Method and device for predicting crowd density in entrepreneurial space

Abstract

The invention relates to the technical field of data management, in particular to an entrepreneurship space people flow density prediction method and device, which can capture people flow characteristics of different areas more meticulously and improve the pertinence and accuracy of prediction. The method comprises the following steps: carrying out functional region division on an entrepreneurial space to obtain an entrepreneurial space functional region cluster; the entrepreneurship space functional region cluster comprises all functional regions in the entrepreneurship space; for each functional area, performing traversal screening in a preset people flow density weight space according to the functional attribute corresponding to the functional area, and extracting a weather influence weight matrix corresponding to the functional attribute; performing date attribute analysis on the to-be-predicted time window to obtain a date attribute vector corresponding to the to-be-predicted time window segment, the date attribute vector comprising a week sequence and a time sequence to which the to-be-predicted time window belongs; wherein the week sequence represents the week corresponding to the to-be-predicted time window, and the time sequence represents a starting timestamp and a deadline timestamp corresponding to the to-be-predicted time window.

Description

Technical Field

[0001] The present invention relates to the technical field of data management, and in particular to a method and device for predicting crowd density in an entrepreneurial space. Background Art

[0002] As a new type of workplace, entrepreneurial space has developed rapidly around the world in recent years, such as co-working space and maker space, providing entrepreneurs, freelancers and small teams with a flexible and dynamic working environment. In the management and operation of entrepreneurial space, accurately predicting the density of people flow in the entrepreneurial space can not only help managers to allocate resources reasonably, but also optimize the user's access experience and improve the overall operational efficiency.

[0003] Entrepreneurial spaces usually contain multiple functional areas, such as office areas, meeting rooms, rest areas, exhibition areas, etc. Each functional area presents different crowd density characteristics due to its specific functional attributes and user needs. However, traditional crowd density prediction methods often treat the entrepreneurial space as a whole for prediction, without considering the differences between functional areas, and also without considering the impact of external weather factors on the crowd density of different functional areas, resulting in low prediction accuracy and difficulty in meeting the needs of practical applications.

[0004] Therefore, there is an urgent need to provide a method and device for predicting the density of people flow in an entrepreneurial space to solve the above technical problems. Summary of the invention

[0005] In order to solve the above technical problems, the present invention provides a method and device for predicting the density of entrepreneurial space pedestrian flow, which can more carefully capture the characteristics of pedestrian flow in different areas and improve the pertinence and accuracy of prediction.

[0006] In a first aspect, the present invention provides a method for predicting the density of people flow in an entrepreneurial space, the method comprising:

[0007] Dividing the entrepreneurial space into functional areas to obtain an entrepreneurial space functional area cluster; the entrepreneurial space functional area cluster includes all functional areas in the entrepreneurial space;

[0008] For each functional area, the functional attributes corresponding to the functional area are traversed and screened in the preset crowd density weight space to extract the corresponding weather impact weight matrix;

[0009] Perform date attribute analysis on the time window to be predicted to obtain a date attribute vector corresponding to the time window segment to be predicted, wherein the date attribute vector includes the week sequence and time sequence to which the time window to be predicted belongs; wherein the week sequence indicates the day of the week corresponding to the time window to be predicted, and the time sequence indicates the start timestamp and end timestamp corresponding to the time window to be predicted;

[0010] For each functional area, from the historical statistical records of entrepreneurial space pedestrian density, extract a preset number of historical time window pedestrian density that is the same as the date attribute vector and closest to the time window to be predicted in terms of time series;

[0011] Arrange the crowd density of the preset number of historical time windows according to time series to obtain a crowd density time series vector;

[0012] Inputting the crowd density time series vector into a preset benchmark crowd density prediction model to obtain a benchmark crowd density of the corresponding functional area within the time window to be predicted;

[0013] Obtaining the weather forecast status corresponding to the time window to be predicted, and extracting the corresponding weather impact weight coefficient in the weather impact weight matrix based on the weather forecast status;

[0014] Based on the weather impact weight coefficient, the reference crowd density is corrected to obtain the crowd density corresponding to the functional area;

[0015] The crowd density corresponding to each functional area in the entrepreneurial space functional area cluster is summarized to obtain the entrepreneurial space crowd density prediction result.

[0016] On the other hand, the present application also provides a device for predicting the density of people flow in an entrepreneurial space, the device comprising:

[0017] A functional area division module is used to divide the entrepreneurial space into functional areas and obtain an entrepreneurial space functional area cluster including all functional areas in the entrepreneurial space;

[0018] The weather impact weight matrix extraction module is used to traverse and screen each functional area in the preset pedestrian density weight space based on the functional attributes corresponding to the functional area, and extract the corresponding weather impact weight matrix;

[0019] A date attribute analysis module, used to perform date attribute analysis on the time window to be predicted, and obtain a date attribute vector including the week sequence and time sequence to which the time window to be predicted belongs;

[0020] A historical pedestrian flow density extraction module is used to extract, for each functional area, from the historical statistical records of entrepreneurial space pedestrian flow density, a preset number of historical time window pedestrian flow densities that are the same as the date attribute vector and closest to the time window to be predicted in terms of time series;

[0021] A crowd density time series vector generation module, used for arranging the crowd density of the preset number of historical time windows according to time series to generate a crowd density time series vector;

[0022] A reference crowd density prediction module, used to input the crowd density time series vector into a preset reference crowd density prediction model to obtain a reference crowd density of a corresponding functional area within a time window to be predicted;

[0023] A weight coefficient extraction module is used to obtain the weather forecast status corresponding to the time window to be predicted, and extract the corresponding weather impact weight coefficient in the weather impact weight matrix based on the weather forecast status;

[0024] A crowd density correction module, used to perform impact correction on the reference crowd density according to the weather impact weight coefficient to obtain the crowd density corresponding to the functional area;

[0025] The crowd density summary module is used to summarize the crowd density corresponding to each functional area in the entrepreneurial space functional area cluster to obtain the entrepreneurial space crowd density prediction result.

