Electromagnetic shielding method and shielding system for power distribution

Abstract

The invention relates to an electromagnetic shielding method and shielding system for power distribution, and belongs to the technical field of electromagnetic shielding. Comprising the following steps: establishing a shielding model corresponding to the shielded room according to volume data of the shielded room; simulating and positioning electromagnetic wave conditions in the shielded room in the shielding model; selecting key detection points; constructing a phase difference matrix; shielding requirements of different waveforms in the same frequency band are distinguished, leakage path positioning based on a waveguide equation is carried out, and a core leakage area and an auxiliary protection area are divided. According to the electromagnetic shielding method and shielding system for power distribution provided by the invention, the dynamic adaptability of electromagnetic shielding and the detection capability of transient electromagnetic pulses are remarkably improved, and a multi-dimensional risk quantification and differentiation reinforcement strategy is realized. Meanwhile, intelligent materials and a dynamic optimization mechanism are adopted, the material utilization rate and the protection effect are improved, and a reinforcing effect feedback closed loop is formed.

Description

Technical Field

[0001] The invention relates to an electromagnetic shielding method and a shielding system for power distribution, belonging to the technical field of electromagnetic shielding. Background Art

[0002] Electromagnetic shielding technology is crucial in the power distribution sector. Its core goal is to isolate external electromagnetic interference and ensure the stable operation of electronic and electrical equipment within the shielded room. Shielded rooms, as a typical facility for electromagnetic shielding, physically block the propagation path of electromagnetic waves and are widely used in scenarios such as power control and communication base stations.

[0003] However, existing magnetic shielding technologies have significant shortcomings in dynamic adaptability, accurate leakage risk assessment, and differentiated reinforcement strategies, making them difficult to meet protection requirements in complex electromagnetic environments. Existing detection technologies lack the accuracy to capture transient electromagnetic pulses, relying solely on single field intensity detection without integrating time-domain parameters such as pulse width and rise time. This makes it difficult to accurately assess leakage risks in complex "high-frequency + fast-changing" scenarios. For example, narrow pulse signals in the 5G millimeter wave band are prone to efficient coupling through gaps. For electromagnetic waves within the same frequency band but with different time-domain characteristics, existing technologies lack a mechanism to correlate waveform characteristics with shielding requirements. For example, narrow pulses and fast-rising edge signals carry richer high-frequency components and pose a higher leakage risk. However, traditional methods only classify risk levels by frequency band, resulting in a "one-size-fits-all" protection strategy that can fail in high-risk scenarios and waste resources in low-risk scenarios. Existing leakage detection relies on single-point or sparse array sampling and lacks a coordinated analysis of field intensity, phase, and standing wave distribution within the shielded room's three-dimensional space. This makes it difficult to accurately locate waveguide leakage paths (such as gaps and seams) and standing wave antinodes (regions of peak field intensity). At the same time, the risk assessment did not integrate structural characteristics such as joint density and material electromagnetic parameters, resulting in a vague division between the core leakage area and the auxiliary protection area, and the reinforcement measures lacked spatial specificity. Summary of the Invention

[0004] The present invention provides an electromagnetic shielding method and shielding system for power distribution, so as to solve the problems in the prior art, such as the limitations of the static protection mode, insufficient transient electromagnetic pulse detection capability, lack of differentiated protection for waveforms in the same frequency band, and rough leakage path positioning and risk assessment.

[0005] The present invention provides an electromagnetic shielding method and system for power distribution, which includes establishing a shielding model corresponding to the shielding room based on the volume data of the shielding room, arranging a 16-channel synchronous sampling electromagnetic wave detection array in the shielding model, simulating and locating the electromagnetic wave conditions in the shielding room, using a density clustering algorithm to identify abnormal areas, selecting key detection points, constructing a phase difference matrix to perform a fast Fourier transform on the time domain phase data, converting it into a frequency domain distribution, extracting the phase periodic variation characteristics at a specific frequency, calculating the standing wave ratio to establish a frequency-distance-attenuation three-dimensional surface model, dividing the reinforcement area into three levels, calculating the deviation between the actual attenuation value and the expected value, and establishing a reinforcement effect feedback closed loop; The 16-channel detection array is used to collect pulse parameters and construct time domain waveform factors, quantify the time domain characteristics of the pulse, distinguish the shielding requirements of different waveforms in the same frequency band, and combine the pulse waveform characteristics with the frequency band risk. By calculating the modified impact index, standing wave ratio and leakage path positioning based on the waveguide equation, an electromagnetic risk heat map is generated, and the high-risk areas of electromagnetic leakage are identified. By calculating the joint density and comprehensive risk score, the core leakage area and auxiliary protection area are divided.

[0006] Preferably, a coordinate system is established with a corner of the shielding room as the origin, and a three-dimensional shielding model corresponding to the shielding room is established in proportion according to the three-dimensional rectangular coordinate system. A 16-channel synchronous sampling electromagnetic wave detection array is arranged, and a signal generation link for emitting standard electromagnetic waves is built.

[0007] A sliding window denoising method is used for the 16-channel synchronous sampling electromagnetic wave detection array data. The median of the five adjacent sampling points is calculated, and the median is used to replace the outliers to remove noise interference. All field strength data are normalized by frequency band and the core points are marked.

[0008] Starting from the core point, the density-connected points are merged into the same cluster, the average field strength of each cluster is calculated, the abnormal clusters are marked, the key detection points in the cluster are selected, the phase difference between each detection point is calculated, and the phase difference matrix is constructed.

