Factory personnel tracking management method and system driven by Bluetooth positioning

Abstract

The invention discloses a factory personnel tracking management method and system driven by Bluetooth positioning, and relates to the technical field of personnel tracking, and the method comprises the steps: recording the spatial positioning at the disconnection moment when a Bluetooth device worn by a factory personnel is disconnected, and automatically tracking the movement track of the Bluetooth device; after the Bluetooth is reconnected, the mobile data during the disconnection period is processed, and a coordinate sequence is generated through the Bluetooth blind area positioning network, so that the accurate tracking management of the personnel position during the disconnection period is realized. The technical problem that an existing factory personnel real-time positioning system based on Bluetooth cannot continuously and accurately track the position of the personnel in the Bluetooth connection disconnection period is solved, and the technical effect of continuously and accurately tracking the position of the factory personnel in the Bluetooth connection disconnection period is achieved.

Description

Technical Field

[0001] The present application relates to the technical field of personnel tracking, and in particular to a factory personnel tracking management method and system driven by Bluetooth positioning. Background Art

[0002] As factories continue to expand and production processes become increasingly complex, the demand for real-time positioning and management of personnel is also increasing. In many manufacturing and industrial scenarios, by having factory employees wear Bluetooth positioning devices, their location and movement trajectory can be dynamically monitored, thereby improving the orderliness and safety of production. However, due to the complex internal structure of the factory, the large number of mechanical equipment, and severe signal obstruction, Bluetooth signal transmission often faces obstacles and interference, which makes the connection between Bluetooth and the server easy to interrupt. During the period of disconnection, the specific location and movement path of employees are difficult to capture accurately, which has a potential impact on the factory's production scheduling, personnel deployment and safety management. The limitations of traditional positioning methods have become more and more obvious in this context. How to maintain effective tracking of personnel locations during periods when Bluetooth signals are unstable or even disconnected has become a technical problem that needs to be solved urgently.

[0003] In the current relevant technologies, there is a technical problem that the Bluetooth-based real-time positioning system for factory personnel cannot continue to accurately track the location of personnel during the period when the Bluetooth connection is disconnected. Summary of the invention

[0004] This application solves the technical problem that the existing Bluetooth-based factory personnel real-time positioning system cannot continuously and accurately track the location of personnel during the period when the Bluetooth connection is disconnected by providing a factory personnel tracking management method and system driven by Bluetooth positioning.

[0005] This application provides a factory personnel tracking management method driven by Bluetooth positioning, including:

[0006] When the first employee positioning Bluetooth is disconnected, the spatial positioning at the time of disconnection is obtained, wherein the first employee positioning Bluetooth will automatically record the spatial movement trajectory with the spatial positioning at the time of disconnection as the origin after disconnection; continuously request to reconnect with the first employee positioning Bluetooth, and when the reconnection is successful, obtain the Bluetooth blind spot movement trajectory and the spatial positioning at the time of successful reconnection from the first employee positioning Bluetooth; according to the factory workshop number, activate the Bluetooth blind spot positioning network to process the spatial positioning at the time of disconnection, the Bluetooth blind spot movement trajectory and the spatial positioning at the time of successful reconnection, and obtain the Bluetooth blind spot positioning coordinate sequence; according to the Bluetooth blind spot positioning coordinate sequence, perform the first employee blind spot coordinate compensation tracking management.

[0007] This application also provides a factory personnel tracking management system driven by Bluetooth positioning, including:

[0008] A spatial positioning acquisition module, wherein the spatial positioning acquisition module is used to obtain the spatial positioning at the time of disconnection when the first employee positioning Bluetooth is disconnected, wherein the first employee positioning Bluetooth will automatically record the spatial movement trajectory with the spatial positioning at the time of disconnection as the origin after disconnection; a reconnection module, wherein the reconnection module is used to continuously request reconnection with the first employee positioning Bluetooth, and when the reconnection is successful, obtain the Bluetooth blind spot movement trajectory and the spatial positioning at the time of successful reconnection from the first employee positioning Bluetooth; a positioning coordinate sequence acquisition module, wherein the positioning coordinate sequence acquisition module is used to activate the Bluetooth blind spot positioning network according to the factory workshop number to process the spatial positioning at the time of disconnection, the Bluetooth blind spot movement trajectory and the spatial positioning at the time of successful reconnection, and obtain the Bluetooth blind spot positioning coordinate sequence; a coordinate compensation tracking management module, wherein the coordinate compensation tracking management module is used to perform the first employee blind spot coordinate compensation tracking management according to the Bluetooth blind spot positioning coordinate sequence.

[0009] The Bluetooth positioning-driven factory personnel tracking management method and system proposed in this application first records the spatial positioning at the time of disconnection and automatically tracks the movement trajectory when the Bluetooth device worn by the factory personnel is disconnected; when the Bluetooth is reconnected, the movement data during the disconnection period is processed, and a coordinate sequence is generated through the Bluetooth blind spot positioning network, thereby realizing accurate tracking and management of the personnel position during the disconnection period, achieving the technical effect of continuously and accurately tracking the position of factory personnel during the period when the Bluetooth connection is disconnected. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] In order to more clearly illustrate the technical solution of the embodiment of the present invention, the accompanying drawings of the embodiment of the present invention will be briefly introduced below. A flow chart is used in the present application to illustrate the operations performed by the system according to the embodiment of the present application. It should be understood that the preceding or following operations are not necessarily performed accurately in order. On the contrary, various steps can be processed in reverse order or simultaneously as needed. At the same time, other operations can also be added to these processes, or one or more operations can be removed from these processes.

[0011] Figure 1 A flowchart of a factory personnel tracking management method driven by Bluetooth positioning provided in an embodiment of the present application;

[0012] Figure 2 A schematic diagram of the structure of a factory personnel tracking management system driven by Bluetooth positioning provided in an embodiment of the present application.

[0013] Description of reference numerals: spatial positioning acquisition module 10 , reconnection module 20 , positioning coordinate sequence acquisition module 30 , coordinate compensation tracking management module 40 . DETAILED DESCRIPTION

[0014] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below.

[0015] In order to make the objectives, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings. The described embodiments should not be regarded as limiting the present application. All other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of this application.

[0016] In the following description, reference is made to "some embodiments", which describe a subset of all possible embodiments, but it is understood that "some embodiments" may be the same subset or different subsets of all possible embodiments, and may be combined with each other without conflict, and the terms "first\second" involved are merely to distinguish similar objects and do not represent a specific ordering of objects. The terms "including" and "having" and any variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or modules that are not clearly listed or inherent to these processes, methods, products, or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those generally understood by technicians in the technical field of this application. The terms used herein are for the purpose of describing the embodiments of the present application only.

[0017] The present application embodiment provides a factory personnel tracking management method driven by Bluetooth positioning, such as Figure 1 As shown, the method includes:

[0018] Step S100, when the first employee positioning Bluetooth is disconnected, the spatial positioning at the time of disconnection is obtained, wherein the first employee positioning Bluetooth will automatically record the spatial movement trajectory with the spatial positioning at the time of disconnection as the origin after the disconnection. Specifically, when the first employee positioning Bluetooth is disconnected from the server, the system continues to receive its real-time positioning information until it is interrupted, so as to accurately determine the spatial positioning at the time of disconnection, and this positioning becomes the key benchmark. Subsequently, the first employee positioning Bluetooth starts the automatic recording mechanism, with the positioning as the origin. Its acceleration sensor captures the acceleration changes in multi-directional motion and the speed information after integral calculation, and the gyroscope measures the rotation angle and direction change in three-dimensional space. The two work together, and the internal positioning algorithm of Bluetooth is comprehensively processed, and mathematical tools are used to decompose and synthesize the speed information, and the displacement components are calculated according to the time interval. The coordinates are updated with the displacement accumulated at the origin and connected in series to form a complete spatial movement trajectory data, which is temporarily stored in the Bluetooth device and transmitted to the system after reconnection, providing a solid basis for restoring the employee's offline action path and ensuring the tracking accuracy and integrity.