[0026] In a third aspect, the present application provides an electronic device, comprising a bus, a transceiver, a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the transceiver, the memory, and the processor are connected via the bus, and the computer program, when executed by the processor, implements the steps of any one of the above methods.

[0027] In a fourth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program implements the steps in any one of the above-mentioned methods when executed by a processor.

[0028] Compared with the prior art, the present invention has the following beneficial effects:

[0029] First, by dividing the entrepreneurial space into functional areas in detail and establishing functional area clusters, this method can accurately capture the unique usage patterns and traffic characteristics of each area, avoiding the problem of insufficient prediction accuracy caused by the traditional method of treating the entire space as a whole. It not only improves the accuracy of the prediction results, but also makes resource allocation more refined, and can implement customized management strategies according to the needs of different areas;

[0030] Secondly, the weather impact weight matrix is ​​introduced to consider the impact of external environmental factors on different functional areas, which further enhances the model's real-world adaptability and prediction effect. Since weather conditions are directly related to people's behavioral choices, such as bad weather may increase the frequency of indoor activities, while good weather encourages outdoor activities, by quantifying the impact of weather changes, the trend of changes in the density of pedestrian traffic in each functional area can be more accurately estimated, which helps to prepare countermeasures in advance and improve user experience and service quality.

[0031] Furthermore, the combination of historical data extraction and date attribute analysis for each functional area ensures that the prediction model can make full use of rich historical information. At the same time, the crowd density vector generated by time series arrangement provides a solid data foundation for the benchmark crowd density prediction. This time window matching technology based on historical similarity can effectively identify past time periods with similar characteristics, thereby improving the reliability of the prediction.

[0032] Finally, the preset benchmark crowd density prediction model is combined with the weather impact weight coefficient for correction, realizing the ability to dynamically adjust the prediction results to reflect immediate changes, which not only ensures the scientific nature of the prediction, but also takes into account flexibility. Ultimately, by summarizing the crowd density of each functional area, a comprehensive and accurate entrepreneurial space crowd density prediction result is obtained, which greatly improves space management and operational efficiency and optimizes the user's access experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0034] Figure 1 This is a flow chart of a method for predicting crowd density in an entrepreneurial space provided by an embodiment of the present invention;

[0035] Figure 2 is a hardware architecture diagram of an electronic device provided by an embodiment of the present invention;

[0036] Figure 3 It is a structural diagram of a device for predicting crowd density in an entrepreneurial space provided by one embodiment of the present invention. DETAILED DESCRIPTION

[0037] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0038] Please refer to Figure 1 , an embodiment of the present invention provides a method for predicting the density of people flow in an entrepreneurial space, the method comprising:

[0039] Step S1, dividing the entrepreneurial space into functional areas to obtain an entrepreneurial space functional area cluster; the entrepreneurial space functional area cluster includes all functional areas in the entrepreneurial space;

[0040] Step S2: for each functional area, the functional attributes corresponding to the functional area are traversed and screened in the preset crowd density weight space to extract the corresponding weather impact weight matrix;

[0041] Step S3, performing date attribute analysis on the time window to be predicted, obtaining a date attribute vector corresponding to the time window segment to be predicted, wherein the date attribute vector includes the week sequence and time sequence to which the time window to be predicted belongs; wherein the week sequence indicates the day of the week corresponding to the time window to be predicted, and the time sequence indicates the start timestamp and end timestamp corresponding to the time window to be predicted;

[0042] Step S4: for each functional area, extract the crowd density of a preset number of historical time windows that are the same as the date attribute vector and closest to the time window to be predicted in terms of time sequence from the historical statistical records of crowd density in the entrepreneurial space;

[0043] Step S5, arranging the preset number of historical time window crowd density according to time series to obtain a crowd density time series vector;

[0044] Step S6: input the crowd density time series vector into a preset reference crowd density prediction model to obtain a reference crowd density of the corresponding functional area within the time window to be predicted;

[0045] Step S7, obtaining the weather forecast status corresponding to the time window to be predicted, and extracting the corresponding weather impact weight coefficient in the weather impact weight matrix based on the weather forecast status;

[0046] Step S8: Based on the weather impact weight coefficient, the reference crowd density is corrected to obtain the crowd density corresponding to the functional area;

[0047] Step S9: Summarize the crowd density corresponding to each functional area in the entrepreneurial space functional area cluster to obtain the entrepreneurial space crowd density prediction result.

[0048] In this embodiment, firstly, by dividing the entrepreneurial space into functional areas in detail and establishing functional area clusters, this method can accurately capture the unique usage patterns and traffic characteristics of each area, avoiding the problem of insufficient prediction accuracy caused by the traditional method of treating the entire space as a whole, which not only improves the accuracy of the prediction results, but also makes resource allocation more refined, and can implement customized management strategies according to the needs of different areas; secondly, the weather impact weight matrix is ​​introduced to consider the impact of external environmental factors on different functional areas, further enhancing the realistic adaptability and prediction effect of the model; since weather conditions are directly related to people's behavioral choices, such as bad weather may increase the frequency of indoor activities, while good weather encourages outdoor activities, so by quantifying the impact of weather changes, the trend of changes in traffic density in each functional area can be more accurately estimated, which helps to prepare for response in advance. measures to improve user experience and service quality; secondly, the combination of historical data extraction and date attribute analysis for each functional area ensures that the prediction model can make full use of rich historical information. At the same time, the crowd density vector generated by time series arrangement provides a solid data foundation for benchmark crowd density prediction; this time window matching technology based on historical similarity can effectively identify past time periods with similar characteristics, thereby improving the reliability of the prediction; finally, the preset benchmark crowd density prediction model is combined with the weather impact weight coefficient for correction to achieve the ability to dynamically adjust the prediction results to reflect immediate changes, which not only ensures the scientific nature of the prediction, but also takes into account flexibility. Finally, by summarizing the crowd density of each functional area, a comprehensive and accurate entrepreneurial space crowd density prediction result is obtained, which greatly improves the space management and operation efficiency and optimizes the user's access experience.

