Safety risk early warning method, system and equipment for constructors working aloft under high-voltage electric tower and medium

Abstract

The invention discloses a safety risk early warning method, system and equipment for constructors working aloft under a high-voltage electric tower and a medium. The method comprises the following steps: acquiring environmental data of the high-voltage electric tower and vital sign data of the constructors; according to the environment data, the vital sign data and a pre-generated working condition determination model, determining a working condition corresponding to the current high-altitude operation; and according to the environment data, the vital sign data and a pre-generated safety early warning model corresponding to the working condition, carrying out early warning on the safety risk of the constructor. According to the invention, the machine learning algorithm (spectral clustering and isolated forest) is utilized to intelligently and dynamically analyze the real-time environment and personnel data of high-voltage electric tower high-altitude operation, so that accurate working condition identification and targeted risk early warning are realized, and the accuracy and efficiency of safety monitoring are remarkably improved; finally, the purposes of actively preventing accidents and guaranteeing life safety of constructors are achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of construction safety early warning, and in particular to a method, system, equipment and medium for early warning of safety risks for construction workers performing high-altitude operations under high-voltage power towers. Background Art

[0002] High-altitude work under high-voltage towers is a high-risk scenario in the power industry. Existing safety risk warning methods face many technical challenges. The main technical issues include:

[0003] High-voltage power towers are exposed to extremely strong power-frequency electromagnetic fields (50 / 60 Hz) and potential transient electromagnetic interference (e.g., from switching operations and lightning strikes). Conventional electronic sensors (such as accelerometers, gyroscopes, proximity sensors, and some gas sensors) are susceptible to interference in this environment, resulting in signal distortion, drift, false alarms, or failure. This makes it impossible to accurately monitor critical safety information, such as personnel posture (to prevent falls), safe distance from live objects (to prevent electric shock), and environmental parameters (e.g., oxygen and hazardous gases).

[0004] Furthermore, high-voltage towers are large metal structures that severely block GPS, Beidou, and other satellite signals, creating multipath effects that significantly reduce or even disable positioning accuracy. Similarly, when radio waves (such as Wi-Fi, Bluetooth, Zigbee, and UWB) propagate through dense metal structures, they experience severe signal attenuation and complex, unpredictable paths, making it difficult to guarantee the accuracy of indoor positioning technologies based on wireless signals.

[0005] Finally, the intensity of the high-voltage electric field varies dynamically with voltage level, phase distance, and human posture and position. Accurately and in real time, measuring the dynamic safety distance between a worker's body parts (especially their hands and feet) and live wires is extremely difficult.

[0006] Existing proximity warning devices (such as those based on electric field strength sensing or ultrasonic / laser ranging) have limited accuracy in complex configurations and are susceptible to interference or misjudgment (e.g., triggering the approach of metal tools), making them difficult to provide reliable electric shock warnings. They are also unable to accurately detect changes in risk caused by subtle movements such as workers stretching their limbs. Summary of the Invention

[0007] The purpose of the present invention is to at least provide a safety risk warning method and equipment for construction workers working at height under high-voltage power towers, so as to solve at least part of the above-mentioned technical problems.

[0008] Specifically, to solve the above technical problems, at least one embodiment of the present application provides a safety risk warning method for construction workers working at height under a high-voltage power tower, comprising:

[0009] Obtain environmental data from high-voltage towers and vital signs data from construction workers;

[0010] Determining a working condition corresponding to the current high-altitude operation based on the environmental data, the vital sign data, and a pre-generated working condition determination model; wherein the working condition determination model is generated based on a spectral clustering algorithm, and the number of connected components of the Laplacian matrix of the working condition determination model is determined by the number of outliers in the vital sign data;

[0011] An early warning of safety risks to the construction workers is issued based on the environmental data, the vital signs data, and a pre-generated safety early warning model corresponding to the working conditions, wherein the safety early warning model is generated based on an isolation forest algorithm; wherein the depth of the tree of the safety early warning model is determined by the number of categories of the vital signs data.

[0012] In some embodiments, the environmental data includes: the height, altitude, air pressure, wind speed, humidity, rainfall, and temperature of the high-voltage tower;

[0013] The vital signs data include: the construction worker's body temperature, pulse, heart rate, respiratory rate and blood pressure.

[0014] In some embodiments, the step of generating the operating condition determination model includes:

[0015] Generating similarity graphs of the height, altitude, air pressure, wind speed, humidity, rainfall, temperature, body temperature, pulse, heart rate, respiratory rate, and blood pressure respectively;

[0016] generating an expected corresponding Laplacian matrix according to the similarity graph;

[0017] performing eigendecomposition on the Laplace matrix;

[0018] generating a space matrix of the Laplacian matrix according to the structure of the eigendecomposition;

[0019] In the feature space corresponding to the spatial matrix, cluster division is performed on the spatial matrix to generate the operating condition determination model.

[0020] In some embodiments, before performing eigendecomposition on the Laplacian matrix, the method further includes:

[0021] performing random walk normalization on the Laplace matrix;

[0022] Generating an expected corresponding Laplacian matrix according to the similarity graph includes:

[0023] generating a degree matrix of the height, altitude, air pressure, wind speed, humidity, rainfall, temperature, body temperature, pulse, heart rate, respiratory rate and blood pressure according to the similarity graph;

[0024] The Laplacian matrix is ​​generated according to the degree matrix.

[0025] In some embodiments, the step of generating the security early warning model includes:

[0026] Randomly extracting characteristic dimensions of the environmental data and the vital sign data under corresponding working conditions;

[0027] Randomly selecting a first segmentation point between the maximum value and the minimum value of the characteristic dimension of the environmental data;

[0028] Randomly selecting a second segmentation point between the maximum value and the minimum value of the characteristic dimension of the vital sign data;

[0029] Dividing the environmental data and the vital sign data into four subtrees according to the first segmentation point and the second segmentation point;

[0030] The above steps are iterated on the four subtrees until preset conditions are met to generate the safety early warning model.

[0031] In some embodiments, the vital sign data further includes: time series data of skeletal points of the construction worker;

[0032] The construction worker safety risk early warning method further includes:

[0033] Generate a graph structure of joint points according to the skeleton point time series data;

[0034] Performing a graph convolution operation on the graph structure to extract spatial topological features of the skeleton point time series data;

[0035] Performing a temporal convolution operation on the spatial topological features to extract motion evolution features of the skeleton point temporal data;

[0036] The safety risk of the construction workers is warned based on the safety warning model of the working conditions corresponding to the skeleton point time series data, the motion evolution characteristics, the height, altitude, air pressure, wind speed, humidity, rainfall, temperature, body temperature, pulse, heart rate, respiratory rate and blood pressure.