[0019] Step S200, continuously request to reconnect with the first employee positioning Bluetooth, and when the reconnection is successful, obtain the Bluetooth blind spot movement trajectory and the spatial positioning at the time of successful reconnection from the first employee positioning Bluetooth. Specifically, after the first employee positioning Bluetooth is disconnected from the server, the system starts the reconnection mechanism, and continuously sends a reconnection request data packet containing an identification code and a connection instruction to it at a preset time interval (such as every 1 second or depending on the system's real-time requirements and device performance). Once the reconnection is successful, the system immediately obtains the Bluetooth blind spot movement trajectory data recorded with the spatial positioning at the time of disconnection as the origin during the Bluetooth connection interruption from the Bluetooth device, as well as the spatial positioning information at the time of successful reconnection. This information helps to accurately connect the Bluetooth blind spot movement trajectory with the real-time positioning trajectory after reconnection, thereby constructing a complete action route map for employees and providing comprehensive and accurate data support for all aspects of factory management.

[0020] Step S300, according to the factory workshop number, activate the Bluetooth blind spot positioning network to process the spatial positioning at the time of disconnection, the movement trajectory of the Bluetooth blind spot, and the spatial positioning at the time of successful reconnection, and obtain a Bluetooth blind spot positioning coordinate sequence. Specifically, according to the factory workshop number, the system activates the corresponding Bluetooth blind spot positioning network, and has previously obtained data such as spatial positioning at the time of disconnection, movement trajectory of the Bluetooth blind spot, and spatial positioning at the time of successful reconnection. After activation, the first to Nth Bluetooth blind spot positioning base networks receive data and analyze and calculate based on their own algorithm models that have been optimized and trained for specific areas or scenes in the workshop. Taking the first Bluetooth blind spot positioning base network as an example, its algorithm model is constructed based on the precise mapping and digital modeling of the workshop space layout, dividing the workshop into many tiny sub-areas, and each sub-area is assigned specific spatial coding and signal propagation characteristic parameters. For obstacle distribution, the location, shape and size information of the obstacles are pre-acquired through laser scanning or other high-precision mapping means, and the information is integrated into the algorithm model. In terms of Bluetooth signal propagation characteristics, based on a large amount of field test data, a signal strength attenuation model is constructed, and a multipath propagation correction algorithm is established taking into account the reflection, refraction and absorption effects of metal equipment, walls, etc. in the workshop. When the data is received, the sub-area where the starting point of the Bluetooth blind spot moving trajectory is located is first determined based on the spatial coding, and the approximate position range of the starting point is preliminarily calculated by combining the signal propagation characteristic parameters and the currently received Bluetooth signal strength information. By analyzing the speed change of the Bluetooth blind spot moving trajectory, the displacement vector between adjacent time points is calculated using the kinematic principle. For steering information, the steering angle and direction of the trajectory are determined by combining the gyroscope data (if any) or by analyzing the direction change of the displacement vector at different time points. Then, based on the relative position relationship with the workshop facilities, for example, when the trajectory is close to large metal equipment, the signal anomaly caused by equipment interference is compensated and corrected based on the signal attenuation model and the multipath propagation correction algorithm, so as to accurately calculate the position coordinates of each point on the trajectory, and finally generate the corresponding positioning coordinate sequence. Similarly, the Nth Bluetooth blind spot positioning base network is also processed according to a similar process and its own unique algorithm model to generate the Nth Bluetooth blind spot positioning coordinate sequence. The fully connected network then evaluates the sequence distance parameters of these sequences pairwise, performs outlier factor analysis based on this, finds out abnormal sequences that are greatly different from other sequences, and finally extracts the positioning coordinate sequence corresponding to the minimum value of the outlier factor as the Bluetooth blind spot positioning coordinate sequence, providing a reliable data basis for employee blind spot coordinate compensation tracking management.

[0021] In a possible implementation, according to the factory workshop number, the Bluetooth blind spot positioning network is activated to process the spatial positioning at the time of disconnection, the Bluetooth blind spot movement trajectory, and the spatial positioning at the time of successful reconnection, and a Bluetooth blind spot positioning coordinate sequence is obtained. Step S300 further includes step S310, and the Bluetooth blind spot positioning network includes the first Bluetooth blind spot positioning base network until the Nth Bluetooth blind spot positioning base network, and a fully connected network. Specifically, the Bluetooth blind spot positioning network is composed of the first Bluetooth blind spot positioning base network until the Nth Bluetooth blind spot positioning base network and the fully connected network. The first to Nth Bluetooth blind spot positioning base networks process positioning data for specific areas or scenarios in the factory workshop. For example, the first Bluetooth blind spot positioning base network divides the open production area into grid units, records the coordinates and signal strength reference values, and estimates the moving distance and direction according to the grid signal strength, fluctuation and kinematic formula of the trajectory starting point after receiving the data, and then generates a positioning coordinate sequence in combination with the position correction results of large equipment and other facilities; the second Bluetooth blind spot positioning base network focuses on the narrow channel area, constructs a one-dimensional coordinate system, and calculates the speed and distance generation sequence based on the propagation characteristics of the signal in the narrow space and the intensity change rate; the other base networks are similar. The fully connected network compares the positioning coordinate sequences output by each base network in pairs, quantifies the differences by the distance difference of the coordinate points, the similarity of the trajectory shape, and then uses weighted average, data fitting and other methods to make comprehensive adjustments, weight the stable and small-difference coordinate points, and correct or discard the large deviations, so as to obtain a more accurate and reliable Bluetooth blind spot positioning coordinate sequence, providing strong data support for personnel tracking management.

[0022] Step S320: Process the spatial positioning at the time of disconnection, the movement trajectory of the Bluetooth blind area, and the spatial positioning at the time of successful reconnection, respectively, according to the first Bluetooth blind area positioning base network to the Nth Bluetooth blind area positioning base network, and obtain the first Bluetooth blind area positioning coordinate sequence to the Nth Bluetooth blind area positioning coordinate sequence. Specifically, for the first Bluetooth blind area positioning base network, after receiving the spatial positioning at the time of disconnection, the movement trajectory of the Bluetooth blind area, and the spatial positioning at the time of successful reconnection, the processing flow is started according to the algorithm system constructed internally. The algorithm system first accurately maps and digitally models the spatial geometry of the factory workshop, divides it into different area units and determines its spatial coordinate range and boundary conditions. At the same time, it constructs a detailed signal interference matrix through preliminary signal strength testing and propagation path simulation to address the impact of internal equipment and obstacles on Bluetooth signal propagation. For example, for large metal equipment, the signal attenuation coefficient, reflection angle and multi-path propagation path set are calculated based on its material, shape, size and relative position to the positioning point. Based on a large amount of historical data and field observations, a speed range model for employees' normal movements in the workshop is established, taking into account factors such as the average, maximum and minimum speed limits of employees in different work tasks and regional environments, as well as a common path pattern library. Through cluster analysis of employees' daily movement trajectories, typical path patterns such as from raw material storage areas to production and processing areas are summarized. When receiving the actual positioning data, the algorithm first determines the starting area unit based on the spatial positioning at the time of disconnection, and gradually calculates the approximate direction of the trajectory in each area unit by combining the signal strength change information in the Bluetooth blind spot moving trajectory and the signal interference matrix, and uses the speed range model to verify the time interval and distance change between adjacent points on the corrected trajectory, and optimizes the trajectory path selection with reference to the common path pattern library. After comprehensive analysis and complex calculations, the input data is converted into a series of preliminary positioning coordinate points and connected in time or logical order to form the first Bluetooth blind spot positioning coordinate sequence. Similarly, the Nth Bluetooth blind spot positioning base network also processes the same input data according to a similar process and algorithm framework, but due to differences in design and optimization direction, some focus on optimizing positioning accuracy in complex equipment-intensive areas, and some focus on rapid response and positioning stability optimization in high-traffic channel areas, resulting in different focuses on processing data, and then generating positioning coordinate sequences with their own characteristics.