[0049] Described below Figure 1 How the various steps are performed.

[0050] For step S1:

[0051] Step S1 is the process of dividing the entrepreneurial space into functional areas to obtain the functional area clusters of the entrepreneurial space, ensuring that the prediction can be carried out for different functional areas in the entrepreneurial space, thereby reflecting the characteristics of the flow of people density presented by each area due to different functional attributes and user needs.

[0052] Specifically, it is necessary to have a comprehensive understanding and analysis of the entrepreneurial space; first identify and divide all functional areas within the entrepreneurial space, including but not limited to:

[0053] Office area: open or enclosed workspace for daily work of individuals or teams;

[0054] Meeting room: a space used for holding meetings, lectures or training;

[0055] Rest area: a place for users to relax, socialize or take a short break;

[0056] Exhibition area: a space used to display products, project results or hold exhibitions;

[0057] Service area: such as cafes, printing rooms and other auxiliary facilities areas.

[0058] Each functional area has its own unique traffic density characteristics, which are closely related to its specific functional attributes; for example, the traffic density in the office area may be higher during peak hours on weekdays, while the rest area is busier during lunch time and afternoon tea time; therefore, when dividing functional areas, the following factors must be fully considered:

[0059] Purpose: What is the main purpose of the area? Is it mainly for work, communication or leisure?

[0060] Visit frequency: When do users typically visit this area? How often?

[0061] Dwell time: How long do users stay in this area on average?

[0062] Activity type: The nature of the activities taking place in the area (e.g., a quiet working environment versus an active social venue).

[0063] After completing the above analysis, the entrepreneurial space is divided into several functional areas, and a functional area cluster that includes all functional areas is formed; this cluster not only covers the physical spatial distribution, but also reflects the logical relationship between the various areas and their impact patterns on the flow of people; in this way, a prediction model can be constructed separately for each functional area, thereby improving the accuracy and practicality of the overall prediction.

[0064] The division of functional areas should be as detailed and accurate as possible in order to better capture the differences between areas and the changing characteristics of pedestrian density; at the same time, as the entrepreneurial space develops and changes, the division of functional areas needs to be adjusted in a timely manner to ensure the accuracy and practicality of the prediction results.

[0065] For step S2:

[0066] Step S2 is to traverse and screen the preset crowd density weight space for each functional area with the functional attributes corresponding to the functional area, and extract the corresponding weather impact weight matrix; the purpose is to quantify the specific impact of various weather conditions on the crowd density of each functional area in different time periods, so as to improve the accuracy and practicality of the prediction, specifically:

[0067] Construct a pedestrian density weight space, which is a multidimensional data structure, which contains the weather impact weight matrix corresponding to all functional areas; the column items of each matrix represent time periods (such as 0-1 o'clock, 1-2 o'clock, etc.), and the row items represent different types of weather conditions (such as temperature, humidity, rainfall, wind speed, etc.); the intersection of rows and columns represents the weight coefficient of the impact of a certain type of weather condition on the pedestrian density of the functional area within a specific time period.

[0068] According to the specific functional attributes of each functional area (such as office area, meeting room, rest area, etc.), determine the response mode of the area under different time periods and weather conditions; for example:

[0069] Office areas: Usually less affected by weather, but in extreme weather conditions (such as snow or heavy rain), more people may choose to stay indoors to work;

[0070] Rest areas: When the weather is good, people tend to rest outdoors, so the flow of people in rest areas may decrease; conversely, bad weather may cause more people to stay in indoor rest areas;

[0071] Display area: If the display area is located in an outdoor or semi-outdoor environment, the weather will have a greater impact on it; while the indoor display area is relatively stable.

[0072] By traversing the pedestrian density weight space defined above, the weather impact weight matrix that matches the functional attributes of each functional area is screened out; determine which time periods are critical periods (such as weekday peak hours, lunch time, afternoon tea time, etc.); assign an impact coefficient to each weather condition in each time period, indicating the degree of influence of the condition on the change in pedestrian density; these coefficients can be obtained through historical data analysis, expert evaluation or machine learning model training.

[0073] A weather impact weight matrix is ​​constructed for each functional area; the matrix not only contains a series of weather conditions and their corresponding impact coefficients, but also clarifies the changing trends of these impacts in different time periods; for example, for a rest area, the weather impact weight matrix is ​​as follows:

[0074]

[0075] Among them, the positive and negative signs represent the impact of weather conditions on the increase or decrease of the pedestrian density in the functional area, and the numerical value represents the degree of impact; the weight coefficients of different time periods reflect the different impacts of weather conditions on the pedestrian density in different periods.

[0076] In summary, step S2 constructs and screens the weather impact weight matrix through scientific methods to provide a personalized weather impact assessment tool for each functional area; the structure of each matrix clearly reflects the impact of various weather conditions on the crowd density in different time periods, improves the accuracy and practicality of the prediction results, and enables entrepreneurial space managers to better cope with the complex and changing operating environment.

