Personnel dynamic positioning and monitoring method based on Internet of Things laboratory

Abstract

The invention relates to the technical field of Internet of Things, in particular to a dynamic personnel positioning and monitoring method based on an Internet of Things laboratory, which comprises the following steps of: performing high-precision real-time tracking and positioning on laboratory personnel by adopting a high-precision positioning technology and combining a laboratory layout; processing video data monitored in the laboratory by using a deep learning algorithm, identifying actions of laboratory personnel, and analyzing a behavior pattern based on a time sequence; environmental parameters in the laboratory are monitored in real time, receipts are collected in a unified mode through the Internet of Things gateway, and the receipts are uploaded to the cloud server after being preliminarily processed through edge calculation. By integrating a high-precision positioning technology and a deep learning algorithm, the scheme realizes real-time tracking positioning and behavior recognition of laboratory personnel. Therefore, the accuracy and efficiency of laboratory personnel management are improved, and potential safety hazards can be found and dealt with in time. The high-precision positioning technology ensures the accuracy and the real-time performance of the personnel position.

Description

Technical Field

[0001] The present invention relates to the technical field of Internet of Things, and in particular to a personnel dynamic positioning and monitoring method based on an Internet of Things laboratory. Background Art

[0002] With the rapid development and popularization of Internet of Things technology, its application in various fields is becoming more and more extensive. Internet of Things technology realizes the intelligent perception, identification, tracking, monitoring and management of the physical world through various technical means such as sensors, RFID, Bluetooth, UWB, etc. In the field of laboratory management, the application of Internet of Things technology provides strong technical support for laboratory safety management, operational efficiency improvement and scientific research innovation.

[0003] As an important place for scientific research and innovation, the safety management and operational efficiency of laboratories are crucial to the smooth progress of scientific research. Traditional laboratory management methods often rely on manual monitoring and paper records, which are not only inefficient but also prone to errors. With the introduction of Internet of Things technology, laboratory management has gradually become intelligent and automated, greatly improving management efficiency and safety.

[0004] In current laboratory management, although there are some IoT-based technical means for personnel monitoring and environmental monitoring, these technologies still have certain limitations and shortcomings. Specifically:

[0005] Positioning technology: Although traditional positioning technologies such as RFID and Bluetooth low energy have been widely used in laboratory personnel positioning, their positioning accuracy is limited and it is difficult to meet the needs of high-precision real-time tracking. In the complex environment of the laboratory, these technologies often cannot accurately obtain the specific location information of personnel, resulting in inefficient management and safety hazards.

[0006] Video surveillance: Laboratories are usually equipped with video surveillance systems to monitor personnel behavior and laboratory safety. However, traditional video surveillance mainly relies on manual monitoring and analysis, which is not only inefficient but also prone to missed detections and misjudgments. At the same time, there is a lack of automated behavior recognition algorithms to accurately identify the movements and behavior patterns of laboratory personnel, and it is impossible to provide timely and effective decision support for managers.

[0007] Environmental monitoring: Real-time monitoring of environmental parameters in the laboratory through IoT sensors is one of the main functions of existing environmental monitoring systems. However, these systems can only collect data and lack real-time early warning and emergency response mechanisms. When the environmental parameters in the laboratory are abnormal, the system cannot issue early warning signals and take emergency measures in time, thus failing to effectively protect the safety of the laboratory and the health of personnel. Summary of the invention

[0008] In view of the technical problems existing in the prior art, the present invention provides a personnel dynamic positioning and monitoring method based on the Internet of Things laboratory to solve the above-mentioned laboratory management. Although there are some technical means based on the Internet of Things for personnel monitoring and environmental monitoring, these technologies still have certain limitations and shortcomings.

[0009] The technical solution of the present invention to solve the above technical problems is as follows: A method for dynamic positioning and monitoring of personnel based on an Internet of Things laboratory comprises the following steps:

[0010] Use high-precision positioning technology, combined with the laboratory layout map, to track and locate laboratory personnel in real time with high precision;

[0011] Use deep learning algorithms to process video data from internal laboratory surveillance, identify the actions of laboratory personnel, and analyze behavioral patterns based on time series;

[0012] Monitor the environmental parameters in the laboratory in real time, collect data uniformly through the IoT gateway, and upload them to the cloud server after preliminary processing through edge computing;

[0013] According to the preset safety rules, the pre-tensioning signal is automatically triggered and pushed to the management personnel;

[0014] Design an emergency evacuation route planning algorithm based on the preload signal; if an emergency is detected, provide the laboratory personnel with the optimal escape route;

[0015] Provide laboratory operation reports based on big data analysis to assist laboratory managers in making management decisions, and use 3D visualization technology to display laboratory personnel distribution, environmental status and early warning information.

[0016] Based on the above technical solution, the present invention can also be improved as follows.

[0017] Furthermore, the high-precision positioning technology includes at least one of UWB (ultra-wideband), RFID (radio frequency identification), Bluetooth low energy or GPS (global positioning system), and is combined with fixed reference points in the laboratory to achieve real-time tracking and positioning through triangulation or fingerprint positioning algorithm.

