A cloud-computing-based anti-electric shock early warning method and system for power work wear

Abstract

The present application relates to the technical field of power safety protection and early warning, in particular to a kind of wearable power operation electric shock prevention and early warning method and system based on cloud computing, comprising the following steps: step S100: through the metal pole plate needle plate electric field sensor built-in in wearable terminal, capture the power frequency electric field distribution in the working environment, obtain original space electric field signal;Step S200: the original space electric field signal is analyzed, and the multi-dimensional electric field characteristic value including electric field maximum, effective value, harmonic content and peak value conversion rate is extracted;The present application effectively eliminates the electromagnetic crosstalk of adjacent operating equipment and the influence of initial static state by using metal pole plate needle plate electric field sensor combined with multi-dimensional feature extraction technology of signal frequency domain and time domain, significantly improves the power frequency electric field detection precision in complex electromagnetic environment;Rely on the electric field inversion algorithm constructed by equivalent charge equation and mirror method, accurately calculate the actual physical distance from the space electric field characteristic value.

Description

Technical Field

[0001] This invention relates to the field of power safety protection and early warning technology, specifically to a wearable power operation anti-electric shock early warning method and system based on cloud computing. Background Technology

[0002] During maintenance work in energized areas such as substations and high-voltage transmission lines, workers are highly susceptible to high-voltage discharge or direct electric shock due to complex situations involving parallel, adjacent, crossing, or shared tower installations, especially if they fail to maintain sufficient safe distances. Currently, portable proximity alarms or alarm helmets are commonly used as auxiliary safety devices. However, these existing technologies have the following core technical shortcomings in practical engineering applications: First, existing non-contact proximity warning devices mainly measure voltage based on the principle of induced charge accumulation. This principle is extremely sensitive to the initial charge state of the equipment, and in complex work sites where multiple circuits are erected on the same tower or multiple devices operate in parallel, it is easily affected by crosstalk from the electromagnetic coupling effect of adjacent normally operating equipment. This causes severe distortion of the potential difference signal acquired by the sensor, resulting in an extremely high false trigger rate or false alarm rate in practical applications.

[0003] Secondly, existing alarm devices mostly use a method of setting a fixed static threshold by comparing a single transient electric field strength. Because the ionization and breakdown characteristics of high-voltage equipment at different voltage levels differ significantly, and the spatial electric field distribution exhibits a non-linear exponential decay law, a simple field strength threshold sensing cannot directly and equivalently reflect the actual physical distance between the worker and the high-voltage live conductor. This static judgment mechanism cannot achieve dynamic matching of safety distance thresholds in substation scenarios with mixed voltage levels, nor can it provide progressive, graded early warnings with sufficient dynamic braking time based on the worker's actual spatial approach rate.

[0004] Third, existing voltage detection and early warning equipment can only provide real-time audio-visual or vibration feedback on-site, and cannot achieve the data evolution from single-machine perception to Internet of Things nodes. Due to the lack of synchronous recording and cloud integration of multi-dimensional dynamic data such as operation trajectory, frequency of exceeding safety distance limits, and transient peak values ​​of extremely high field strength over long-term time series, the management platform cannot build a temporal characteristic model of the operator's operating habits, nor can it accurately isolate and cluster high-risk areas with abnormal electromagnetic coupling or clearance design defects from massive and noisy spatial coordinates. As a result, power safety management has long remained at the post-event tracking level, and cannot achieve pre-event regional hidden danger investigation and predictive control. Summary of the Invention

[0005] To achieve the above objectives, the present invention provides the following technical solution: This invention provides a cloud-based wearable electric shock prevention early warning method for power operations, comprising the following steps: Step S100: Capturing the power frequency electric field distribution in the working environment and obtaining the original spatial electric field signal through the metal plate needle plate electric field sensor built into the wearable terminal; Step S200: Performing feature analysis on the original spatial electric field signal to extract multi-dimensional electric field feature values, including the maximum electric field value, effective value, harmonic content, and peak conversion rate; Step S300: Calculating the electric field strength of the charged body using a preset electric field inversion algorithm based on the multi-dimensional electric field feature values, and calculating the real-time spatial distance between the worker and the charged body; Step S400: Comparing the real-time spatial distance with the safety distance threshold of the corresponding voltage level, and triggering the wearable terminal to perform graded vibration alarms of different intensities based on the comparison result; Step S500: Uploading the work data, including the work location, time, and alarm frequency, to the cloud platform, and using a deep learning model to generate evaluation results of the worker's operating habits and a safety risk level plan for the work site.

[0006] Furthermore, the metal electrode needle-plate electric field sensor includes: The inner layer consists of two thin metal plates of equal area; A tetrafluoroethylene separator is disposed between the two thin metal plates to generate a pressure difference signal between the plate layers when the original spatial electric field passes through the two thin metal plates. The metal electrode needle plate electric field sensor is externally wrapped with spherical mica filler to capture the power frequency electric field distribution in the working environment and reduce the influence of the metal electrode plate on the surrounding electric field intensity distribution.