[0037] In some embodiments, providing an early warning of the safety risk of the construction personnel based on the environmental data, the vital sign data, and a pre-generated safety early warning model corresponding to the working condition includes:

[0038] Generate time series data having the environmental data and the vital sign data; wherein each time point in the time series data has a feature vector of the environmental data and a feature vector of the vital sign data;

[0039] Determining a risk value corresponding to a current time point according to the safety warning model, the characteristic vector of the environmental data, and the characteristic vector of the vital sign data;

[0040] An early warning is issued to the construction workers based on the risk value and a preset early warning threshold.

[0041] At least one embodiment of the present application further provides a safety risk warning device for construction workers performing high-altitude operations under high-voltage power towers, comprising:

[0042] Input data acquisition module, used to obtain environmental data of high-voltage towers and vital signs data of construction workers;

[0043] a working condition determination module, configured to determine the working condition corresponding to the current high-altitude operation based on the environmental data, the vital sign data, and a pre-generated working condition determination model; wherein the working condition determination model is generated based on a spectral clustering algorithm, and the number of connected components of the Laplacian matrix of the working condition determination model is determined by the number of outliers in the vital sign data;

[0044] The first safety risk warning module is used to warn the construction workers of safety risks based on the environmental data, the vital signs data and a pre-generated safety warning model corresponding to the working conditions, wherein the safety warning model is generated based on the isolation forest algorithm; wherein the depth of the tree of the safety warning model is determined by the number of categories of the vital signs data.

[0045] In some embodiments, the environmental data includes: the height, altitude, air pressure, wind speed, humidity, rainfall, and temperature of the high-voltage tower;

[0046] The vital signs data include: the construction worker's body temperature, pulse, heart rate, respiratory rate and blood pressure.

[0047] In some embodiments, a safety risk warning device for construction workers performing high-altitude operations under a high-voltage power tower further includes:

[0048] The operating condition determination model generation module is used to generate the operating condition determination model; the operating condition determination model generation module includes:

[0049] a similarity graph generating unit, configured to respectively generate similarity graphs of the height, altitude, air pressure, wind speed, humidity, rainfall, temperature, body temperature, pulse, heart rate, respiratory rate, and blood pressure;

[0050] A Laplace matrix generating unit, configured to generate an expected corresponding Laplace matrix according to the similarity graph;

[0051] a Laplace matrix eigendecomposition unit, configured to perform eigendecomposition on the Laplace matrix;

[0052] a spatial matrix generating unit, configured to generate a spatial matrix of the Laplace matrix according to the structure of the eigendecomposition;

[0053] The operating condition determination model generating unit is used to perform cluster division on the spatial matrix in the feature space corresponding to the spatial matrix to generate the operating condition determination model.

[0054] In some embodiments, before performing eigendecomposition on the Laplace matrix, a safety risk warning device for construction workers performing high-altitude operations under a high-voltage power tower further includes:

[0055] a Laplace matrix normalization unit, configured to perform random walk normalization on the Laplace matrix;

[0056] The Laplacian matrix generation unit includes:

[0057] a degree matrix generating unit, configured to generate a degree matrix of the height, altitude, air pressure, wind speed, humidity, rainfall, temperature, body temperature, pulse, heart rate, respiratory rate, and blood pressure according to the similarity graph;

[0058] The Laplacian matrix generating subunit is configured to generate the Laplacian matrix according to the degree matrix.

[0059] In some embodiments, a safety risk warning device for construction workers performing high-altitude operations under a high-voltage power tower further includes:

[0060] A security warning model generation module is used to generate the security warning model; the security warning model generation module includes:

[0061] A feature dimension random extraction unit, configured to randomly extract the feature dimensions of the environmental data and the feature dimensions of the vital sign data under corresponding working conditions;

[0062] A first segmentation point selection unit, configured to randomly select a first segmentation point between a maximum value and a minimum value of a characteristic dimension of the environmental data;

[0063] a second segmentation point selection unit, configured to randomly select a second segmentation point between a maximum value and a minimum value of a characteristic dimension of the vital sign data;

[0064] a subtree division unit, configured to divide the environmental data and the vital sign data into four subtrees according to the first division point and the second division point;

[0065] The subtree iteration unit is used to iterate the above steps on the four subtrees until a preset condition is met to generate a safety warning model.

[0066] In some embodiments, the vital sign data further includes: time series data of skeletal points of the construction worker;

[0067] The construction worker safety risk early warning device further includes:

[0068] A graph structure generation module, configured to generate a graph structure of joint points based on the skeleton point time series data;

[0069] A graph structure convolution module, configured to perform a graph convolution operation on the graph structure to extract spatial topological features of the skeleton point time series data;

[0070] A spatial topological feature temporal convolution module, configured to perform a temporal convolution operation on the spatial topological feature to extract the motion evolution features of the skeleton point temporal data;

[0071] The second safety risk warning module is used to warn the construction workers of safety risks based on the safety warning model of the working conditions corresponding to the skeleton point time series data, the motion evolution characteristics, the height, altitude, air pressure, wind speed, humidity, rainfall, temperature, body temperature, pulse, heart rate, respiratory rate and blood pressure.

[0072] In some embodiments, the first security risk warning module includes:

[0073] A time series data generating unit, configured to generate time series data having the environmental data and the vital sign data; wherein each time point in the time series data has a feature vector of the environmental data and a feature vector of the vital sign data;

[0074] a risk value determining unit, configured to determine a risk value corresponding to a current time point based on the safety warning model, the characteristic vector of the environmental data, and the characteristic vector of the vital sign data;

[0075] The safety risk warning unit is used to warn the construction workers of the safety risks based on the risk value and a preset warning threshold.

[0076] The present invention provides an electronic device, comprising:

[0077] memory and processor;

[0078] The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, the steps of a safety risk warning method for construction workers working at high altitude under a high-voltage power tower are implemented.

[0079] The present invention provides a computer-readable storage medium storing computer-executable instructions. When the computer-executable instructions are executed by a processor, the steps of a safety risk warning method for construction workers performing high-altitude operations under a high-voltage power tower are implemented.