[0018] Further, the triangulation algorithm includes: using the signal propagation time difference (Δt_A, Δt_B, Δt_C) or phase difference (Φ_A, Φ_B, Φ_C) between at least three fixed reference points (A, B, C) and the mobile tag (M), combined with the speed of light (c) to calculate the distance (d_A, d_B, d_C), and then determining the coordinates (x, y, z) of the mobile tag by the distance intersection formula in three-dimensional space;

[0019] The calculation formula is as follows:

[0020]

[0021] Among them, (x_A, y_A, z_A), (x_B, y_B, z_B), (x_C, y_C, z_C) are the coordinates of fixed reference points A, B, and C respectively;

[0022] Furthermore, the fingerprint positioning algorithm includes: collecting wireless signal characteristics at multiple locations in the laboratory in advance and building a signal characteristic database;

[0023] During positioning, the real-time signal features collected by the mobile tag are matched with the features in the fingerprint library, and the most likely location of the mobile tag is determined through machine learning algorithms such as the nearest neighbor algorithm, weighted K-nearest neighbor algorithm, or support vector machine.

[0024] Further, the deep learning algorithm includes: convolutional neural network (CNN), recurrent neural network (RNN) or long short-term memory network (LSTM);

[0025] The convolutional neural network (CNN) is used for feature extraction in video frames;

[0026] The recurrent neural network (RNN) or long short-term memory network (LSTM) is used for time series analysis.

[0027] Further, the convolutional neural network (CNN) is used to extract spatial features from video frames, and the process includes:

[0028] Convolutional neural network (CNN) extracts features and classifies input video frames through multiple layers of convolutional layers, pooling layers, and fully connected layers;

[0029] The convolution layer slides the convolution kernel over the input image, calculates the weighted sum of the local area, and applies an activation function (such as ReLU) to increase nonlinearity;

[0030] The pooling layer downsamples the output of the convolutional layer to reduce the dimension of the data and the amount of computation;

[0031] The fully connected layer maps the convolution and pooling features to the final classification or regression task;

[0032] The convolution operation formula is as follows:

[0033]

[0034] in, is the output of the kth layer at position (i,j), is the weight of the convolution kernel, is the input of the k-1th layer, b k is the bias term, f is the activation function;

[0035] The Long Short-Term Memory Network (LSTM) formula is as follows:

[0036] f t =σ(W f ·[h t-1 , x t ]+b f )(Forget Gate);

[0037] i t =σ(W i ·[h t-1 , x t ]+b f )(input gate);

[0038]

[0039] o t =σ(W o ·[h t-1 , x t ]+b o )(output gate);

[0040] h t =o t tanh(C t )(Hidden state);

[0041] Among them, f t 、i t , o t They are the outputs of the forget gate, input gate, and output gate respectively; is the candidate cell state, C t is the cell state, h t is the hidden state, σ is the sigmoid activation function, and W and b are the weight and bias terms.

[0042] Furthermore, it also includes a wearable device for identifying the behavior of laboratory personnel, and the wearable device includes one or more of an accelerometer, a gyroscope, a magnetometer, a temperature sensor or a heart rate monitor.

[0043] Furthermore, the real-time monitoring of environmental parameters in the laboratory includes one or more of temperature, humidity, light intensity, air quality, and noise level;

[0044] The environmental parameters are collected through the Internet of Things sensor network and uniformly managed and transmitted through the Internet of Things gateway.

[0045] Further, the edge computing preliminary processing includes the following steps:

[0046] The original data is preliminarily screened by setting a threshold, and the sliding window filtering method is used to remove noise data or invalid data;

[0047] The sliding window filtering formula is:

[0048]

[0049] Among them, x i is the i-th data point in the original data sequence; W is the size of the sliding window, y t is the tth data point after filtering. t When the data point exceeds the preset valid data range [L,U], it is considered as noise or invalid data and is removed;

[0050] Use the statistical-based anomaly detection Z-score method to identify outliers in the data;

[0051] The calculation formula of Z-score is:

[0052]

[0053] Among them, x i is the data point, μ is the mean of the data set; σ is the standard deviation of the data set;

[0054] If Z i If it is greater than the preset threshold Zthresh, then x is determined i is an outlier;

[0055] Conduct preliminary statistics and analysis on the filtered data to calculate the movement speed and residence time of personnel;

[0056] The calculation formula of personnel moving speed v is as follows:

[0057]

[0058] Among them, (x1, y1) and (x1, y1) are the coordinate positions of the person at time t1 and time t2 respectively.

[0059] Furthermore, the emergency evacuation path planning algorithm takes into account the laboratory layout, personnel location, obstacle distribution and environmental parameters, uses a graph theory algorithm or a heuristic search algorithm to generate the optimal escape route, and provides escape guidance to laboratory personnel through display screens, wearable devices or mobile applications in the laboratory.

[0060] Furthermore, the laboratory operation report based on big data analysis includes personnel flow analysis, environmental parameter change trends, safety incident statistics or resource utilization evaluation.

[0061] Furthermore, the 3D visualization technology uses WebGL, Unity3D or virtual reality (VR) technology, so that managers can intuitively view laboratory personnel distribution, environmental status and warning information.