[0007] Furthermore, the grounding of the needle electrode of the metal electrode needle plate electric field sensor through a series resistor specifically includes: using the voltage drop formed on the series resistor by the capacitive current or corona current generated by the loop voltage to obtain the original spatial electric field signal.

[0008] Furthermore, in step S200, before performing feature analysis on the original spatial electric field signal, a signal preprocessing process is also included: The extremely weak original spatial electric field signal was sequentially amplified, filtered, and compared. The filtering process removes high-frequency noise from the environment to obtain a three-dimensional spatial electric field intensity signal at power frequency.

[0009] Furthermore, the process of extracting multidimensional electric field feature values ​​in step S200 specifically includes: The acquired three-dimensional spatial electric field intensity signal is subjected to peak detection and average value detection using the built-in signal processing algorithm; Combining the results of peak detection and average detection, the maximum value, effective value, harmonic content and peak conversion rate of the electric field are calculated and extracted. The electric field strength in three-dimensional space is evaluated from multiple angles to eliminate the interference of adjacent operating equipment and prevent false triggering.

[0010] Furthermore, the process of calculating the real-time spatial distance using the electric field inversion algorithm in step S300 includes: Based on the ground voltage matrix of each conductor and the preset potential coefficient matrix The equivalent charge density matrix per unit length of each conductor is calculated using matrix equations. : ; Based on Gauss's law and the superposition theorem, the spatial sampling points where the workers are located are calculated according to the equivalent charge density of each conductor. Horizontal electric field component and vertical electric field components ; The sampling points Comprehensive electric field strength value : ; The combined electric field strength value is obtained. The calculation process then includes: Based on a pre-defined mathematical model of exponential decay between the electric field intensity distribution of a charged body and spatial distance, the comprehensive electric field intensity value is... The inversion is used to determine the real-time spatial distance between the worker and the live conductor. The exponential decay mathematical model is modified based on the line-to-ground operating voltage, unit distance, and line-to-ground height.

[0011] Furthermore, in step S400, before the comparison is performed, a dynamic setting process for the safety distance threshold is also included: Obtain the operating voltage level of live conductors within the current work area; Based on the stated operating voltage level, match the corresponding standard safety distance; A preset warning margin is superimposed on the standard safe distance to generate the safe distance threshold used to trigger an alarm.

[0012] Furthermore, the graded vibration alarm process in step S400 specifically includes: When the real-time spatial distance is less than the safe distance threshold and greater than the set limit safety value, the wearable terminal is controlled to emit a continuous vibration signal to remind the operator to maintain a safe distance. When the real-time spatial distance is less than the limit safety value, as the real-time spatial distance gradually decreases, the wearable terminal is controlled to emit a gradually stronger vibration signal to warn the operator to stay away from the live conductor.

[0013] Furthermore, in step S500, before generating the evaluation results using the deep learning model, a process for monitoring the work trajectory and attendance based on the operation ticket is also included: Assign a unique identification ID number to each worker; The location information of the workers is monitored in real time, and the time of their arrival and departure from the work area is recorded to generate a work trajectory; When it is detected that the operator has not arrived at the designated work area as required by the operation ticket, or has changed the work scope without authorization, an boundary crossing warning is triggered, and the boundary crossing warning information, work trajectory, duration of strong electric field environment and alarm frequency are integrated into the input data of the deep learning model. The method of using deep learning models to generate evaluation results of workers' operating habits and planning of safety risk levels for work sites specifically includes: Extract the frequency of alarms indicating insufficient safe distance from live conductors and the duration of exposure to strong electric fields from the input data. Use the deep learning model to perform feature classification, assign safety operation levels to the operators, and output the evaluation results of their operating habits. The system extracts geographic coordinates from the input data of areas where there are frequent occurrences of insufficient safe distance from charged objects or long-term high field strength. It then uses the deep learning model to perform spatial clustering analysis to delineate high-risk work areas prone to safety accidents, outputs a safety risk level plan for the work site, and sends a hazard investigation reminder to the competent authority.

[0014] On the other hand, a cloud-based wearable electric shock prevention and early warning system for power operations includes: an electric field signal acquisition module and a multi-dimensional feature extraction module, a safe distance inversion module, a local hierarchical early warning module, and a cloud-based risk assessment module.

[0015] This invention effectively eliminates electromagnetic crosstalk and the influence of initial electrostatic state from adjacent operating equipment by employing a metal electrode needle plate electric field sensor combined with multi-dimensional feature extraction technology in the frequency and time domains of the signal, significantly improving the detection accuracy of power frequency electric fields in complex electromagnetic environments. Based on an electric field inversion algorithm constructed using the equivalent charge equation and the image method, it accurately calculates spatial electric field characteristic values ​​into actual physical distances, and achieves adaptive dynamic hierarchical early warning by combining voltage level specifications and dynamic compensation margins. Simultaneously, relying on a cloud-based long short-term memory network and a density-based spatial clustering algorithm, it performs habit evaluation and geometric clustering analysis on massive dynamic temporal features and three-dimensional abnormal spatial data point sets. Based on objectively quantifying the safe operation performance of workers, it achieves proactive investigation of structurally high-risk areas in the physical site of substations, effectively improving the overall proactive safety management level of power operations. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating the overall process of the method of the present invention. Figure 2 This is a block diagram of the hardware structure of the wearable terminal of the present invention; Figure 3 This is a logical block diagram of the cloud-based risk assessment system of the present invention. Detailed Implementation

