Intelligent label management system based on data analysis

Abstract

The invention relates to the technical field of smart city information release, in particular to a smart label management system based on data analysis, which comprises a data acquisition module, a feature fusion module, a behavior analysis module, a strategy generation module and an execution control module. Wherein the data acquisition module is used for acquiring residence time data and interaction frequency data of users in a target area covered by a label; the feature fusion module is used for generating a user attention index; the behavior analysis module is used for identifying a high-attention user group and calculating an effective service range boundary of the label; and the strategy generation module is used for generating an optimization strategy set. According to the invention, through accurate perception and hierarchical strategy control of the user behavior data, orientation, hierarchy and linkage of intelligent label information delivery are realized, and the response efficiency and delivery precision of the system are effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of smart city information publishing, and particularly to a smart sign management system based on data analysis. Background Art

[0002] With the rapid development of smart cities and information visualization technologies, the digital sign systems deployed in public places are gradually evolving from traditional static displays to intelligent interaction and dynamic perception directions; currently, most smart signs mainly rely on preset time carousel or regional push mechanisms for information publishing. Although some systems have touch response functions, they still lack the ability to actively perceive user behavior characteristics and data-based strategy feedback; especially in transportation hubs, shopping mall advertising areas, and indoor guiding environments, users have short stay times, unstable interactions, and scattered behavior patterns, and traditional sign systems based on static configurations are difficult to effectively adapt to real-time information needs and user response behaviors in different scenarios.

[0003] The existing technologies generally have the following problems: First, there is a lack of a real-time fusion mechanism for multi-modal user behavior data, and it is impossible to accurately obtain the attention behavior characteristics of users; second, in terms of behavior analysis and regional strategy generation, it is impossible to achieve the division of placement priorities based on spatial aggregation rules, resulting in low information coverage efficiency; third, there is a lack of an executable control mechanism with closed-loop linkage, and the information placement process lacks the ability of dynamic adjustment. Therefore, there is an urgent need to construct a smart sign management system based on data analysis to solve the above problems. Summary of the Invention

[0004] Based on the above purpose, the present invention provides a smart sign management system based on data analysis.

[0005] A smart sign management system based on data analysis includes a data collection module, a feature fusion module, a behavior analysis module, a strategy generation module, and an execution control module; where: Data collection module: used to collect the stay time data and interaction frequency data of users in the target area covered by the sign in real time through an infrared thermal imaging sensor array and a touch interaction panel deployed on the sign body; Feature fusion module: used to receive the stay time data and interaction frequency data, generate a user behavior feature package through spatio-temporal correlation matching, calculate the corresponding weight coefficient according to the scene type to which the sign belongs, and then weighted fusion to generate a user attention index; Behavior analysis module: used to identify high-attention user groups according to the comparison result between the user attention index and a preset threshold, and calculate the boundary of the effective service range of the sign using the grid heat zone division method based on their spatial aggregation density; Strategy generation module: Based on the boundary coordinates of the effective service range and the real-time distribution coordinates of the high-attention user group, it generates an optimization strategy set including multi-level information push priority areas, optimal sign orientation angles, and multi-sign collaborative control parameters; Execution control module: used to analyze and optimize the strategy set, and control the sign content update, posture adjustment and information coordination through the IoT terminal to achieve information delivery.

[0006] Optionally, the data acquisition module includes an infrared image acquisition unit, a trajectory recognition unit, a touch event recording unit and a data synchronization unit; wherein: Infrared image acquisition unit: used to continuously acquire thermal image frame data within the area covered by the sign through an infrared thermal imaging sensor array, and to capture the two-dimensional coordinates of the user's body heat source in real time based on a set frame rate; Trajectory recognition unit: used to track the heat source points in the infrared image frame in time series, by calculating the continuous displacement vector of each user in the image coordinate system, to determine whether the user is in a stationary state, and to convert the user's stay time at the specified location based on the number of stay frames and the image frame rate ; Touch event recording unit: used to monitor each touch event on the touch interaction panel and record the corresponding timestamp to count the interaction frequency within a given time window ; Data synchronization unit: used to synchronize the residence time Corresponding touch frequency Perform timestamp alignment and package to generate structured original interaction behavior records.