[0062] In addition, the personnel dynamic positioning and monitoring method based on the Internet of Things laboratory provided by the present invention has at least the following beneficial effects compared with the prior art:

[0063] By integrating high-precision positioning technologies (such as UWB, RFID, etc.) with deep learning algorithms, this solution achieves real-time tracking and positioning of laboratory personnel and behavior recognition. This not only improves the accuracy and efficiency of laboratory personnel management, but also helps to promptly discover and respond to potential safety hazards. High-precision positioning technology ensures the accuracy and real-time nature of personnel locations, while deep learning algorithms can intelligently analyze personnel behavior patterns, providing strong technical support for laboratory safety management.

[0064] This solution uses the IoT sensor network to monitor the environmental parameters in the laboratory in real time, such as temperature, humidity, light intensity, air quality, and noise level, and performs preliminary processing through edge computing, effectively reducing the processing burden of the cloud server. At the same time, the use of big data analysis technology to conduct in-depth mining and analysis of environmental parameters can timely detect abnormal changes in environmental parameters and trigger early warning signals to ensure the safety and normal operation of the laboratory. This real-time monitoring and early warning mechanism provides laboratory managers with timely and accurate information support, helping them make scientific and reasonable management decisions.

[0065] In an emergency, this solution can generate the optimal escape route based on the laboratory layout, personnel location, obstacle distribution, and environmental parameters using graph theory algorithms or heuristic search algorithms, and provide escape guidance to laboratory personnel through multiple channels (such as display screens, wearable devices, mobile applications, etc.). This not only improves the self-rescue ability of laboratory personnel in emergencies, but also ensures their life safety. In addition, by continuously optimizing the emergency evacuation path planning algorithm, this solution can also adapt to changes in different laboratory environments and personnel distribution, providing more comprehensive and reliable support for laboratory safety management. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Figure 1 It is a flow chart of a method for dynamic positioning and monitoring of personnel based on an Internet of Things laboratory of the present invention;

[0067] Figure 2 The figure is a block diagram of the overall principle of the personnel dynamic positioning and monitoring method based on the Internet of Things laboratory of the present invention. DETAILED DESCRIPTION

[0068] The principles and features of the present invention are described below in conjunction with the accompanying drawings. The examples given are only used to explain the present invention and are not used to limit the scope of the present invention.

[0069] The dynamic positioning and monitoring system for laboratory personnel based on the Internet of Things is an important application of the Internet of Things technology in the field of laboratory management. The system integrates multiple technical fields such as high-precision positioning technology, deep learning algorithms, Internet of Things gateways, edge computing, cloud servers, and big data analysis to achieve real-time tracking and positioning of laboratory personnel, behavior recognition, environmental monitoring, data analysis and visualization, and other functions, providing strong technical support for laboratory safety management and operational efficiency improvement.

[0070] In order to further improve the safety management level and operational efficiency of the laboratory, as well as solve customer limitations and shortcomings, the present invention also designs a personnel dynamic positioning and monitoring method based on the Internet of Things laboratory to solve the above problems.

[0071] like Figure 1 and Figure 2 As shown, the personnel dynamic positioning and monitoring method based on the Internet of Things laboratory involved in this embodiment includes the following steps:

[0072] Use high-precision positioning technology, combined with the laboratory layout map, to track and locate laboratory personnel in real time with high precision;

[0073] Use deep learning algorithms to process video data from internal laboratory monitoring, identify the actions of laboratory personnel, and analyze behavioral patterns based on time series; (Laboratory personnel can use wearable devices to improve the accuracy and robustness of behavior recognition)

[0074] Monitor the environmental parameters in the laboratory in real time, collect data uniformly through the IoT gateway, and upload them to the cloud server after preliminary processing through edge computing;

[0075] According to the preset safety rules, the pre-tensioning signal is automatically triggered and pushed to the management personnel;

[0076] Design an emergency evacuation route planning algorithm based on the preload signal; if an emergency is detected, provide the laboratory personnel with the optimal escape route;

[0077] Provide laboratory operation reports based on big data analysis to assist laboratory managers in making management decisions, and use 3D visualization technology to display laboratory personnel distribution, environmental status and early warning information.

[0078] By integrating multiple technical fields such as high-precision positioning technology, deep learning algorithms, IoT gateways, edge computing, cloud servers, and big data analysis, we can achieve real-time tracking and positioning of laboratory personnel, behavior recognition, environmental monitoring, early warning signal processing, emergency evacuation route planning, and laboratory operation data analysis and visualization.

[0079] Specifically:

[0080] High-precision real-time tracking and positioning: Use high-precision positioning technology (such as UWB, RFID, etc.) combined with the laboratory layout map to track and accurately locate laboratory personnel in real time to ensure the accuracy and real-time nature of personnel location information.

[0081] Behavior recognition and pattern analysis: Deep learning algorithms are used to process the video data of internal laboratory monitoring, identify the actions of laboratory personnel, and analyze behavioral patterns based on time series to achieve intelligent monitoring and analysis of personnel behavior. At the same time, the combination of wearable devices can improve the accuracy and robustness of behavior recognition.

[0082] Environmental monitoring and data collection: Real-time monitoring of environmental parameters in the laboratory (such as temperature, humidity, gas concentration, etc.), and unified data collection through the IoT gateway. After preliminary processing by edge computing, the data is uploaded to the cloud server for further analysis and processing.

[0083] Early warning signal processing and push: According to the preset safety rules, the pre-tensioning signal is automatically triggered and the early warning information is pushed to the management personnel. This helps the management personnel to timely discover potential safety hazards and take corresponding measures to deal with them.