[0017] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0018] It should be noted that the terms "first," "second," etc., used in this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but includes other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0019] The present invention will now be described in further detail with reference to the accompanying drawings: Example: Figures 1-3As shown, a cloud-based wearable method for preventing electric shock during power operations includes the following steps: Step S100: Capture the power frequency electric field distribution in the working environment and obtain the original spatial electric field signal by using the metal electrode needle plate electric field sensor built into the wearable terminal; Step S200: Perform feature analysis on the original spatial electric field signal to extract multi-dimensional electric field feature values, including the maximum electric field value, effective value, harmonic content, and peak conversion rate; Step S300: Based on the multidimensional electric field characteristic values, calculate the electric field strength of the charged body using a preset electric field inversion algorithm, and calculate the real-time spatial distance between the worker and the charged body; Step S400: Compare the real-time spatial distance with the safety distance threshold of the corresponding voltage level, and trigger the wearable terminal to perform graded vibration alarms of different intensities based on the comparison result; Step S500: Upload the operation data, including operation location, time and alarm frequency, to the cloud platform, and use a deep learning model to generate the operation habit evaluation results of the operators and the safety risk level planning of the operation site.

[0020] Furthermore, the specific implementation process of step S100 is as follows: Step S100 specifically includes the process of capturing the power frequency electric field distribution in the working environment through the metal electrode needle plate electric field sensor built into the wearable terminal, thereby obtaining the original spatial electric field signal.

[0021] Since the operating frequency of my country's power system is 50Hz, a typical low-frequency alternating electric field is formed around high-voltage charged bodies. To achieve contactless and accurate acquisition of this frequency band's electric field within the extremely small space of the human wrist, the wearable terminal in this embodiment incorporates a metal electrode needle-plate electric field sensor with a specific structure. The entire outer surface of this electric field sensor is encased in spherical mica filler. Mica material possesses extremely high dielectric strength and a stable dielectric constant; fabricating it as a spherical filler layer effectively captures the power frequency electric field distribution lines dispersed in the air. More importantly, because metal conductors placed in an applied electric field will experience surface charge redistribution due to electrostatic induction, leading to severe distortion of the original spatial electric field, the outer spherical mica filler acts as a physical buffer layer, minimizing the distortion impact of the internal metal electrode on the surrounding original electric field intensity distribution and ensuring the objectivity and authenticity of the acquired signal.

[0022] Inside the spherical mica filler, an upper metal electrode plate, a tetrafluoroethylene (PTFE) spacer, and a lower metal electrode plate are stacked sequentially. The upper and lower metal electrode plates are two thin, sheet-like metal conductors with identical surface areas, positioned parallel to each other. The PTFE spacer serves as an insulating medium between the two sheet-like metal electrode plates. When an external low-frequency alternating electric field passes through the support layer of this sensor structure and perpendicularly through the two metal electrode plates, the alternating electric field induces bound charges on the surfaces of the two electrode plates whose polarity changes over time. Due to the blocking effect of the intermediate PTFE spacer, the induced charges cannot move freely in the intermediate medium, thus forming a weak potential difference between the two metal electrode plate layers that varies in real time with the spatial electric field strength. This potential difference is typically in the microvolt range. Magnitude.

[0023] In the electrical connection circuit, the sensor employs a typical needle-plate non-uniform electric field structure model for signal extraction. Specifically, two thin metal plates act as equivalent "plate electrodes" to decouple the spatial electric field, while the sensor's "needle electrode" is grounded through a series resistor of a predetermined resistance. When the sensor is near a high-voltage charged body, a coupling capacitance is formed between the sensor and the high-voltage charged body. Under the influence of a high-voltage alternating electric field, a weak capacitive displacement current will be generated in the circuit. If the local field strength reaches the corona induction field strength, an extremely weak corona current will also be generated near the needle electrode with extremely high curvature. Let the induced current generated on the sensor by the time-varying alternating electric field be... The resistance of the series resistor is According to Ohm's law, when this induced current flows through the series resistor to ground, a voltage drop will be generated across the series resistor. Their basic relationship is This weak voltage drop signal contains all the characteristic information reflecting the intensity of the external space electric field. The system acquisition unit will extract the voltage drop signal across the series resistor, thus completing the acquisition of the original space electric field signal.

[0024] Furthermore, the specific implementation process of step S200 is as follows: Step S200 specifically includes hardware-level preprocessing and software-level multidimensional feature extraction of the acquired raw spatial electric field signal.