[0080] Compared with the prior art, the present invention has the following beneficial effects: the present invention obtains environmental data of high-voltage towers and vital sign data of construction workers; then, determines the working condition corresponding to the current high-altitude operation based on the environmental data, vital sign data and a pre-generated working condition determination model; wherein, the working condition determination model is generated based on a spectral clustering algorithm, and the number of connected components of the Laplace matrix of the working condition determination model is determined by the number of outliers in the vital sign data; finally, based on the environmental data, vital sign data and a pre-generated safety warning model corresponding to the working condition, a warning is issued to the safety risks of the construction workers, wherein the safety warning model is generated based on an isolation forest algorithm; wherein, the depth of the tree of the safety warning model is determined by the number of categories of the vital sign data.

[0081] In summary, the present invention uses machine learning algorithms (spectral clustering and isolation forest) to perform intelligent and dynamic analysis of the real-time environmental and personnel data of high-voltage tower high-altitude operations, achieving accurate working condition identification and targeted risk warnings, significantly improving the accuracy and efficiency of safety monitoring, and ultimately achieving the goal of proactively preventing accidents and protecting the lives of construction workers. BRIEF DESCRIPTION OF THE DRAWINGS

[0082] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0083] Figure 1 This is a flow chart of a method for early warning of safety risks for construction workers working at heights under high-voltage power towers, provided by an embodiment of the present application;

[0084] Figure 2 This is another flow chart of a method for early warning of safety risks for construction workers working at heights under high-voltage power towers provided in an embodiment of the present application;

[0085] Figure 3 This is a flow chart of step 400 provided in one embodiment of the present application;

[0086] Figure 4 is another flowchart of step 400 provided in one embodiment of the present application;

[0087] Figure 5is a flowchart of step 402 provided by an embodiment of the present application;

[0088] Figure 6 This is another flow chart of a method for early warning of safety risks for construction workers working at heights under high-voltage power towers provided in an embodiment of the present application;

[0089] Figure 7 An embodiment of the present application provides a flowchart of step 500;

[0090] Figure 8 This is a third flow chart of a method for early warning of safety risks for construction workers working at heights under high-voltage power towers provided by an embodiment of the present application;

[0091] Figure 9 An embodiment of the present application provides a flowchart of step 300;

[0092] Figure 10 This is a flowchart of a safety risk warning method for construction workers working at height under a high-voltage power tower, provided by a specific embodiment of the present application;

[0093] Figure 11 This is a block diagram of a safety risk warning device for construction workers performing high-altitude operations under high-voltage power towers, provided in an embodiment of the present application;

[0094] Figure 12 It is a structural diagram of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0095] To make the above-mentioned objects, features, and advantages of the present invention more clearly understood, the following detailed description of the specific embodiments of the present invention is given in conjunction with the accompanying drawings. It is obvious that the described embodiments are only part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in this field without creative work should fall within the scope of protection of the present invention.

[0096] Example 1, with reference to Figure 1 , which is an embodiment of the present invention, provides a safety risk warning method for construction workers working at height under a high-voltage power tower, comprising:

[0097] The safety risk warning method for construction workers working at height under a high-voltage power tower in this embodiment can be applied to electronic devices with communication, computing and data storage capabilities. The specific process can be as follows: Figure 1 Shown, including:

[0098] Step 100: Obtain environmental data of the high-voltage tower and vital sign data of construction workers;

[0099] Step 200: Determine a working condition corresponding to the current high-altitude operation based on the environmental data, the vital sign data, and a pre-generated working condition determination model; wherein the working condition determination model is generated based on a spectral clustering algorithm, and the number of connected components of the Laplacian matrix of the working condition determination model is determined by the number of outliers in the vital sign data;

[0100] Step 300: Issue a warning of safety risks to the construction workers based on the environmental data, the vital signs data, and a pre-generated safety warning model corresponding to the working conditions, wherein the safety warning model is generated based on an isolation forest algorithm; wherein the depth of the tree of the safety warning model is determined by the number of categories of the vital signs data.

[0101] An embodiment of the present application provides a safety risk warning method for construction workers working at height under high-voltage power towers, which dynamically determines the key parameters of the model (number of clusters, tree depth) according to actual data (number of outliers, number of categories), thereby enhancing the system's adaptability.

[0102] Example 2:

[0103] In some examples, the environmental data of step 100 includes: height of the high-voltage tower, altitude, air pressure, wind speed, humidity, rainfall, and temperature;

[0104] Tower height: 15 to 100 meters (river-crossing and valley-crossing towers for ultra-high / ultra-high voltage lines can exceed 200 meters). The impacts of tower height include: The higher the height, the greater the risk of falling and the more difficult rescue. The higher the height from the ground, the greater the average wind speed (reduced ground friction). Wind load is the primary load in tower design and the primary threat to high-altitude operations. The temperature, humidity, and wind speed at the top of the tower and on the ground can differ significantly (e.g., it's colder and windier at the top). Furthermore, climbing up and down the tower increases the time it takes for personnel to travel, consumes more energy, and is prone to fatigue.

[0105] Effects of altitude include:

[0106] Air pressure and oxygen content: The higher the altitude, the lower the air pressure and the lower the oxygen content (at 3,000 meters above sea level, the oxygen content is approximately 70% of that at sea level). This can easily lead to altitude sickness (dizziness, fatigue, and difficulty breathing) in workers working at height, impairing their physical fitness and judgment, and increasing the risk of accidents.

[0107] Temperature and radiation: High altitude areas usually have lower temperatures and stronger ultraviolet radiation, so you need to prevent frostbite and sunburn.

[0108] Equipment Performance: Low air pressure may affect the performance of some pneumatic or unsealed electronic devices (e.g., drones, certain sensors).

[0109] Meteorological complexity: The weather in high-altitude mountainous areas is changeable, and the risk of sudden strong winds, thunderstorms, and hail is higher.

[0110] The effects of air pressure include: Low air pressure is directly linked to altitude sickness (see Effects of Altitude). Airtight equipment manufactured at low altitudes (such as some instruments) may be damaged or malfunction at high altitudes due to the pressure differential between the internal and external pressures. This can also affect gas sensor calibration (for example, oxygen sensor readings must be corrected based on air pressure).

[0111] Wind speed: High-altitude wind speed is 1.5-2 times higher than ground-level wind speed. Safe operating wind speed is usually required to be ≤ Level 8 (approximately 17.2-20.7 m / s). Operations must be stopped if wind speed exceeds this limit.

[0112] The impacts of wind speed include: Strong winds directly threaten personnel's balance, potentially leading to falls or being struck by falling tools. High winds can cause tower vibration or swaying, increasing operational difficulty and causing panic. They can also increase the wind chill index, accelerating heat loss (in low-temperature environments). Strong wind noise can mask warning signals, impacting the safety of drone inspections and temporary lifting equipment.

