Rfid-based restricted space personnel positioning and behavior prediction analysis method and system

Abstract

The application discloses a kind of personnel positioning and behavior prediction analysis method and system in restricted space based on RFID, method includes: obtaining RFID signal data in space, according to RFID signal data, the real-time signal strength data of operating personnel in space is obtained;The historical signal strength data and sensor data are input into recurrent neural network, the real position of operating personnel is used as output to train, to construct path prediction model;According to signal strength data and sensor layout position, the moving track and behavior mode are predicted by path prediction model;Crowding degree analysis is carried out, if unsafe moving track, behavior mode or crowding degree is higher than set threshold, then early warning is carried out.The application carries out signal strength optimization calculation, combines behavior prediction algorithm and crowding degree evaluation, can find potential safety hazard in time, real-time feedback operating personnel's position and behavior information, improve operating efficiency and safety, through real-time monitoring and intelligent analysis, it is helpful to reduce accident risk.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of industrial safety monitoring, and in particular to a method and system for personnel positioning and behavior prediction analysis in a restricted space based on RFID. BACKGROUND

[0002] During power production and maintenance, workers often need to work in restricted spaces that are narrow and potentially dangerous. These spaces usually have small entrances, poor ventilation, and a risk of harmful gas or substance accumulation. Therefore, real-time monitoring of the positions of workers, predicting their behavior, and analyzing the degree of space congestion are crucial for ensuring the safety of workers.

[0003] Traditional RFID technology often cannot achieve ideal positioning accuracy due to environmental factors, blockage of items, and multipath effects. In the complex environment of a power plant restricted space, this can result in an inability to accurately determine the specific location of workers. Existing technologies can only perform simple behavior analysis and congestion assessment, and cannot perform deep prediction of worker behavior based on historical and real-time data, nor can they accurately determine the degree of space congestion and its impact on work. RFID tags in power plant restricted spaces generate a large amount of data, but existing technologies have limited processing capabilities for these data, making it difficult to extract useful information such as personnel movement paths and behavior patterns. Traditional monitoring devices and manual observation often cannot provide real-time feedback on the positions and behaviors of workers, which is very disadvantageous for safety management. In power plant restricted spaces, timely safety warnings are crucial. However, the above problems often cannot provide accurate positioning information, and lack effective behavior prediction and congestion analysis functions. Therefore, there is an urgent need for a method that can efficiently and accurately position information for behavior prediction and warning. SUMMARY

[0004] In view of the above existing problems, the present application is proposed.

[0005] Therefore, the present application provides a method and system for personnel positioning and behavior prediction analysis in a restricted space based on RFID to solve the problem of being unable to accurately determine the specific location of workers, unable to perform deep prediction of worker behavior based on historical and real-time data, and unable to accurately determine the degree of space congestion and its impact on work.

[0006] To solve the above technical problems, the present application provides the following technical solutions:

[0007] In a first aspect, the present application provides an RFID-based personnel positioning and behavior prediction analysis method in a restricted space, comprising: obtaining RFID signal data in the space, and obtaining real-time signal strength data of workers in the space according to the RFID signal data;

[0008] inputting historical signal strength data and sensor data into a recurrent neural network, training the recurrent neural network with real positions of the workers as output, and constructing a path prediction model;

[0009] performing time series analysis on continuous positions of the workers by the path prediction model according to the signal strength data and sensor layout positions, to predict moving tracks and behavior patterns;

[0010] performing congestion analysis, the analysis being obtained by distances and proximities between the workers;

[0011] if an unsafe moving track, behavior pattern or congestion is predicted to be higher than a set threshold, performing a warning.

[0012] As a preferred scheme of the RFID-based personnel positioning and behavior prediction analysis method in a restricted space, wherein: obtaining real-time signal strength data of workers in the space according to the RFID signal data comprises:

[0013] when real-time signal strength data in the RFID signal data is obtained, performing real-time signal strength data optimization, the optimization including data filtering, dynamically adjusting path loss model parameters and adjusting an angle between a tag and a sensor;

[0014] wherein adjusting the angle between the tag and the sensor comprises:

[0015] calculating a tag direction by an antenna array according to an angle between an RFID signal tag and an RFID reader, the antenna array including four antenna elements, signal strength data received by an i-th antenna element being represented as RSSI_i, and an angle of the tag direction relative to the antenna array being represented as:

[0016] θ=arctan2(ΔRSSI_y,ΔRSSI_x)

[0017] wherein ΔRSSI_y and ΔRSSI_x respectively represent RSSI difference values between adjacent antenna elements in x-axis and y-axis directions, and the arctan2 function returns an angle formed by a vector from a positive direction of the x-axis to a point (ΔRSSI_x, ΔRSSI_y), the angle being in a range of -π to π.