[0007] Optionally, the feature fusion module includes a space aggregation unit, a feature package assembly unit, a weight coefficient calculation unit and an attention generation unit; wherein: Spatial aggregation unit: divide the grid units based on the user coordinate information in the original interaction behavior record, and aggregate the behavior data in the same grid and with overlapping time periods into the same user instance; Feature package assembly unit: used to extract the residence time, interaction frequency and time period label from each user instance and encapsulate it into a user behavior feature package; Weight coefficient calculation unit: used to call the preset scene weight table according to the scene type to which the sign belongs, and assign weight coefficients of determined values ​​to the dwell time, interaction frequency and time period labels respectively. The scene types include public transportation station scenes, supermarket advertising space scenes and indoor guide scenes; Attention generation unit: performs weighted fusion on the user behavior feature package based on the corresponding weight coefficient to generate the corresponding user attention index.

[0008] Optionally, the spatial aggregation unit includes: Grid division sub - unit: Used to equally divide the coverage area at equal intervals according to a fixed grid side length with the sign reference origin as the benchmark, and calculate the grid index where any user coordinate is located. ; Time overlap determination sub - unit: Used to compare the recorded time intervals of two behavior data belonging to the same grid index . When and are satisfied, it is determined that there is an effective time overlap between the two behavior data; where are the start and end timestamps of the corresponding behavior data respectively; . User instance aggregation sub - unit: Used to merge the behavior data that is in the same grid index and meets the time overlap determination condition into a single user instance.

[0009] Optionally, the attention degree generation unit includes: Feature extraction sub - unit: Used to extract feature parameters from the user behavior feature package, including user stay time , user interaction frequency and the label of the time period when the behavior occurs ; Weight loading sub - unit: Used to read the corresponding weight coefficients of each feature from the preset weight configuration table according to the scene type to which the sign belongs, namely the stay time weight , the interaction frequency weight , the time period weight , satisfying ; Weighted calculation sub - unit: Used to perform a linear weighted operation on the extracted feature parameters and the loaded weight coefficients to generate the final user attention degree index, and the calculation formula is: , where represents the user attention degree index.

[0010] Optionally, the behavior analysis module includes a threshold comparison unit, a group recognition unit, a hot zone division unit, and a service boundary calculation unit; among them: Threshold comparison unit: Used to receive the user attention degree index output by the feature fusion module, compare it item by item with the attention degree threshold built in the behavior analysis module, and generate a determination label. The determination label distinguishes two types of states: high attention degree and normal attention degree; Group recognition unit: Used to perform a connectivity scan on user instances that are in the same attention degree state and spatially adjacent according to the determination label output by the threshold comparison unit, and aggregate the high - attention - degree users in the continuous area into a high - attention - degree user group; Hot Zone Division Unit: It is used to calculate the spatial aggregation density of high-concern user groups, divide the coverage area into multi-level hot zone grids, and classify the grids according to the number of high-concern users in each grid to obtain a set of hot spot grids; Service Boundary Calculation Unit: It is used to fit the circumscribed polygon of the set of hot spot grids and output the vertex coordinates of the fitted polygon as the boundary of the effective service range of the sign.

[0011] Optionally, the Hot Zone Division Unit includes: Density Calculation Sub-unit: It is used to traverse and count the user coordinates within each high-concern user group, and calculate the spatial aggregation density within the basic grid with a side length of ; ; Grid Classification Sub-unit: Based on the grid density , perform multi-level grid subdivision processing on the coverage area. In the basic grid where the density is higher than the threshold , continue to use binary refinement of the side length to divide dense sub-grids, forming a multi-level hot zone grid system; Heat Calibration Sub-unit: It is used to classify and calibrate the heat levels of all grids according to the number of high-concern users inside. The heat levels are divided into three levels: first-level heat, second-level heat, and third-level heat according to the threshold interval; specifically expressed as: ; Among them, is the threshold of the number of high-concern users; is the heat level, taking values ; Hot Spot Extraction Sub-unit: It is used to collect the grids with the heat level and combine them to obtain a set of hot spot grids.

[0012] Optionally, the Policy Generation Module includes a Region Priority Division Unit, an Orientation Optimization Unit, and a Cooperative Control Unit; among them: Region Priority Division Unit: It is used to receive the set of service boundary vertex coordinates and the real-time distribution coordinates of high-concern user groups, divide the boundary area into multi-level information push priority regions according to the minimum distance from the user location to the service boundary, and assign corresponding priority identifiers to each region; Orientation Optimization Unit: Based on the coordinates of all high-concern users in each priority region, calculate the centroid coordinates of the region, and take the sign installation center point as the rotation center, and determine the optimal orientation angle for each region according to the following formula: , where is the optimal orientation angle corresponding to the -level push region; and are the The horizontal and vertical coordinates of the center of gravity of highly concerned users within the level push priority area, in meters; Are the coordinates of the sign installation position; Is the two-variable arctangent function; Collaborative control sub-unit: Used in the scenario of multi-sign deployment, according to the service boundary overlap relationship and push area priority of each sign, generate the sign content update timing and communication relay strategy.