[0077] For step S3:

[0078] Step S3 is to understand the characteristics of the time window to be predicted from the time dimension, including the day of the week to which it belongs and the specific time range (start timestamp and end timestamp); the flow density corresponding to different dates and times may be significantly different; for example, the peak and non-peak hours on weekdays, the flow difference between weekends and weekdays, etc.; in order to improve the accuracy of the prediction, it is necessary to perform a detailed date attribute analysis on each time window to be predicted;

[0079] Step S31, clarify the time range to be predicted, that is, the time window to be predicted; this time window can be a period of time of any length, such as one day, half a day or several hours; after determining the time window, it is necessary to further analyze the specific date attributes of the time period;

[0080] Step S32: for each time window to be predicted, extract the corresponding weekday information, that is, the day of the week on which the time period falls; for example, if the time window to be predicted is from 9 am to 5 pm on Monday, the weekday is "Monday"; if it is from 6 pm on Friday to 8 am on Saturday, two weekdays are involved: "Friday" and "Saturday"; the weekday information helps to identify the difference between weekdays and weekends, because different types of days usually have different traffic patterns; for example, the office area may be busier on weekdays, while the rest area may have more visitors on weekends;

[0081] Step S33, extract the time sequence information of the time window to be predicted, specifically represented by the start timestamp and end timestamp of the time period; this step is to accurately define the time boundary of the prediction and ensure the matching accuracy of historical data; for example: for a time window from 9 am to 5 pm on Monday, the time sequence is "09:00-17:00"; for an overnight time window, such as 6 pm on Friday to 8 am on Saturday, the time sequence is "18:00-08:00"; the time sequence information not only helps to refine to a specific hourly level, but also captures the changing patterns of passenger flow in different time periods within a day; for example, lunch time and rush hour are usually periods with higher passenger flow density;

[0082] Step S34, combining the above extracted information, generating a date attribute vector including the week order and the time order; the vector is used to identify the unique time features of the time window to be predicted; for example: for the time window from 9 am to 5 pm on Monday, the date attribute vector can be expressed as [Monday, 09:00-17:00]; for the time window from 6 pm on Friday to 8 am on Saturday, the date attribute vector can be expressed as [Friday, 18:00-24:00] and [Saturday, 00:00-08:00].

[0083] Step S3 improves the accuracy of crowd density prediction by analyzing the date attributes of the time window to be predicted in detail. By considering the two dimensions of week order and time order, it is possible to capture the difference in crowd density between weekdays and weekends and different time periods, which helps to more accurately screen historical data, build a prediction model, and correct the prediction results.

[0084] For step S4:

[0085] In order to improve the accuracy of crowd density prediction, historical data needs to be used for training and calibration. The goal of step S4 is to extract records that are similar to the time window to be predicted in date attributes and close in time sequence from the historical data, so as to more accurately reflect the changing trend of future crowd density. The specific implementation is as follows:

[0086] Step S41, clarify the historical time window pedestrian density that needs to be extracted for each functional area; find the historical time period with the same week sequence (day of the week) and time sequence (start timestamp and end timestamp) as the time window to be predicted; select the historical time period closest to the time window to be predicted in terms of time sequence to ensure the relevance and timeliness of the data;

[0087] Step S42: Based on the above objectives, search from the historical statistical records of the density of people in the entrepreneurial space; use the date attribute vector as the query condition to filter out the time window that meets the condition from the historical database; for example, if the time window to be predicted is "Monday, 09:00-17:00", then search for all historical records that are also "Monday, 09:00-17:00"; for the retrieved historical time periods, calculate their time distances relative to the time window to be predicted, and arrange them in ascending order; the most recent historical data can be given priority because these data can better reflect the user behavior patterns in the current situation;

[0088] Step S43: According to a preset number (such as the most recent 5 or 10), select a corresponding number of historical time windows from the sorted results; the selection of this preset number can be adjusted according to actual needs, and usually enough data points are selected to ensure statistical validity, but not too many to introduce noise;

[0089] Step S44: For each selected historical time window, extract the crowd density value of the corresponding functional area; these values ​​will be used for analysis and modeling in subsequent steps; for example, if 5 historical time windows are selected, 5 sets of crowd density data on different dates but in the same time period will be obtained.

[0090] In this step, by explicitly extracting the pedestrian density of historical time windows with the same date attribute and close time series for each functional area, the characteristic differences of different functional areas are fully considered to make the data more targeted; accurate retrieval and sorting with date attribute vector as query condition can ensure that the selected data is highly relevant to the time window to be predicted and has strong timeliness, which can more accurately reflect the current trend; the selection of preset quantity is flexible, which not only ensures the statistical validity but also avoids excessive noise interference; at the same time, the most recent historical data is given priority to better reflect the current user behavior pattern; the pedestrian density value of the selected historical time window is extracted to provide an accurate and reliable data basis for subsequent analysis and modeling, which helps to improve the accuracy and reliability of pedestrian density prediction.