[0018] As a preferred scheme of the RFID-based personnel positioning and behavior prediction analysis method in a confined space according to the present application, wherein: according to the signal strength data and the sensor layout position, the continuous position of the working personnel is analyzed in time series through a path prediction model to predict the moving track and the behavior mode, including:

[0019] The historical and real-time signal strength data are acquired, integrated, converted into time series, and the time series features are acquired.

[0020] The sensor layout position is acquired, the sensor spatial coordinate information is encoded, converted into a one-dimensional vector to represent the position, and the spatial features are constructed.

[0021] As a preferred scheme of the RFID-based personnel positioning and behavior prediction analysis method in a confined space according to the present application, wherein: according to the signal strength data and the sensor layout position, the continuous position of the working personnel is analyzed in time series through a path prediction model to predict the moving track and the behavior mode, further including:

[0022] A joint input containing time series features and spatial features is constructed, wherein each time step contains a plurality of sensor signal strength values and a one-dimensional vector sequence of corresponding coordinate information;

[0023] The trend of the signal strength changing over time and the probability of the sensor being read subsequently are captured through an RNN network to obtain the moving track and the behavior mode of the working personnel.

[0024] As a preferred scheme of the RFID-based personnel positioning and behavior prediction analysis method in a confined space according to the present application, wherein: the analysis is obtained through the distance and the proximity between the working personnel, including:

[0025] The distance between the working personnel is calculated and represented as:

[0026] dij(t) = ||pi(i) - pj(t) ||

[0027] Wherein, pi(t) - pj(t) represents the Euclidean distance between two position vectors.

[0028] As a preferred scheme of the RFID-based personnel positioning and behavior prediction analysis method in a confined space according to the present application, wherein: the analysis is obtained through the distance and the proximity between the working personnel, further including:

[0029] The proximity between the working personnel is calculated and represented as:

[0030] Proximityij(t) = 1 + exp(σA - dij(t))

[0031] where Proximity_{ij}(t)∈(0,1) represents the proximity between the ith and jth workers; A represents a safety distance threshold, and σ is a parameter controlling the rate of change of proximity, and dij(t) is the distance between the workers;

[0032] When the distance dij(t) between the workers is less than the safety distance threshold A, it indicates that the two workers are close; when the distance dij(t) between the workers is greater than A, it indicates that the two workers are far away.

[0033] As a preferred scheme of the RFID-based restricted space personnel positioning and behavior prediction analysis method, the analysis is obtained through the distance and proximity between the workers, and further comprises: calculating the overall crowding degree, represented as:

[0034] C(t)=N1i=1∑Nj=i+1∑NProximityij(t)·SppedFactorij(t)

[0035] where SpeedFactor_{ij}(t) is a dynamic factor based on the speed of the workers.

[0036] In a second aspect, the present application provides an RFID-based restricted space personnel positioning and behavior prediction analysis system, comprising:

[0037] An acquisition module is configured to acquire RFID signal data in the space, and obtain real-time signal strength data of workers in the space according to the RFID signal data;

[0038] A model construction module is configured to input historical signal strength data and sensor data into a recurrent neural network, and train the recurrent neural network with real positions of the workers as output, so as to construct a path prediction model;

[0039] A prediction module is configured to perform time series analysis on continuous positions of the workers by the path prediction model according to real-time signal strength data and sensor layout positions, so as to predict a moving track and a behavior mode;

[0040] An analysis module is configured to perform crowding degree analysis, and the analysis is obtained through the distance and proximity between the workers;

[0041] A warning module is configured to perform warning if an unsafe moving track, behavior mode or crowding degree higher than a set threshold is predicted.

[0042] In a third aspect, the present application provides a computing device, comprising:

[0043] a memory and a processor;

[0044] The memory is used for storing computer executable instructions, and the processor is used for executing the computer executable instructions, which realize the steps of the RFID-based personnel positioning and behavior prediction analysis method in a restricted space.

[0045] In a fourth aspect, the present application provides a computer readable storage medium storing computer executable instructions, which realize the steps of the RFID-based personnel positioning and behavior prediction analysis method in a restricted space when executed by a processor.