[0023] Step S330, the first Bluetooth blind spot positioning coordinate sequence is merged to the Nth Bluetooth blind spot positioning coordinate sequence through the fully connected network to obtain the Bluetooth blind spot positioning coordinate sequence. Specifically, after the first Bluetooth blind spot positioning coordinate sequence to the Nth Bluetooth blind spot positioning coordinate sequence are generated. The fully connected network first performs a pairwise sequence distance evaluation on multiple positioning coordinate sequences, and obtains a series of sequence distance parameters by calculating the spatial distances of corresponding coordinate points between different sequences. The parameters can intuitively reflect the degree of similarity and difference between each positioning coordinate sequence. Based on the sequence distance parameters, the fully connected network conducts outlier factor analysis. Sequences with large differences in coordinate point positions, overall sequence trends, etc. compared with other sequences are identified. The sequence may be affected by factors such as local environmental interference or data anomalies during the processing of a certain base network and produce deviations. The fully connected network extracts the positioning coordinate sequence corresponding to the minimum value of the outlier factor of the first Bluetooth blind spot positioning coordinate sequence to the outlier factor of the Nth Bluetooth blind spot positioning coordinate sequence, and determines it as the final Bluetooth blind spot positioning coordinate sequence. The final sequence combines the advantages of each base network, eliminates possible abnormal situations, and can more accurately reflect the actual location change trajectory of employees during the Bluetooth blind area, providing a reliable positioning data basis for subsequent personnel tracking management.

[0024] In a possible implementation, the Bluetooth blind spot positioning network includes a first Bluetooth blind spot positioning base network until the Nth Bluetooth blind spot positioning base network, and a fully connected network, and step S310 further includes step S311, based on the factory workshop number, constructing a factory workshop three-dimensional model, and based on the preset spatial grid side length, constructing a factory workshop three-dimensional grid model of the factory workshop three-dimensional model. Specifically, according to the factory workshop number, an accurate factory workshop three-dimensional model is constructed using the factory's architectural design drawings, field measurement data, and three-dimensional modeling technology. The model can truly reflect the spatial structure, size, layout of internal facilities, and other information of the workshop. Based on the preset spatial grid side length, the factory workshop three-dimensional model is further divided into a factory workshop three-dimensional grid model. For example, if the preset spatial grid side length is 1 meter, then every 1 cubic meter of space in the workshop will be divided into an independent grid unit, and the grid units together constitute the three-dimensional grid model of the entire workshop, laying the basic framework for the subsequent construction of the neural network.

[0025] Step S312, configure each grid of the three-dimensional grid model of the factory workshop as an input node and an output node at the same time, the input information is 0 or 1, and the output information is 0 or 1, to obtain an initial graph neural network, wherein the grid input information and output information of the initial graph neural network are 0 by default. Specifically, for the constructed three-dimensional grid model of the factory workshop, each grid therein is specially configured so that it has the functions of both input node and output node, and the input information and output information are limited to only 0 or 1, to obtain the initial graph neural network. In the initial state, all grid input information and output information of the graph neural network are set to 0 by default. At this time, the neural network has not been trained and does not have the ability to effectively process employee location information.

[0026] Step S313, collecting the employee historical monitoring information of the three-dimensional grid model of the factory workshop, wherein the employee historical monitoring information includes the location record information at the time of disconnection, the Bluetooth blind area movement trajectory record information, the reconnection time location record information and the Bluetooth blind area positioning coordinate sequence identification data, and the Bluetooth blind area movement trajectory record information is the movement trajectory recorded with the location record information at the time of disconnection as the origin. Specifically, the employee historical monitoring information corresponding to the three-dimensional grid model of the factory workshop is collected. The information covers multiple key aspects, including the location record information of the employee positioning Bluetooth disconnection time, which accurately records the position coordinates of the employee at the moment of Bluetooth connection interruption; Bluetooth blind area movement trajectory record information, which takes the location record information at the time of disconnection as the origin, and records in detail the employee's movement path in the workshop during the Bluetooth connection interruption, such as the employee's movement distance in different directions, turning conditions, etc.; the location record information at the time of reconnection clarifies the employee's location when the Bluetooth reconnection is successful; and the Bluetooth blind area positioning coordinate sequence identification data, which is used for the subsequent training of the neural network to verify and supervise the output results.

[0027] Step S314, configure the 1-value input grid of the initial graph neural network according to the positioning record information at the time of disconnection, the Bluetooth blind spot movement trajectory record information, and the positioning record information at the time of reconnection, configure the 1-value output grid of the initial graph neural network with the Bluetooth blind spot positioning coordinate sequence identification data, as the 1-value output grid supervision data, train the initial graph neural network, and obtain the first Bluetooth blind spot positioning base network. Specifically, according to the collected positioning record information at the time of disconnection, the Bluetooth blind spot movement trajectory record information, and the positioning record information at the time of reconnection, the 1-value input grid is configured in the initial graph neural network. For example, if a Bluetooth disconnection event occurs to an employee in a specific grid area, the input information of the corresponding grid is set to 1 to indicate the event of the grid being associated with the employee's position. At the same time, the 1-value output grid of the initial graph neural network is configured with the Bluetooth blind spot positioning coordinate sequence identification data, and is used as the 1-value output grid supervision data. Then, a large number of employee historical monitoring information samples are used to train the initial graph neural network through a machine learning algorithm, such as a back propagation algorithm. During the training process, the weights and parameters of the neural network are continuously adjusted so that the network output gradually approaches the Bluetooth blind spot positioning coordinate sequence identification data. After multiple rounds of training and optimization, when the network output reaches a certain accuracy and stability requirement, the first Bluetooth blind spot positioning base network is successfully obtained.

[0028] Step S315, the construction steps of the second Bluetooth blind spot positioning base network until the Nth Bluetooth blind spot positioning base network are the same as the construction steps of the first Bluetooth blind spot positioning base network. Specifically, the construction steps of the second Bluetooth blind spot positioning base network until the Nth Bluetooth blind spot positioning base network are exactly the same as the construction steps of the first Bluetooth blind spot positioning base network. That is, first construct a three-dimensional grid model based on the factory workshop number and the preset spatial grid side length, initialize the graph neural network, collect employee historical monitoring information, and then configure the input and output grids according to the corresponding information and train them, and finally obtain independent Bluetooth blind spot positioning base networks with similar construction processes and functions. Multiple Bluetooth blind spot positioning base networks will jointly provide multi-dimensional analysis and processing capabilities for the determination of subsequent Bluetooth blind spot positioning coordinate sequences, thereby improving the accuracy and reliability of the entire Bluetooth blind spot positioning system.