[0025] Because the original spatial electric field signal obtained through step S100 is extremely weak, and the electromagnetic environment at the actual work site is extremely complex, the system first performs hardware-level preprocessing on the signal, including protection, amplification, filtering, and comparison circuitry. Specifically, the system front-end uses a high-input-impedance operational amplifier circuit to amplify the microvolt-level potential difference signal to a voltage range that the microprocessor's built-in analog-to-digital converter can recognize. Subsequently, the amplified signal is sent to a bandpass filter circuit. Due to the presence of high-frequency electromagnetic interference in the substation working environment caused by high-voltage switch opening, corona discharge from side equipment, or transient discharge, the core hardware parameters of this filter circuit are set to a low-frequency bandpass for the 50 Hz power frequency. Its physical purpose is to filter out high-frequency noise exceeding this frequency band and retain the 50 Hz power frequency signal, thereby accurately measuring and acquiring the three-dimensional spatial electric field intensity signal at the power frequency.

[0026] After acquiring the power frequency three-dimensional spatial electric field intensity signal and completing analog-to-digital conversion, the microcontroller uses its built-in digital signal processing algorithm to perform peak detection and average value detection. Based on this data, the system further calculates and extracts four core characteristic values ​​of the spatial electric field: maximum value, effective value, harmonic content, and peak conversion rate.

[0027] First, the system extracts the maximum value of the electric field by traversing all discrete sampling points within a set time window. This maximum value reflects the absolute peak value of the spatial electric field and represents the highest level of physical danger in the transient state.

[0028] Second, the system uses the root mean square algorithm to perform mathematical integration on discrete sampling points to obtain the effective value of the electric field. This effective value eliminates the interference of instantaneous spike signals and truly reflects the equivalent continuous energy intensity of the alternating electric field within a complete power frequency cycle.

[0029] Third, the system uses the Fast Fourier Transform algorithm to convert the time-domain electric field signal to the frequency domain for expansion, extracting high-frequency harmonic components other than the fundamental wave, i.e., harmonic content. Since the electric field signal of a normal high-voltage transmission line is dominated by a single 50 Hz fundamental wave, while crosstalk from adjacent equipment or sudden environmental interference is often accompanied by broadband harmonics of a specific spectrum, extracting harmonic content can serve as a spectral fingerprint feature to effectively distinguish the target high-voltage charged body from harmless interference sources on the side.

[0030] Fourth, the system calculates the peak conversion rate by performing discrete difference operations on the electric field peaks of adjacent time windows. Specifically, the system microprocessor sets continuous and fixed sampling time window intervals. Let the current number be... The peak electric field intensity detected and extracted within each time window is In the adjacent preceding number The peak electric field intensity detected and extracted within each time window is The system's microprocessor-built-in arithmetic logic unit executes the discrete difference formula. ; Peak conversion rate is calculated in real time. The peak conversion rate In mathematical terms, it represents the discrete derivative of the peak electric field intensity as a function of time. Its practical engineering physics significance lies in: when the calculated... A consistently positive value not only indicates that the worker is moving towards the high-voltage live conductor, but also precisely quantifies the worker's spatial approach rate relative to the high-voltage live conductor through the absolute magnitude of the value; when When the value approaches zero, it indicates that the worker is in a relatively static state or in motion parallel to the equipotential surface of the charged body. By incorporating this time-series-based dynamic rate characteristic into a multi-dimensional feature matrix, the system can effectively filter out static electric field step abrupt changes caused by instantaneous start-up and shutdown of adjacent equipment or spatial electromagnetic pulse interference. Since such abrupt changes lack continuous approximate derivative characteristics that conform to the laws of human movement, the system can effectively eliminate them based on this peak conversion rate characteristic, thereby greatly improving the accuracy of predicting approaching dangers.

[0031] Through the aforementioned multi-dimensional feature extraction mechanism, the system no longer relies on the single transient voltage threshold of traditional electroscopes for judgment. Instead, it combines the maximum value, effective value, harmonic content in the frequency domain, and peak slew rate in the time domain to form a multi-dimensional feature matrix. The system utilizes this multi-dimensional feature matrix to comprehensively and from multiple perspectives evaluate the electric field strength in three-dimensional space. It can effectively identify and eliminate the cross-electromagnetic interference effects of adjacent operating equipment through spectral differences and time-domain slew rate changes. This fundamentally eliminates the risk of false alarms or missed alarms from the software algorithm level, increasing the system's reliability and stability.

[0032] Furthermore, the specific implementation process of step S300 is as follows: Step S300 specifically includes the complete mathematical derivation and physical calculation process of calculating the electric field strength of the charged body based on the multidimensional electric field characteristic value using a preset electric field inversion algorithm, and calculating the real-time spatial distance between the operator and the charged body.

[0033] Because the electric field distribution at substations and transmission lines is extremely complex, a single field strength value cannot be directly equated to an absolute safe distance. This embodiment abandons the traditional empirical lookup table method and adopts an electric field inversion algorithm based on the equivalent charge method and the image method to rigorously abstract the spatial electric field distribution into a computable mathematical analytical model. First, the system microprocessor constructs the equivalent charge equation for the spatial conductor.