[0046] Compared with the prior art, the present application has the following beneficial effects: the present application utilizes RFID technology and signal strength optimization calculation, combines behavior prediction algorithm and congestion evaluation, can timely find potential safety hazards, real-time feedback of the position and behavior information of the operating personnel, improves the operating efficiency and safety, and through real-time monitoring and intelligent analysis, helps to reduce the risk of accidents. BRIEF DESCRIPTION OF DRAWINGS

[0047] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can be obtained without creative labor on the basis of these drawings.

[0048] Figure 1 The overall flowchart of the RFID-based personnel positioning and behavior prediction analysis method in a restricted space according to an embodiment of the present application is shown. DETAILED DESCRIPTION

[0049] In order to make the above-mentioned purposes, features and advantages of the present application more apparent and easy to understand, the specific embodiments of the present application will be described in detail below with reference to the drawings of the specification. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor should be within the scope of protection of the present application.

[0050] Embodiment 1

[0051] Reference Figure 1 For an embodiment of the present application, an RFID-based personnel positioning and behavior prediction analysis method in a restricted space is provided, which comprises:

[0052] S1: obtaining RFID signal data in the space, and obtaining real-time signal strength data of operating personnel in the space according to the RFID signal data;

[0053] It should be noted that the RFID tag is attached to the safety helmet or uniform of the worker, and the RFID reader network is deployed in the restricted space to capture the signal of the RFID tag.

[0054] Further, the real-time signal strength data of the worker in the space is obtained according to the RFID signal data, including: when the real-time signal strength data in the RFID signal data is obtained, the real-time signal strength data (RSSI) is optimized, and the optimization includes data filtering, dynamically adjusting the path loss model parameters and adjusting the angle between the tag and the sensor.

[0055] Wherein, adjusting the angle between the tag and the sensor includes:

[0056] According to the angle between the RFID signal tag and the RFID reader, the tag direction is calculated by using an antenna array, the antenna array includes four antenna elements, the signal strength data received by the i-th antenna element is represented as RSSI_i, and the angle of the signal tag relative to the direction of the antenna array is represented as:

[0057] θ=arctan2(ΔRSSI_y,ΔRSSI_x)

[0058] Wherein, ΔRSSI_y and ΔRSSI_x respectively represent the RSSI difference between adjacent antenna elements in the x-axis and y-axis directions; the arctan2 function returns the angle formed by the vector from the positive direction of the x-axis to the point (ΔRSSI_x, ΔRSSI_y), and the range is between -π and π.

[0059] Specifically, the angles of the four antennas in the antenna array point to the southeast, southwest and northwest directions respectively.

[0060] It should be noted that the RSSI value may fluctuate greatly in the actual environment, and therefore the positioning algorithm of the RSSI is optimized by the above steps.

[0061] Wherein, for data filtering, a sliding average filter or a Kalman filter can be used to smooth the RSSI value to reduce the influence of noise and burst interference, and by setting appropriate window size and weight, the random error can be reduced while retaining the dynamic change of the signal.

[0062] For dynamically adjusting the path loss model parameters: according to the real-time signal strength data and the corresponding distance information, the signal strength and the environmental attenuation factor parameters are adjusted.

[0063] Specifically, RSSI=-(10\cdotn\cdot\log_{10}(d)+A)

[0064] RSSI is the received signal strength (usually in dBm), d is the distance between the transmitter and the receiver; n is the environmental attenuation factor (usually determined by experiment); A is the signal strength at the reference distance (usually 1 meter).

[0065] It should also be noted that the above use of directional antennas to focus energy can reduce interference in non-target areas and enhance signal reception capability in the target direction; beamforming technology can dynamically adjust the directivity and gain of the signal through the joint operation of multiple antenna units to avoid or penetrate obstacles.

[0066] S2: input historical signal strength data and sensor data into a recurrent neural network, and train the real position of the worker as output to build a path prediction model;

[0067] It should be noted that for complex behavior patterns, a recurrent neural network RNN is used to predict behavior. The principle of the behavior prediction algorithm is to use the RNN model to receive an input vector x_t at each time step t and produce an output vector y_t. The internal state h_t of the RNN is updated according to the current input x_t and the previous state h_{t-1}. The input data includes the historical and current RSSI data of the person and the relationship between the RFID readers.

[0068] First is the conversion from the input layer to the hidden layer, the formula is as follows:

[0069] zt=W {xh} ·x_t+W{h}{h}·h_{t-1}+b_h

[0070] Where W_{xh} is the weight matrix from the input layer to the hidden layer, W{h}{h} is the weight matrix from the hidden layer to the hidden layer (recurrent weight), and b_h is the bias vector of the hidden layer.

