area Flow to functional area The migration weight of Indicates functional area To functional area The path capacity of Represents the corresponding tourist flow ratio in the diffusion matrix.

[0129] The regression algorithm is used to optimize the migration weight. The optimization goal is to minimize the error between the predicted weight and the actual weight. The objective function is defined as: ;in, represents the migration weight predicted by the tourist flow propagation matrix and regression algorithm, Indicates the actual migration weight calculated based on real-time monitoring data.

[0130] According to the optimization results, the migration weights are updated in real time, and the updated migration weights are stored in the database for use in subsequent prediction steps; the comparison analysis, propagation matrix and optimized migration weight results are recorded in the database by time period and functional area to ensure that subsequent steps are traceable.

[0131] It can generate a visualization chart of tourist flows between functional areas and provide a real-time update interface for migration weights to support scenic area management decisions.

[0132] Based on the updated tourist flow propagation matrix and the dynamic distribution parameters of the functional areas, the flow of tourists in the scenic area at different time periods in the future is predicted, and the time-based distribution data of the flow of tourists in the functional areas is generated, including:

[0133] A prediction model is constructed based on the updated tourist flow propagation matrix. The prediction model takes the tourist flow relationship between functional areas as the input variable and the number of tourists in the functional areas in the future as the output variable:

[0134] The updated tourist flow propagation matrix is ​​extracted from the database, and each element of the tourist flow propagation matrix represents the tourist flow ratio between functional areas; the dynamic distribution parameters of the functional areas are extracted, which reflect the current real-time tourist density, tourist flow ratio and average stay time of the functional areas through comprehensive weighted processing.

[0135] A prediction model is defined, which takes the tourist flow relationship between functional areas as the input variable and the number of tourists in the functional areas at different time periods in the future as the output variable.

[0136] The input variables include: the flow ratio between each area in the tourist flow propagation matrix of the functional area; the dynamic distribution parameters of the functional area, which serve as the internal state characteristics of the area.

[0137] Output variables include: the number of tourists in the functional area at different time periods in the future.

[0138] The core relationship of the prediction model is defined as: ;in, Indicates functional area In time period The predicted number of tourists Indicates from the functional area Flow to functional area The proportion of tourist flows, Indicates functional area In time period The number of tourists, Indicates functional area The dynamic distribution parameters of Indicates the current time period (specifically the time point or time interval currently being analyzed).

[0139] The dynamic distribution parameters here can be regarded as weighting factors, reflecting the functional area The comprehensive status of the functional area includes the real-time tourist density, tourist flow ratio and average length of stay. As a weighting factor, the dynamic distribution parameter adjusts the tourist flow ratio, increases or decreases the Flow to functional area The number of tourists can be more realistically reflected in order to better reflect the actual impact of functional areas on flow relations.

[0140] Through the formula of the core relationship of the above prediction model, the prediction model can comprehensively consider the tourist flow relationship between functional areas and the impact of dynamic distribution parameters on the number of tourists.

[0141] The historical passenger flow data and real-time monitoring data of the functional area are input into the prediction model as time series data, and the dynamic distribution parameters of the functional area and the tourist flow propagation matrix are also input:

[0142] The historical pedestrian flow data and real-time monitoring data of the functional areas are extracted from the database and input into the prediction model as time series data.

[0143] Historical passenger flow data: including the number of tourists, flow paths and flow ratios in functional areas at different time periods in the past;

[0144] Real-time monitoring data: including the current number of tourists in the functional area and the flow trends of entering and leaving.

[0145] Time series data is processed uniformly to ensure data format consistency, including: aligning historical traffic data and real-time monitoring data by time period and unifying the time step; filling missing values ​​and estimating missing data by linear interpolation or time series model.

[0146] The time series data are grouped by functional areas and time periods, and the number of tourists in each functional area is taken as part of the input variables and input into the prediction model together with the tourist flow propagation matrix and dynamic distribution parameters.

[0147] The prediction model is used to calculate the passenger flow data of functional areas at different time periods in the future, including the future number of tourists and flow trends in each functional area:

[0148] The prediction model is used to calculate the number of tourists in the functional areas at different time periods in the future. The calculation process is recursive: ;in, Indicates functional area In the future The predicted number of tourists for each time period, Indicates functional area The number of tourists in the previous period, Represents the time step of the prediction.