[0009] The phase difference between each key detection point is calculated, and a phase difference matrix is constructed. Combined with the three-dimensional coordinates of the detection points, the detection point pairs that meet the node and antinode conditions are mapped into space, and the node areas and antinode areas of the standing wave are marked. The standing wave pattern is classified using a clustering algorithm, and the phase difference of each segment is calculated segmentally. The complete leakage path is located through the phase accumulation effect. In the three-dimensional model of the shielded room, the confirmed waveguide leakage path is marked, and the phase difference of each segment is calculated segmentally. The complete leakage path is located through the phase accumulation effect. In the three-dimensional model of the shielded room, the confirmed waveguide leakage path is marked.

[0010] Quantify the time domain characteristics of the pulse, distinguish the shielding requirements of different waveforms in the same frequency band, set the basic frequency band weight, preset the basic risk level according to the electromagnetic wave frequency band, and integrate the basic frequency band weight with the time domain waveform factor.

[0011] The shielded room space is refined into a grid, and parameters such as seam density and comprehensive risk score are calculated. The core leakage area and auxiliary protection area are divided. The reflection phase is adjusted in real time using an adjustable phase metasurface. The synthetic field strength is optimized through a three-dimensional electromagnetic coupling model. The GPS clock source is used to ensure the synchronization of detection data of each unit group. The scenario parameters, optimal reinforcement parameters and weight coefficients of previous leakage incidents are recorded.

[0012] Three types of heterogeneous attenuation nodes are deployed in the core leakage area and auxiliary protection area of the three-dimensional model of the shielded room. Through the standing wave interference effect, the attenuation increases exponentially with the number of nodes, so that the reflection phase difference between adjacent nodes in the ring topology is fixed at 180°, forming standing wave nodes.

[0013] A 16-channel detection array is used to collect field strength, phase, and pulse parameters in each node area in real time. The connection order of nodes in the ring topology is adjusted. The phase difference matrix and PID algorithm are used to fix the reflection phase difference between adjacent nodes at 180°. A gradient prediction model and a long short-term memory (LSTM) network are constructed to calculate differentials in real time and predict future threat index trends. The preset topology configuration is automatically selected according to the interference type to form a new ring topology.

[0014] An electromagnetic shielding system for power distribution includes a modeling and positioning module, a data acquisition and signal generation module, a data processing and algorithm module, a shielding reinforcement and dynamic control module, and an intelligent optimization and feedback module. The modeling and positioning module establishes a three-dimensional rectangular coordinate system with the corner of the shielding room as the origin, generates a shielding model in proportion, locates key locations of the shielding room structure, and refines the spatial discretization. The data acquisition and signal generation module uses a 16-channel synchronous sampling electromagnetic wave detection array to collect field strength and phase in real time, generate standard test signals and transient pulses, and simulate the real electromagnetic environment; The data processing and algorithm module is used to purify the raw data, unify the data format, perform sliding window denoising, and remove outliers; The shield reinforcement and dynamic control module implements differentiated reinforcement according to risk level, suppresses leakage and standing waves, dynamically adjusts the shield structure, and enhances multipath interference attenuation; The intelligent optimization and feedback module iteratively optimizes risk assessment parameters through historical data, verifies reinforcement effects, drives system self-optimization, synchronizes unit group numbers with timestamps, and tracks the spatiotemporal distribution of leakage events. Beneficial effects of the present invention: The present invention provides an electromagnetic shielding method and shielding system for power distribution, which upgrades the electromagnetic shielding technology from "static protection" to a new paradigm of "dynamic regulation" by combining dynamic impact assessment, phase interference analysis and asymmetric reinforcement algorithm. This enables the electromagnetic shielding system to adjust the shielding strategy in real time according to changes in the electromagnetic environment, more effectively suppress electromagnetic interference, and improve electromagnetic compatibility. The intelligent optimization and feedback module iteratively optimizes the risk assessment parameters through historical data, verifies the reinforcement effect, and drives the system to self-optimize. It adopts a 16-channel synchronous sampling electromagnetic wave detection array, combined with a dynamic impact assessment algorithm, which can accurately capture transient electromagnetic pulses and solve the problem of insufficient sensitivity of rising edge pulse detection in the existing technology. Through the newly added pulse parameter sensor, the pulse width and rise time are synchronously collected, the time domain waveform factor is constructed, and the time domain characteristics of the pulse are quantified, thereby distinguishing the shielding requirements of different waveforms in the same frequency band, and improving the identification accuracy of electromagnetic leakage risks. Through the 16-channel synchronous sampling electromagnetic wave detection array, the electromagnetic wave detection system can accurately capture the electromagnetic wave detection parameters, and solve the problem of insufficient sensitivity of rising edge pulse detection in the existing technology. Through the newly added pulse parameter sensor, the pulse width and rise time are synchronously collected, the time domain waveform factor is constructed, and the time domain characteristics of the pulse are quantified, thereby distinguishing the shielding requirements of different waveforms in the same frequency band, and improving the identification accuracy of electromagnetic leakage risks. The magnetic wave detection array realizes the coordinated analysis of the field intensity, phase, and standing wave distribution in the three-dimensional space of the shielded room, accurately locates the waveguide leakage path (such as gaps and joints) and the standing wave antinodes, and combines the frequency domain, time domain and spatial dimensions to form a five-dimensional comprehensive weight to achieve multi-dimensional risk quantification, making the reinforcement area division more accurate and scientific. According to the risk assessment results, the shielded room is divided into level 1 high-risk, level 2 medium-risk and level 3 low-risk areas, and different reinforcement measures are taken to ensure the protection effect and avoid resource waste. The GPS clock source is used to ensure the synchronization of the detection data of each unit group, and the scene parameters, optimal reinforcement parameters and weight coefficients of previous leakage events are recorded to form a reinforcement effect feedback loop to achieve system self-learning and optimization. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 The figure is a flow chart of an electromagnetic shielding method and shielding system for power distribution according to the present invention.