[0071] The hyperbolic tangent function (tanh) can be selected as the activation function in the hidden layer.

[0072] The conversion from the hidden layer to the output layer, the formula is as follows:

[0073] ot=W{hy}·ht+by

[0074] Where W{hy} is the weight matrix from the hidden layer to the output layer, and by is the bias vector of the output layer.

[0075] The activation function of the output layer uses the softmax function to convert the output to a probability distribution, the formula is as follows:

[0076] yt=softmax(o_t)

[0077] For regression tasks, no additional activation function is needed, and (o_t) is used directly as the output.

[0078] S3: According to real-time signal strength data and sensor layout position, the continuous position of the worker is analyzed by a path prediction model to predict the moving track and behavior mode;

[0079] Further, according to real-time signal strength data and sensor layout position, the continuous position of the worker is analyzed by a path prediction model to predict the moving track and behavior mode, including:

[0080] The historical and real-time signal strength data are obtained, integrated, converted into time series, and the time series features are obtained,

[0081] The sensor layout position is obtained, the sensor spatial coordinate information is encoded, converted into a one-dimensional vector to represent the position, and the spatial features are constructed.

[0082] Further, according to real-time signal strength data and sensor layout position, the continuous position of the worker is analyzed by a path prediction model to predict the moving track and behavior mode, further including:

[0083] A joint input containing time series features and spatial features is constructed, wherein each time step contains a plurality of sensor signal strength values and a one-dimensional vector sequence of corresponding coordinate information;

[0084] The trend of signal strength change over time and the probability of sensor being read subsequently are captured by the RNN network to obtain the moving track and behavior mode of the worker.

[0085] It should be noted that the probability of the next reading of the several RFID readers with the highest correlation degree is output, and the higher the probability, the greater the probability of the worker passing through the RFID reader.

[0086] It should also be noted that by selecting RNN to convert each suitable input data into time series analysis, the effect of automatic analysis and rapid prediction can be achieved in the simplest form, rather than using more complex image prediction methods, and by using signal strength for track prediction, the problem of image or video being blocked can also be avoided.

[0087] S4: Crowding degree analysis is performed, and the distance and proximity between workers are analyzed to obtain;

[0088] Further, the distance and proximity between workers are analyzed to obtain, including:

[0089] The distance between workers is calculated and represented as:

[0090] dij(t) = ||pi(t) - pj(t) ||

[0091] wherein, pi(t) - pj(t) represents the Euclidean distance between two position vectors.

[0092] Further, the analysis through the distance and proximity between workers also includes:

[0093] The proximity between workers is calculated and represented as:

[0094] Proximityij(t) = 1 + exp(σA - dij(t))

[0095] wherein, Proximity_{ij}(t) ∈ (0, 1) represents the proximity between the i-th and j-th worker; A represents the safety distance threshold, σ is a parameter that controls the rate of change of proximity, and dij(t) is the distance between workers;

[0096] When the distance between workers dij(t) is less than the safety distance threshold A, it means that the two workers are close; when the distance between workers dij(t) is greater than A, it means that the two workers are far away.

[0097] Further, the analysis through the distance and proximity between workers also includes: calculating the overall congestion, represented as:

[0098] C(t) = N1i=1∑Nj=i+1∑NProximityij(t) · SpeedFactorij(t)

[0099] wherein, SpeedFactor_{ij}(t) is a dynamic factor based on the speed of the worker.

[0100] It should be noted that this dynamic factor is used to consider the influence of speed on congestion.

[0101] It should also be noted that the congestion analysis algorithm comprehensively considers the distance, speed, and spatial layout between workers. By calculating the distance and proximity between workers, the algorithm can evaluate the relative position relationship between workers and determine whether there is a potential congestion risk. At the same time, by introducing the speed factor, the influence of the predicted moving direction and speed of the worker on the congestion is considered, so that the algorithm can more accurately reflect the actual congestion situation.

[0102] S5: If the predicted unsafe moving trajectory, behavior pattern, or congestion is higher than the set threshold, a warning is given.

[0103] Specifically, the unsafe movement trajectory can be manifested as: each construction personnel going to the site belongs to a specific project, each project has a limited work range, when a construction personnel has entered or is about to enter a hot work card area or a limited space area, it is judged whether the project where the construction personnel is located contains the area, if not, it belongs to an unsafe movement trajectory, and an alarm information is generated; the setting of the crowding threshold can be: when the number of people in an area exceeds 30, it means that the area is very crowded, and an alarm information is generated.