[0013] Optionally, the area priority division unit includes: Distance calculation sub-unit: For the spatial coordinates of each highly concerned user, find the minimum Euclidean distance on the entire edge set of the service scope boundary polygon The calculation formula is: , where Is the user coordinate; Is the coordinate of the point on the boundary polygon closest to the user; Is the minimum distance from the user to the service boundary; Ring division sub-unit: Used to divide the service scope into three-level ring areas in the form of concentric rings from the boundary inward according to the comparison between the Output by the distance calculation sub-unit and the preset distance threshold . Specifically, The first-level information push area, satisfying ; The second-level information push area, satisfying ; The third-level information push area, satisfying .

[0014] Optionally, the execution control module includes a policy parsing unit, a display control unit, an attitude driving unit, and an information synchronization unit; among them: Policy parsing unit: Used to receive the optimized policy set output by the policy generation module, extract the information push priority area data, the optimal orientation angle instruction of the sign, and the collaborative control parameters contained therein, and perform structured parsing in the execution order, and output to the corresponding control unit; Display control unit: Used to screen the content push list issued by the sign main control system according to the priority area identifier contained in the policy parsing unit, and push the corresponding content to the display screen in the order from high to low priority; Attitude driving unit: Used to generate a rotation control instruction according to the parsed optimal orientation angle, and drive the built-in attitude adjustment motor to rotate the sign body to the specified direction angle; Information synchronization unit: Used to establish a master-backup sign communication link according to the collaborative control parameters, and schedule the communication relay to synchronously forward the content update and attitude adjustment instructions according to the preset timing.

[0015] Advantages of the present invention: In the present invention, by deploying an infrared thermal imaging sensor array and a touch interaction panel, multi-modal real-time acquisition of user staying behavior and interaction behavior is achieved; combined with spatio-temporal association and hierarchical aggregation technologies, user behavior characteristics can be accurately extracted and an attention index can be quantitatively generated, greatly improving the accuracy of user behavior capture and data consistency.

[0016] In the present invention, based on the aggregation law of high-attention users, algorithms such as threshold comparison, hot zone division, and annular zone priority division are adopted to automatically generate a hierarchical information push area and a targeted placement strategy, and content update, posture adjustment, and multi-sign board collaborative control are completed through an Internet of Things terminal, realizing the closed-loop linkage between strategy and execution; thus significantly improving the accuracy, coverage efficiency, and system adaptability of information placement. Brief description of the drawings

[0017] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following will briefly introduce the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings in the following description are only those of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0018] Figure 1 Schematic diagram of the intelligent sign management system according to an embodiment of the present invention; Figure 2 Schematic diagram of the behavior analysis module according to an embodiment of the present invention. Detailed implementation manners

[0019] The following will describe the present invention in detail with reference to the drawings and specific embodiments. At the same time, it should be noted here that in order to make the embodiments more detailed, the following embodiments are the best and preferred embodiments. For some well-known technologies, those skilled in the art can also adopt other alternative methods for implementation; and the drawing part is only for more specific description of the embodiments, and is not intended to specifically limit the present invention.

[0020] As Figure 1 - Figure 2 shown, an intelligent sign management system based on data analysis includes a data acquisition module, a feature fusion module, a behavior analysis module, a strategy generation module, and an execution control module; wherein: Data acquisition module: used to deploy an infrared thermal imaging sensor array and a touch interaction panel on the sign body to real-time collect the staying time data and interaction frequency data of users in the target area covered by the sign; Feature Fusion Module: It is used to receive the residence time data and interaction frequency data, generate user behavior feature packets through spatio-temporal correlation matching, calculate the corresponding weight coefficients according to the scene type to which the sign belongs, and then generate the user attention index through weighted fusion; Behavior Analysis Module: It is used to identify high-attention user groups according to the comparison result between the user attention index and the preset threshold, and calculate the boundary of the effective service range of the sign by using the grid heat zone division method based on their spatial aggregation density; Strategy Generation Module: Based on the coordinates of the effective service range boundary and the real-time distribution coordinates of high-attention user groups, generate an optimization strategy set including multi-level information push priority areas, the optimal orientation angle of the sign, and multi-sign collaborative control parameters; Execution Control Module: It is used to parse the optimization strategy set, and control the sign content update, attitude adjustment and information collaboration through the Internet of Things terminal to realize information delivery.