[0091] For step S5:

[0092] Step S5 arranges the crowd density of the preset number of historical time windows extracted in step S4 in time series to obtain a crowd density time series vector; the purpose is to provide orderly and structured input data for the subsequent benchmark crowd density prediction model to ensure the accuracy and reliability of the prediction results; the specific implementation is as follows:

[0093] Step S51: The crowd density data of a preset number (such as the most recent 5 or 10) of historical time windows obtained in step S4; these data have been screened to ensure that they are similar in date attributes and close in time sequence to the time window to be predicted;

[0094] Step S52: Arrange the crowd density data of the above historical time windows in chronological order; specifically, sort the crowd density data from small to large according to the time distance of each historical time window relative to the time window to be predicted; this ensures that the closest time window is placed in front, reflecting the timeliness of the data; ensure that the data arrangement of all functional areas is consistent to maintain the consistency and comparability of comparisons between different areas; for example, if 5 historical time windows are selected, the following arrangement will be obtained:

[0095] In the most recent historical time window, the crowd density value is A (Monday, 09:00-17:00);

[0096] The second most recent historical time window has a flow density value of B (Monday, 09:00-17:00);

[0097] The third most recent historical time window has a flow density value of C (Monday, 09:00-17:00);

[0098] The fourth most recent historical time window has a flow density value of D (Monday, 09:00-17:00);

[0099] The fifth most recent historical time window has a flow density value of E (Monday, 09:00-17:00);

[0100] Step S53, combine the crowd density values ​​of the historical time windows arranged in chronological order into a vector, namely, the crowd density time series vector; the vector not only contains the crowd density information in multiple historical time periods, but also maintains the time series relationship between them; for example, for the above five historical time windows, a vector of the following form can be constructed: crowd density time series vector = [A, B, C, D, E]; wherein A, B, C, D, and E respectively represent the crowd density values ​​of the functional areas corresponding to each historical time window.

[0101] In this step, by arranging the crowd density of a preset number of historical time windows in chronological order, the timeliness of the data is ensured, and the data closest to the time window to be predicted is given priority, which can more accurately reflect the current trend; it provides ordered and structured input data for the subsequent benchmark crowd density prediction model, so that the model can better learn and capture the law of changes in crowd density over time, thereby improving the accuracy of the prediction; it ensures the consistency of data arrangement in different functional areas, maintains the consistency and comparability of comparisons between different areas, and helps to analyze the crowd density characteristics of each functional area more comprehensively and accurately; this ordered time series vector construction method makes the data organization clearer and more standardized, facilitates subsequent data processing and analysis operations, and improves the efficiency and reliability of the entire prediction process.

[0102] For step S6:

[0103] Step S6 inputs the crowd density time series vector obtained in step S5 into a preset benchmark crowd density prediction model to obtain the benchmark crowd density of the corresponding functional area within the time window to be predicted; to achieve this goal, the benchmark crowd density prediction model is constructed based on time series analysis and machine learning algorithms, aiming to capture patterns in historical data and perform extrapolated predictions; specifically;

[0104] Decomposing the time series vector of passenger flow density into three parts: trend, seasonality, and residual helps to identify long-term trends, cyclical changes, and random fluctuations; the calculation formula for the benchmark passenger flow density is:

[0105] Y t =T t +S t +R t

[0106] Among them, Y t represents the baseline passenger flow density at the tth time point; T t Represents the trend component, reflecting the long-term trend of change; S t Represents seasonal components, reflecting the periodic change pattern; R t represents the residual component, which represents random fluctuations or noise.

[0107] Among them, trend prediction is used to capture long-term growth or decline trends; for example, during peak working hours, the density of people in the office area may increase year by year because the company expands or more people choose co-working space; linear regression, exponential smoothing and other methods can be used for modeling; for example, for linear trends, a linear regression model is used:

[0108] T t =α+βt+δ

[0109] Among them, α is the intercept term, representing the initial value; β is the slope term, representing the trend growth rate per unit time; δ is the error term, representing variability.

[0110] Seasonal forecasting is used to capture periodic changes; for example, the density of people in the rest area will increase significantly during lunch time and afternoon tea time, and the display area may have more visitors on weekends; Fourier series expansion or ARIMA model can be used to capture periodic changes; for data with a fixed period (such as a day or a week), use Fourier series:

[0111]

[0112] Among them, K represents the order of the Fourier series, which determines the precision of the fitting; P represents the length of the cycle, for example, there are 24 hours in a day and 7 days in a week; a k and b k are the Fourier coefficients, which can be estimated by the least squares method.

[0113] In order to improve the prediction accuracy, the residual component needs to be processed; one of the following methods can be used:

[0114] White noise assumption: If the residual is white noise (i.e., it has zero mean, constant variance, and no autocorrelation), it can be ignored or simplified by weighted averaging.

[0115] ARIMA Model: Use ARIMA models to fit and predict residual components, especially when the residuals exhibit autocorrelation;

[0116] Machine learning methods: Use machine learning algorithms (such as random forests, neural networks) to capture complex patterns in the residuals; for example, train a random forest model to predict the residuals:

[0117] R t =g(X t )

[0118] Among them, X t Including time series features (such as lagged variables, moving averages, etc.) and other external features (such as weather information, holiday signs, etc.); g(·) represents the prediction function obtained by random forest training.

[0119] In this step, by decomposing the time series vector of pedestrian density into three parts: trend, seasonality and residual, the various changing characteristics of pedestrian density can be fully captured to improve the accuracy of prediction; by using a variety of methods (such as linear regression, Fourier series, ARIMA model, machine learning, etc.) for modeling, the most appropriate method can be selected according to the specific data characteristics to enhance the adaptability of the model; by processing the residual component, the prediction error can be further reduced, making the prediction result closer to the true value; targeted prediction can be made according to the characteristics of different functional areas (such as office areas, rest areas, etc.) to meet the diverse needs in practical applications.