[0084] Emergency evacuation route planning: When an emergency situation (such as fire, gas leakage, etc.) is detected, an emergency evacuation route planning algorithm is designed based on the preload signal and laboratory layout diagram to provide the laboratory personnel with the optimal escape route and ensure the safe evacuation of personnel.

[0085] Laboratory operation data analysis and visualization: Use big data analysis technology to process and analyze laboratory operation data to provide valuable insights. At the same time, use 3D visualization technology to display laboratory personnel distribution, environmental status and early warning information, so that managers can intuitively understand the real-time status and problems of the laboratory, so as to make more scientific and reasonable management decisions.

[0086] As an implementation method, the high-precision positioning technology includes at least one of UWB (ultra-wideband), RFID (radio frequency identification), Bluetooth low energy or GPS (global positioning system), and is combined with fixed reference points in the laboratory to achieve real-time tracking and positioning through triangulation or fingerprint positioning algorithm.

[0087] Specifically, the triangulation algorithm includes: using the signal propagation time difference (Δt_A, Δt_B, Δt_C) or phase difference (Φ_A, Φ_B, Φ_C) between at least three fixed reference points (A, B, C) and the mobile tag (M), combined with the speed of light (c) to calculate the distance (d_A, d_B, d_C), and then determining the coordinates (x, y, z) of the mobile tag through the distance intersection formula in three-dimensional space;

[0088] The calculation formula is as follows:

[0089]

[0090] Among them, (x_A, y_A, z_A), (x_B, y_B, z_B), (x_C, y_C, z_C) are the coordinates of fixed reference points A, B, and C respectively;

[0091] The fingerprint positioning algorithm includes: collecting wireless signal features (such as signal strength, phase information, etc.) at multiple locations in the laboratory in advance, and building a signal feature database (i.e., "fingerprint library"). During positioning, the real-time signal features collected by the mobile tag are matched with the features in the fingerprint library, and the most likely location of the mobile tag is determined through machine learning algorithms such as the nearest neighbor algorithm, weighted K-nearest neighbor algorithm, or support vector machine.

[0092] To further illustrate the above-mentioned real-time tracking and positioning by triangulation or fingerprint positioning algorithm, the following embodiments are also provided:

[0093] Example 1

[0094] In a typical IoT laboratory environment, four UWB fixed reference points (A, B, C, D, where D is used as a backup or to improve positioning accuracy) are deployed, and the precise coordinates of each reference point are known. Laboratory personnel wear UWB mobile tags.

[0095] Triangulation algorithm implementation: When the laboratory personnel move, the UWB tag continuously sends signals to four reference points. After receiving the signal, the reference point records the timestamp and calculates the signal propagation time difference. Using the above triangulation formula, combined with the speed of light c, the distance between the tag and each reference point can be calculated. Through the intersection of distances in three-dimensional space, the precise location of the laboratory personnel can be determined in real time.

[0096] Implementation of fingerprint positioning algorithm: At different locations in the laboratory (such as setting a sampling point every 1 meter), the strength and phase of the UWB signal are collected in advance to build a fingerprint library. During positioning, the UWB tag collects the signal characteristics of the current location and matches them with the characteristics in the fingerprint library. The weighted K nearest neighbor algorithm is used to calculate the weighted average position as the current location of the laboratory personnel based on the positions of the K fingerprint points with the highest matching degree.

[0097] As an implementation, the deep learning algorithm includes: a convolutional neural network (CNN), a recurrent neural network (RNN), or a long short-term memory network (LSTM);

[0098] The convolutional neural network (CNN) is used for feature extraction in video frames;

[0099] The recurrent neural network (RNN) or long short-term memory network (LSTM) is used for time series analysis.

[0100] Specifically, the convolutional neural network (CNN) is used to extract spatial features from video frames, and the process includes:

[0101] Convolutional neural network (CNN) extracts features and classifies input video frames through multiple layers of convolutional layers, pooling layers, and fully connected layers;

[0102] The convolutional layer slides over the input image through the convolution kernel (also called filter), calculates the weighted sum of the local area, and applies an activation function (such as ReLU) to increase nonlinearity;

[0103] The pooling layer downsamples the output of the convolutional layer to reduce the dimension of the data and the amount of computation;

[0104] The fully connected layer maps the convolution and pooling features to the final classification or regression task;

[0105] The convolution operation formula is as follows:

[0106]

[0107] in, is the output of the kth layer at position (i,j), is the weight of the convolution kernel, is the input of the k-1th layer, b k is the bias term, f is the activation function;

[0108] Recurrent Neural Network (RNN) and its variant Long Short-Term Memory (LSTM): used for time series analysis, especially for processing time-dependent data. RNN introduces recurrent connections, allowing the network to remember previous states, thereby processing time dependencies in sequence data. However, traditional RNNs are prone to gradient vanishing or gradient exploding problems when processing long sequences. LSTM solves these problems by introducing input gates, forget gates, and output gates, as well as cell states, and can capture long-term dependencies more effectively.