[0029] In a possible implementation, the 1-value input grid of the initial graph neural network is configured according to the positioning record information at the disconnection time, the Bluetooth blind spot movement trajectory record information, and the positioning record information at the reconnection time, and the 1-value output grid of the initial graph neural network is configured with the Bluetooth blind spot positioning coordinate sequence identification data as the 1-value output grid supervision data, and the initial graph neural network is trained to obtain the first Bluetooth blind spot positioning base network. Step S314 further includes step S3141, constructing a blind spot positioning loss function: Among them, X1 represents the 1-value output grid supervision data of the Bluetooth blind spot positioning coordinate sequence identification data to configure the initial graph neural network, X 10 Characterize the 1-valued output grid of the initial graph neural network, n(X1∪X 10 ) represents X1 and X 10 The number of intersection 1-value grids, n(X1∩X 10 ) represents X1 and X 10 The number of 1-valued grids of the union, e represents the natural constant, and LOSS represents the blind spot positioning loss. Specifically, the blind spot positioning loss function is first constructed, and its expression is: In this function, X1 represents the 1-value output grid supervision data of the initial graph neural network configured by the Bluetooth blind spot positioning coordinate sequence identification data. It is the ideal result mode that we expect the neural network to output in the end. It is the grid identification information corresponding to the precise positioning coordinate sequence obtained based on a large amount of historical data and actual positioning conditions. 10 It represents the 1-value output grid actually generated by the initial graph neural network during the training process, that is, the current output result of the network. 10 ) is calculated for X1 and X 10 The number of grids with a union value of 1, which reflects the total number of grids involved in the comparison; n(X1∩X 10 ) calculates the number of grids with 1 values ​​in the intersection of the two, which represents the number of grids whose network output matches the ideal output. e is a natural constant. By constructing such a function, the difference between the network output and the ideal output can be quantified in the form of a loss value LOSS. When the network output is closer to the ideal output, the ratio of the number of grids with 1 values ​​in the intersection to the number of grids with 1 values ​​in the union is higher, the value of the exponential part is larger, and the loss value LOSS is smaller; conversely, the loss value is larger, which provides a clear optimization direction for subsequent neural network training.

[0030] Step S3142, based on the blind spot positioning loss function, configure the 1-value input grid of the initial graph neural network according to the positioning record information at the time of disconnection, the Bluetooth blind spot movement trajectory record information, and the positioning record information at the time of reconnection, and configure the 1-value output grid of the initial graph neural network with the Bluetooth blind spot positioning coordinate sequence identification data as the 1-value output grid supervision data, train the initial graph neural network, and obtain the first Bluetooth blind spot positioning base network. Specifically, the training of the initial graph neural network is carried out based on the constructed blind spot positioning loss function. According to the collected positioning record information at the time of disconnection, the grid position where the employee is located at the moment of Bluetooth connection disconnection is determined, and the input information of the grid in the initial graph neural network is configured as 1 to mark this position as the starting point related information of the employee's movement trajectory. According to the Bluetooth blind spot movement trajectory record information, the grid path that the employee passes through during the Bluetooth loss of connection is analyzed, and the grids involved in the path are configured as 1-value input grids in sequence according to their logical relationship in the trajectory to simulate the stimulation of the employee's movement process in the workshop on the input layer of the neural network. Then, based on the location record information at the time of reconnection, determine the grid where the employee is located when he reconnects to Bluetooth, and also make corresponding configurations in the input layer. Use the Bluetooth blind spot positioning coordinate sequence identification data to configure the 1-value output grid of the initial graph neural network as supervision data. During the training process, the actual output of the network and the supervision data are substituted into the blind spot positioning loss function to calculate the loss value, and then the back propagation algorithm is used to adjust the weights and connection parameters of the neural network according to the loss value, so that the loss value gradually decreases. After multiple iterations of training, the parameters of the network are continuously optimized until the output of the network can stably approach the Bluetooth blind spot positioning coordinate sequence identification data. At this time, the first Bluetooth blind spot positioning base network is successfully obtained. The network has the ability to accurately predict the Bluetooth blind spot positioning coordinate sequence based on the input information related to the disconnection and reconnection of the employee's Bluetooth connection, providing reliable basic positioning services for subsequent actual factory personnel positioning and tracking scenarios.

[0031] In a possible implementation, the first Bluetooth blind spot positioning coordinate sequence to the Nth Bluetooth blind spot positioning coordinate sequence are merged through the fully connected network to obtain the Bluetooth blind spot positioning coordinate sequence, and step S330 further includes step S331, configuring a fully connected rule factor to construct the fully connected network, wherein the fully connected rule factor is: performing a pairwise sequence distance evaluation on the first Bluetooth blind spot positioning coordinate sequence to the Nth Bluetooth blind spot positioning coordinate sequence to obtain a number of sequence distance parameters. Specifically, the fully connected rule factor is configured to construct a fully connected network. The core operation of the fully connected rule factor is to perform a pairwise sequence distance evaluation on the first Bluetooth blind spot positioning coordinate sequence to the Nth Bluetooth blind spot positioning coordinate sequence. For each two different Bluetooth blind spot positioning coordinate sequences, the distance difference between the corresponding coordinate points is calculated. For example, for a certain coordinate point in the first Bluetooth blind spot positioning coordinate sequence and a corresponding coordinate point in the second Bluetooth blind spot positioning coordinate sequence (the correspondence here can be based on a match in a time sequence or a spatial logical sequence), the distance value between them is calculated by a specific distance calculation formula, such as the Euclidean distance formula. After calculating all such pairwise combinations, several sequence distance parameters can be obtained. The parameters constitute the basic data for subsequent analysis and can intuitively reflect the similarity and difference between different Bluetooth blind spot positioning coordinate sequences. The distance parameter value of the sequence with high similarity is relatively small, and vice versa.

[0032] Step S332, based on the several sequence distance parameters, traverse the first Bluetooth blind spot positioning coordinate sequence until the Nth Bluetooth blind spot positioning coordinate sequence to perform outlier factor analysis, and obtain the outlier factor of the first Bluetooth blind spot positioning coordinate sequence until the Nth Bluetooth blind spot positioning coordinate sequence. Specifically, based on the several sequence distance parameters obtained previously, start traversing the first Bluetooth blind spot positioning coordinate sequence until the Nth Bluetooth blind spot positioning coordinate sequence to perform outlier factor analysis. For each Bluetooth blind spot positioning coordinate sequence, analyze the distance relationship between it and other sequences. The calculation of the outlier factor will comprehensively consider the distance parameter of the sequence and all other sequences. For example, if the distance parameter of a Bluetooth blind spot positioning coordinate sequence is significantly larger than that of most other sequences, that is, it has a low consistency with other sequences in spatial position or trajectory morphology, then its outlier factor will be higher; conversely, if the distance parameter of a sequence with other sequences is relatively balanced and small, its outlier factor will be lower. Through the analysis process, the outlier factors of the first Bluetooth blind area positioning coordinate sequence to the Nth Bluetooth blind area positioning coordinate sequence can be obtained, and the outlier factor can quantify the abnormality of each sequence in the entire sequence set.