[0149] By recursively calculating each time period, the number of tourists in the functional area in multiple time periods in the future is generated.

[0150] According to the changing trend of the number of tourists in the functional areas, the flow trend of each functional area is calculated in combination with the flow ratio in the diffusion matrix.

[0151] The predicted functional area flow data is recorded as time period flow distribution data:

[0152] The predicted functional area traffic data is sorted by time period, and the number of tourists in each functional area in different time periods is recorded. The data format is:

[0153] Functional area number: uniquely identifies the functional area;

[0154] Time period: the predicted time point or time interval;

[0155] Predicted number of tourists: The predicted number of tourists in the functional area in the corresponding time period.

[0156] The above-mentioned time-based crowd distribution data is stored in the database to ensure that it can be queried by functional area or time period.

[0157] Analyze the density of tourists in the functional area based on the time-based distribution data of passenger flow in the functional area, and mark potential high-risk areas for passenger flow, including:

[0158] According to the area of ​​the functional area and the distribution data of the flow of people in different time periods, the density of tourists in the functional area is calculated:

[0159] The formula for tourist density within a functional area is: ;in, Indicates functional area In time period density of tourists; Indicates functional area In time period the number of tourists; Indicates functional area area.

[0160] The tourist density is calculated for all time periods of each functional area in turn, and the calculation results are stored in the following format: functional area number: uniquely identifies the functional area; time period: the corresponding time point or time interval; tourist density: the tourist density of the functional area within the time period.

[0161] According to the management requirements of the scenic area, the high-risk tourist density threshold of the functional area is set, and the calculated tourist density of the functional area is compared with the high-risk density threshold. The functional area with a tourist density greater than the high-risk density threshold is marked as a potential high-risk area for human flow:

[0162] Combined with the scenic area's safety management standards and historical data, the high-risk tourist density threshold of the functional area is set and recorded as The high-risk tourist density threshold can be dynamically adjusted according to the characteristics of the scenic area (such as the area of ​​the functional area and the characteristics of tourist activities).

[0163] like , then mark the functional area as a potential high-risk area for human flow.

[0164] like , then the functional area is not marked as a potential high-risk area for human flow.

[0165] The functional areas marked as potential high-risk areas are recorded in the database, and a distribution map of high-risk areas is generated based on the marked data, including the tourist density and high-risk marking status of each functional area. The distribution map is displayed in the form of a heat map.

[0166] Embodiment 2: Embodiment 2 of the present invention differs from Embodiment 1 in that this embodiment introduces a scenic spot passenger flow prediction system.

[0167] Figure 2 A structural schematic diagram of a scenic spot passenger flow prediction system of the present invention is given, which includes a multi-source data processing module, a functional area modeling module, a dynamic parameter generation module, a propagation matrix optimization module, a future passenger flow prediction module and a risk area marking module.

[0168] Multi-source data processing module: collects multi-source tourist behavior data and pre-processes the multi-source tourist behavior data.

[0169] Functional area modeling module: Based on the principles of group behavior and geographic information system data, a node relationship model of each functional area in the scenic area is constructed.

[0170] Dynamic parameter generation module: Calculates the tourist congestion parameters of each functional area based on the node relationship model and multi-source tourist behavior data, and generates dynamic distribution parameters of the functional area by combining the time dimension and space dimension.

[0171] Propagation matrix optimization module: compare and analyze historical passenger flow data with real-time monitoring data, construct a tourist flow propagation matrix between functional areas, and dynamically optimize and update the migration weights between functional areas through regression algorithms.

[0172] Future passenger flow prediction module: Based on the updated tourist flow propagation matrix and the dynamic distribution parameters of the functional areas, the passenger flow of the scenic area at different time periods in the future is predicted, and the time period passenger flow distribution data within the functional area is generated.

[0173] Risk area marking module: Analyze the density of tourists in functional areas based on the time-based passenger flow distribution data in the functional areas, and mark potential high-risk passenger flow areas.

[0174] Embodiment 3: A device terminal comprises: a processor, a memory, and a program or instruction stored in the memory and executable on the processor, wherein the program or instruction implements a method for predicting the flow of people in a scenic spot when the program or instruction is executed by the processor.