[0113] Effects of humidity include:

[0114] Electrical safety: High humidity (especially when accompanied by dirt) significantly reduces the insulation strength of air, increases the risk of arc flash (electric shock), and is a major threat to high-voltage live work.

[0115] Equipment reliability: This can cause metal parts to rust, electronic equipment to short-circuit due to moisture, and the performance of insulating tools to deteriorate.

[0116] Personal discomfort: High humidity exacerbates the discomfort of high or low temperatures (stuffy or damp), affecting physical fitness and concentration.

[0117] Climbing safety: Excessive humidity can cause tower materials, tools, and gloves to become slippery, increasing the risk of falling.

[0118] Visibility: Heavy fog (humidity 100%) seriously reduces vision, affecting operations and monitoring.

[0119] The impacts of rainfall (in the form of rain, freezing rain, hail, and snow) include the risk of lightning strikes (often accompanied by heavy convective rainfall), slippery falls, and hail injuries. Rainwater moistens the surfaces of insulating tools and equipment, sharply reducing insulation performance and potentially causing electric shock. Live work in the rain is generally prohibited. Prolonged work in the rain can easily lead to hypothermia. Heavy rainfall can pose geological risks around the tower base (particularly in mountainous areas). Moderate to heavy rain generally requires immediate cessation of overhead work.

[0120] The effects of temperature include:

[0121] Extreme heat: Heatstroke, dehydration, and exhaustion. Risk of burns from metal tower materials. Equipment (especially batteries) overheating and failure.

[0122] Extreme low temperatures: Risk of frostbite (instant freezing of exposed skin upon contact with metal). Stiff movements, decreased flexibility, and slowed reactions. Material embrittlement (metal, plastic), failure of hydraulic / lubrication systems, and a sudden decrease in battery capacity. Snow / ice accumulation increases climbing difficulty and structural loads.

[0123] Temperature changes: Metal expansion and contraction may cause fasteners to loosen (requires regular inspection).

[0124] Vital signs data include: construction workers' body temperature, pulse, heart rate, respiratory rate and blood pressure.

[0125] Physiological parameters such as body temperature, pulse, heart rate, respiratory rate, and blood pressure for workers working at height on high-voltage towers are key indicators of their physical condition and fatigue level, directly impacting operational safety. Real-time monitoring of these parameters and enabling risk warnings in the harsh environment of strong electromagnetic interference and high altitude present unique technical challenges and application requirements, as detailed in Table 1.

[0126] Table 1

[0127]

[0128] Key risk scenarios:

[0129] Accumulated fatigue: Continuous high heart rate and rapid breathing lead to physical exhaustion, thereby increasing the risk of falling.

[0130] Altitude sickness: In a low-oxygen environment, faster breathing and heart rate lead to decreased judgment, thereby increasing the risk of incorrect operation.

[0131] Extreme temperatures: Heatstroke (increased body temperature and decreased blood pressure) or hypothermia (decreased body temperature and decreased heart rate) may cause sudden loss of consciousness.

[0132] For some examples, see Figure 2 , a safety risk warning method for construction workers working at height under a high-voltage power tower, further comprising:

[0133] Step 400: Generate the working condition determination model; further, see Figure 3 , step 400 includes:

[0134] Step 401: Generate similarity graphs for the height, altitude, air pressure, wind speed, humidity, rainfall, temperature, body temperature, pulse, heart rate, respiratory rate, and blood pressure, respectively;

[0135] Specifically, a similarity graph is constructed for each parameter k (height, altitude, air pressure, wind speed, humidity, rainfall, temperature, body temperature, pulse, heart rate, respiratory rate, blood pressure) to represent the similarity between samples.

[0136] First, for parameter k, the similarity S between samples i and j is k (i,j) is calculated using the following formula:

[0137]

[0138] Among them, x i,k : the value of parameter k for sample i; σ k : Bandwidth parameter (preferably the standard deviation of parameter k).

[0139] Next, the similarity matrix is ​​calculated based on the similarities and converted into a similarity graph.

[0140] Step 402: Generate an expected corresponding Laplacian matrix according to the similarity graph;

[0141] Convert each similarity graph into a Laplacian matrix L k , used to capture the structural characteristics of the graph.

[0142] Step 403: performing eigendecomposition on the Laplace matrix;

[0143] Decompose the Laplacian matrix L k Extract the characteristic vectors for dimensionality reduction. Specifically, for each L k Solution:

[0144] L k v k,i =λ k,i v k,i ;

[0145] λ k,i : eigenvalue λ k,1 ≤λ k,2 ≤…≤λ k,n ;v k,i : corresponding eigenvector.

[0146] Take the eigenvectors corresponding to the first m smallest non-zero eigenvalues ​​and construct the characteristic matrix:

[0147]

[0148] Step 404: Generate a space matrix of the Laplacian matrix according to the structure of the eigendecomposition;

[0149] The feature matrices of all parameters (height, altitude, air pressure, wind speed, humidity, rainfall, temperature, body temperature, pulse, heart rate, respiratory rate, and blood pressure) are fused to generate a unified spatial representation.

[0150] First, vertically stitch all V k :

[0151]

[0152] Next, dimensionality reduction is performed:

[0153] where d: the target dimension (d << m, retaining the main features).

[0154] Finally, a spatial matrix V is output final , where each row corresponds to the coordinates of a sample in the low-dimensional space.

[0155] Step 405: In the feature space corresponding to the spatial matrix, perform cluster partitioning on the spatial matrix to generate the working condition determination model.

[0156] Specifically, cluster the samples in the feature space to generate the working condition determination model.

[0157] Use the K-means algorithm for clustering:

[0158]

[0159] where C = {c1,..., c K} is the cluster center; z i is the cluster label of sample i; finally, model output is performed:

[0160] Cluster center C; cluster label mapping function: f: Working condition category (by projecting onto the V final space and calculating the nearest cluster center). <6000410>In some examples, refer to Figure 4 , before step 403, step 400 further includes:

[0162] Step 406: Perform random walk normalization on the Laplacian matrix;

[0163] Specifically, L rw = D -1 L k = I - D -1 W; where D is the degree matrix.