[0034] The design calculates the space available. For multi-phase or multi-circuit high-voltage conductors, the microprocessor obtains the ground operating voltage of each conductor and constructs a column vector matrix of ground voltages. Simultaneously, based on the physical arrangement and spatial geometry of the conductors on site, a potential coefficient matrix is ​​constructed. This matrix is ​​composed of self-potential coefficients representing the geometric parameters of the conductors themselves and mutual potential coefficients representing the spatial relative positions between the conductors. The microprocessor performs matrix inversion and multiplication operations. The equivalent charge density matrix on each unit length of wire was accurately calculated. This includes the equivalent line charge density variable of each conductor. .

[0035] After obtaining the equivalent line charge density of each conductor, the system calculates the spatial sampling points of the wearable terminal based on Gauss's law and the principle of superposition of electric fields. The electric field intensity components. To eliminate the boundary effect of the ground on the electric field distribution, the algorithm introduces the image charge method, setting the sampling points of the conductor and its underground image. The geometric distance parameters are calculated. The system's underlying computational logic calculates the sampling points based on the following discrete accumulation formula. Horizontal electric field component and the vertical electric field component : ； ; In the two sets of physical analytical formulas mentioned above, The absolute permittivity of air. and Representing the first The x and y coordinates of a high-voltage conductor in a two-dimensional Cartesian coordinate system. Representing the From the actual physical center of the conductor to the sampling point Euclidean distance, Representing the The underground mirror image of the conductor to this sampling point The Euclidean distance is calculated. Based on this calculation, the microprocessor performs a two-dimensional vector composition operation based on the Pythagorean theorem. ; This allows us to calculate the sampling point where the wearable device is located. Analytical value of comprehensive electric field strength at the location .

[0036] After completing the forward analytical calculation of the electric field intensity at various points in space, the system enters the final spatial distance inversion calculation process. Within the range from the high-voltage conductor to the ground, the electric field intensity exhibits a nonlinear exponential decay distribution along the conductor to the ground. The system's microprocessor has a built-in mathematical model representing the precise relationship between the electric field intensity distribution and the physical distance in space, characterizing this decay law. The microprocessor substitutes the characteristic parameters such as the maximum and effective values ​​of the actual environmental electric field extracted in step S200 into the electric field intensity isosurface equation established by the aforementioned forward analytical calculation as approximation target values. Simultaneously, based on the specific line-to-ground operating voltage, the preset unit distance variable, and the absolute height of the line space above the ground, the model attenuation parameters are dynamically corrected in real time. The microprocessor then calls a numerical approximation algorithm (such as Newton's iteration method) to solve for the spatial coordinate independent variable that satisfies the electric field intensity isosurface equation at this time. The three-dimensional Euclidean geometric difference between this coordinate independent variable and the known physical coordinates of the high-voltage charged body is confirmed as the accurate and real-time spatial distance between the operator and the high-voltage charged body.

[0037] Furthermore, the specific implementation process of step S400 is as follows: In a specific embodiment, step S400 specifically includes a dynamic setting process for the safety distance threshold and a hardware-level driving process for multi-level vibration alarm.

[0038] Because the electric field ionization and breakdown characteristics of high-voltage equipment at different voltage levels differ significantly, a fixed spatial physical distance cannot be used as a universal safety assessment boundary. The system microprocessor first obtains the operating voltage level of the high-voltage live conductors within the current working area. After obtaining the operating voltage level, the microprocessor calls the pre-loaded safety distance mapping table for each voltage level in its internal non-volatile memory to match the corresponding standard safety distance.

[0039] For example, when the operating voltage level is matched to 10kV, the retrieved standard safety distance parameter is 0.7 meters; when the operating voltage level is matched to 500kV, the retrieved standard safety distance parameter is 5.0 meters. To ensure that personnel have sufficient physical reaction braking time during dynamic construction operations, the microprocessor, based on the extracted standard safety distance value, superimposes a preset spatial warning margin through linear addition, thereby calculating and generating the absolute safety distance threshold used to trigger a hardware-level alarm under the current equipment state.

[0040] In a specific embodiment, the spatial warning margin The numerical setting is not a static empirical value, but is calculated by the microprocessor based on the system processing latency, the time consumed by human physiological reactions, and the preset maximum approximation rate through a dynamic compensation model.

[0041] Specifically, to ensure that operators have sufficient physical space to complete the braking response after receiving a vibration warning, the microprocessor establishes the following margin calculation equation: ; In the formula, This represents the maximum approximate movement speed of the operator in a power work environment, typically set according to the work specifications. to between. The discrete sampling period of the system represents the time interval required for the sensor to acquire one frame of the original electric field signal. This represents the closed-loop logic computation delay required for the microprocessor to execute the multidimensional feature extraction in step S200 and the electric field inversion algorithm in step S300. The average physiological delay time for the worker's skin nerve sensors to receive vibration signals and transmit them to the limbs via the central nervous system to produce a physical braking response is represented by this parameter. In this embodiment, this parameter is preferably set to [value missing]. to .