[0175] Embodiment 4: A computer-readable storage medium stores a program or instruction, and when the program or instruction is executed by a processor, a method for predicting the flow of people in a scenic spot is implemented.

[0176] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters and thresholds in the formula are set by technicians in this field according to actual conditions.

[0177] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination thereof. When implemented by software, the above embodiments may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium, or may be transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions may be transmitted from one website, computer, server or data center to another website, computer, server or data center by wired (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium may be any available medium that a computer can access or a data storage device such as a server or data center that contains one or more available media sets. The available medium may be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium may be a solid-state hard disk.

[0178] Those of ordinary skill in the art will appreciate that the modules and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0179] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and modules described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0180] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the modules is only a logical function division. There may be other division methods in actual implementation, such as multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or modules, which can be electrical, mechanical or other forms.

[0181] The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical modules, and may be located in one place or distributed on multiple network modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0182] In addition, each functional module in each embodiment of the present application may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module.

[0183] If the functions are implemented in the form of software function modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in the various embodiments of the present application. The aforementioned storage media include: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, and other media that can store program codes.

[0184] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0185] Finally: 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 in the protection scope of the present invention.

Claims

1. A method for predicting the flow of people in a scenic spot, characterized in that: The steps include: S1: Collect multi-source tourist behavior data and pre-process the multi-source tourist behavior data; S2: Based on the principles of group behavior and geographic information system data, a node relationship model of each functional area of ​​the scenic area is constructed, including: Obtaining geographic information system data of the scenic area, which includes geographic boundary information of functional areas, traffic path information and location data of important landmarks in the scenic area; Based on the GIS data, the scenic area is divided into multiple functional areas, which are connected by access paths, and the adjacency relationship and path attributes between the functional areas are recorded; The node relationship model of functional areas is constructed based on the principle of group behavior. The node relationship model includes the following contents: each functional area is defined as a node in the node relationship model; the traffic paths between functional areas are defined as edges connecting nodes, and the attributes of edges include path length, traffic capacity and congestion coefficient; the geographical boundaries, entrance and exit locations and accessible directions of each functional area are recorded; S3: Based on the node relationship model and multi-source tourist behavior data, the tourist crowding parameters of each functional area are calculated, and the dynamic distribution parameters of the functional area are generated by combining the time dimension and the space dimension, including: Calculate the real-time tourist density of the functional area, which is the ratio of the current number of tourists in the functional area to the area of ​​the functional area; By analyzing the trajectory data of tourists in the functional areas, the average length of stay of tourists in the functional areas is calculated; Divide the tourist data by time period to generate the distribution of the number of tourists in each functional area at different time periods; Based on the adjacency relationship and path attributes between functional areas, the tourist flow ratio between functional areas and adjacent areas is calculated; Based on the real-time tourist density, tourist flow ratio and average length of stay in the functional area, the dynamic distribution parameters are obtained by comprehensive calculation: the real-time tourist density, tourist flow ratio and average length of stay are dimensionless, the real-time tourist density, tourist flow ratio and average length of stay are assigned weight coefficients respectively, the real-time tourist density, the inverse of the tourist flow ratio and the average length of stay are multiplied by their corresponding weight coefficients and then added to obtain the dynamic distribution parameters; S4: Compare and analyze historical passenger flow data with real-time monitoring data, construct a tourist flow propagation matrix between functional areas, and dynamically optimize and update the migration weights between functional areas through regression algorithms, including: Obtain historical passenger flow data and real-time monitoring data for functional areas, and compare the number of tourists and flow changes by time period and functional area; The number of tourists flowing between functional areas is counted based on real-time monitoring data, and the proportion of tourists flowing is calculated in combination with historical passenger flow data; A tourist flow propagation matrix is ​​constructed based on the adjacency relationship between functional areas and the tourist flow ratio. The elements in the tourist flow propagation matrix represent the tourist flow relationship between adjacent functional areas. Based on the tourist flow propagation matrix, the regression algorithm is used to calculate the migration weights between functional areas, and the weights are dynamically optimized according to real-time monitoring data. The optimization goal is to minimize the prediction error of the migration weights. S5: Based on the updated tourist flow propagation matrix and the dynamic distribution parameters of the functional areas, the tourist flow of the scenic spot at different time periods in the future is predicted to generate the time-based tourist flow distribution data within the functional areas; S6: Analyze the density of tourists in the functional areas based on the time-based crowd distribution data in the functional areas, and mark potential high-risk areas for crowd flow.