[0109] Preferably, the long short-term memory network (LSTM) formula is as follows:

[0110] ft =σ(W f ·[h t-1 , x t ]+b f )(Forget Gate);

[0111] i t =σ(W i ·[h t-1 , x t ]+b f )(input gate);

[0112]

[0113] o t =σ(W o ·[h t-1 , x t ]+b o )(output gate);

[0114] h t =o t tanh(C t )(Hidden state);

[0115] Among them, f t 、i t , o t They are the outputs of the forget gate, input gate, and output gate respectively; is the candidate cell state, C t is the cell state, h t is the hidden state, σ is the sigmoid activation function, and W and b are the weight and bias terms.

[0116] In order to further explain the above deep learning algorithm in detail, the following embodiments are also provided:

[0117] Embodiment 2:

[0118] In an IoT laboratory, in order to achieve dynamic positioning and monitoring of personnel, we combined video surveillance and sensor data. Video surveillance is used to capture real-time images in the laboratory, while sensor data (such as UWB, RFID, etc.) provides personnel location information.

[0119] CNN application: First, we use CNN to process the video frames and extract the spatial features of people, such as posture, position, etc. These features can be used for preliminary person detection and tracking.

[0120] RNN / LSTM application: We then input the feature sequence extracted by CNN and the sensor data into the RNN or LSTM network for time series analysis. RNN / LSTM can capture the temporal dependency of personnel movement, such as speed changes, dwell time, etc., thereby achieving a deeper understanding of personnel behavior.

[0121] By combining CNN and RNN / LSTM, we can achieve dynamic positioning and monitoring of laboratory personnel. We can not only track the location of personnel in real time, but also analyze their behavior patterns, providing strong support for laboratory safety management and efficiency improvement.

[0122] As an embodiment, a wearable device for identifying laboratory personnel's behavior is also included, and the wearable device includes one or more of an accelerometer, a gyroscope, a magnetometer, a temperature sensor, or a heart rate monitor. The combined use of these sensors can more comprehensively capture and record various physiological and behavioral data of laboratory personnel.

[0123] Accelerometer: used to measure the acceleration changes of wearable devices (i.e. laboratory personnel) in three-dimensional space, so as to infer the movement state of the person, such as walking, running, stillness, etc.

[0124] Gyroscope: Provides information about changes in orientation by measuring the rate of rotation around an axis. Combined with an accelerometer, it can more accurately track a person's movement and direction.

[0125] Magnetometer: Used to measure the direction and strength of the Earth's magnetic field, thereby helping to determine the absolute direction of the wearable device and further improving the accuracy of position and behavior recognition.

[0126] Temperature sensor: used to monitor the body temperature or ambient temperature of laboratory personnel and provide data support for health monitoring and environmental adaptability analysis.

[0127] Heart rate monitor: Real-time measurement of lab personnel’s heart rate helps assess personnel’s health and ability to cope with stress.

[0128] The data collected by these wearable devices can be combined with deep learning algorithms to more accurately identify the movements and behavior patterns of laboratory personnel. For example, by analyzing the data of accelerometers and gyroscopes, the gait, posture, and possible abnormal behaviors (such as falls, sudden acceleration, etc.) of personnel can be identified. At the same time, combined with magnetometers and temperature sensors, the scenarios and contexts of behavior recognition can be further refined to improve the accuracy and robustness of recognition.

[0129] In addition, these wearable devices can also be linked with other IoT devices in the laboratory (such as tags of high-precision positioning systems, cameras of video surveillance systems, etc.) to form a more comprehensive dynamic monitoring system for laboratory personnel. By integrating data from multiple sources, it is possible to achieve all-round and multi-angle monitoring and analysis of laboratory personnel behavior, providing more powerful support for laboratory safety management and operational efficiency improvement.

[0130] As an embodiment, the real-time monitoring of environmental parameters in the laboratory includes one or more of temperature, humidity, light intensity, air quality, and noise level; these parameters have an important impact on the health of laboratory personnel, the normal operation of experimental equipment, and the accuracy of experimental results.

[0131] The environmental parameters are collected through the Internet of Things sensor network and uniformly managed and transmitted through the Internet of Things gateway.

[0132] For example, temperature sensors are used to measure the temperature in the laboratory, humidity sensors are used to monitor humidity levels, light intensity sensors are used to assess indoor light conditions, air quality sensors are used to detect the concentration of harmful substances in the air, and noise sensors are used to measure the noise level in the laboratory.

[0133] The collected environmental parameter data is uniformly managed and transmitted through the IoT gateway. As a bridge connecting the sensor network and the cloud server, the IoT gateway is responsible for receiving data from sensors and performing preliminary processing and integration. The data is then uploaded to the cloud server for further analysis, storage and visualization.

[0134] On cloud servers, we can use big data analysis technology to conduct in-depth mining and analysis of these environmental parameter data. By comparing historical data, setting thresholds, and applying machine learning algorithms, we can identify abnormal changes in environmental parameters and trigger early warning signals in a timely manner. These early warning signals can be pushed to the mobile devices or workstations of laboratory managers or relevant personnel so that they can take timely countermeasures to ensure the safety and normal operation of the laboratory.

[0135] As an implementation method, the preliminary edge computing processing includes data filtering, anomaly detection or preliminary data analysis to reduce the processing burden of the cloud server and improve the real-time and efficiency of data processing.