[0164] In this case, refer to Figure 5 , step 402 includes:

[0165] Step 4021: Generate the degree matrix of the height, altitude, air pressure, wind speed, humidity, rainfall, temperature, body temperature, pulse, heart rate, respiratory rate, and blood pressure according to the similarity graph;

[0166] Specifically, generate the degree matrix according to the following formula:

[0167]

[0168] Step 4022: Generate the Laplacian matrix according to the degree matrix.

[0169] When the Laplace matrix in step 4022 is the unnormalized Laplace matrix L k hour:

[0170] L k =D k -S k ;

[0171] When the Laplacian matrix in step 4022 is a normalized Laplacian matrix:

[0172]

[0173] For some examples, see Figure 6 , a safety risk warning method for construction workers working at height under a high-voltage power tower, further comprising:

[0174] Step 500: Generate the security warning model; then, see Figure 7 , step 500 includes:

[0175] Step 501: randomly extracting characteristic dimensions of the environmental data and the vital sign data under corresponding working conditions;

[0176] Randomly extract a dimension from all environmental data feature dimensions (such as temperature, humidity, and light intensity), and record it as D e ; Randomly extract a dimension from all vital sign feature dimensions (such as heart rate, blood pressure, blood oxygen saturation), denoted as D v .

[0177] Step 502: randomly selecting a first segmentation point between the maximum value and the minimum value of the characteristic dimension of the environmental data;

[0178] Step 503: randomly selecting a second segmentation point between the maximum value and the minimum value of the characteristic dimension of the vital sign data;

[0179] In step 502 to step 503, D e Dimension, calculate all samples of the current node in D e The maximum value max e and minimum value min e , in the interval [min e , max e ] Randomly select the segmentation point P e ; for D v Dimension, calculate all samples of the current node in D v The maximum value maxv and minimum value minv, in the interval [min v , max v ] Randomly select the segmentation point P v .

[0180] Step 504: Divide the environmental data and the vital sign data into four subtrees according to the first segmentation point and the second segmentation point;

[0181] According to (P e , P v ) Divide the current node data into four subsets, see Table 2:

[0182] Table 2

[0183] Subtree Split conditions Subtree 1 <![CDATA[D e ≤P e and D v ≤P v ]]> Subtree 2 <![CDATA[D e ≤P e and D v >P v ]]> Subtree 3 <![CDATA[D e >P e and D v ≤P v ]]> Subtree 4 <![CDATA[D e >P e and D v >P v ]]>

[0184] Step 505: Iterate the above steps on the four subtrees until the preset conditions are met to generate the security warning model.

[0185] Repeat the above steps for each subtree until any of the following termination conditions is met:

[0186] 1. Depth limit: the tree depth reaches a preset threshold (e.g. 10 layers);

[0187] 2. Sample purity: All sample category labels within a node are consistent;

[0188] 3. Sample sparseness: the number of node samples is less than the threshold (e.g. 5);

[0189] 4. Feature exhaustion: Insufficient feature dimensions available.

[0190] In some examples, the vital sign data further includes: time series data of skeletal points of the construction worker;

[0191] The time series data of construction workers' skeleton points includes the following key dimensions:

[0192] Spatial dimension: 17-25 key points (such as head, neck, shoulders, elbows, wrists, hips, knees, ankles);

[0193] Time dimension: continuous time series (usually 30-60fps sampling rate)

[0194] Coordinate data: 2D coordinates (x, y) + confidence; 3D coordinates (x, y, z) (depth camera).

[0195] Derived features: joint angles (such as elbow flexion angle); limb length ratios; movement speed and acceleration.

[0196] For some examples, see Figure 8, a safety risk warning method for construction workers working at height under a high-voltage power tower, further comprising:

[0197] Step 600: Generate a graph structure of joint points according to the skeleton point time series data;

[0198] Specifically, the 17 key human joints (head, neck, shoulder, elbow, wrist, hip, knee, ankle, etc.) are converted into graph nodes, each containing three-dimensional spatial coordinates (x, y, z) and motion state information. The connection process is based on the human physiological structure (for example, the shoulder, elbow, and wrist form an arm chain). Joints that move frequently are given stronger connections (for example, the knee joint gains weight during climbing). The relative distances and angles between joints are calculated as edge attributes. The states of each joint at consecutive time points form a time series, and the states of the same joint at different time points form time-series edges.

[0199] Step 700: performing a graph convolution operation on the graph structure to extract spatial topological features of the skeleton point time series data;

[0200] Each joint node aggregates information from adjacent nodes (such as integrated shoulder and wrist data from the elbow joint) and performs hierarchical propagation (local joint → limb part → whole body posture), highlighting high-risk joints (such as the waist during carrying), adaptively adjusting the connection strength according to the amplitude of the movement, and ultimately outputting a 128-dimensional feature vector that represents the coordination of the whole body posture.

[0201] Step 800: performing a temporal convolution operation on the spatial topological features to extract motion evolution features of the skeleton point temporal data;

[0202] Specifically, we capture action patterns of different durations (1 second to fall vs. 10 minutes to fatigue), ensure temporal directionality (the order of actions is irreversible), and perform multi-scale feature extraction:

[0203] Short-term features include sudden movements (slips, falls) and long-term features include progressive risks (fatigue accumulation, elevated body temperature). The final output is a 128-dimensional time series feature that represents the evolution of the movement.

[0204] Step 900: Provide a warning of the safety risks of the construction workers based on the safety warning model of the working conditions corresponding to the skeleton point time series data, the motion evolution characteristics, the height, altitude, air pressure, wind speed, humidity, rainfall, temperature, body temperature, pulse, heart rate, respiratory rate and blood pressure.

[0205] For some examples, see Figure 9 , step 300 includes:

[0206] Step 301: Generate time series data having the environmental data and the vital sign data; wherein each time point in the time series data has a feature vector of the environmental data and a feature vector of the vital sign data;

[0207] Step 302: determining a risk value corresponding to the current time point based on the safety warning model, the characteristic vector of the environmental data, and the characteristic vector of the vital sign data;

[0208] Step 303: issuing an early warning of the safety risk of the construction workers according to the risk value and a preset early warning threshold.

[0209] Example 3: To further illustrate the solution, see Figure 10 The present invention also provides a specific implementation method of a safety risk warning method for construction workers working at high altitude under a high-voltage power tower, which specifically includes the following contents.

[0210] S1: Obtain the height, altitude, air pressure, wind speed, humidity, rainfall, and temperature of the high-voltage tower working environment, as well as the body temperature, pulse, heart rate, respiratory rate, and blood pressure of the construction workers.