[0033] Step S333, extract the positioning coordinate sequence of the minimum value of the outlier factor of the first Bluetooth blind spot positioning coordinate sequence until the outlier factor of the Nth Bluetooth blind spot positioning coordinate sequence, and output it as the Bluetooth blind spot positioning coordinate sequence. Specifically, after obtaining the outlier factor of the first Bluetooth blind spot positioning coordinate sequence until the outlier factor of the Nth Bluetooth blind spot positioning coordinate sequence, extract the positioning coordinate sequence corresponding to the minimum value. The sequence corresponding to the minimum value has the smallest difference with other sequences among all sequences, that is, it is the most consistent with the overall trend and most sequence characteristics. Output it as the final Bluetooth blind spot positioning coordinate sequence. The sequence has been fused and optimized by the fully connected network, and can eliminate abnormal positioning information caused by factors such as local environmental interference or individual Bluetooth blind spot positioning base network errors to the greatest extent, thereby providing a more reliable and accurate positioning coordinate data basis for subsequent employee blind spot coordinate compensation tracking management, ensuring that the actual action trajectory of employees during the Bluetooth blind spot can be accurately restored in the factory personnel tracking management system, and improving the positioning accuracy and stability of the entire system.

[0034] In a possible implementation, based on the blind spot positioning loss function, the 1-value input grid of the initial graph neural network is configured according to the positioning record information at the time of disconnection, the Bluetooth blind spot movement trajectory record information, and the positioning record information at the time of reconnection, and the 1-value output grid of the initial graph neural network is configured with the Bluetooth blind spot positioning coordinate sequence identification data, as the 1-value output grid supervision data, the initial graph neural network is trained to obtain the first Bluetooth blind spot positioning base network, and step S3142 further includes step S31421, based on the preset trajectory length, the Bluetooth blind spot movement trajectory record information is trimmed to obtain the Bluetooth blind spot movement trajectory pruning information, wherein the Bluetooth blind spot movement trajectory pruning information is a non-continuous trajectory. Specifically, the Bluetooth blind spot movement trajectory record information is trimmed according to the preset trajectory length. The preset trajectory length is determined based on a combination of factors such as the actual size of the factory workshop, the effective transmission range of the Bluetooth signal, and the normal movement speed range of employees in the workshop. During the clipping process, some key trajectory points are selected from the original Bluetooth blind spot movement trajectory record information according to certain time or space interval standards, so that the final Bluetooth blind spot movement trajectory pruning information becomes a discontinuous trajectory. For example, if the preset trajectory length is to select a trajectory point every 3 meters, then the original continuous Bluetooth blind spot movement trajectory will be simplified into a discontinuous trajectory composed of a series of discrete trajectory points spaced 3 meters apart. The purpose of this is to reduce the amount of data, highlight the key features of the trajectory, reduce noise interference in the data, and improve the efficiency and accuracy of subsequent neural network training.

[0035] Step S31422, based on the blind spot positioning loss function, the 1-value input grid of the initial graph neural network is configured according to the positioning record information at the time of disconnection, the pruning information of the Bluetooth blind spot movement trajectory, and the positioning record information at the time of reconnection, and the 1-value output grid of the initial graph neural network is configured with the Bluetooth blind spot positioning coordinate sequence identification data as the 1-value output grid supervision data, and the initial graph neural network is trained to obtain the first Bluetooth blind spot positioning base network. Specifically, the initial graph neural network is trained based on the blind spot positioning loss function constructed previously. According to the positioning record information at the time of disconnection, the grid corresponding to the position of the employee at the moment of Bluetooth disconnection is set as the 1-value input grid in the input layer of the initial graph neural network to mark the starting point of the trajectory. Using the Bluetooth blind spot movement trajectory pruning information, the grids corresponding to each discrete trajectory point are set as the 1-value input grid in sequence according to their logical order in the trajectory, simulating the stimulation of the employee's movement path during the Bluetooth blind spot on the neural network. Then, according to the positioning record information at the time of reconnection, the grid where the employee is when he reconnects to Bluetooth is also set as the 1-value input grid. At the same time, the 1-value output grid of the initial graph neural network is configured with the Bluetooth blind spot positioning coordinate sequence identification data, and it is used as the 1-value output grid supervision data. During the training process, the actual output of the network and the supervision data are substituted into the blind spot positioning loss function to calculate the loss value, and the back propagation algorithm is used to adjust the weights and connection parameters of the neural network according to the loss value. Through multiple iterative training, the network parameters are continuously optimized until the network output result can stably approach the Bluetooth blind spot positioning coordinate sequence identification data. At this time, the first Bluetooth blind spot positioning base network is successfully trained. The network can accurately predict the Bluetooth blind spot positioning coordinate sequence based on the input related positioning information, providing strong support for the positioning and tracking of factory personnel in Bluetooth blind spots.

[0036] In a possible implementation, according to the factory workshop number, the Bluetooth blind spot positioning network is activated to process the spatial positioning at the time of disconnection, the Bluetooth blind spot movement trajectory and the spatial positioning at the time of successful reconnection to obtain a Bluetooth blind spot positioning coordinate sequence, and step S300 further includes step S340, obtaining the trajectory start positioning and trajectory end positioning of the Bluetooth blind spot movement trajectory. Specifically, the trajectory start positioning and trajectory end positioning of the Bluetooth blind spot movement trajectory are obtained. The two positioning information are key elements for judging the integrity and accuracy of the Bluetooth blind spot movement trajectory. By extracting and parsing the relevant data recorded by the employee positioning Bluetooth when the connection is disconnected and reconnected successfully, the specific coordinate values ​​of the trajectory start positioning and the trajectory end positioning are determined.

[0037] Step S350, when the starting position of the trajectory is the same as the spatial position at the time of disconnection, and the ending position of the trajectory is the same as the spatial position at the time of successful reconnection, and the Bluetooth blind spot movement trajectory is a continuous trajectory, the Bluetooth blind spot movement trajectory is positioned based on the spatial position at the time of disconnection and the spatial position at the time of successful reconnection, and the Bluetooth blind spot movement trajectory is obtained. Specifically, when the starting position of the trajectory is exactly the same as the spatial position at the time of disconnection obtained previously, and the ending position of the trajectory is also consistent with the spatial position at the time of successful reconnection, and the Bluetooth blind spot movement trajectory presents a continuous trajectory, it means that the Bluetooth blind spot movement trajectory has good consistency and accuracy in time and space. At this time, it is relatively simple and direct to perform positioning processing on the Bluetooth blind spot movement trajectory based on the spatial position at the time of disconnection and the spatial position at the time of successful reconnection. Since the starting and ending points of the trajectory are clear and continuous, it can be regarded as a complete and reliable movement path. The Bluetooth blind spot movement trajectory can be directly determined as the actual movement trajectory of the employee during the Bluetooth connection interruption. Subsequently, relevant personnel tracking management analysis can be performed based on this trajectory, such as calculating the employee's movement speed during the Bluetooth blind spot, the area passed through, and other information, to provide data support for the factory's production management, safety monitoring, etc.

[0038] Step S360, otherwise, according to the factory workshop number, activate the Bluetooth blind spot positioning network to process the spatial positioning at the time of disconnection, the Bluetooth blind spot movement trajectory and the spatial positioning at the time of successful reconnection, and obtain the Bluetooth blind spot positioning coordinate sequence. Specifically, if the above ideal situation is not met, that is, the trajectory start positioning does not match the spatial positioning at the time of disconnection, or the trajectory end positioning is inconsistent with the spatial positioning at the time of successful reconnection, or the Bluetooth blind spot movement trajectory is a discontinuous trajectory, it is necessary to use a more complex Bluetooth blind spot positioning network to process. According to the factory workshop number, activate the pre-built Bluetooth blind spot positioning network, and input the spatial positioning at the time of disconnection, the problematic Bluetooth blind spot movement trajectory and the spatial positioning at the time of successful reconnection into the network. Each component in the Bluetooth blind spot positioning network, such as the first Bluetooth blind spot positioning base network to the Nth Bluetooth blind spot positioning base network and the fully connected network, will analyze and fuse the data according to their respective algorithms and processing procedures. Through multi-dimensional processing of data, such as preliminary positioning calculation of the base network, sequence optimization of the fully connected network and other operations, the Bluetooth blind spot positioning coordinate sequence is finally obtained. The coordinate sequence can restore the actual location change trajectory of employees during the Bluetooth blind area as accurately as possible under complex Bluetooth positioning anomalies, ensuring that the factory personnel tracking management system can maintain high positioning accuracy and reliability under various circumstances, providing solid technical support for the efficient operation and management of the factory.