[0021] The data acquisition module includes an infrared image acquisition unit, a trajectory recognition unit, a touch event recording unit, and a data synchronization unit; among them: Infrared Image Acquisition Unit: It is used to continuously obtain the thermal image frame data within the sign coverage area through an infrared thermal imaging sensor array, and capture the two-dimensional coordinates of the user's body heat source in real time based on the set frame rate; Trajectory Recognition Unit: It is used to perform time-series tracking on the heat source points in the infrared image frames, judge whether the user is in a stationary state by calculating the continuous displacement vector of each user in the image coordinate system, and convert the residence time of the user at the specified position according to the number of frames stayed and the image frame rate ; Its calculation formula is: , where, represents the user residence time, in seconds; represents the number of frames that the user continuously appears in the same spatial grid; represents the infrared image frame rate, in frames per second; Touch Event Recording Unit: It is used to monitor each touch event on the touch interaction panel and record the corresponding time stamp to count the interaction frequency within a given time window ; Data Synchronization Unit: It is used to align the time stamps of the residence time and the corresponding touch frequency , and pack and generate a structured original interaction behavior record as the data input of the feature fusion module; by setting the infrared image acquisition unit and the trajectory recognition unit to work together, non-contact residence behavior capture of users in the area around the sign can be realized. Combining with the interaction behavior recorded by the touch event recording unit, the spatio-temporal fusion of thermal imaging and touch data can be effectively realized, improving the reduction accuracy and real-time performance of the data acquisition module for users' actual behaviors.

[0022] The feature fusion module includes a spatial aggregation unit, a feature packet assembly unit, a weight coefficient calculation unit, and an attention degree generation unit; among which: Spatial aggregation unit: Divide grid cells based on the user coordinate information in the original interaction behavior record, and aggregate the behavior data located in the same grid and with overlapping time periods into the same user instance; Feature packet assembly unit: Extract the stay time, interaction frequency, and time period label from each user instance, and encapsulate them into a user behavior feature packet; Weight coefficient calculation unit: Call the preset scenario weight table according to the scenario type to which the sign belongs, and assign weight coefficients with determined values to the stay time, interaction frequency, and time period label respectively. The scenario types include public transportation station scenarios, shopping mall advertisement position scenarios, and indoor wayfinding scenarios; Table 1 Preset Scenario Weight Example

[0023] In Table 1 above, the stay time weight is used to reflect the importance of the stay duration of the user in front of the sign for the attention degree; the interaction frequency weight is used to reflect the evaluation weight of the user's touch or operation behavior in this scenario; the time period level weight is used to reflect the influence intensity of the time period when the user behavior occurs on the attention degree in this scenario.

[0024] Attention degree generation unit: Perform weighted fusion on the user behavior feature packet based on the corresponding weight coefficients, generate the corresponding user attention degree index, and output it to the behavior analysis module; through the above units, accurate spatio-temporal pairing of stay behavior and interaction behavior, scenario-based weight assignment, and index representation can be realized, and the user attention degree index can be quickly generated on the premise of ensuring data consistency, improving the recognition accuracy of the feature fusion module for user behavior differences in different scenarios, so as to provide a reliable basis for subsequent analysis.

[0025] The spatial aggregation unit includes: Grid division sub-unit: Use the sign reference origin as a benchmark, equally divide the coverage area according to a fixed grid side length, and calculate the grid index where any user coordinate is located , and the specific calculation formula is: , where and respectively represent the grid numbers in the horizontal and vertical directions (dimensionless); and are the real-time coordinate values of the user in the sign coordinate system (unit: meter); is the coordinate of the sign installation position (unit: meter); is the grid side length (unit: meter); Time overlap determination sub-unit: For belonging to the same grid index Two pieces of behavior data, and compare their recorded time intervals and , when meeting , it is determined that there is an effective time overlap between the two pieces of behavior data; where are the start and end timestamps (unit: seconds) of the corresponding behavior data respectively; User instance aggregation subunit: used to merge the behavior data that is in the same grid index and meets the time overlap determination condition into a single user instance, and assign a unique number to this instance for subsequent reference by the feature package assembly subunit; by setting the grid division subunit to achieve unified spatial quantization and subdivision of the coverage area, combined with the second-level interval comparison of the time overlap determination subunit, the user data at the same location and with intersecting behavior time periods can be accurately aggregated. The user instance aggregation subunit further guarantees the consistency and integrity of the behavior data, thus significantly improving the subsequent feature fusion accuracy and operation efficiency.