[0211] S2: Generate similarity graphs for height, altitude, air pressure, wind speed, humidity, rainfall, temperature, body temperature, pulse, heart rate, respiratory rate, and blood pressure respectively;

[0212] S3: Generate the expected corresponding Laplacian matrix based on the similarity graph;

[0213] S4: perform eigendecomposition on the Laplacian matrix;

[0214] S5: Generate the space matrix of the Laplacian matrix according to the structure of the eigendecomposition;

[0215] S6: In the feature space corresponding to the spatial matrix, cluster the spatial matrix to generate a spatial determination model.

[0216] S7: Determine the working condition corresponding to the current high-altitude operation based on the height, altitude, air pressure, wind speed, humidity, rainfall, temperature, body temperature, pulse, heart rate, respiratory rate, blood pressure and working condition determination model;

[0217] S8: Provide early warnings of safety risks to construction workers based on a pre-generated safety warning model corresponding to height, altitude, air pressure, wind speed, humidity, rainfall, temperature, body temperature, pulse, heart rate, respiratory rate, blood pressure, and working conditions, wherein the safety warning model is generated based on the isolation forest algorithm; wherein the depth of the tree of the safety warning model is determined by the number of categories of the vital signs data.

[0218] Compared with the prior art, the specific application example of this application provides a safety risk warning method for construction workers working at height under high-voltage power towers, which has the following beneficial effects:

[0219] Achieve comprehensive real-time monitoring: Simultaneously collect environmental data (such as wind speed, temperature, humidity, lightning risk, etc.) and construction workers' vital signs data (such as heart rate, blood oxygen, body movement, fatigue, etc.), providing a comprehensive real-time data foundation for safety assessment.

[0220] Intelligent identification of working conditions: Utilizing a working condition determination model built based on a spectral clustering algorithm, the system automatically categorizes the current complex environment and personnel status data into predefined, representative working condition categories. The model's adaptability is reflected in the fact that its core parameter (the number of connected components in the Laplace matrix, which directly affects the number of clusters) is dynamically determined by the number of outliers in the vital sign data. This means that the system can automatically adjust the precision of working condition classification based on the degree of abnormality in the personnel status (higher anomalies may lead to the identification of more detailed hazardous working conditions), improving the model's ability to respond to actual risks.

[0221] Precise safety risk warning: For specific identified operating conditions, a pre-generated safety warning model (based on the isolation forest algorithm) is used to perform risk assessment. The adaptability of the warning model is reflected in the fact that its core structure (tree depth) is determined by the number of vital sign data categories. This ensures that the model complexity matches the diversity of the vital sign data types to be processed, helping to more effectively detect anomalies (potential risks) related to the current operating conditions that deviate from the "normal" pattern.

[0222] Improving the relevance and accuracy of early warnings: By first identifying the operating conditions and then applying the appropriate early warning model, risk warnings are now more precise and contextualized. Normal and abnormal patterns vary under different operating conditions (such as high winds, high temperatures, and nighttime operations). This approach avoids a one-size-fits-all approach to early warning thresholds, significantly improving the relevance and accuracy of early warning signals and reducing false positives and missed alerts.

[0223] Data-driven dynamic risk management: The entire process (data collection -> working condition identification -> risk warning) relies heavily on real-time data and utilizes machine learning algorithms (spectral clustering and isolation forest) for intelligent analysis. The system dynamically adjusts the granularity of working condition classification (based on abnormal vital signs) and the complexity of the warning model (based on vital sign categories) based on real-time changes in the on-site environment and personnel status (reflected in the data), achieving dynamic and adaptive safety risk management.

[0224] Preventing safety accidents: Promptly identify potential safety risks (such as sudden environmental changes, physical discomfort, or abnormal behavior) and issue early warnings before an accident occurs. This provides a critical window for management to intervene and take measures (such as suspending operations, evacuating personnel, and providing assistance), effectively preventing safety accidents such as falls from heights, electric shocks, and heatstroke, and protecting the lives of construction workers.

[0225] Example 3: Another embodiment of the present application relates to a safety risk warning device for construction personnel working at high altitude under a high-voltage power tower. The following is a detailed description of the implementation details of the safety risk warning device for construction personnel working at high altitude under a high-voltage power tower in this embodiment. The following content is only for the convenience of understanding the implementation details and is not necessary for the implementation of this solution. The schematic diagram of the safety risk warning device for construction personnel working at high altitude under a high-voltage power tower in this embodiment can be as follows: Figure 11 As shown, the device includes: an input data acquisition module 801, a working condition determination module 802, and a first safety risk warning module 803.

[0226] Input data acquisition module 801, used to obtain environmental data of high-voltage towers and vital sign data of construction workers;

[0227] a working condition determination module 802 for determining the working condition corresponding to the current high-altitude operation based on the environmental data, the vital sign data, and a pre-generated working condition determination model; wherein the working condition determination model is generated based on a spectral clustering algorithm, and the number of connected components of the Laplacian matrix of the working condition determination model is determined by the number of outliers in the vital sign data;

[0228] The first safety risk warning module 803 is used to warn the construction workers of safety risks based on the environmental data, the vital signs data and a pre-generated safety warning model corresponding to the working conditions, wherein the safety warning model is generated based on the isolation forest algorithm; wherein the depth of the tree of the safety warning model is determined by the number of categories of the vital signs data.

[0229] In some embodiments, the environmental data includes: the height, altitude, air pressure, wind speed, humidity, rainfall, and temperature of the high-voltage tower;

[0230] The vital signs data include: the construction worker's body temperature, pulse, heart rate, respiratory rate and blood pressure.

[0231] In some embodiments, a safety risk warning device for construction workers performing high-altitude operations under a high-voltage power tower further includes:

[0232] The operating condition determination model generation module is used to generate the operating condition determination model; the operating condition determination model generation module includes:

[0233] a similarity graph generating unit, configured to respectively generate similarity graphs of the height, altitude, air pressure, wind speed, humidity, rainfall, temperature, body temperature, pulse, heart rate, respiratory rate, and blood pressure;

[0234] A Laplace matrix generating unit, configured to generate an expected corresponding Laplace matrix according to the similarity graph;

[0235] a Laplace matrix eigendecomposition unit, configured to perform eigendecomposition on the Laplace matrix;

[0236] a spatial matrix generating unit, configured to generate a spatial matrix of the Laplace matrix according to the structure of the eigendecomposition;

[0237] The operating condition determination model generating unit is used to perform cluster division on the spatial matrix in the feature space corresponding to the spatial matrix to generate the operating condition determination model.