[0136] The edge computing preliminary processing includes the following steps:

[0137] The original data is preliminarily screened by setting a threshold, and the sliding window filtering method is used to remove noise data or invalid data;

[0138] The sliding window filtering formula is:

[0139]

[0140] Among them, x i is the i-th data point in the original data sequence; W is the size of the sliding window, y t is the tth data point after filtering. t When the data point exceeds the preset valid data range [L,U], it is considered as noise or invalid data and is removed;

[0141] Use the statistical-based anomaly detection Z-score method to identify outliers in the data;

[0142] The calculation formula of Z-score is:

[0143]

[0144] Among them, x i is the data point, μ is the mean of the data set; σ is the standard deviation of the data set;

[0145] If Z i If it is greater than the preset threshold Zthresh, then x is determined i is an outlier;

[0146] Conduct preliminary statistics and analysis on the filtered data to calculate the movement speed and residence time of personnel;

[0147] The calculation formula of personnel moving speed v is as follows:

[0148]

[0149] Among them, (x1, y1) and (x1, y1) are the coordinate positions of the person at time t1 and time t2 respectively.

[0150] In order to explain the preliminary processing of edge computing in detail, the following embodiments are also provided:

[0151] Example 3

[0152] In an IoT-based laboratory environment, multiple sensors are deployed to collect personnel location information. In order to improve the real-time and efficiency of data processing, edge computing devices are used for preliminary processing.

[0153] Data filtering: Set the sliding window size W = 5, and perform sliding window filtering on the original position data. If the filtered data point yt is not in the valid data range [L = 0, U = 100] (assuming that the position data is expressed in a certain unit, and data outside this range is considered invalid), then the data point is removed.

[0154] Anomaly detection: Calculate the Z-score of each position data and set the threshold Zthresh = 3. If the absolute value of the Z-score of a data point is greater than 3, the data point is judged to be an outlier, which may be caused by sensor failure or data transmission error and should be ignored.

[0155] Preliminary data analysis: For the filtered and anomaly-detected data, calculate the movement speed of the person. For example, if a person is located at (x1=10, y1=20) at time t1=0 seconds and at (x2=30, y2=40) at time t2=5 seconds, then the movement speed is calculated according to the formula:

[0156] Through the above-mentioned preliminary edge computing processing, the processing burden of the cloud server is effectively reduced, the real-time and efficiency of data processing are improved, and a more accurate and reliable data foundation is provided for the subsequent dynamic positioning and monitoring of personnel.

[0157] As an implementation method, the emergency evacuation path planning algorithm takes into account the laboratory layout, personnel location, obstacle distribution and environmental parameters, uses a graph theory algorithm or a heuristic search algorithm to generate the optimal escape route, and provides escape guidance to laboratory personnel through display screens, wearable devices or mobile applications in the laboratory.

[0158] Specifically, laboratory layout and obstacle modeling:

[0159] The nodes and edges in graph theory are used to represent the spatial structure of the laboratory, where nodes represent key points (such as doorways, corridor intersections, etc.) and edges represent channels.

[0160] Obstacles are abstracted as impassable edges or nodes and are removed from the graph or given extremely large weights.

[0161] Integration of personnel location and environmental parameters:

[0162] The location of personnel is obtained in real time through IoT devices and marked as dynamic nodes in the graph.

[0163] Environmental parameters (such as smoke concentration, fire source location, etc.) are integrated through sensor data to affect the weight of the edge. The weight function can be defined as:

[0164] w(e)=w base (e)+α·Smoke(e)+β·FireProximity(e)+γ·

[0165] Other Factors(e);

[0166] Among them, w(e) is the weight of edge e, w base (e) is the basic weight of edge e, α is the weight coefficient of smoke concentration, Smoke(e) is the smoke concentration in the area where edge e is located, β is the weight coefficient of the fire source location, FireProximity(e) is the distance or proximity from edge e to the nearest fire source, γ is the weight coefficient of other environmental factors (including temperature, humidity, toxic gas concentration, etc., depending on the specific conditions of the laboratory and the sensor configuration. By adjusting the value of γ, the importance of these factors in path planning can be emphasized or ignored), and OtherFactors(e) is a comprehensive measure of other environmental factors in the area where edge e is located.

[0167] Path planning algorithm:

[0168] Use Dijkstra algorithm or A* algorithm to search for paths.

[0169] Dijkstra algorithm: Applicable to the case where the weights of all edges are non-negative, and gradually expands the shortest path tree.

[0170] A* algorithm: Use heuristic functions (such as Manhattan distance) to guide the search direction and speed up the search for the optimal path.

[0171] The objective function is to find the shortest path from the current position of the person to the nearest safe exit (target node), that is, to minimize the total weight of the path.

[0172] Escape instructions released:

[0173] The calculated optimal path is displayed to laboratory personnel in real time through display screens, wearable devices (such as smart bracelets, AR glasses) or mobile applications (such as mobile phone APP) in the laboratory.

[0174] In order to explain the above emergency evacuation path planning algorithm in detail, the following embodiments are also provided:

[0175] Example 4

[0176] Emergency evacuation route planning based on Dijkstra algorithm

[0177] Laboratory layout: Assume that the laboratory is a 5x5 grid, each grid represents an area, and some grids are marked as obstacles.

[0178] Personnel location: Real-time acquisition of a person’s location at (2,2).

[0179] Environmental parameters: The smoke sensor detects that the smoke density in areas (3,3) and (4,4) is high, and the fire source is located at (4,3).