[0026] The attention degree generation unit includes: Feature extraction subunit: used to extract feature parameters from the user behavior feature package, including user stay time , user interaction frequency and the label of the time period when the behavior occurs , where the unit is seconds, is the number of touches per unit time (times), represents the level encoding of the time period where the behavior is located; Weight loading subunit: used to read the corresponding weight coefficients of each feature from the preset weight configuration table according to the scene type to which the sign belongs, which are the stay time weight , interaction frequency weight , time period weight , where each weight coefficient is a dimensionless positive real number, satisfying ; Weighted calculation subunit: used to perform a linear weighted operation on the extracted feature parameters and the loaded weight coefficients to generate the final user attention degree index, and the calculation formula is: , where, represents the user attention degree index, and the unit is seconds (keeping the same dimension as to reflect the comprehensive stay behavior intensity of the user towards the sign); the above subunits combine different feature dimensions (stay, interaction, time period) with the preset scene weight strategy, and use the linear weighted method to generate a unified attention degree index, which not only realizes the unified evaluation of multiple behavior indicators, but also has good scene adaptability, providing quantitative support for subsequent user group screening and sign strategy optimization.

[0027] The behavior analysis module includes a threshold comparison unit, a group identification unit, a hot zone division unit, and a service boundary calculation unit; among them: The threshold comparison unit: is used to receive the user attention index output by the feature fusion module, compare it item by item with the attention threshold built in the behavior analysis module, and generate a determination label, where the determination label distinguishes between two states of high attention and normal attention; The specific determination formula is: ; Among them, is the user attention index; is the attention threshold, with the unit of second; is the user determination label, 1 indicates high attention, and 0 indicates normal attention; The group identification unit: is used to perform a connectivity scan on user instances that are in the same attention state and adjacent in spatial position according to the determination label output by the threshold comparison unit, and aggregate high-attention users in the continuous area into a high-attention user group; The hot zone division unit: is used to calculate the spatial aggregation density of the high-attention user group, divide the coverage area into multiple levels of hot zone grids, and classify the grids according to the number of high-attention users in each grid to obtain a set of hot spot grids; The service boundary calculation unit: is used to fit the circumscribed polygon of the set of hot spot grids, output the vertex coordinates of the fitted polygon as the effective service range boundary of the sign, and send the boundary coordinates to the policy generation module; through the above units, high-attention users can be quickly screened and the service boundary can be determined based on their aggregation density, which not only improves the accuracy of high-attention user group identification, but also accurately defines the effective service range of the sign, providing a reliable spatial basis for subsequent policy optimization.

[0028] The hot zone division unit includes: The density calculation sub-unit: is used to traverse and count the user coordinates within each high-attention user group, and calculate the spatial aggregation density within the basic grid with a side length of The calculation formula is: , where Among them, represents the number of high-attention users (dimensionless) located within the grid number ; represents the area of a single basic grid, with the unit of square meter; is the spatial aggregation density of the basic grid, with the unit of person / square meter; is the grid side length; The grid classification sub-unit: based on the grid density perform multi-level grid subdivision processing on the coverage area, and continue to use binary refinement of the side length within the basic grid where the density is higher than the threshold ​ Divide the dense sub - grids to form a multi - level hot - zone grid system; Heat - level calibration sub - unit: Used to classify and calibrate the heat levels of all grids (basic grids and sub - grids) according to the number of highly - concerned users inside. The heat levels are divided into three levels: first - level heat, second - level heat, and third - level heat according to the threshold range; specifically expressed as: ; Among them, is the threshold of the number of highly - concerned users, both are dimensionless positive integers; is the heat level, taking values ; Hot - spot extraction sub - unit: Used to collect the grids with heat level , combine to obtain the hot - spot grid set, and pass the set index to the service - boundary calculation sub - unit; The above - mentioned sub - units can accurately depict the degree of user aggregation on the premise of ensuring dimensional consistency through four - level processing processes of density calculation, hierarchical refinement, heat - level calibration, and hot - spot extraction, realize the visualization of multi - scale heat distribution in the coverage area, and provide high - resolution hot - spot data support for subsequent service - boundary fitting.