[0039] Step S400, performing blind spot coordinate compensation tracking management for the first employee according to the Bluetooth blind spot positioning coordinate sequence. Specifically, based on the Bluetooth blind spot positioning coordinate sequence, in the background data processing module of the factory personnel tracking management system, the fitted discrete coordinate points are connected in time or logical order, and the employee's Bluetooth blind spot action trajectory is restored using spline interpolation and other methods and integrated with the positioning trajectory during normal connection to form a complete action route data set. Based on this, the action characteristics of employees in Bluetooth blind spots are analyzed, and information such as production processes and equipment layouts are associated to evaluate work efficiency and abnormal behaviors, provide support for management decisions, and provide timely warnings if employees stay abnormally in key equipment or dangerous areas. At the same time, this management can achieve real-time tracking of employees, managers can view their locations and trajectories at any time, determine the distribution locations and formulate plans in case of emergency, and update the trajectories in real time with the employee activity system, dynamically adjust the analysis results and management strategies, and ensure efficient and safe operation of the factory.

[0040] The embodiment of the present application adopts a method of recording the spatial location at the time of disconnection and automatically tracking the movement trajectory when the Bluetooth device worn by the factory personnel is disconnected; when the Bluetooth is reconnected, the movement data during the disconnection period is processed, and a coordinate sequence is generated through the Bluetooth blind spot positioning network, thereby realizing accurate tracking and management of the personnel's position during the disconnection period, achieving the technical effect of continuously and accurately tracking the position of factory personnel during the period when the Bluetooth connection is disconnected.

[0041] In the above, refer to Figure 1 The factory personnel tracking management method driven by Bluetooth positioning according to an embodiment of the present invention is described in detail. Figure 2 A factory personnel tracking management system driven by Bluetooth positioning according to an embodiment of the present invention is described.

[0042] The factory personnel tracking management system driven by Bluetooth positioning according to the embodiment of the present invention is used to solve the technical problem that the existing factory personnel real-time positioning system based on Bluetooth cannot continuously and accurately track the personnel position during the period when the Bluetooth connection is disconnected, and achieves the technical effect of continuously and accurately tracking the factory personnel position during the period when the Bluetooth connection is disconnected. The factory personnel tracking management system driven by Bluetooth positioning includes: a spatial positioning acquisition module 10, a reconnection module 20, a positioning coordinate sequence acquisition module 30, and a coordinate compensation tracking management module 40.

[0043] The spatial positioning acquisition module 10 is used to obtain the spatial positioning at the time of disconnection when the first employee positioning Bluetooth is disconnected, wherein the first employee positioning Bluetooth will automatically record the spatial movement trajectory with the spatial positioning at the time of disconnection as the origin after disconnection.

[0044] The reconnection module 20 is used to continuously request reconnection with the first employee positioning Bluetooth. When the reconnection is successful, the Bluetooth blind area movement trajectory and the spatial positioning at the time of successful reconnection are obtained from the first employee positioning Bluetooth.

[0045] The positioning coordinate sequence acquisition module 30 is used to activate the Bluetooth blind spot positioning network according to the factory workshop number to process the spatial positioning at the disconnection time, the Bluetooth blind spot movement trajectory and the spatial positioning at the reconnection success time to obtain the Bluetooth blind spot positioning coordinate sequence.

[0046] The coordinate compensation tracking management module 40 is used to perform the first employee's blind spot coordinate compensation tracking management according to the Bluetooth blind spot positioning coordinate sequence.

[0047] The specific configuration of the positioning coordinate sequence acquisition module 30 will be described in detail below. As described above, according to the factory workshop number, the Bluetooth blind spot positioning network is activated to process the spatial positioning at the disconnection time, the Bluetooth blind spot movement trajectory and the spatial positioning at the time of successful reconnection, and the Bluetooth blind spot positioning coordinate sequence is obtained. The positioning coordinate sequence acquisition module 30 further includes: a Bluetooth blind spot positioning network component unit, the Bluetooth blind spot positioning network component unit is used for the Bluetooth blind spot positioning network including the first Bluetooth blind spot positioning base network until the Nth Bluetooth blind spot positioning base network, and a fully connected network; a spatial positioning processing unit, the spatial positioning processing unit is used to process the spatial positioning at the disconnection time, the Bluetooth blind spot movement trajectory and the spatial positioning at the time of successful reconnection according to the first Bluetooth blind spot positioning base network until the Nth Bluetooth blind spot positioning base network, respectively, to obtain the first Bluetooth blind spot positioning coordinate sequence until the Nth Bluetooth blind spot positioning coordinate sequence; a coordinate sequence fusion unit, the coordinate sequence fusion unit is used to fuse the first Bluetooth blind spot positioning coordinate sequence until the Nth Bluetooth blind spot positioning coordinate sequence through the fully connected network to obtain the Bluetooth blind spot positioning coordinate sequence.

[0048] Among them, the Bluetooth blind spot positioning network includes the first Bluetooth blind spot positioning base network to the Nth Bluetooth blind spot positioning base network, and a fully connected network. The Bluetooth blind spot positioning network component unit further includes: a factory workshop three-dimensional model construction subunit, the factory workshop three-dimensional model construction subunit is used to construct a factory workshop three-dimensional model based on the factory workshop number, and based on the preset spatial grid side length, a factory workshop three-dimensional grid model of the factory workshop three-dimensional model; a three-dimensional grid model configuration subunit, the three-dimensional grid model configuration subunit is used to configure each grid of the factory workshop three-dimensional grid model as an input node and an output node at the same time, the input information is 0 or 1, and the output information is 0 or 1, to obtain an initial graph neural network, wherein the grid input information and output information of the initial graph neural network are both 0 by default; a historical monitoring information collection subunit, the historical monitoring information collection subunit is used to collect employee historical monitoring information of the factory workshop three-dimensional grid model, wherein the employee historical monitoring information includes disconnection time setting The present invention relates to a method for preparing a graph neural network comprising: a first graph neural network training subunit, a second graph neural network training subunit, a second graph neural network training subunit and a second graph neural network training subunit. The first graph neural network training subunit comprises: a first ...

[0049] Among them, the 1-value input grid of the initial graph neural network is configured according to the positioning record information at the time of disconnection, the Bluetooth blind spot movement trajectory record information, and the positioning record information at the time of reconnection, and the 1-value output grid of the initial graph neural network is configured with the Bluetooth blind spot positioning coordinate sequence identification data, as the 1-value output grid supervision data, the initial graph neural network is trained to obtain the first Bluetooth blind spot positioning base network, and the initial graph neural network training subunit further includes: a blind spot positioning loss function construction micro unit, and the blind spot positioning loss function construction micro unit is used to construct a blind spot positioning loss function: Among them, X1 represents the 1-value output grid supervision data of the Bluetooth blind spot positioning coordinate sequence identification data to configure the initial graph neural network, X 10Characterize the 1-valued output grid of the initial graph neural network, n(X1∪X 10 ) represents X1 and X 10 The number of intersection 1-value grids, n(X1∩X 10 ) represents X1 and X 10 The number of 1-value grids in the union, e represents a natural constant, and LOSS represents a blind spot positioning loss value. Specifically, the first Bluetooth blind spot positioning base network obtains a micro-unit, and the first Bluetooth blind spot positioning base network obtains a micro-unit for configuring the 1-value input grid of the initial graph neural network based on the blind spot positioning loss function, according to the positioning record information at the time of disconnection, the Bluetooth blind spot movement trajectory record information, and the positioning record information at the time of reconnection, and configuring the 1-value output grid of the initial graph neural network with the Bluetooth blind spot positioning coordinate sequence identification data as the 1-value output grid supervision data, training the initial graph neural network, and obtaining the first Bluetooth blind spot positioning base network.