[0180] Weight setting: The basic weight is 1. For every unit increase in smoke density, the weight increases by 0.5. For every unit increase in the proximity of the fire source (calculated by Manhattan distance), the weight increases by 1. Other factors are not considered for the time being.

[0181] Path planning: Using Dijkstra algorithm, starting from (2,2), calculate the shortest path to all reachable nodes, and finally find the shortest path to the nearest safe exit (0,0).

[0182] Escape Guidance: Path guidance is displayed on the display screen, such as "follow the left-up-left direction to reach the safe exit."

[0183] Example 5

[0184] Emergency evacuation route planning based on A* algorithm

[0185] Laboratory layout and personnel positions: Same as Example 4.

[0186] Environmental parameters: The smoke and fire source locations are also detected, but the temperature factor is added, and the weight of the high temperature area (4,2) is increased by an additional 1.

[0187] Heuristic function: Use Manhattan distance as the heuristic function to estimate the straight-line distance from the current node to the target node.

[0188] Path planning: Use the A* algorithm, combined with heuristic functions and weight functions, to quickly find the optimal path from (2,2) to (0,0).

[0189] Escape guidance: Wearable devices (such as AR glasses) use virtual arrows to guide people to escape along the optimal path in real time.

[0190] The above examples show how to use graph theory algorithms or heuristic search algorithms to plan and guide emergency evacuation routes based on the specific conditions and real-time data of the laboratory.

[0191] As an implementation mode, the laboratory operation report based on big data analysis includes personnel flow analysis, environmental parameter change trends, safety incident statistics or resource utilization evaluation.

[0192] The Internet of Things technology is used to collect multi-dimensional data in the laboratory in real time (including information on personnel flow, environmental parameters, safety incidents, and resource utilization), and then the data is deeply mined, integrated, and displayed using big data analysis and visualization technology. This process aims to reveal key indicators and trends in laboratory operations and provide laboratory managers with comprehensive and accurate operational insights. Based on these reports, managers can evaluate the effectiveness of the laboratory, identify potential problems, and formulate and optimize management strategies to improve the laboratory's operational efficiency, safety, and resource utilization.

[0193] As an implementation method, the 3D visualization technology uses WebGL, Unity3D or virtual reality (VR) technology, so that managers can intuitively view laboratory personnel distribution, environmental status and warning information.

[0194] Using advanced graphics rendering and interactive technologies such as WebGL, Unity3D or virtual reality (VR), a three-dimensional virtual laboratory environment is constructed. In this environment, the actual layout of the laboratory, personnel distribution, environmental status (such as temperature, humidity, etc.) and early warning information are accurately simulated and displayed. Managers can intuitively browse and interact with this three-dimensional model by wearing VR equipment or watching it directly on the computer screen, thereby obtaining a comprehensive and three-dimensional perception of the current status of the laboratory.

[0195] Specifically, WebGL technology allows three-dimensional graphics to be rendered directly in a web browser without installing additional software, facilitating cross-platform access; Unity3D is a powerful game engine that supports complex three-dimensional scene construction and real-time rendering, making it ideal for creating interactive laboratory simulations; and VR technology provides an immersive experience, allowing managers to feel as if they are in a real laboratory environment, allowing them to have a deeper understanding and experience of the laboratory's operating status.

[0196] Through these technologies, managers can easily view the location of personnel in the laboratory, real-time changes in environmental parameters, and any potential safety hazards or warning information. This intuitive, real-time monitoring method helps them make decisions faster and improve laboratory management efficiency and safety.

[0197] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0198] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A personnel dynamic positioning and monitoring method based on the Internet of Things laboratory, characterized in that: The following steps are involved: Use high-precision positioning technology, combined with the laboratory layout map, to track and locate laboratory personnel in real time with high precision; Use deep learning algorithms to process video data from internal laboratory surveillance, identify the actions of laboratory personnel, and analyze behavioral patterns based on time series; Monitor the environmental parameters in the laboratory in real time, collect data uniformly through the IoT gateway, and upload them to the cloud server after preliminary processing through edge computing; According to the preset safety rules, the pre-tensioning signal is automatically triggered and pushed to the management personnel; Based on the preload signal, an emergency evacuation route planning algorithm is designed; Provide optimal escape routes for laboratory personnel if an emergency is detected; Provide laboratory operation reports based on big data analysis to assist laboratory managers in making management decisions, and use 3D visualization technology to display laboratory personnel distribution, environmental status and early warning information.

2. The method for dynamic positioning and monitoring of personnel based on the Internet of Things laboratory according to claim 1 is characterized in that: The high-precision positioning technology includes at least one of UWB (ultra-wideband), RFID (radio frequency identification), Bluetooth low energy or GPS (global positioning system), and is combined with fixed reference points in the laboratory to achieve real-time tracking and positioning through triangulation or fingerprint positioning algorithm.

3. The personnel dynamic positioning and monitoring method based on the Internet of Things laboratory according to claim 2 is characterized in that: The triangulation algorithm includes: using the signal propagation time difference (Δt_A, Δt_B, Δt_C) or phase difference (Φ_A, Φ_B, Φ_C) between at least three fixed reference points (A, B, C) and the mobile tag (M), combined with the speed of light (c) to calculate the distance (d_A, d_B, d_C), and then determining the coordinates (x, y, z) of the mobile tag through the distance intersection formula in three-dimensional space; The calculation formula is as follows: Among them, (x_A,y_A,z_A),(x_B,y_B,z_B),(x_C,y_C,z_C) are the coordinates of fixed reference points A, B, and C respectively.