[0029] The policy - generation module includes a region - priority division unit, an orientation - optimization unit, and a collaborative - control unit; Among them: Region - priority division unit: Used to receive the set of service - boundary vertex coordinates and the real - time distribution coordinates of the highly - concerned user group, divide the boundary area into multi - level information - push priority regions according to the minimum distance from the user location to the service boundary, and assign corresponding priority identifiers to each region; Orientation - optimization unit: Based on the coordinates of all highly - concerned users in each priority region, calculate the centroid coordinates of the region, and take the sign - board installation center point as the rotation center, and determine the optimal orientation angle for each region according to the following formula: , where, is the optimal orientation angle corresponding to the - level push region; and are the abscissa and ordinate of the centroid of highly - concerned users in the - level push - priority region, with the unit of meter; is the coordinate of the sign - board installation position; is the two - variable arctangent function; Collaborative - control sub - unit: Used in the scenario of multi - sign - board deployment, generate the sign - board content update timing sequence and communication relay strategy according to the service - boundary overlap relationship and push - region priority of each sign - board, including: Determine the content - push order according to the region priority; Allocate mutually - backup communication relay parameters for adjacent sign - boards to ensure the sequential transmission of key - area information between the main sign - board and the backup sign - board; Output a multi-sign board collaborative control parameter set for driving by the execution control module.

[0030] The area priority division unit includes: The distance calculation sub-unit: For the spatial coordinates of each highly concerned user, calculate the minimum Euclidean distance on the entire edge set of the service scope boundary polygon. The calculation formula is: , where is the user coordinate, in meters; is the coordinate of the point on the boundary polygon closest to the user, in meters; is the minimum distance from the user to the service boundary, in meters; The ring division sub-unit: For the output by the distance calculation sub-unit and the preset distance threshold compare, and divide the service scope into three-level ring areas in the form of concentric rings from the boundary inward. Specifically, The first-level information push area satisfies ; The second-level information push area satisfies ; The third-level information push area satisfies ; The above-mentioned sub-units can quickly generate hierarchical push areas based on the spatial distance from the user to the boundary, realize the fine control of the information delivery priority, and improve the sign board resource allocation efficiency and user reception experience.

[0031] The execution control module includes a policy analysis unit, a display control unit, an attitude drive unit, and an information synchronization unit; among them: The policy analysis unit: Used to receive the optimized policy set output by the policy generation module, extract the information push priority area data, the optimal orientation angle instruction of the sign board, and the collaborative control parameters contained therein, and perform structured analysis according to the execution order, and output to the corresponding control unit; The display control unit: Used to screen the content push list sent by the sign board main control system according to the priority area identifier contained in the policy analysis unit, and push the corresponding content to the display screen in the order from high to low priority, ensuring that the users closest to the inner circle of the boundary receive the target information content first; The attitude drive unit: Used to generate a rotation control instruction according to the analyzed optimal orientation angle, and drive the built-in attitude adjustment motor to rotate the sign board body to the specified direction angle to achieve directional delivery to the target user group; Information synchronization unit: It is used to establish a master-backup sign communication link according to the collaborative control parameters, schedule the communication repeater to synchronously forward the content update and attitude adjustment instructions according to the predetermined sequence, and realize the multi-sign information collaboration; through the above unit, the complete link conversion from strategy to action can be realized, so that the sign can accurately publish the content, intelligently adjust the attitude and synchronously cooperate with multiple devices after receiving the optimized strategy, thereby ensuring the unified effectiveness of information delivery in the dimensions of space, time and users.

[0032] The present invention covers any alternatives, modifications, equivalent methods and solutions made on the essence and scope of the present invention. In order to enable the public to have a thorough understanding of the present invention, specific details are described in detail in the following preferred embodiments of the present invention, and those skilled in the art can fully understand the present invention without the description of these details. In addition, well-known methods, processes, procedures, components and circuits are not described in detail to avoid unnecessary confusion to the essence of the present invention.

[0033] The above are only the preferred embodiments of the present invention. It should be pointed out that for those of ordinary skill in the art, without departing from the principle of the present invention, several improvements and refinements can be made, and these improvements and refinements should also be regarded as the protection scope of the present invention.