[0016] Figure 2 The figure is a schematic structural diagram of an electromagnetic shielding method and shielding system for power distribution according to the present invention. DETAILED DESCRIPTION

[0017] The preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0018] Example 1: The present invention provides an electromagnetic shielding method for power distribution, which includes obtaining volume data of a shielding room, establishing a three-dimensional rectangular coordinate system according to the volume data of the shielding room, establishing a shielding model corresponding to the shielding room in proportion according to the three-dimensional rectangular coordinate system, marking the position where an electromagnetic wave experiment needs to be performed in the shielding model, recording the marked point as a simulation point, drawing a perpendicular line from the simulation point to the six faces of the shielding model, and recording the intersection of the shortest perpendicular line and the face as a face point; drawing a perpendicular line from the electromagnetic simulation point to the twelve edges of the shielding model, and recording the intersection of the shortest perpendicular line and the edge as an edge point; connecting the electromagnetic simulation point with the eight vertices of the shielding model, and recording the vertex corresponding to the shortest connection line as a vertex point, evenly arranging 16-channel synchronous sampling electromagnetic wave detection arrays at the corresponding face points, edge points and vertex points in the shielding room, performing a standard electromagnetic wave experiment, building a signal generation link, scanning the carrier frequency in 1% steps, maintaining each frequency point for 60 seconds, alternating vertical and horizontal polarization of the antenna, synchronously recording 16-channel field strength values, and generating a frequency domain field strength distribution diagram; A dual-exponential pulse source was constructed with the following pulse parameters: rise time tr: 1 ns; pulse width tw: 50 ns (full width at half maximum); and peak voltage: 6 kV. A transverse electromagnetic wave chamber was used to inject the pulses, and a GPS-disciplined clock source was employed to ensure that the multi-channel sampling time deviation was ≤ 1 ns. Sliding window denoising was employed: for each sensor data point, the median of the five adjacent sampling points was calculated to replace outliers, and all field strength data were normalized to the [0, 1] interval by frequency band. Specifically:

[0019] in: : minimum field strength of the current frequency band; : Maximum field strength of the current frequency band; If there are ≥MinPts points in the ε neighborhood of a certain point, they are marked as core points. Starting from the core point, the density-connected points are merged into the same cluster. The average field strength of each cluster is calculated. If it exceeds 3 times the global average value, it is marked as an abnormal area, that is, an abnormal cluster. For the abnormal cluster, the average value of the coordinates of all points in the cluster and the two points farthest from the centroid in the cluster are selected as key detection points. A three-dimensional electromagnetic field distribution model is established. The shielding room is initially divided into 1m³ cubes and octree space segmentation is performed. The field strength of unmeasured locations is predicted based on the data of neighboring detection points. 10% of the detection points not used for modeling are randomly selected, and the prediction error is calculated. If the error is greater than 2dB, the grid of the area is refined and re-interpolated. Construct the phase difference matrix, specifically:

[0020] in: : the instantaneous phase of the i-th detection point; : The instantaneous phase of the jth detection point; N: total number of detection points; If the phase difference The two points are located in the same half-wavelength region of the standing wave, and there is a maximum point of field intensity (antinode) in the middle; if the phase difference between the two points is , then the two points are located in different half-wavelength regions of the standing wave, and there is a minimum field intensity point (node) in the middle; Perform a fast Fourier transform to convert the time-domain phase data into a frequency-domain distribution, extract the periodic phase variation characteristics at a specific frequency, and calculate the standing wave ratio (SWR). Specifically:

[0021] in: : reflection coefficient; Combined with the three-dimensional coordinates of the detection points, the detection points that meet the node and antinode conditions are mapped into space, and the node areas (field intensity troughs) and antinode areas (field intensity peaks) of the standing waves are marked. Based on the spatial distribution and characteristic frequencies of the nodes and antinodes, a clustering algorithm is used to classify the standing wave patterns. For each type of standing wave pattern, the corresponding high-risk area is marked. Calculate the impact index of each detection point, specifically:

[0022] in: : peak field strength; : field strength change rate; : Distance to the interference source, obtained through three-dimensional model positioning; The waveguide equation is used to analyze the leakage path, where the leakage path is the gap, hole or seam in the shielded room. Specifically:

[0023] in: : Maximum size of the gap; : magnetic permeability; : dielectric constant; The geometric parameters of the waveguide structure are associated with the three-dimensional coordinates of the detection points to establish a spatial position mapping relationship. Starting from the potential interference source outside the shielded room, the phase change is tracked along the waveguide equivalent path. For complex structures, the phase difference of each segment is calculated segment by segment. The complete leakage path is located through the phase accumulation effect. The confirmed waveguide leakage path is marked in the three-dimensional model of the shielded room. Establish a frequency-distance-attenuation three-dimensional surface model, select detection points in the three-dimensional space of the shielded room and define three orthogonal directions with each point as the center: X-axis (length direction), Y-axis (width direction), and Z-axis (height direction); For each detection point, electromagnetic wave attenuation data is collected under different conditions. Specifically, the frequency range is: 0.1-40GHz (step 0.1GHz, covering 5G millimeter wave and traditional communication frequency bands); the distance range is: the distance d from the detection point to the shielding wall; the polarization direction is: vertical polarization, horizontal polarization, circular polarization; For each direction (X / Y / Z) and polarization state, calculate the attenuation decibel value. For a single polarization direction, calculate the omnidirectional average attenuation value at each frequency f and distance d. Convert the attenuation decibel value to a linear attenuation multiple. For the target direction (such as the X axis) and polarization direction (such as vertical polarization), calculate the anisotropy correction factor. Specifically:

[0024] Comprehensive weight calculation, specifically:

[0025] in: 、 、 is the adaptive weight coefficient; Divide the reinforcement area into three levels according to the weight W: Level 1 high risk: , (such as standing wave antinodes and waveguide leakage entrances) adopt three-level reinforcement: base layer: apply conductive glue to fill gaps and enhance conductive continuity, middle layer: lay high-density copper mesh to suppress high-frequency leakage, surface layer: paste absorbing material to eliminate standing wave field intensity peak; Level 2 medium risk: , (such as ordinary seams, non-critical waveguide structures), base layer: conductive glue, middle layer: low-density copper mesh balance cost and high-frequency attenuation; Level 3 low risk: , (such as the center area of the complete shielding plate) use basic reinforcement and only apply conductive glue to ensure basic conductive continuity; in: is the maximum weight of all detection points in the current experiment; After reinforcement, the field strength data of each point is collected in real time through the 16-channel detection array, the deviation between the actual attenuation value and the expected value is calculated, a reinforcement effect feedback loop is established, historical reinforcement data is collected, and the weight coefficient is optimized using the Q-learning algorithm. 、 、 , after completing 5 experiments, the weight coefficient is automatically updated.

[0026] During specific use, the length, width and height of the shielding room are measured to determine its volume. A coordinate system is established with one corner of the shielding room as the origin and the length, width and height directions as the X-axis, Y-axis and Z-axis respectively. According to the three-dimensional rectangular coordinate system, a three-dimensional shielding model corresponding to the shielding room is established in equal proportion. In the shielding model, the positions where electromagnetic wave experiments need to be carried out are marked, and these marked points are recorded as simulation points. From each simulation point, perpendicular lines are drawn to the six faces of the shielding model, and the intersection points of the shortest perpendicular lines and the faces are recorded as face points; from each simulation point, perpendicular lines are drawn to the twelve edges of the shielding model, and the intersection points of the shortest perpendicular lines and the edges are recorded as edge points; from each simulation point, the eight vertices of the shielding model are connected, and the vertex corresponding to the shortest line is recorded as the vertex point. At the actual positions corresponding to the face points, edge points and vertex points determined in the model in the shielding room, 16-channel synchronous sampling electromagnetic wave detection arrays are evenly arranged to build a signal generation link for transmitting standard electromagnetic waves. The frequency was scanned in 1% steps to cover the relevant frequency band, and each frequency point was maintained for 60 seconds. The antenna polarization mode was alternating vertically and horizontally. During the experiment, a 16-channel synchronous sampling electromagnetic wave detection array was used to synchronously record the field strength values of each channel to generate a frequency domain field strength distribution diagram. The pulse parameters were set as follows: rise time tr was 1 ns, pulse width tw was 50 ns (half-maximum width), and peak voltage was 6 kV. A transverse electromagnetic wave chamber was used to inject pulses, and a GPS disciplined clock source was used to ensure that the multi-channel sampling time deviation was ≤ 1 ns, ensuring the time consistency of data acquisition in each channel. For the data collected by each sensor, a sliding window denoising method was used to calculate the median of the five adjacent sampling points, and the median was used to replace the outliers to remove noise interference. All field strength data were normalized according to the frequency band. Using the density clustering algorithm, the points with ≥MinPts points in the ε neighborhood of a certain point were marked as core points. Starting from the core point, the points with density connections were merged into the same cluster. Calculate the average field strength of each cluster. If it exceeds 3 times the global average value, mark the cluster as an abnormal area (abnormal cluster). For the abnormal cluster, select the average value of the coordinates of all points in the cluster as the centroid, and the two points farthest from the centroid in the cluster as key detection points. Initially divide the shielded room into 1m³ cubes and use the octree space segmentation method. Based on the data of the neighboring detection points, predict the field strength of the unmeasured position. Randomly select 10% of the detection points that are not used for modeling and calculate the prediction error. If the error is greater than 2dB, refine the grid of the area and re-interpolate to improve the prediction accuracy. Calculate the phase difference between each detection point and construct a phase difference matrix. If the phase difference is greater than 2dB, the phase difference matrix is constructed. The two points are located in the same half-wavelength region of the standing wave, and there is a maximum point of field intensity (antinode) in the middle; if the phase difference between the two points is , the two points are located in different half-wavelength regions of the standing wave, with a point (node) of minimum field intensity between them. A fast Fourier transform is performed on the time-domain phase data, converting it into a frequency-domain distribution. The periodic phase variation characteristics at specific frequencies are extracted and the standing wave ratio is calculated. Combined with the three-dimensional coordinates of the detection points, the detection points that meet the node and antinode conditions are mapped into space, and the node areas (field intensity troughs) and antinode areas (field intensity peaks) of the standing wave are marked. A clustering algorithm is used to classify standing wave patterns, and for each type of standing wave pattern, the corresponding high-risk area is marked. An impact index is calculated for each detection point to assess the impact of transient electromagnetic pulses. The waveguide equation is used to analyze the leakage paths of gaps, holes, or seams in the shielded room. The geometric parameters of the waveguide structure are associated with the three-dimensional coordinates of the test points to establish a spatial position mapping relationship. Starting from the potential interference source outside the shielded room, phase changes are tracked along the waveguide equivalent path. For complex structures, the phase difference of each segment is calculated segment by segment, and the complete leakage path is located through the phase accumulation effect. The confirmed waveguide leakage path is annotated in the shielded room 3D model. Test points are selected within the shielded room 3D space, and three orthogonal directions (X-axis (length), Y-axis (width), and Z-axis (height)) are defined around each point. Electromagnetic wave attenuation data is collected under different conditions. The attenuation decibel value is calculated for each direction (X / Y / Z) and polarization state and converted into a linear attenuation factor. For a single polarization direction, the omnidirectional average attenuation value is calculated for each frequency f and distance d. Anisotropy correction coefficients are calculated for the target direction (such as the X-axis) and polarization direction (such as vertical polarization). The comprehensive weight of each test point is calculated, and the reinforcement area is divided into three levels, with the first-level high-risk area receiving the third level of reinforcement. Conductive glue is applied to the base layer to fill gaps and enhance conductive continuity; high-density copper mesh is laid in the middle layer to suppress high-frequency leakage; absorbing material is pasted on the surface layer to eliminate the peak of the standing wave field strength; conductive glue is used for the base layer in the second-level medium-risk area, and low-density copper mesh is laid in the middle layer to balance the cost and high-frequency attenuation; the third-level low-risk area adopts foundation reinforcement and only applies conductive glue to ensure basic conductive continuity. After reinforcement, the field strength data of each point is collected in real time through the 16-channel detection array, the deviation between the actual attenuation value and the expected value is calculated, and a reinforcement effect feedback loop is established. The historical reinforcement data is collected, and the Q-learning algorithm is used to optimize the weight coefficient. After completing 5 experiments, the weight coefficient is automatically updated to continuously improve the reinforcement effect and system performance.