[0042] The microprocessor obtains the result by superimposing the above linear addition operations. Its essential physical significance lies in quantifying the inertial displacement distance generated within the entire cycle from the system detecting an electric field anomaly, completing distance inversion, triggering hardware vibration, to the worker completing physical braking. Through this dynamic calculation method based on time delay compensation, the generated absolute safe distance threshold can ensure that even in the extreme case where a person approaches a live conductor at the highest speed, their final physical stopping position after braking is still outside the standard safe distance protection line for the corresponding voltage level.

[0043] After generating the safety distance threshold, the digital comparator module built into the system microprocessor performs a high-frequency, real-time closed-loop logic comparison between the real-time spatial distance output from the forward calculation in step S300 and the safety distance threshold, as well as a hardware-defined limit safety value. The limit safety value is numerically equivalent to the physical safety distance line at the corresponding voltage level. When the underlying logic comparison result indicates that the real-time spatial distance is less than the safety distance threshold but greater than the limit safety value, the microprocessor determines that the operator has entered the warning buffer zone. At this time, the microprocessor's pulse width modulation peripheral unit outputs a pulse width modulation control signal with a fixed duty cycle to the micro-vibration motor drive circuit inside the wearable terminal, activating the motor's operating circuit and driving the micro-vibration motor to generate continuous, constant-frequency physical mechanical vibration to remind the operator to stop the current approach and maintain the existing safety distance.

[0044] When the logical comparison result indicates that the real-time spatial distance has exceeded the physical limit, i.e., the value of the real-time spatial distance is less than the ultimate safety value, the system determines that the worker has entered an extremely high-risk gas breakdown discharge area. At this time, the microprocessor immediately activates the dynamic negative feedback drive adjustment mechanism, establishing a strict inverse proportional algebraic function mapping relationship between the duty cycle parameter of the pulse width modulation control signal and the value of the real-time spatial distance. As the real-time spatial distance between the human body and the high-voltage charged body gradually decreases, the microprocessor calculates in real time and synchronously linearly increases the duty cycle of the pulse width modulation control signal. By changing the duty cycle of the pulse width modulation signal, the system increases the equivalent effective value of the driving voltage at both ends of the vibration motor, forcing the micro vibration motor inside the wearable terminal to output a mechanical vibration signal with gradually increasing amplitude and vibration frequency. Through the dynamic and continuous adjustment of the duty cycle parameter of the underlying drive circuit, the system accurately and without delay converts the reduction in the microscopic spatial distance difference into an increasing mechanical vibration intensity perceived by the worker's skin nerves, thereby forcibly warning the worker to immediately move away from the high-voltage charged body with the highest priority physical stress feedback.

[0045] Furthermore, the specific implementation process of step S500 is as follows: In a specific embodiment, step S500 specifically includes the process of monitoring the work trajectory and attendance based on the operation ticket, as well as the specific calculation process of generating evaluation results of the operator's operating habits and planning the safety risk level of the work site using a cloud-based deep learning model.

[0046] First, the system assigns a unique identity ID to each power operation and maintenance worker through a pre-set database, and uniquely binds this ID to the MAC address of the wearable terminal. During power maintenance operations, the positioning module in the wearable terminal synchronously collects the worker's real-time geographic coordinates at set time intervals, and uploads it to the cloud platform via a low-power wide-area network (such as NB-IoT or 5G network), along with the current real-time electric field strength data, alarm status, and corresponding timestamps. After receiving the data packet, the cloud platform concatenates it in the time domain to generate the worker's real-time work trajectory. Subsequently, the cloud platform calls the electronic operation ticket file in the power system business database and extracts the polygonal geographic boundary coordinates of the authorized designated work area in the operation ticket. The system's computing unit uses the ray-crossing method to determine in real time whether the worker's real-time geographic coordinates are contained within the polygonal geographic boundary coordinates. When the calculation result shows that the real-time geographic coordinates are outside the boundary coordinates and the duration exceeds the preset drift tolerance time limit, the cloud platform determines that the worker has changed the operating range without authorization or mistakenly entered the energized area, and immediately triggers an out-of-bounds warning. The cloud platform then integrates the boundary crossing warning information, historical operation trajectory, duration of operation in a strong electric field environment, and frequency of distance alarm triggering into a multi-dimensional dynamic temporal feature vector, which serves as the basic input data for subsequent deep learning models.

[0047] In evaluating operational habits, since the safety awareness and operational habits of workers are a continuous process that evolves dynamically over time, conventional static threshold classification cannot accurately characterize this dynamic temporal feature. Therefore, a Long Short-Term Memory (LSTM) deep learning model is deployed in the cloud platform of this embodiment. The cloud platform inputs a multi-dimensional dynamic temporal feature vector, including alarm frequency, duration of strong electric field environment, and number of boundary violations, into the input gate of the LSTM network after slicing it according to the time series. The LSTM network uses its internal forget gate and update gate mechanism to weight and fuse the long-term safety performance of workers with recent violations. For example, when a worker frequently triggers near-electric alarms recently, the forget gate in the network will reduce the weight of his early standardized operation history features, thereby outputting a feature expression that tends towards high risk. The final fully connected layer of the LSTM network will output a normalized safety score quantification value. The system maps this quantification value to a preset safety operation level range (e.g., divided into four levels: excellent, good, medium, and high risk), thereby outputting the operational habit evaluation result of the worker. Through this deep learning mechanism, the system can accurately identify "high-risk personnel" who are constantly on the verge of electric shock, so that the competent authorities can conduct targeted safety training or job adjustments.