[0120] For step S7:

[0121] In step S6, the benchmark crowd density of the corresponding functional area within the time window to be predicted has been obtained; however, the actual crowd density may be affected by external weather factors; in order to further improve the accuracy of the prediction, it is necessary to make corrections based on the benchmark crowd density and the weather forecast status; the specific operation steps include:

[0122] Step S71, obtaining the weather forecast status of the time window to be predicted from a reliable weather forecast service; the weather forecast status generally includes information such as temperature, humidity, rainfall, wind speed, weather type (sunny, cloudy, rainy, etc.); for example, if the time window to be predicted is from 9 am to 5 pm on Monday, it is necessary to obtain hourly weather forecast data within the time period;

[0123] Step S72: According to the weather impact weight matrix extracted in step S2, each functional area has a specific matrix, and the intersection of its rows and columns represents the weight coefficient of the impact of a certain type of weather conditions on the flow density of people in the functional area in a specific time period;

[0124] Step S73, matching the obtained weather forecast status with the weather impact weight matrix; specifically: ensuring that the time period of the weather forecast is consistent with the time window to be predicted; finding the corresponding row and column intersections in the matrix according to the specific values ​​of the weather forecast, and extracting the corresponding weather impact weight coefficients; for example, if the weather forecast shows that there will be light rain and high humidity from 9 am to 5 pm on Monday, the weight coefficients of the corresponding time period and weather conditions in the matrix can be found;

[0125] Step S74: for each functional area, extract the weather impact weight coefficient corresponding to the time window to be predicted; these coefficients reflect the specific impact of specific weather conditions on the density of human traffic; for example, for the rest area, if the weather forecast shows rain, it may be found that the weight coefficients of high humidity and heavy rainfall are -0.3 and -0.6 respectively, indicating that bad weather will significantly reduce the density of human traffic in the rest area;

[0126] Step S75: In order to more comprehensively evaluate the impact of weather on pedestrian density, it is necessary to comprehensively consider a variety of weather conditions; a weighted summation method can be used to combine the weight coefficients of various weather conditions to form a comprehensive weather impact coefficient.

[0127] In this step, by obtaining data from reliable weather forecast services, it is ensured that the forecast is based on the latest weather conditions, thereby improving the timeliness and accuracy of the forecast; each functional area has a specific weather impact weight matrix, which can accurately reflect the impact of different weather conditions on the density of pedestrian flow in the area, thereby improving the refinement of the forecast; a variety of weather conditions and their interactions are comprehensively considered, and a weighted summation method is used to form a comprehensive impact coefficient, making the forecast result more comprehensive and reliable; it can adapt to changes in different time periods and weather types, and has strong flexibility and scalability.

[0128] For step S8:

[0129] In step S6, the benchmark crowd density of the corresponding functional area within the time window to be predicted has been obtained; however, the actual crowd density may be affected by external weather factors; in order to further improve the accuracy of the prediction, it is necessary to make corrections based on the benchmark crowd density and the weather forecast status; the specific operation steps are as follows:

[0130] Step S81, obtaining the reference flow density of people in the corresponding functional area within the time window to be predicted from step S6; the reference flow density of people is a preliminary prediction based on historical data and time series analysis;

[0131] Step S82: According to step S7, the weather forecast status corresponding to the time window to be predicted has been obtained, and the corresponding comprehensive weather impact coefficient is extracted from it, reflecting the impact of specific weather conditions (such as temperature, humidity, rainfall, etc.) on the flow density of people in each functional area;

[0132] Step S83: Use the comprehensive weather impact coefficient to correct the reference crowd density; specifically, the adjustment can be made using the following formula:

[0133]

[0134] Among them, represents the corrected crowd density in the time window t to be predicted; represents the baseline crowd density in the time window t to be predicted; represents the comprehensive weather impact coefficient corresponding to the time window t to be predicted. A positive value indicates an increase in the crowd density, and a negative value indicates a decrease in the crowd density.

[0135] In this step, the benchmark passenger flow density is corrected by introducing the weather impact weight coefficient, taking into account the actual impact of external weather factors, making the prediction result closer to the actual situation and improving the accuracy of the prediction; the weight coefficients of multiple weather conditions are combined for correction, which can more finely reflect the specific impact of different weather conditions on the passenger flow density and improve the refinement of the prediction; adjustments are made based on the latest weather forecast status to ensure that the prediction results can promptly reflect the current weather conditions and enhance the timeliness of the prediction.

[0136] For step S9:

[0137] In the previous steps, the crowd density of each functional area in the time window to be predicted has been obtained; in order to obtain the crowd density prediction result of the entire entrepreneurial space in the time window to be predicted, the crowd density of each functional area needs to be summarized; the specific operation steps are as follows: obtain the crowd density value corresponding to each functional area from step S8; each functional area occupies a different area in the entrepreneurial space, and its contribution to the overall crowd density is also different; therefore, it is necessary to calculate the weight of each functional area according to its area; for example, if the office area occupies 60% of the total area, the rest area occupies 20%, the conference room occupies 10%, and the exhibition area occupies 10%, the corresponding weights are 0.6, 0.2, 0.1, and 0.1 respectively; use the area weights of each functional area to perform weighted average on the crowd density to obtain the overall crowd density of the entire entrepreneurial space; use the calculated overall crowd density as the final crowd density prediction result of the entrepreneurial space in the time window to be predicted;

[0138] Step S9 provides a comprehensive and accurate prediction result of the comprehensive pedestrian density of the entrepreneurial space by weighted summary of the pedestrian density of each functional area; this method not only improves the scientificity and practicality of the prediction, but also enhances the adaptability and flexibility of the model, enabling managers to better cope with the complex and changing operating environment, optimize resource allocation, and improve user experience and overall operational efficiency.