[0027] Compared with the existing technology, by establishing a three-dimensional rectangular coordinate system and a proportional shielding model, it is possible to accurately simulate and locate the electromagnetic wave situation in the shielding room. The arrangement of the 16-channel synchronous sampling electromagnetic wave detection array can comprehensively and synchronously record the field strength value in the shielding room, generate a frequency domain field strength distribution map, improve detection accuracy and efficiency, and expand the detection dimension from a single intensity to a four-dimensional fusion of intensity, phase, time domain, and spectrum, which significantly improves the standing wave positioning accuracy and helps to more accurately identify and solve electromagnetic interference problems. The density clustering algorithm is used to identify abnormal areas (abnormal clusters) and select key detection points for abnormal clusters, which helps to accurately identify and deal with weak links in electromagnetic wave shielding. According to the impact index and standing wave of the detection point Based on the characteristics of the signal, high-risk areas can be further divided and corresponding reinforcement measures can be taken to improve the shielding effect. The octree space segmentation method and the data of adjacent detection points are used to predict the field strength of unmeasured locations, which can improve the prediction accuracy and reduce the experimental cost. By establishing a reinforcement effect feedback loop, collecting historical reinforcement data, and optimizing the weight coefficient, the reinforcement plan can be continuously optimized to improve system performance. The electromagnetic shielding technology is upgraded to a "dynamic control" paradigm, which can adjust the shielding strategy in real time according to changes in the electromagnetic environment, thereby more effectively suppressing electromagnetic interference and improving electromagnetic compatibility. The dynamic impact assessment algorithm can accurately capture transient electromagnetic pulses, solving the problem of insufficient sensitivity in rising edge pulse detection in existing technologies. This helps to improve the response speed and accuracy of the electromagnetic shielding system to transient electromagnetic pulses.

[0028] Example 2: In the above embodiment, by combining dynamic impact assessment, phase interference analysis and asymmetric reinforcement algorithm, the electromagnetic shielding technology is upgraded from "static protection" to a new paradigm of "dynamic regulation". However, only a fixed frequency band weight factor F is used, which cannot distinguish electromagnetic waves with different time domain characteristics in the same frequency band, resulting in inaccurate protection strategy. The embodiment of the present application makes certain optimizations based on the above embodiment.

[0029] The pulse parameter sensor newly added to the 16-channel detection array in this embodiment extracts the pulse width (PW) and rise time (RT) from the pulse source parameters, where PW is the half-maximum pulse width and RT is the rise time from 10% to 90% of the amplitude. The time domain waveform factor is constructed based on the pulse width and rise time, specifically:

[0030] Narrow pulses (smaller PW) and short rise times (smaller RT) correspond to higher high-frequency components, exacerbating the risk of electromagnetic leakage. TWF is inversely proportional to PW and RT, increasing sensitivity to narrow pulses and fast-rising-edge waveforms. The basic weight F is set according to the electromagnetic wave frequency band: traditional communication frequency band F=1.0; 5G millimeter wave frequency band F=1.5; ultra-high frequency band F=2.0; Multiply the frequency band base weight by the time domain waveform factor to form a joint weight , specifically:

[0031] Among them, if the high frequency band (such as millimeter wave) has both narrow PW and short RT (such as 5G data transmission pulse), It will significantly increase the ability to accurately identify high-frequency + fast-changing compound high-risk scenarios; The joint weight is introduced into the modified impact index, specifically:

[0032]

[0033] Add pulse parameter sensors to the 16-channel detection array to synchronously collect PW and RT; calculate the corrected impact index in real time for each detection point

[0034] Added a revised impact index to the comprehensive weight W , forming a five-dimensional comprehensive weight, specifically:

[0035] in: : is the impact index weight coefficient (is the impact index weight coefficient);

[0036] By quantifying the time domain characteristics of the pulse through TWF, the shielding requirements of different waveforms in the same frequency band are distinguished. The dynamic weight F' couples the frequency band and time domain, amplifies the risk of narrow pulse width and fast rising edge characteristics of high frequency bands such as millimeter waves, and adapts to the physical characteristics of "high frequency leakage".