[0104] The above is a schematic scheme of the method for personnel positioning and behavior prediction analysis in a restricted space based on RFID. It should be noted that the technical scheme of the system for personnel positioning and behavior prediction analysis in a restricted space based on RFID is the same as the technical scheme of the method for personnel positioning and behavior prediction analysis in a restricted space based on RFID described above, and the technical scheme of the system for personnel positioning and behavior prediction analysis in a restricted space based on RFID in the present embodiment is not described in detail, and the description of the technical scheme of the method for personnel positioning and behavior prediction analysis in a restricted space based on RFID described above can be referred to.

[0105] The system for personnel positioning and behavior prediction analysis in a restricted space based on RFID in the present embodiment comprises:

[0106] The acquisition module is configured to acquire RFID signal data in the space, and obtain real-time signal strength data of the working personnel in the space according to the RFID signal data.

[0107] The model construction module is configured to input the historical signal strength data and the sensor data into a recurrent neural network, and train the recurrent neural network with the real position of the working personnel as the output, so as to construct a path prediction model.

[0108] The prediction module is configured to perform time series analysis on the continuous positions of the working personnel by the path prediction model according to the real-time signal strength data and the sensor layout position, so as to predict the movement trajectory and the behavior mode.

[0109] The analysis module is configured to perform crowding analysis, and the analysis is obtained by the distance and proximity between the working personnel.

[0110] The early warning module is configured to perform early warning if the unsafe movement trajectory, the behavior mode or the crowding degree is higher than the set threshold.

[0111] The present embodiment further provides a computing device suitable for personnel positioning and behavior prediction analysis in a restricted space based on RFID, comprising:

[0112] The memory is configured to store computer-executable instructions, and the processor is configured to execute the computer-executable instructions to implement the method for realizing personnel positioning and behavior prediction analysis in a restricted space based on RFID.

[0113] The embodiment also provides a storage medium, which stores a computer program. The computer program is executed by a processor to implement the method for realizing personnel positioning and behavior prediction analysis in a restricted space based on RFID.

[0114] The storage medium provided by the embodiment belongs to the same inventive concept as the method for realizing personnel positioning and behavior prediction analysis in a restricted space based on RFID provided by the above embodiment. The technical details not described in the embodiment can be referred to the above embodiment, and the embodiment has the same beneficial effects as the above embodiment.

[0115] From the above description about the embodiments, those skilled in the art can clearly understand that the present application can be realized by means of software and necessary universal hardware, and of course can also be realized by hardware, but in many cases, the former is a better embodiment. Based on such understanding, the technical solutions of the present application can be embodied in the form of a software product, and the computer software product can be stored in a computer readable storage medium, such as a floppy disk, a read-only memory (ROM), a random access memory (RAM), a FLASH memory, a hard disk, or an optical disc, and includes a number of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods of various embodiments of the present application.

[0116] Embodiment 2

[0117] Referring to Table 1, based on the previous embodiment, the present embodiment provides a verification comparison example of the method for realizing personnel positioning and behavior prediction analysis in a restricted space based on RFID. The verification case is based on the data collected and analyzed during the construction of a certain power plant project of Huaneng. The traditional scheme only observes and warns through RFID and video images.

[0118] Table 1 Comparison of parameters between the present scheme and the traditional scheme

[0119]

[0120] From table 1, the scheme can find abnormal situation and trigger alarm in time, improve the safety and management level of the personnel in the restricted space, at the same time, the system runs normally for 3 months, the method is implemented, and 170 alarms of construction personnel are found, and part of the non-standard construction on site is found, and the optimization of the project and construction is improved.

[0121] It should be noted that the above examples are only used to illustrate the technical solutions of the present application and are not limiting. Although the present application has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that the technical solutions of the present application can be modified or replaced equivalently without departing from the spirit and scope of the present application, and they should be covered in the scope of the claims of the present application.