[0139] like Figure 2 , Figure 3 As shown, an embodiment of the present invention provides a device for predicting the density of people flow in an entrepreneurial space. The device embodiment can be implemented by software, or by hardware or a combination of software and hardware. From the hardware level, Figure 2 As shown, it is a hardware architecture diagram of an electronic device in which a device for predicting the density of people flowing into an entrepreneurial space provided by an embodiment of the present invention is located. Figure 2 In addition to the processor, memory, network interface, and non-volatile memory shown, the electronic device in the embodiment may also include other hardware, such as a forwarding chip responsible for processing messages, etc. Taking software implementation as an example, Figure 3 As shown, as a device in a logical sense, the CPU of the electronic device in which it is located reads the corresponding computer program in the non-volatile memory into the internal memory and runs it.

[0140] like Figure 3 As shown, this embodiment provides a device for predicting crowd density in an entrepreneurial space, including:

[0141] A functional area division module is used to divide the entrepreneurial space into functional areas and obtain an entrepreneurial space functional area cluster including all functional areas in the entrepreneurial space;

[0142] The weather impact weight matrix extraction module is used to traverse and screen each functional area in the preset pedestrian density weight space based on the functional attributes corresponding to the functional area, and extract the corresponding weather impact weight matrix;

[0143] A date attribute analysis module, used to perform date attribute analysis on the time window to be predicted, and obtain a date attribute vector including the week sequence and time sequence to which the time window to be predicted belongs;

[0144] A historical pedestrian density extraction module is used to extract, for each functional area, from the historical statistical records of entrepreneurial space pedestrian density, a preset number of historical time window pedestrian densities that are the same as the date attribute vector and closest to the time window to be predicted in terms of time series;

[0145] A crowd density time series vector generation module, used for arranging the crowd density of the preset number of historical time windows according to time series to generate a crowd density time series vector;

[0146] A reference crowd density prediction module, used to input the crowd density time series vector into a preset reference crowd density prediction model to obtain a reference crowd density of a corresponding functional area within a time window to be predicted;

[0147] A weight coefficient extraction module is used to obtain the weather forecast status corresponding to the time window to be predicted, and extract the corresponding weather impact weight coefficient in the weather impact weight matrix based on the weather forecast status;

[0148] A crowd density correction module, used to perform impact correction on the reference crowd density according to the weather impact weight coefficient to obtain the crowd density corresponding to the functional area;

[0149] The crowd density summary module is used to summarize the crowd density corresponding to each functional area in the entrepreneurial space functional area cluster to obtain the entrepreneurial space crowd density prediction result.

[0150] It is to be understood that the structure illustrated in the embodiment of the present invention does not constitute a specific limitation on a device for predicting the density of people flow in an entrepreneurial space. In other embodiments of the present invention, a device for predicting the density of people flow in an entrepreneurial space may include more or fewer components than shown in the figure, or combine some components, or split some components, or arrange the components differently. The components shown in the figure may be implemented in hardware, software, or a combination of software and hardware.

[0151] The information interaction, execution process and other contents between the modules in the above-mentioned device are based on the same concept as the embodiment of the method of the present invention. For the specific contents, please refer to the description in the embodiment of the method of the present invention, and no further description is given here.

[0152] An embodiment of the present invention further provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, a method for predicting the density of people flow in an entrepreneurial space in any embodiment of the present invention is implemented.

[0153] An embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the processor executes a method for predicting the density of people flow in an entrepreneurial space in any embodiment of the present invention.

[0154] Specifically, a system or device equipped with a storage medium can be provided, on which software program code that implements the functions of any of the above-mentioned embodiments is stored, and a computer (or CPU or MPU) of the system or device can be enabled to read and execute the program code stored in the storage medium.

[0155] In this case, the program code itself read from the storage medium can realize the function of any one of the above-mentioned embodiments, and thus the program code and the storage medium storing the program code constitute a part of the present invention.

[0156] The storage medium embodiments for providing the program code include a floppy disk, a hard disk, a magneto-optical disk, an optical disk (such as CD-ROM, CD-R, CD-RW, DVD-ROM, DVD-RAM, DVD-RW, DVD+RW), a magnetic tape, a non-volatile memory card, and a ROM. Alternatively, the program code can be downloaded from a server computer by a communication network.

[0157] In addition, it should be clear that the functions of any of the above embodiments can be implemented not only by executing the program code read by the computer, but also by enabling an operating system operating on the computer to complete part or all of the actual operations based on instructions from the program code.

[0158] In addition, it can be understood that the program code read from the storage medium is written to a memory provided in an expansion board inserted into the computer or to a memory provided in an expansion module connected to the computer, and then based on the instructions of the program code, a CPU installed on the expansion board or expansion module is enabled to perform part or all of the actual operations, thereby realizing the functions of any of the above-mentioned embodiments.

[0159] It should be noted that, in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0160] A person of ordinary skill in the art can understand that all or part of the steps of implementing the above method embodiments can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above method embodiments; and the aforementioned storage medium includes: ROM, RAM, magnetic disk or optical disk, etc., various media that can store program codes.

[0161] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for predicting the density of people flow in an entrepreneurial space, characterized in that: The method comprises: Dividing the entrepreneurial space into functional areas to obtain an entrepreneurial space functional area cluster; the entrepreneurial space functional area cluster includes all functional areas in the entrepreneurial space; For each functional area, the functional attributes corresponding to the functional area are traversed and screened in the preset crowd density weight space to extract the corresponding weather impact weight matrix; Perform date attribute analysis on the time window to be predicted to obtain a date attribute vector corresponding to the time window segment to be predicted, wherein the date attribute vector includes the week sequence and time sequence to which the time window to be predicted belongs; For each functional area, from the historical statistical records of entrepreneurial space pedestrian density, extract a preset number of historical time window pedestrian density that is the same as the date attribute vector and closest to the time window to be predicted in terms of time series; Arrange the crowd density of the preset number of historical time windows according to time series to obtain a crowd density time series vector; Inputting the crowd density time series vector into a preset benchmark crowd density prediction model to obtain a benchmark crowd density of the corresponding functional area within the time window to be predicted; Obtaining the weather forecast status corresponding to the time window to be predicted, and extracting the corresponding weather impact weight coefficient in the weather impact weight matrix based on the weather forecast status; Based on the weather impact weight coefficient, the reference crowd density is corrected to obtain the crowd density corresponding to the functional area; The crowd density corresponding to each functional area in the entrepreneurial space functional area cluster is summarized to obtain the prediction result of the entrepreneurial space crowd density.