[0037] When in use, by utilizing the newly added pulse parameter sensor in the 16-channel detection array, the half-height pulse width (PW) and 10% to 90% amplitude rise time (RT) of the pulse signal are synchronously collected. The sensor works synchronously with the 16-channel electromagnetic wave detection array to ensure that the PW and RT data are time-aligned with the field strength, phase and other data, and to construct the time domain waveform factor. The smaller the PW (narrow pulse) and the smaller the RT (fast rising edge), the larger the TWF value, the richer the high-frequency components, and the higher the risk of electromagnetic leakage. The basic weight of the frequency band is set, and the basic risk level is preset according to the electromagnetic wave frequency band. The basic weight of the frequency band is integrated with the time domain waveform factor. Domain waveform factor, a joint weight is introduced into the impact index formula in Example 1. For each detection point, the revised impact index is dynamically calculated based on the PW, RT and field strength data collected in real time. The pulse waveform characteristics (PW / RT) are combined with the frequency band risk (F) through TWF and F' to solve the problem of inaccurate protection strategies for different waveforms in the same frequency band. On the basis of the three-dimensional weights in Example 1, a revised impact index is added to form a five-dimensional weight to realize the "frequency domain + time domain + space" multi-dimensional risk quantification. The reinforcement area division continues to use the three-level reinforcement strategy, but the weight calculation includes the newly added revised impact index.

[0038] Compared with existing designs, this design upgrades electromagnetic shielding from a "static protection" paradigm to a "dynamic control" paradigm by combining dynamic impact assessment, phase interference analysis, and an asymmetric reinforcement algorithm. This makes electromagnetic shielding more flexible and efficient, adapting to electromagnetic interference in diverse environments and conditions. A pulse parameter sensor has been added to the 16-channel detection array to simultaneously acquire the pulse width at half maximum (PW) and the rise time (RT) between 10% and 90% of the amplitude of the pulse signal. A time-domain waveform factor (TWF) is constructed based on these parameters. The TWF quantifies the pulse's time-domain characteristics and differentiates the shielding requirements for different waveforms within the same frequency band, thereby improving the accuracy of electromagnetic leakage risk identification. The frequency band base weight is multiplied by the time-domain waveform factor to form a joint weight (F'), which is then incorporated into the modified impact index. Each detection point can dynamically calculate the modified impact index based on real-time PW, RT, and field strength data. The TWF and F' combine the pulse waveform characteristics (PW / RT) with the frequency band risk (F), resolving the issue of misaligned protection strategies for different waveforms within the same frequency band. At the same time, a modified impact index is added on the basis of the three-dimensional weight to form a five-dimensional weight, realizing the multi-dimensional risk quantification of "frequency domain + time domain + space", making the reinforcement area division more accurate and scientific.

[0039] Example 3: In the above embodiment, the pulse waveform characteristics (PW / RT) are combined with the frequency band risk (F) through TWF and F', which solves the problem of misalignment of protection strategies for different waveforms in the same frequency band. However, it cannot suppress waveguide leakage and cavity resonance. The embodiment of the present application makes certain optimizations based on the above embodiment.

[0040] In this embodiment, a 16-channel detection array is used to synchronously collect the field strength E and phase of each detection point. , pulse width PW, rise time RT, calculate the modified shock index , combining the standing wave ratio and the leakage path location results based on the waveguide equation to generate the electromagnetic risk heat map of each area; Calculate the total length of the seams per unit volume to reflect the structural weakness of the shielding room. The degree of overlap with the operating frequency range divides the waveguide risk level: high risk ( <10)GHz), medium risk ( =10-20GHz), low risk ( >20GHz); Taking the corner of the shielding room as the origin, divide the cubic grid along the X, Y, and Z axes at a resolution of 1m³ to cover the entire shielding room space, generate an initial grid set, and perform the octree segmentation algorithm on the initial grid, then divide it into 8 sub-grids until the resolution is refined to 0.25m³, forming a refined grid set, and collect the total length of the seam in each grid. , Maximum gap size , material magnetic permeability , dielectric constant ; Use 16-channel detection array to collect field strength , Phase , pulse width , rise time , synchronously record the distance to the interference source

[0041] The integrity of the shielding structure within the grid is reflected by calculating the seam density. Specifically:

[0042] in: : is the grid volume; The higher the seam density, the denser the waveguide leakage paths; Calculate the single-test point modified impact index as follows: ,

[0043] in: : is the basic weight of the frequency band (traditional communication frequency band mmWave = 1.5, UHF = 2.0); Calculate the grid mean, specifically:

[0044] in: : is the number of detection points in the grid; Calculate the standing wave ratio of a single detection point, specifically:

[0045] in: : is the reflection coefficient; Grid extremes: , reflecting the extreme risk of the peak value of the standing wave field strength within the grid; Calculate the comprehensive risk score as follows:

[0046] in: : : The weight is 3:4:3; The joint density and comprehensive risk score are used as the two-dimensional feature space, and the neighborhood radius is set , the minimum number of core points MinPts=3; Mark Satisfaction The area with ≥MinPts high-risk grids in the neighborhood is the core leakage area. With the core area as the center, one grid is expanded outward to include adjacent areas with a joint density ≥20cm / m³ as auxiliary protection areas, forming a unit group structure of 1 core + 3-5 auxiliary areas. The unit group structure is checked to see if it contains a complete leakage path. If there is a break, the adjacent grids are automatically merged based on the spatial connectivity of the 3D model until the path is closed. For each unit group, a three-dimensional electromagnetic coupling model is established, where the coupling strength matrix is specifically:

[0047] in: : Indicates area arrive The electromagnetic wave coupling efficiency, the larger the value, the stronger the coupling; Phase difference between the core area and the auxiliary area within the group By deploying adjustable phase metasurface in the auxiliary area, the reflection phase is adjusted in real time to make the synthetic field strength within the group ; The 16-channel detection array is used to synchronously collect the field strength E and phase of each detection point in the unit group. , pulse parameters (PW / RT), data is updated every 1μs; Constructing a three-dimensional protection matrix model: ,in is the spatial coordinate, is the frequency, For time, a three-level reinforcement layer (conductive glue + high-density copper mesh + absorbing material) is deployed in the core area of the unit group, and a two-level reinforcement layer (conductive glue + low-density copper mesh) is deployed in the auxiliary area to form a cascade structure of core strong attenuation and auxiliary weak attenuation. The GPS clock source is used to ensure the synchronization of the detection data of each unit group, and a timestamp is added to the data of each detection point. The field strength E and phase are matched through time window matching. , pulse parameters (PW / RT) are bound to the unit group number in real time, recording the scene parameters of previous leakage events, optimal reinforcement parameters, and saving the weight coefficients during training 、 、 , the change curve, when a certain type of scene appears more than 5 times, the corresponding weight is automatically adjusted.

[0048] When in use, the 16-channel detection array is used to synchronously collect the field strength E and phase of each detection point in the shielded room. , pulse width PW, rise time RT, and record the distance D between each point and the interference source, and synchronously calculate the waveguide cutoff frequency ,according to The waveguide risk level is divided according to the degree of overlap with the operating frequency. With the corner of the shielded room as the origin, a cubic grid is divided at a resolution of 1m³ to generate an initial grid set. The initial grid is then divided into 8 equal subgrids by the octree segmentation algorithm until the resolution is refined to 0.25m³. This forms a refined grid set and collects the structural parameters of each grid: total length of the seam , Maximum gap size , material magnetic permeability , dielectric constant , calculate the grid joint density, single detection point modified impact index, single detection point standing wave ratio, grid extreme standing wave ratio, the comprehensive risk score integrates the joint density, average impact index, extreme standing wave ratio, and uses the joint density and comprehensive risk score as two-dimensional features. The density clustering algorithm is used to divide the core leakage area and the auxiliary protection area. With the core area as the center, one grid is expanded outward, and the adjacent area is included as the auxiliary protection area to form a "1 core + 3-5 auxiliary" unit group structure. Based on the spatial connectivity of the three-dimensional model, the adjacent grids are automatically merged to ensure the closure of the leakage path. A three-dimensional electromagnetic coupling model of the unit group is established. The adjustable phase metasurface deployed in the auxiliary area is used to adjust the reflection phase in real time to make the phase difference between the core area and the auxiliary area , achieving synthetic field strength To suppress electromagnetic leakage, the GPS clock source is used to ensure the synchronization of detection data of each unit group. A timestamp is added to each detection point, and the unit group number is bound. The detection array updates the data every 1μs and feeds back to the protection system in real time. The scenario parameters, optimal reinforcement parameters and weight coefficients of previous leakage events are recorded.

[0049] Compared with the existing technology, the 16-channel detection array synchronously collects the field intensity E and phase , pulse width PW, rise time RT and other parameters can more comprehensively evaluate electromagnetic risks. By calculating the modified impact index, standing wave ratio and leakage path positioning based on the waveguide equation, an electromagnetic risk heat map is generated, making risk assessment more intuitive and accurate. The octree segmentation algorithm is used to refine the shielding room space into a 0.25m³ grid, which can more accurately identify high-risk areas for electromagnetic leakage. By calculating parameters such as joint density and comprehensive risk score, combined with density clustering algorithm, the core leakage area and auxiliary protection area can be accurately divided, providing a scientific basis for subsequent reinforcement measures. The phase metasurface adjusts the reflection phase in real time, optimizes the synthetic field strength through a three-dimensional electromagnetic coupling model, and realizes intelligent suppression of electromagnetic leakage. The GPS taming clock source ensures the synchronization of detection data of each unit group, and records the scene parameters, optimal reinforcement parameters and weight coefficients of previous leakage events, so that the system can self-learn and optimize protection strategies. It adopts a "core strong attenuation-auxiliary weak attenuation" cascade structure for reinforcement. Compared with existing designs, it can more effectively reduce electromagnetic leakage and save reinforcement costs. By intelligently identifying high-risk areas and prioritizing reinforcement, it improves reinforcement efficiency and reduces unnecessary waste of resources.

[0050] Example 4: In the above embodiment, by adopting a cascade structure of "core strong attenuation-auxiliary weak attenuation" for reinforcement, electromagnetic leakage is effectively reduced and reinforcement costs are saved. However, the attenuation efficiency in the millimeter wave band and dynamic scenarios make it difficult to cope with transient strong interference. The embodiment of the present application makes certain optimizations based on the above embodiment.