[0048] When planning the safety risk level of a work site, the cloud platform not only focuses on the dynamic temporal evaluation of individual worker behaviors, but also emphasizes the static investigation of structural hazards in the physical "environment" through spatial characteristics. Specifically, the cloud platform's underlying database pre-constructs a high-precision GIS (Geographic Information System) or BIM (Building Information Modeling) benchmark coordinate system for the entire work area. Based on the set spatiotemporal filtering criteria, the cloud platform's data processing engine extracts abnormal spatiotemporal feature clusters that meet specific abnormal threshold conditions from the massive multidimensional dynamic temporal feature vectors uploaded by all workers.

[0049] This anomalous spatiotemporal feature cluster is defined in the mathematical model as a set of multidimensional anomalous spatial data points. Each discrete outlier data point Each is structured into a tuple containing spatial location and physical field strength. This includes the high-precision three-dimensional geographic coordinates (longitude) of that point. ,latitude Elevation (and the peak absolute electric field intensity recorded at that spatial location) .

[0050] In order to construct this multidimensional anomaly space data point set During the data cleaning phase, the cloud platform established a dual extraction and filtering criterion that is independent yet complementary: the first criterion is the "over-limit alarm criterion," which involves the cloud platform traversing the database and extracting the three-dimensional geographic coordinates corresponding to all vibration alarm records that have been triggered according to step S400 and are below the specified safe distance threshold; the second criterion is the "absolute field strength criterion," which, considering the possibility of missed alarms in complex electromagnetic environments, involves the cloud platform independently setting a static physical threshold constant that characterizes extremely high dangerous field strength. (For example, the safe conversion value of the critical electric field strength for air breakdown at this voltage level can be taken), and the peak electric field strength in all uploaded data can be forcibly retrieved and extracted. The three-dimensional geographic coordinates of the data point are used regardless of whether the local alarm of the wearable terminal was successfully triggered when the data point was generated. Through the above dual extraction criteria, the cloud platform maps isolated high-risk data scattered in the feature vectors of different times and different workers into a unified set of high-precision three-dimensional anomaly coordinates with clear physical meaning.

[0051] After completing the above multidimensional anomaly spatial data point set After extraction and construction, in order to extract the physical regions with real structural security risks from this massive set of coordinates containing individual positional errors, the cloud platform uses a density-based spatial clustering algorithm to cluster the set. Perform spatial geometric analysis.

[0052] To extract truly structurally hazardous physical areas from a massive set of coordinates containing individual location errors, the cloud platform employs a density-based spatial clustering algorithm for spatial geometric analysis. The computational unit sets parameters for the cluster neighborhood radius and the minimum number of coordinate points required to form a high-density cluster. When the density of alarm coordinate points within a substation's spatial area exceeds the threshold set by the MinPts parameter, the DBSCAN algorithm automatically aggregates these coordinate points into a high-risk spatial cluster and automatically filters out isolated noise coordinate points caused by accidental individual misoperations. The cloud platform extracts the outer envelope contour of these high-risk spatial clusters and delineates them on the system's electronic map as high-risk work areas prone to safety accidents. This area often corresponds to specific equipment bays in actual substations with insufficient safety clearance design margins or abnormally strong electromagnetic coupling effects. The cloud platform then automatically generates a hazard investigation work order based on the area planning results and sends it to the relevant authorities, thus achieving a technological leap from passive personnel alarms to proactive investigation of power grid equipment and environmental hazards.

[0053] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions will not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A wearable power operation electric shock prevention early warning method based on cloud computing, characterized in that, Includes the following steps: Step S100: Capture the power frequency electric field distribution in the working environment and obtain the original spatial electric field signal by using the metal electrode needle plate electric field sensor built into the wearable terminal; Step S200: Perform feature analysis on the original spatial electric field signal to extract multi-dimensional electric field feature values, including the maximum electric field value, effective value, harmonic content, and peak conversion rate; Step S300: Based on the multidimensional electric field characteristic values, calculate the electric field strength of the charged body using a preset electric field inversion algorithm, and calculate the real-time spatial distance between the worker and the charged body; Step S400: Compare the real-time spatial distance with the safety distance threshold of the corresponding voltage level, and trigger the wearable terminal to perform graded vibration alarms of different intensities based on the comparison result; Step S500: Upload the operation data, including operation location, time and alarm frequency, to the cloud platform, and use a deep learning model to generate the operation habit evaluation results of the operators and the safety risk level planning of the operation site.

2. The wearable power operation electric shock prevention early warning method based on cloud computing according to claim 1, characterized in that, The metal electrode needle plate electric field sensor includes: The inner layer consists of two thin metal plates of equal area; A tetrafluoroethylene separator is disposed between the two thin metal plates to generate a pressure difference signal between the plate layers when the original spatial electric field passes through the two thin metal plates. The metal electrode needle plate electric field sensor is externally wrapped with spherical mica filler to capture the power frequency electric field distribution in the working environment and reduce the influence of the metal electrode plate on the surrounding electric field intensity distribution.