Claims

1. A smart sign management system based on data analysis, characterized in that, It includes a data acquisition module, a feature fusion module, a behavior analysis module, a policy generation module, and an execution control module; among which: The data acquisition module: is used to collect the residence time data and interaction frequency data of users in the target area covered by the sign in real time through the infrared thermal imaging sensor array and the touch interaction panel deployed on the sign body; The feature fusion module: is used to receive the residence time data and interaction frequency data, generate a user behavior feature package through spatio-temporal correlation matching, calculate the corresponding weight coefficient according to the scene type to which the sign belongs, and then generate a user attention index through weighted fusion; The behavior analysis module: is used to identify the high-attention user group according to the comparison result between the user attention index and the preset threshold, and calculate the boundary of the effective service range of the sign by using the grid hot zone division method based on its spatial aggregation density; The policy generation module: generates an optimization policy set including multi-level information push priority areas, the optimal orientation angle of the sign, and multi-sign collaborative control parameters based on the boundary coordinates of the effective service range and the real-time distribution coordinates of the high-attention user group; The execution control module: is used to parse the optimization policy set, and control the sign content update, attitude adjustment, and information collaboration through the Internet of Things terminal to realize information delivery.

2. The intelligent sign management system based on data analysis according to claim 1, wherein The data acquisition module includes an infrared image acquisition unit, a trajectory recognition unit, a touch event recording unit, and a data synchronization unit; among which: The infrared image acquisition unit: is used to continuously obtain the thermal image frame data in the area covered by the sign through the infrared thermal imaging sensor array, and capture the two-dimensional coordinates of the user's body heat source in real time based on the set frame rate; Trajectory recognition unit: used to perform time-series tracking on the heat source points in the infrared image frame, judge whether each user is in a stationary state by calculating the continuous displacement vector of each user in the image coordinate system, and calculate the residence time of the user at the specified position according to the number of frames stayed and the image frame rate ; Touch event recording unit: used to monitor each touch event on the touch interaction panel and record the corresponding timestamp to count the interaction frequency within a given time window ; Data synchronization unit: used to align the sojourn time with the corresponding touch frequency for timestamp alignment and pack them to generate a structured original interaction behavior record.

3. The intelligent sign management system based on data analysis according to claim 2, wherein, The feature fusion module includes a spatial aggregation unit, a feature package assembly unit, a weight coefficient calculation unit, and an attention generation unit; among which: The spatial aggregation unit: divides grid cells based on the user coordinate information in the original interaction behavior record, and aggregates the behavior data located in the same grid and with overlapping time periods into the same user instance; The feature package assembly unit: is used to extract the residence time, interaction frequency, and time period label from each user instance, and encapsulate them into a user behavior feature package; The weight coefficient calculation unit: is used to call the preset scene weight table according to the scene type to which the sign belongs, and assign definite weight coefficients to the residence time, interaction frequency, and time period label respectively. The scene types include public transportation station scenes, shopping mall advertising space scenes, and indoor wayfinding scenes; The attention generation unit: performs weighted fusion on the user behavior feature package based on the corresponding weight coefficient to generate the corresponding user attention index.

4. The intelligent sign management system based on data analysis according to claim 3, wherein The spatial aggregation unit includes: Grid division sub-unit: used to equally divide the coverage area at a fixed grid side length with the sign reference origin as the benchmark, and calculate the grid index where any user coordinate is located ; Time overlap determination sub-unit: used for two pieces of behavior data belonging to the same grid index to compare their recorded time intervals and When is satisfied, it is determined that there is an effective time overlap between the two pieces of behavior data; where are the start and end timestamps of the corresponding behavior data respectively; The user instance aggregation subunit: is used to merge the behavior data in the same grid index and meeting the time overlap determination condition into a single user instance.

5. The intelligent sign management system based on data analysis according to claim 4, characterized in that, The attention generation unit includes: Feature extraction sub-unit: used to extract feature parameters from the user behavior feature package, including user stay time , user interaction frequency and the time period label when the behavior occurs ; Weight loading subunit: configured to read the corresponding weight coefficient of each feature from a preset weight configuration table according to the scene type to which the sign belongs, including the sojourn time weight , interaction frequency weight , time period weight , satisfying ; Weighted calculation subunit: used to perform a linear weighted operation on the extracted feature parameters and the loaded weight coefficients to generate the final user attention index. The calculation formula is: , where represents the user attention index.