2. The method for predicting the flow of people in a scenic spot according to claim 1, characterized in that: Collect multi-source tourist behavior data and pre-process the multi-source tourist behavior data, including: Multi-source tourist behavior data includes image data collected by video surveillance equipment, location data from tourist smart terminals, and historical transaction data from the ticket sales system; Preprocess multi-source tourist behavior data: identify and remove outliers in multi-source tourist behavior data; filter noise in multi-source tourist behavior data to eliminate irrelevant information; synchronize timestamps of multi-source tourist behavior data to ensure temporal consistency of data from different sources; standardize multi-source tourist behavior data to unify the format and scope of the data.

3. The method for predicting the flow of people in a scenic spot according to claim 1, characterized in that: Based on the updated tourist flow propagation matrix and the dynamic distribution parameters of the functional areas, the flow of tourists in the scenic area at different time periods in the future is predicted, and the time-based distribution data of the flow of tourists in the functional areas is generated, including: A prediction model is constructed based on the updated tourist flow propagation matrix. The prediction model takes the tourist flow relationship between functional areas as the input variable and the number of tourists in the functional areas in the future as the output variable. The historical passenger flow data and real-time monitoring data of the functional area are input into the prediction model as time series data, and the dynamic distribution parameters of the functional area and the tourist flow propagation matrix are also input; Calculate the passenger flow data of functional areas at different time periods in the future through the prediction model, including the future number of tourists and flow trends in each functional area; The predicted functional area flow data is recorded as periodic flow distribution data according to time periods.

4. The method for predicting the flow of people in a scenic spot according to claim 1, characterized in that: Analyze the density of tourists in the functional area based on the time-based distribution data of passenger flow in the functional area, and mark potential high-risk areas for passenger flow, including: Calculate the tourist density in the functional area based on the area of ​​the functional area and the time-based crowd distribution data; The high-risk tourist density threshold of the functional area is set according to the management needs of the scenic spot, the calculated tourist density of the functional area is compared with the high-risk density threshold, and the functional area where the tourist density is greater than the high-risk density threshold is marked as a potential high-risk area for human flow.

5. A scenic spot passenger flow prediction system, used to implement the scenic spot passenger flow prediction method according to any one of claims 1 to 4, characterized in that: include: Multi-source data processing module: collects multi-source tourist behavior data and pre-processes the multi-source tourist behavior data; Functional area modeling module: Based on the principles of group behavior and geographic information system data, the node relationship model of each functional area of ​​the scenic area is constructed; Dynamic parameter generation module: calculates the tourist crowding parameters of each functional area based on the node relationship model and multi-source tourist behavior data, and generates dynamic distribution parameters of the functional area by combining the time dimension and space dimension; Propagation matrix optimization module: compares and analyzes historical passenger flow data with real-time monitoring data, constructs a tourist flow propagation matrix between functional areas, and dynamically optimizes and updates the migration weights between functional areas through regression algorithms; Future passenger flow prediction module: Based on the updated tourist flow propagation matrix and the dynamic distribution parameters of the functional areas, the passenger flow of the scenic area at different time periods in the future is predicted, and the time period passenger flow distribution data within the functional area is generated; Risk area marking module: Analyze the density of tourists in functional areas based on the time-based passenger flow distribution data in functional areas, and mark potential high-risk passenger flow areas.

6. A device terminal, characterized in that: The device terminal includes: a processor, a memory, and a program or instruction stored in the memory and executable on the processor. When the program or instruction is executed by the processor, the scenic spot flow prediction method as described in any one of claims 1 to 4 is implemented.

7. A computer-readable storage medium, characterized in that: A program or instruction is stored on a computer-readable storage medium, and when the program or instruction is executed by a processor, the scenic spot passenger flow prediction method as described in any one of claims 1-4 is implemented.

Images