3. The wearable power operation electric shock prevention early warning method based on cloud computing according to claim 2, characterized in that, The grounding of the needle electrode of the metal electrode plate needle plate electric field sensor through a series resistor specifically includes: using the voltage drop formed on the series resistor by the capacitive current or corona current generated by the loop voltage to obtain the original spatial electric field signal.

4. The wearable power operation electric shock prevention early warning method based on cloud computing according to claim 3, characterized in that, In step S200, before performing feature analysis on the original spatial electric field signal, a signal preprocessing process is also included: The extremely weak original spatial electric field signal was sequentially amplified, filtered, and compared. The filtering process removes high-frequency noise from the environment to obtain a three-dimensional spatial electric field intensity signal at power frequency.

5. A wearable power operation electric shock prevention early warning method based on cloud computing according to claim 4, characterized in that, The process of extracting multidimensional electric field feature values ​​in step S200 specifically includes: The acquired three-dimensional spatial electric field intensity signal is subjected to peak detection and average value detection using the built-in signal processing algorithm; Combining the results of peak detection and average detection, the maximum value, effective value, harmonic content and peak conversion rate of the electric field are calculated and extracted. The electric field strength in three-dimensional space is evaluated from multiple angles to eliminate the interference of adjacent operating equipment and prevent false triggering.

6. A wearable power operation electric shock prevention early warning method based on cloud computing according to claim 5, characterized in that, Step S300, which involves calculating the real-time spatial distance using the electric field inversion algorithm, includes: Based on the ground voltage matrix of each conductor and the preset potential coefficient matrix The equivalent charge density matrix per unit length of each conductor is calculated using matrix equations. : ; Based on Gauss's law and the superposition theorem, the spatial sampling points where the workers are located are calculated according to the equivalent charge density of each conductor. Horizontal electric field component and vertical electric field components ; The sampling points Comprehensive electric field strength value : ; The comprehensive electric field strength value is obtained. The subsequent calculation process also includes: Based on a pre-defined mathematical model of exponential decay between the electric field intensity distribution of a charged body and spatial distance, the comprehensive electric field intensity value is... The inversion is used to determine the real-time spatial distance between the worker and the live conductor. The exponential decay mathematical model is modified based on the line-to-ground operating voltage, unit distance, and line-to-ground height.

7. A wearable power operation electric shock prevention early warning method based on cloud computing according to claim 6, characterized in that, In step S400, before the comparison is performed, a dynamic setting process for the safety distance threshold is also included: Obtain the operating voltage level of live conductors within the current work area; Based on the stated operating voltage level, match the corresponding standard safety distance; A preset warning margin is superimposed on the standard safe distance to generate the safe distance threshold used to trigger an alarm.

8. A wearable power operation electric shock prevention early warning method based on cloud computing according to claim 7, characterized in that, The graded vibration alarm process in step S400 specifically includes: When the real-time spatial distance is less than the safe distance threshold and greater than the set limit safety value, the wearable terminal is controlled to emit a continuous vibration signal to remind the operator to maintain a safe distance. When the real-time spatial distance is less than the limit safety value, as the real-time spatial distance gradually decreases, the wearable terminal is controlled to emit a gradually stronger vibration signal to warn the operator to stay away from the live conductor.

9. A wearable power operation electric shock prevention early warning method based on cloud computing according to claim 8, characterized in that, In step S500, before generating evaluation results using a deep learning model, a process for monitoring work trajectories and attendance based on operation tickets is also included: Assign a unique identification ID number to each worker; The location information of the workers is monitored in real time, and the time of their arrival and departure from the work area is recorded to generate a work trajectory; When it is detected that the operator has not arrived at the designated work area as required by the operation ticket, or has changed the work scope without authorization, an boundary crossing warning is triggered, and the boundary crossing warning information, work trajectory, duration of strong electric field environment and alarm frequency are integrated into the input data of the deep learning model. The method of using deep learning models to generate evaluation results of workers' operating habits and planning of safety risk levels for work sites specifically includes: Extract the frequency of alarms indicating insufficient safe distance from live conductors and the duration of exposure to strong electric fields from the input data. Use the deep learning model to perform feature classification, assign safety operation levels to the operators, and output the evaluation results of their operating habits. The system extracts geographic coordinates from the input data of areas where there are frequent occurrences of insufficient safe distance from charged objects or long-term high field strength. It then uses the deep learning model to perform spatial clustering analysis, delineates high-risk work areas prone to safety accidents, outputs a safety risk level plan for the work site, and sends a hazard investigation reminder to the competent authority.

10. A cloud-based wearable electric shock prevention and early warning system for power operations, and a cloud-based wearable electric shock prevention and early warning method for power operations according to any one of claims 1-9, characterized in that, include: The system includes an electric field signal acquisition module, a multi-dimensional feature extraction module, a safe distance inversion module, a local hierarchical early warning module, and a cloud-based risk assessment module.

Images