2. The method for predicting the density of people flow in an entrepreneurial space as claimed in claim 1, characterized in that: The functional areas include an office functional area, a meeting functional area, a rest functional area, a display functional area and a service functional area.

3. The method for predicting the density of people flow in an entrepreneurial space as claimed in claim 2, characterized in that: The pedestrian density weight space includes weather impact weight matrices corresponding to all functional areas; the column items of each matrix represent time periods, and the row items represent different types of weather conditions; the intersection of rows and columns represents the weight coefficient of the impact of the corresponding weather conditions in the corresponding time period on the pedestrian density of the functional area.

4. The method for predicting the density of people flow in an entrepreneurial space as claimed in claim 3, characterized in that: In the historical statistical records of the crowd density of the entrepreneurial space, a preset number of historical time window crowd densities that are the same as the date attribute vector and closest to the time window to be predicted in terms of time series are extracted, including: Based on the date attribute vector, searching is performed from the historical statistical records of the density of people in the entrepreneurial space to filter out the historical time windows that meet the conditions; For the filtered historical time windows, calculate their time distances relative to the time window to be predicted and arrange them in ascending order; Select a corresponding number of historical time windows from the sorted results according to a preset number; For each selected historical time window, the crowd density value of the corresponding functional area is extracted.

5. The method for predicting the density of people flow in an entrepreneurial space as claimed in claim 4, characterized in that: The calculation formula of the benchmark crowd density prediction model is: Y t =T t +S t +R t Among them, Y t T represents the baseline passenger flow density in the time window t to be predicted; t Represents the trend component, which is used to reflect the long-term trend of change; S t Represents seasonal components, which are used to reflect the periodic change pattern; R t represents the residual component.

6. The method for predicting the density of people flow in an entrepreneurial space as claimed in claim 5, characterized in that: The trend component is expressed as: T t =α+βt+δ Among them, α is the intercept term, representing the initial value; β is the slope term, representing the trend growth rate per unit time; δ is the error term; The seasonal component is expressed as: Among them, K represents the order of the Fourier series, which determines the precision of the fitting; P represents the period length; a k and b k are the Fourier coefficients, estimated by the least squares method; The residual component is expressed as: R t =g(X t ) Among them, X t Including time series features and external features; g(·) represents the prediction function obtained by random forest training.

7. The method for predicting the density of people flow in an entrepreneurial space as claimed in claim 6, characterized in that: Obtaining the weather forecast status corresponding to the time window to be predicted, and extracting the corresponding weather impact weight coefficient in the weather impact weight matrix based on the weather forecast status, including: Obtain the weather forecast status for the time window to be predicted from the weather forecast service; Match the obtained weather forecast status with the corresponding weather impact weight matrix: according to the specific value of the weather forecast status, find the corresponding row and column intersection in the weather impact weight matrix and extract the corresponding weather impact weight coefficient; For each functional area, extract the weather impact weight coefficient corresponding to the time window to be predicted; For each functional area, a weighted summation method is adopted to comprehensively consider various weather conditions, combine the weight coefficients of various weather conditions, and obtain the comprehensive weather impact coefficient.

8. A device for predicting the density of people flow in an entrepreneurial space, characterized in that: include: A functional area division module is used to divide the entrepreneurial space into functional areas and obtain an entrepreneurial space functional area cluster including all functional areas in the entrepreneurial space; The weather impact weight matrix extraction module is used to traverse and screen each functional area in the preset pedestrian density weight space based on the functional attributes corresponding to the functional area, and extract the corresponding weather impact weight matrix; A date attribute analysis module, used to perform date attribute analysis on the time window to be predicted, and obtain a date attribute vector including the week sequence and time sequence to which the time window to be predicted belongs; A historical pedestrian density extraction module is used to extract, for each functional area, from the historical statistical records of entrepreneurial space pedestrian density, a preset number of historical time window pedestrian densities that are the same as the date attribute vector and closest to the time window to be predicted in terms of time series; A crowd density time series vector generation module, used for arranging the crowd density of the preset number of historical time windows according to time series to generate a crowd density time series vector; A reference crowd density prediction module, used to input the crowd density time series vector into a preset reference crowd density prediction model to obtain a reference crowd density of a corresponding functional area within a time window to be predicted; A weight coefficient extraction module is used to obtain the weather forecast status corresponding to the time window to be predicted, and extract the corresponding weather impact weight coefficient in the weather impact weight matrix based on the weather forecast status; A crowd density correction module, used to perform impact correction on the reference crowd density according to the weather impact weight coefficient to obtain the crowd density corresponding to the functional area; The crowd density summary module is used to summarize the crowd density corresponding to each functional area in the entrepreneurial space functional area cluster to obtain the entrepreneurial space crowd density prediction result.

9. An electronic device, characterized in that: The method comprises a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the method according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and when the computer program is executed in a computer, the computer is caused to execute the method according to any one of claims 1 to 7.

Images