[0238] In some embodiments, before performing eigendecomposition on the Laplace matrix, a safety risk warning device for construction workers performing high-altitude operations under a high-voltage power tower further includes:

[0239] a Laplace matrix normalization unit, configured to perform random walk normalization on the Laplace matrix;

[0240] The Laplacian matrix generation unit includes:

[0241] a degree matrix generating unit, configured to generate a degree matrix of the height, altitude, air pressure, wind speed, humidity, rainfall, temperature, body temperature, pulse, heart rate, respiratory rate, and blood pressure according to the similarity graph;

[0242] The Laplacian matrix generating subunit is configured to generate the Laplacian matrix according to the degree matrix.

[0243] In some embodiments, a safety risk warning device for construction workers performing high-altitude operations under a high-voltage power tower further includes:

[0244] A security warning model generation module is used to generate the security warning model; the security warning model generation module includes:

[0245] A feature dimension random extraction unit, configured to randomly extract the feature dimensions of the environmental data and the feature dimensions of the vital sign data under corresponding working conditions;

[0246] A first segmentation point selection unit, configured to randomly select a first segmentation point between a maximum value and a minimum value of a characteristic dimension of the environmental data;

[0247] a second segmentation point selection unit, configured to randomly select a second segmentation point between a maximum value and a minimum value of a characteristic dimension of the vital sign data;

[0248] a subtree division unit, configured to divide the environmental data and the vital sign data into four subtrees according to the first division point and the second division point;

[0249] The subtree iteration unit is used to iterate the above steps on the four subtrees until a preset condition is met to generate a safety warning model.

[0250] In some embodiments, the vital sign data further includes: time series data of skeletal points of the construction worker;

[0251] The construction worker safety risk early warning device further includes:

[0252] A graph structure generation module, configured to generate a graph structure of joint points based on the skeleton point time series data;

[0253] A graph structure convolution module, configured to perform a graph convolution operation on the graph structure to extract spatial topological features of the skeleton point time series data;

[0254] A spatial topological feature temporal convolution module, configured to perform a temporal convolution operation on the spatial topological feature to extract the motion evolution features of the skeleton point temporal data;

[0255] The second safety risk warning module is used to warn the construction workers of safety risks based on the safety warning model of the working conditions corresponding to the skeleton point time series data, the motion evolution characteristics, the height, altitude, air pressure, wind speed, humidity, rainfall, temperature, body temperature, pulse, heart rate, respiratory rate and blood pressure.

[0256] In some embodiments, the first security risk warning module includes:

[0257] A time series data generating unit, configured to generate time series data having the environmental data and the vital sign data; wherein each time point in the time series data has a feature vector of the environmental data and a feature vector of the vital sign data;

[0258] a risk value determining unit, configured to determine a risk value corresponding to a current time point based on the safety warning model, the characteristic vector of the environmental data, and the characteristic vector of the vital sign data;

[0259] The safety risk warning unit is used to warn the construction workers of the safety risks based on the risk value and a preset warning threshold.

[0260] It is worth mentioning that all modules involved in this embodiment are logical modules. In actual applications, a logical unit can be a physical unit, a part of a physical unit, or a combination of multiple physical units. In addition, to highlight the innovation of this application, this embodiment does not include units that are not closely related to solving the technical problem proposed by this application. However, this does not mean that other units do not exist in this embodiment.

[0261] Embodiment 4: Another embodiment of the present application relates to an electronic device, such as Figure 12 As shown, including: the electronic equipment specifically includes the following contents:

[0262] Processor (processor) 1201, memory (memory) 1202, communication interface (CommunicationsInterface) 1203 and bus 1204;

[0263] The processor 1201, the memory 1202, and the communication interface 1203 communicate with each other via the bus 1204; the communication interface 1203 is used to implement information transmission between the server-side device and the user-side device and other related devices;

[0264] The processor 1201 is configured to call a computer program in the memory 1202. When the processor executes the computer program, all steps of the safety risk warning method for construction workers working at height under a high-voltage power tower in the above embodiment are implemented. For example, when the processor executes the computer program, the following steps are implemented:

[0265] Obtain environmental data from high-voltage towers and vital signs data from construction workers;

[0266] Determining a working condition corresponding to the current high-altitude operation based on the environmental data, the vital sign data, and a pre-generated working condition determination model; wherein the working condition determination model is generated based on a spectral clustering algorithm, and the number of connected components of the Laplacian matrix of the working condition determination model is determined by the number of outliers in the vital sign data;

[0267] An early warning of safety risks to the construction workers is issued based on the environmental data, the vital signs data, and a pre-generated safety early warning model corresponding to the working conditions, wherein the safety early warning model is generated based on an isolation forest algorithm; wherein the depth of the tree of the safety early warning model is determined by the number of categories of the vital signs data.

[0268] The memory and processor are connected using a bus, which can include any number of interconnected buses and bridges. The bus connects various circuits of one or more processors and memories. The bus can also connect various other circuits such as peripheral devices, voltage regulators, and power management circuits. These are all well known in the art and are therefore not described further herein. The bus interface provides an interface between the bus and the transceiver. The transceiver can be a single component or multiple components, such as multiple receivers and transmitters, providing a unit for communicating with various other devices over a transmission medium. Data processed by the processor is transmitted over a wireless medium via an antenna. Furthermore, the antenna receives data and transmits it to the processor.

[0269] The processor is responsible for managing the bus and general processing, and can also provide various functions, including timing, peripheral interfaces, voltage regulation, power management, and other control functions. Memory can be used to store data used by the processor when performing operations.

[0270] Embodiment 5: Another embodiment of the present application relates to a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, the steps of the above-mentioned embodiment of the method for early warning of safety risks for construction workers working at height under a high-voltage power tower are implemented, including:

[0271] Obtain environmental data from high-voltage towers and vital signs data from construction workers;

[0272] Determining a working condition corresponding to the current high-altitude operation based on the environmental data, the vital sign data, and a pre-generated working condition determination model; wherein the working condition determination model is generated based on a spectral clustering algorithm, and the number of connected components of the Laplacian matrix of the working condition determination model is determined by the number of outliers in the vital sign data;

[0273] An early warning of safety risks to the construction workers is issued based on the environmental data, the vital signs data, and a pre-generated safety early warning model corresponding to the working conditions, wherein the safety early warning model is generated based on an isolation forest algorithm; wherein the depth of the tree of the safety early warning model is determined by the number of categories of the vital signs data.