[0050] The first Bluetooth blind spot positioning coordinate sequence to the Nth Bluetooth blind spot positioning coordinate sequence are fused through the fully connected network to obtain the Bluetooth blind spot positioning coordinate sequence, and the coordinate sequence fusion unit further includes: a fully connected rule factor configuration subunit, the fully connected rule factor configuration subunit is used to configure the fully connected rule factor and construct the fully connected network, wherein the fully connected rule factor is: performing pairwise sequence distance evaluation on the first Bluetooth blind spot positioning coordinate sequence to the Nth Bluetooth blind spot positioning coordinate sequence to obtain a number of sequence distance parameters; an outlier factor analysis subunit, the outlier factor analysis subunit is used to traverse the first Bluetooth blind spot positioning coordinate sequence to the Nth Bluetooth blind spot positioning coordinate sequence based on the several sequence distance parameters to perform outlier factor analysis, and obtain the outlier factor of the first Bluetooth blind spot positioning coordinate sequence to the outlier factor of the Nth Bluetooth blind spot positioning coordinate sequence; a positioning coordinate sequence extraction subunit, the positioning coordinate sequence extraction subunit is used to extract the positioning coordinate sequence of the minimum value of the outlier factor of the first Bluetooth blind spot positioning coordinate sequence to the Nth Bluetooth blind spot positioning coordinate sequence, and output it as the Bluetooth blind spot positioning coordinate sequence.

[0051] Among them, based on the blind spot positioning loss function, the 1-value input grid of the initial graph neural network is configured according to the positioning record information at the time of disconnection, the Bluetooth blind spot movement trajectory record information, and the positioning record information at the time of reconnection, and the 1-value output grid of the initial graph neural network is configured with the Bluetooth blind spot positioning coordinate sequence identification data, as the 1-value output grid supervision data, the initial graph neural network is trained to obtain the first Bluetooth blind spot positioning base network, and the first Bluetooth blind spot positioning base network acquisition micro unit further includes: a record information clipping nano unit, and the record information clipping nano unit is used to clip the Bluetooth blind spot movement trajectory record information based on a preset trajectory length. , obtain Bluetooth blind spot movement trajectory pruning information, wherein the Bluetooth blind spot movement trajectory pruning information is a discontinuous trajectory; the first Bluetooth blind spot positioning base network obtains a nano unit, and the first Bluetooth blind spot positioning base network obtains a nano unit for configuring the 1-value input grid of the initial graph neural network based on the blind spot positioning loss function, according to the positioning record information at the disconnection moment, the Bluetooth blind spot movement trajectory pruning information, and the positioning record information at the reconnection moment, and configuring the 1-value output grid of the initial graph neural network with the Bluetooth blind spot positioning coordinate sequence identification data as the 1-value output grid supervision data, training the initial graph neural network, and obtaining the first Bluetooth blind spot positioning base network.

[0052] Among them, the positioning coordinate sequence acquisition module 30 further includes: a trajectory positioning unit, the trajectory positioning unit is used to obtain the trajectory starting positioning and trajectory ending positioning of the Bluetooth blind spot moving trajectory; a Bluetooth blind spot moving trajectory acquisition unit, the Bluetooth blind spot moving trajectory acquisition unit is used to locate the Bluetooth blind spot moving trajectory based on the spatial positioning at the disconnection moment and the spatial positioning at the successful reconnection moment when the trajectory starting positioning is the same as the spatial positioning at the disconnection moment, and the trajectory ending positioning is the same as the spatial positioning at the successful reconnection moment, and the Bluetooth blind spot moving trajectory is a continuous trajectory, so as to obtain the Bluetooth blind spot moving trajectory; a Bluetooth blind spot positioning coordinate sequence acquisition unit, the Bluetooth blind spot positioning coordinate sequence acquisition unit is used to otherwise, according to the factory workshop number, activate the Bluetooth blind spot positioning network to process the spatial positioning at the disconnection moment, the Bluetooth blind spot moving trajectory and the spatial positioning at the successful reconnection moment, so as to obtain the Bluetooth blind spot positioning coordinate sequence.

[0053] The Bluetooth positioning-driven factory personnel tracking management system provided in the embodiment of the present invention can execute the Bluetooth positioning-driven factory personnel tracking management method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0054] Although the present application makes various references to certain modules in the system according to the embodiments of the present application, any number of different modules may be used and run on the user terminal and / or server, and the various units and modules included are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional units are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of the present invention.

[0055] The above specific implementation manner does not constitute a limitation to the protection scope of the present application. It should be understood by those skilled in the art that various modifications, combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present application should be included in the protection scope of the present application. In some cases, the actions or steps recorded in the present application can be performed in an order different from that in the embodiment and can still achieve the desired results. In addition, the process depicted in the accompanying drawings does not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

Claims

1. A factory personnel tracking management method driven by Bluetooth positioning, characterized in that: include: When the first employee positioning Bluetooth is disconnected, the spatial positioning at the time of disconnection is obtained, wherein the first employee positioning Bluetooth automatically records the spatial movement trajectory with the spatial positioning at the time of disconnection as the origin after the disconnection; Continuously requesting to reconnect with the first employee positioning Bluetooth, and when the reconnection is successful, obtaining the Bluetooth blind area movement trajectory and the spatial positioning at the time of successful reconnection from the first employee positioning Bluetooth; According to the factory workshop number, the Bluetooth blind spot positioning network is activated to process the spatial positioning at the disconnection time, the Bluetooth blind spot movement trajectory and the spatial positioning at the time of successful reconnection to obtain a Bluetooth blind spot positioning coordinate sequence; The blind spot coordinate compensation tracking management of the first employee is performed according to the Bluetooth blind spot positioning coordinate sequence.

2. The method according to claim 1, characterized in that According to the factory workshop number, the Bluetooth blind spot positioning network is activated to process the spatial positioning at the disconnection time, the Bluetooth blind spot movement trajectory and the spatial positioning at the reconnection success time to obtain the Bluetooth blind spot positioning coordinate sequence, including: The Bluetooth blind area positioning network includes a first Bluetooth blind area positioning base network to an Nth Bluetooth blind area positioning base network, and a fully connected network; According to the first Bluetooth blind spot positioning base network to the Nth Bluetooth blind spot positioning base network, the spatial positioning at the disconnection moment, the Bluetooth blind spot movement trajectory and the spatial positioning at the reconnection success moment are processed respectively to obtain a first Bluetooth blind spot positioning coordinate sequence to the Nth Bluetooth blind spot positioning coordinate sequence; The first Bluetooth blind area positioning coordinate sequence to the Nth Bluetooth blind area positioning coordinate sequence are merged through the fully connected network to obtain the Bluetooth blind area positioning coordinate sequence.