6. The intelligent sign management system based on data analysis according to claim 1, wherein The behavior analysis module includes a threshold comparison unit, a group recognition unit, a hot zone division unit, and a service boundary calculation unit; among which: The threshold comparison unit: is used to receive the user attention index output by the feature fusion module, compare it item by item with the attention threshold built in the behavior analysis module, and generate a determination label. The determination label distinguishes two states: high attention and normal attention; Group identification unit: Used to perform connectivity scanning on user instances in the same attention state and adjacent in spatial position according to the determination tags output by the threshold comparison unit, and aggregate high-attention users in the continuous area into high-attention user groups; Hot zone division unit: Used to calculate the spatial aggregation density of high-attention user groups, divide the coverage area into multi-level hot zone grids, and classify the grids according to the number of high-attention users in each grid to obtain a set of hot spot grids; Service boundary calculation unit: Used to fit the circumscribed polygon of the set of hot spot grids, and output the vertex coordinates of the fitted polygon as the effective service range boundary of the sign; 7. The intelligent sign management system based on data analysis according to claim 6, characterized in that The hot zone division unit includes: Density calculation subunit: used to traverse and count the user coordinates within each highly concerned user group, and calculate the spatial aggregation density within the basic grid with a side length of ; ; Grid grading sub-unit: Based on grid density Perform multi-level grid subdivision processing on the covered area. Within the basic grid with a density higher than the threshold Continue to use binary refinement of the side length to divide dense sub-grids and form a multi-level hot zone grid system; Heat level calibrating sub-unit: Used to classify and calibrate the heat levels of all grids according to the number of high-attention users inside, and the heat levels are divided into three levels: primary heat level, secondary heat level and tertiary heat level according to the threshold range; Specifically expressed as: ； Among them, is the threshold value of the number of highly concerned users; is the heat level, and the value is ; Hotspot extraction subunit: used to collect grids with heat levels and combine them to obtain a set of hotspot grids.

8. An intelligent sign management system based on data analysis according to claim 1, characterized in that, The strategy generation module includes a regional priority division unit, an orientation optimization unit, and a collaborative control unit; among them: Regional priority division unit: Used to receive the set of service boundary vertex coordinates and the real-time distribution coordinates of high-attention user groups, divide the boundary area into multi-level information push priority areas according to the minimum distance from the user location to the service boundary, and assign corresponding priority identifiers to each area; Orientation optimization unit: Based on the coordinates of all highly concerned users in each priority area, calculate the centroid coordinates of the area, and take the sign installation center point as the rotation center. For each area, determine the optimal orientation angle according to the following formula: , where is the optimal orientation angle corresponding to the -level push area; and are the horizontal and vertical coordinates of the centroid of highly concerned users in the -level push priority area, in meters; is the coordinate of the sign installation position; is the two-variable arctangent function; Collaborative control sub-unit: Used in the scenario of multi-sign deployment, generate the sign content update timing and communication relay strategy according to the service boundary overlap relationship and push area priority of each sign; 9. The intelligent sign management system based on data analysis according to claim 8, characterized in that The regional priority division unit includes: Distance calculation sub-unit: For the spatial coordinates of each highly concerned user, calculate the minimum Euclidean distance on the entire edge set of the service scope boundary polygon The calculation formula is as follows: , where is the user coordinate; is the coordinate of the point on the boundary polygon closest to the user; is the minimum distance from the user to the service boundary; Ring division sub-unit: used to calculate the output of the sub-unit according to the distance and a preset distance threshold for comparison, and divide the service area into three-level ring areas in the form of concentric rings from the boundary inward. Specifically, The first-level information push area satisfies ; The second-level information push area, satisfying ; The third-level information push area satisfies .

10. A smart sign management system based on data analysis according to claim 1, characterized in that, The execution control module includes a strategy parsing unit, a display control unit, an attitude drive unit, and an information synchronization unit; among them: Strategy parsing unit: Used to receive the optimized strategy set output by the strategy generation module, extract the information push priority area data, the optimal sign orientation angle instruction, and the collaborative control parameters contained therein, and perform structured parsing in the execution order, and output to the corresponding control unit; Display control unit: Used to screen the content push list sent by the sign main control system according to the priority area identifier contained in the strategy parsing unit, and push the corresponding content to the display screen in the order from high to low priority; Attitude drive unit: Used to generate a rotation control instruction according to the parsed optimal orientation angle, and drive the built-in attitude adjustment motor to rotate the sign body to the specified direction angle; Information synchronization unit: Used to establish a master-backup sign communication link according to the collaborative control parameters, and schedule the communication relay to synchronously forward the content update and attitude adjustment instructions according to the predetermined timing.

Images