[0274] The various embodiments in this specification are described in a progressive manner. Similar portions between the various embodiments can be referenced to each other. Each embodiment focuses on the differences between the other embodiments. In particular, the hardware + program embodiments are generally similar to the method embodiments, so their description is relatively simple. For relevant portions, refer to the description of the method embodiments.

[0275] The foregoing description of this specification describes specific embodiments. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims can be performed in an order different from that described in the embodiments and still achieve the desired results. Furthermore, the processes depicted in the accompanying drawings do not necessarily require the specific order shown or the sequential order to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0276] Although the present application provides method operation steps such as embodiments or flowcharts, more or fewer operation steps may be included based on conventional or non-creative work. The order of steps listed in the embodiments is only one way of executing the steps among many steps and does not represent the only execution order. When an actual device or client product is executed, it can be executed in the order shown in the embodiments or the drawings or in parallel (for example, in a parallel processor or multi-threaded processing environment).

[0277] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0278] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0279] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 The steps for the function specified in one or more boxes.

[0280] Specific embodiments are used in the present invention to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core ideas. At the same time, for those skilled in the art, according to the ideas of the present invention, there may be changes in the specific implementation methods and application scopes. In summary, the contents of this specification should not be understood as limiting the present invention.

Claims

1. A safety risk early warning method for construction workers working at height under a high-voltage power tower, characterized in that: include: Obtain environmental data from high-voltage towers and vital signs data from construction workers; Determining a working condition corresponding to the current high-altitude operation based on the environmental data, the vital sign data, and a pre-generated working condition determination model; wherein the working condition determination model is generated based on a spectral clustering algorithm, and the number of connected components of the Laplacian matrix of the working condition determination model is determined by the number of outliers in the vital sign data; An early warning of safety risks to the construction workers is issued based on the environmental data, the vital signs data, and a pre-generated safety early warning model corresponding to the working conditions, wherein the safety early warning model is generated based on an isolation forest algorithm; wherein the depth of the tree of the safety early warning model is determined by the number of categories of the vital signs data.

2. The method for early warning of safety risks for construction workers working at height under a high-voltage power tower according to claim 1, characterized in that: The environmental data includes: Height of high-voltage towers, altitude, air pressure, wind speed, humidity, rainfall, and temperature; The vital signs data include: The construction workers' body temperature, pulse, heart rate, respiratory rate and blood pressure.

3. The method for early warning of safety risks for construction workers working at height under a high-voltage power tower according to claim 2, characterized in that: Generating the operating condition determination model includes: Generating similarity graphs of the height, altitude, air pressure, wind speed, humidity, rainfall, temperature, body temperature, pulse, heart rate, respiratory rate, and blood pressure respectively; generating an expected corresponding Laplacian matrix according to the similarity graph; performing eigendecomposition on the Laplace matrix; generating a space matrix of the Laplacian matrix according to the structure of the eigendecomposition; In the feature space corresponding to the spatial matrix, cluster division is performed on the spatial matrix to generate the operating condition determination model.

4. A safety risk early warning method for construction workers working at height under a high-voltage power tower as claimed in claim 3, characterized in that: Before performing eigendecomposition on the Laplace matrix, the method further includes: performing random walk normalization on the Laplace matrix; Generating an expected corresponding Laplacian matrix according to the similarity graph includes: generating a degree matrix of the height, altitude, air pressure, wind speed, humidity, rainfall, temperature, body temperature, pulse, heart rate, respiratory rate and blood pressure according to the similarity graph; The Laplacian matrix is ​​generated according to the degree matrix.

5. The method for early warning of safety risks for construction workers working at height under a high-voltage power tower according to claim 4, characterized in that: The steps of generating the security early warning model include: Randomly extracting characteristic dimensions of the environmental data and the vital sign data under corresponding working conditions; Randomly selecting a first segmentation point between the maximum value and the minimum value of the characteristic dimension of the environmental data; Randomly selecting a second segmentation point between the maximum value and the minimum value of the characteristic dimension of the vital sign data; Dividing the environmental data and the vital sign data into four subtrees according to the first segmentation point and the second segmentation point; The above steps are iterated on the four subtrees until preset conditions are met to generate the safety early warning model.

6. A safety risk warning method for construction workers working at height under a high-voltage power tower as claimed in claim 5, characterized in that: Skeletal point time series data of the construction workers; The construction worker safety risk early warning method further includes: Generate a graph structure of joint points according to the skeleton point time series data; Performing a graph convolution operation on the graph structure to extract spatial topological features of the skeleton point time series data; Performing a temporal convolution operation on the spatial topological features to extract motion evolution features of the skeleton point temporal data; The safety risk of the construction workers is warned based on the safety warning model of the working conditions corresponding to the skeleton point time series data, the motion evolution characteristics, the height, altitude, air pressure, wind speed, humidity, rainfall, temperature, body temperature, pulse, heart rate, respiratory rate and blood pressure.

7. A safety risk warning method for construction workers working at height under a high-voltage power tower as claimed in claim 6, characterized in that: Providing an early warning of the safety risk of the construction personnel according to the environmental data, the vital sign data, and a pre-generated safety early warning model corresponding to the working condition, including: Generate time series data having the environmental data and the vital sign data; wherein each time point in the time series data has a feature vector of the environmental data and a feature vector of the vital sign data; Determining a risk value corresponding to a current time point according to the safety warning model, the characteristic vector of the environmental data, and the characteristic vector of the vital sign data; An early warning is issued to the construction workers based on the risk value and a preset early warning threshold.

8. A safety risk warning system for construction workers working at height under a high-voltage power tower, applying the method according to any one of claims 1 to 7, characterized in that: include: Input data acquisition module, used to obtain environmental data of high-voltage towers and vital signs data of construction workers; a working condition determination module, configured to determine the working condition corresponding to the current high-altitude operation based on the environmental data, the vital sign data, and a pre-generated working condition determination model; wherein the working condition determination model is generated based on a spectral clustering algorithm, and the number of connected components of the Laplacian matrix of the working condition determination model is determined by the number of outliers in the vital sign data; A safety risk warning module is used to warn the construction workers of safety risks based on the environmental data, the vital signs data and a pre-generated safety warning model corresponding to the working conditions, wherein the safety warning model is generated based on the isolation forest algorithm; wherein the depth of the tree of the safety warning model is determined by the number of categories of the vital signs data.

9. An electronic device comprising: memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, the steps of the safety risk warning method for construction workers working at high altitude under a high-voltage tower as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the steps of a safety risk warning method for construction workers working at height under a high-voltage power tower as described in any one of claims 1 to 7.

Images