3. The method according to claim 2, characterized in that The first Bluetooth blind area positioning base network construction step includes: Based on the factory workshop number, construct a three-dimensional model of the factory workshop, and based on a preset spatial grid side length, construct a three-dimensional grid model of the factory workshop three-dimensional model; Each grid of the three-dimensional grid model of the factory workshop is configured as an input node and an output node at the same time, the input information is 0 or 1, and the output information is 0 or 1, to obtain an initial graph neural network, wherein the grid input information and output information of the initial graph neural network are 0 by default; Collecting employee historical monitoring information of the three-dimensional grid model of the factory workshop, wherein the employee historical monitoring information includes disconnection time positioning record information, Bluetooth blind area movement trajectory record information, reconnection time positioning record information and Bluetooth blind area positioning coordinate sequence identification data, and the Bluetooth blind area movement trajectory record information is a movement trajectory recorded with the disconnection time positioning record information as the origin; According to the positioning record information at the time of disconnection, the moving trajectory record information of the Bluetooth blind spot, and the positioning record information at the time of reconnection, the 1-value input grid of the initial graph neural network is configured, and the 1-value output grid of the initial graph neural network is configured with the Bluetooth blind spot positioning coordinate sequence identification data as the 1-value output grid supervision data, and the initial graph neural network is trained to obtain the first Bluetooth blind spot positioning base network; The steps for constructing the second Bluetooth blind area positioning base network until the Nth Bluetooth blind area positioning base network are the same as the steps for constructing the first Bluetooth blind area positioning base network.

4. The method according to claim 3, characterized in that The 1-value input grid of the initial graph neural network is configured according to the positioning record information at the disconnection time, the Bluetooth blind spot movement trajectory record information, and the positioning record information at the reconnection time, and the 1-value output grid of the initial graph neural network is configured with the Bluetooth blind spot positioning coordinate sequence identification data as the 1-value output grid supervision data, and the initial graph neural network is trained to obtain the first Bluetooth blind spot positioning base network, including: Construct blind spot positioning loss function: Among them, X1 represents the 1-value output grid supervision data of the Bluetooth blind spot positioning coordinate sequence identification data to configure the initial graph neural network, X 10 Characterize the 1-valued output grid of the initial graph neural network, n(X1∪X 10 ) represents X1 and X 10 The number of intersection 1-value grids, n(X1∩X 10 ) represents X1 and X 10 The number of grids with a union value of 1, e represents the natural constant, and LOSS represents the blind area positioning loss value; Based on the blind spot positioning loss function, the 1-value input grid of the initial graph neural network is configured according to the positioning record information at the time of disconnection, the Bluetooth blind spot movement trajectory record information, and the positioning record information at the time of reconnection. The 1-value output grid of the initial graph neural network is configured with the Bluetooth blind spot positioning coordinate sequence identification data as the 1-value output grid supervision data, and the initial graph neural network is trained to obtain the first Bluetooth blind spot positioning base network.

5. The method according to claim 2, characterized in that The method of fusing the first Bluetooth blind area positioning coordinate sequence to the Nth Bluetooth blind area positioning coordinate sequence through the fully connected network to obtain the Bluetooth blind area positioning coordinate sequence includes: The fully connected rule factor is configured to construct the fully connected network, wherein the fully connected rule factor is: Performing pairwise sequence distance evaluation on the first Bluetooth blind area positioning coordinate sequence to the Nth Bluetooth blind area positioning coordinate sequence to obtain a plurality of sequence distance parameters; Based on the plurality of sequence distance parameters, traverse the first Bluetooth blind spot positioning coordinate sequence until the Nth Bluetooth blind spot positioning coordinate sequence to perform outlier factor analysis, and obtain the outlier factor of the first Bluetooth blind spot positioning coordinate sequence until the Nth Bluetooth blind spot positioning coordinate sequence outlier factor; A positioning coordinate sequence having a minimum value of the outlier factor of the first Bluetooth blind area positioning coordinate sequence until the outlier factor of the Nth Bluetooth blind area positioning coordinate sequence is extracted, and output as the Bluetooth blind area positioning coordinate sequence.

6. The method according to claim 4, characterized in that Based on the blind spot positioning loss function, the 1-value input grid of the initial graph neural network is configured according to the positioning record information at the disconnection time, the Bluetooth blind spot movement trajectory record information, and the positioning record information at the reconnection time, and the 1-value output grid of the initial graph neural network is configured with the Bluetooth blind spot positioning coordinate sequence identification data as the 1-value output grid supervision data, and the initial graph neural network is trained to obtain the first Bluetooth blind spot positioning base network, and further includes: Based on a preset trajectory length, the Bluetooth blind area movement trajectory record information is trimmed to obtain Bluetooth blind area movement trajectory pruning information, wherein the Bluetooth blind area movement trajectory pruning information is a discontinuous trajectory; Based on the blind spot positioning loss function, the 1-value input grid of the initial graph neural network is configured according to the positioning record information at the time of disconnection, the Bluetooth blind spot movement trajectory pruning information, and the positioning record information at the time of reconnection. The 1-value output grid of the initial graph neural network is configured with the Bluetooth blind spot positioning coordinate sequence identification data as the 1-value output grid supervision data, and the initial graph neural network is trained to obtain the first Bluetooth blind spot positioning base network.

7. The method according to claim 1, characterized in that According to the factory workshop number, the Bluetooth blind spot positioning network is activated to process the spatial positioning at the disconnection time, the Bluetooth blind spot movement trajectory and the spatial positioning at the reconnection success time to obtain the Bluetooth blind spot positioning coordinate sequence, which includes: Obtaining the trajectory start location and trajectory end location of the Bluetooth blind area moving trajectory; When the starting position of the trajectory is the same as the spatial position at the time of disconnection, and the ending position of the trajectory is the same as the spatial position at the time of successful reconnection, and the Bluetooth blind area movement trajectory is a continuous trajectory, the Bluetooth blind area movement trajectory is positioned based on the spatial position at the time of disconnection and the spatial position at the time of successful reconnection to obtain the Bluetooth blind area movement trajectory; Otherwise, according to the factory workshop number, the Bluetooth blind spot positioning network is activated to process the spatial positioning at the disconnection time, the Bluetooth blind spot movement trajectory and the spatial positioning at the successful reconnection time to obtain the Bluetooth blind spot positioning coordinate sequence.

8. Bluetooth positioning driven factory personnel tracking management system, characterized by: The system is used to implement the factory personnel tracking management method driven by Bluetooth positioning according to any one of claims 1 to 7, and the system comprises: A spatial positioning acquisition module, wherein the spatial positioning acquisition module is used to obtain the spatial positioning at the time of disconnection when the first employee positioning Bluetooth is disconnected, wherein the first employee positioning Bluetooth automatically records a spatial movement trajectory with the spatial positioning at the time of disconnection as the origin after the disconnection; A reconnection module, the reconnection module is used to continuously request to reconnect with the first employee positioning Bluetooth, and when the reconnection is successful, obtain the Bluetooth blind area movement trajectory and the spatial positioning at the time of successful reconnection from the first employee positioning Bluetooth; A positioning coordinate sequence acquisition module, the positioning coordinate sequence acquisition module is used to activate the Bluetooth blind spot positioning network according to the factory workshop number to process the spatial positioning at the disconnection time, the Bluetooth blind spot movement trajectory and the spatial positioning at the time of successful reconnection, and obtain the Bluetooth blind spot positioning coordinate sequence; A coordinate compensation tracking management module is used to perform coordinate compensation tracking management of the first employee's blind spot according to the Bluetooth blind spot positioning coordinate sequence.

Images