4. The method for dynamic positioning and monitoring of personnel based on the Internet of Things laboratory according to claim 2 is characterized in that: The fingerprint positioning algorithm includes: collecting wireless signal features at multiple locations in the laboratory in advance and building a signal feature database; During positioning, the real-time signal features collected by the mobile tag are matched with the features in the fingerprint library, and the most likely location of the mobile tag is determined through machine learning algorithms such as the nearest neighbor algorithm, weighted K-nearest neighbor algorithm, or support vector machine.

5. The method for dynamic positioning and monitoring of personnel based on the Internet of Things laboratory according to claim 1 is characterized in that: The deep learning algorithm includes: convolutional neural network (CNN), recurrent neural network (RNN) or long short-term memory network (LSTM); The convolutional neural network (CNN) is used for feature extraction in video frames; The recurrent neural network (RNN) or long short-term memory network (LSTM) is used for time series analysis.

6. The method for dynamic positioning and monitoring of personnel based on the Internet of Things laboratory according to claim 5 is characterized in that: The convolutional neural network (CNN) is used to extract spatial features from video frames, and the process includes: Convolutional neural network (CNN) extracts features and classifies input video frames through multiple layers of convolutional layers, pooling layers, and fully connected layers; The convolution layer slides the convolution kernel over the input image, calculates the weighted sum of the local area, and applies an activation function (such as ReLU) to increase nonlinearity; The pooling layer downsamples the output of the convolutional layer to reduce the dimension of the data and the amount of computation; The fully connected layer maps the convolution and pooling features to the final classification or regression task; The convolution operation formula is as follows: in, is the output of the kth layer at position (i,j), is the weight of the convolution kernel, is the input of the k-1th layer, b k is the bias term, f is the activation function; The Long Short-Term Memory Network (LSTM) formula is as follows: f t =σ(W f ·[h t-1 , x t ]+b f )(Forget Gate); i t =σ(W i ·[h t-1 , x t ]+b f )(input gate); o t =σ(W o ·[h t-1 , x t ]+b o )(output gate); h t =o t tanh(C t )(Hidden state); Among them, f t 、i t , o t They are the outputs of the forget gate, input gate, and output gate respectively; is the candidate cell state, C t is the cell state, h t is the hidden state, σ is the sigmoid activation function, and W and b are the weight and bias terms.

7. The method for dynamic positioning and monitoring of personnel based on the Internet of Things laboratory according to claim 1 is characterized in that: It also includes a wearable device for identifying the behavior of laboratory personnel, wherein the wearable device includes one or more of an accelerometer, a gyroscope, a magnetometer, a temperature sensor or a heart rate monitor.

8. The method for dynamic positioning and monitoring of personnel based on the Internet of Things laboratory according to claim 1 is characterized in that: The real-time monitoring of environmental parameters in the laboratory includes one or more of temperature, humidity, light intensity, air quality, and noise level; The environmental parameters are collected through the Internet of Things sensor network and uniformly managed and transmitted through the Internet of Things gateway.

9. The method for dynamic positioning and monitoring of personnel based on the Internet of Things laboratory according to claim 1 is characterized in that: The edge computing preliminary processing includes the following steps: The original data is preliminarily screened by setting a threshold, and the sliding window filtering method is used to remove noise data or invalid data; The sliding window filtering formula is: Among them, x i is the i-th data point in the original data sequence; W is the size of the sliding window, y t is the tth data point after filtering. t When the data point exceeds the preset valid data range [L,U], it is considered as noise or invalid data and is removed; Use the statistical-based anomaly detection Z-score method to identify outliers in the data; The calculation formula of Z-score is: Among them, x i is the data point, μ is the mean of the data set; σ is the standard deviation of the data set; If Z i If it is greater than the preset threshold Zthresh, then x is determined i is an outlier; Conduct preliminary statistics and analysis on the filtered data to calculate the movement speed and residence time of personnel; The calculation formula of personnel moving speed v is as follows: Among them, (x1, y1) and (x1, y1) are the coordinate positions of the person at time t1 and time t2 respectively.

10. The method for dynamic positioning and monitoring of personnel based on the Internet of Things laboratory according to claim 7 is characterized in that: The emergency evacuation path planning algorithm takes into account the laboratory layout, personnel location, obstacle distribution and environmental parameters, uses a graph theory algorithm or a heuristic search algorithm to generate the optimal escape route, and provides escape guidance to laboratory personnel through display screens, wearable devices or mobile applications in the laboratory.

11. The method for dynamic positioning and monitoring of personnel based on the Internet of Things laboratory according to claim 1 is characterized in that: The laboratory operation report based on big data analysis includes personnel flow analysis, environmental parameter change trends, safety incident statistics or resource utilization evaluation.

12. The method for dynamic positioning and monitoring of personnel based on the Internet of Things laboratory according to claim 1 is characterized in that: The 3D visualization technology uses WebGL, Unity3D or virtual reality (VR) technology, so that managers can intuitively view laboratory personnel distribution, environmental status and warning information.

Images