Claims

1. A method for RFID-based personnel positioning and behavior prediction analysis in a restricted space, characterized in that, Comprise: Obtaining RFID signal data in the space, obtaining real-time signal strength data of workers in the space according to the RFID signal data, comprising: When the real-time signal strength data in the RFID signal data is obtained, the real-time signal strength data is optimized, and the optimization comprises data filtering, dynamically adjusting path loss model parameters and adjusting the angle between the tag and the sensor; Wherein, adjusting the angle between the tag and the sensor comprises: According to the angle between the RFID signal tag and the RFID reader, the direction of the tag is calculated by using an antenna array, the antenna array comprises four antenna elements, the signal strength data received by the i-th antenna element is represented as RSSI_i, and the angle of the direction of the signal tag relative to the antenna array is represented as: θ = arctan2(ΔRSSI_y, ΔRSSI_x) Wherein, ΔRSSI_y and ΔRSSI_x respectively represent the RSSI difference value between adjacent antenna elements in the x-axis and y-axis directions; the arctan2 function returns the angle formed by the vector from the positive direction of the x-axis to the point (ΔRSSI_x, ΔRSSI_y), which is in the range of -π to π; Inputting historical signal strength data and sensor data into a recurrent neural network, training the real position of the worker as output to construct a path prediction model; According to the signal strength data and the layout position of the sensor, the time series analysis of the continuous position of the worker is carried out by the path prediction model to predict the moving track and behavior mode, comprising: A joint input containing time series features and spatial features is constructed, wherein each time step contains a one-dimensional vector sequence of multiple sensor signal strength values and corresponding coordinate information; The trend of signal strength change over time and the probability of sensor being read subsequently are captured by the RNN network to obtain the moving track and behavior mode of the worker; Crowding degree analysis is carried out, and the analysis is obtained by the distance and proximity between workers; The analysis is obtained by the distance and proximity between workers, comprising: The proximity between workers is calculated and represented as: Proximityij(t) = 1 + exp(σA-dij(t)) Wherein, Proximity_{ij}(t) ∈ (0, 1) represents the proximity between the i-th and j-th workers; A represents a safety distance threshold, σ is a parameter controlling the change rate of proximity, and dij(t) is the distance between workers; When the distance dij(t) between workers is less than the safety distance threshold A, it means that the two workers are close; when the distance dij(t) between workers is greater than A, it means that the two workers are far away; The analysis is obtained by the distance and proximity between workers, and further comprises: calculating the overall crowding degree, represented as: C(t) = N1i=1∑Nj=i+1∑NProximityij(t)·SpeedFactorij(t) Wherein, SpeedFactor_{ij}(t) is a dynamic factor based on the speed of the worker; If an unsafe moving trajectory, behavior pattern or crowding degree is predicted to be higher than a set threshold, a warning is given.

2. The RFID-based restricted space personnel positioning and behavior prediction analysis method of claim 1, wherein, According to the signal strength data and the sensor layout position, a time series analysis is performed on the continuous positions of the workers by a path prediction model to predict the moving trajectory and the behavior pattern, including: The historical and real-time signal strength data are obtained, integrated, converted into time series, and the time series features are obtained, The sensor layout position is obtained, the sensor spatial coordinate information is encoded, and the position is converted into a one-dimensional vector to represent the position, and the spatial features are constructed.

3. The RFID-based restricted space personnel positioning and behavior prediction analysis method of claim 2, wherein, The analysis is obtained by the distance and proximity between the workers, including: The distance between the workers is calculated and represented as: dij(t)=||pi(t)-pj(t)| Wherein, pi(t)-pj(t) represents the Euclidean distance between two position vectors.

4. A system for applying the RFID-based restricted space personnel positioning and behavior prediction analysis method according to claim 1, characterized in that, Including: An acquisition module is configured to acquire RFID signal data in a space and obtain real-time signal strength data of workers in the space according to the RFID signal data; A model construction module is configured to input historical signal strength data and sensor data into a recurrent neural network and train the recurrent neural network with real positions of the workers as output to construct a path prediction model; A prediction module is configured to perform a time series analysis on continuous positions of the workers by the path prediction model according to real-time signal strength data and sensor layout positions to predict moving trajectories and behavior patterns; An analysis module is configured to perform crowding degree analysis, and the analysis is obtained by the distance and proximity between the workers; A warning module is configured to give a warning if an unsafe moving trajectory, behavior pattern or crowding degree is predicted to be higher than a set threshold.

5. An electronic device, comprising: a memory and a processor; The memory is configured to store computer executable instructions, and the processor is configured to execute the computer executable instructions, which, when executed by the processor, implement the steps of the RFID-based positioning and behavior prediction analysis method for workers in a restricted space according to any one of claims 1 to 3.

6. A computer readable storage medium storing computer executable instructions, which, when executed by a processor, implement the steps of the RFID-based positioning and behavior prediction analysis method for workers in a restricted space according to any one of claims 1 to 3.

Images