[0051] In this embodiment, three types of heterogeneous attenuation nodes are deployed in the core leakage area and auxiliary protection area of the shielded room three-dimensional model: for high-frequency millimeter wave bands such as 28 GHz, a 5μm-thick graphene conductive film is deployed at the apex of the ring topology, covering 50% of the area. Its high conductivity is used to reduce surface reflection and skin effect loss of high-frequency electromagnetic waves; Ferrite magnetic sheets are deployed at the edges and joints with a coverage area of 33%, absorbing low-frequency interference below 1 GHz through magnetic loss. Plasma discharge units are embedded in the antinode area of the standing wave, covering 17% of the area, and respond to transient pulses such as EMP by ionizing the gas; With the core leakage area as the center, three adjacent graphene nodes, two ferrite nodes, and one plasma node are connected into a hexagonal ring topology through conductive copper tape. Each node is equipped with an adjustable phase shifter to achieve multipath reflection of electromagnetic waves in the ring path and calculate the total attenuation. Specifically:

[0052] in: : is the number of ring nodes; : is the reflection coefficient of each node; Through the standing wave interference effect, the attenuation increases exponentially with the number of nodes. The GPS servo clock source is used to synchronize the phase shifters of each node, so that the reflection phase difference between adjacent nodes in the ring topology is fixed at 180°, forming a standing wave node. Frequency of real-time collection of detection points , through the waveguide cutoff frequency Determine the dominant leakage mode when >10GHz: The proportion of graphene nodes increases to 60%, ferrite nodes decreases to 25%, and plasma nodes decreases to 15%, enhancing high-frequency reflection. ≤1GHz (low frequency band): the proportion of ferrite nodes is increased to 50%, graphene to 30%, and plasma to 20%, with enhanced magnetic loss; when transient pulses with a rise time RT < 5ns are detected: the proportion of plasma nodes is increased to 40%, graphene to 40%, and ferrite to 20%, with priority given to responding to fast pulses; Using phase difference matrix For any two nodes in a ring topology Make a judgment, if , it is determined to be a valid interference pair, and the phase shifter is adjusted by PID algorithm to make Approaching 180°, maximizing standing wave node area coverage; Calculate the joint density of each grid for SD>30cm / m³ and For high-risk grids >0.8, the graphene node density is increased by 20% (coverage area increased from 50% to 60%), and the ferrite node density is reduced by 10%, with priority given to strengthening high-frequency leakage paths; For SD≤10cm / m³ and For low-risk grids with a risk ratio of ≤0.5, the proportion of plasma nodes is reduced to 10%, redundant deployment is reduced, and material utilization is increased from 60% to 85%; The 16-channel detection array collects the field intensity E and phase of each node area every 1μs , pulse parameters (PW / RT), calculate the actual attenuation specific A measured = ; If |A measured - A total |> 3dB, the self-optimization mechanism is triggered to adjust the reflection coefficients of the graphene, ferrite, and plasma nodes, correct the node connection order in the ring topology, and automatically update the optimal ratio parameters after every 5 feedbacks; Obtain the current frequency band f, pulse width PW, and rise time RT through the 16-channel detection array, and calculate the frequency band-time domain joint weight Building a gradient prediction model , through the sliding window (window width 1μs) to calculate the differential in real time, triggering the dynamic reconstruction mechanism. The long short-term memory network (LSTM) is used to train the history Sequence, predict the threat index trend of the next 5μs, trigger reconstruction preparation in advance, and preset three typical topology configurations: high-frequency mode: 6-node full-graphene ring connection, phase difference 180°, attenuation 82dB; low-frequency mode: 3 ferrite + 2 graphene + 1 plasma ring connection, phase difference 150°, attenuation 65dB; transient mode: 4 plasma + 2 graphene ring connection, phase difference 180°, response time 5μs, and anti-interference improvement 4 times; when detected If the threshold is exceeded or the LSTM predicts an increase in the threat level, reconstruction is immediately initiated, the existing ring topology link is cut off, and the preset topology configuration is automatically selected based on the interference type to form a new ring topology. The optimal parameters of all previous reconstructions are recorded, and the FPGA configuration file is automatically updated after every 10 valid reconstructions.

[0053] When in use, with the area with the most serious electromagnetic leakage in the three-dimensional model of the shielded room as the core, auxiliary protection areas are divided radiating outward to the surrounding areas. A 5μm thick graphene conductive film is deployed at the vertex of the ring topology, with an initial coverage area of 50%. For high-frequency millimeter wave bands such as 28GHz, high conductivity is used to achieve surface reflection and skin effect loss. Ferrite magnetic sheets are deployed in areas with dense edges and seams, with an initial coverage area of 33%. Low-frequency interference below 1GHz is absorbed through magnetic loss. Plasma discharge units are embedded in the standing wave antinode area, with an initial coverage area of 17%. By ionizing gas to respond to transient pulses such as EMP, with the core leakage area as the center, three adjacent graphene nodes, two ferrite nodes, and one plasma node are connected into a hexagonal ring structure through conductive copper tapes. Each node is equipped with an adjustable phase shifter. Access the GPS service clock source to ensure that the reflection phase difference between adjacent nodes is initially fixed at 180° to form a standing wave node. Use a 16-channel detection array to collect parameters of each node area in real time with a period of 1μs, calculate the actual attenuation and the seam density of each grid, and use the phase difference matrix to determine the phase difference between any two nodes. Use the PID algorithm to adjust the phase shifter to make the phase difference approach 180°, maximize the coverage of the standing wave node area, enhance multi-path reflection attenuation, compare the measured attenuation with the theoretical calculated value in real time, automatically adjust the reflection coefficient of graphene, ferrite, and plasma nodes, correct the node connection order in the ring topology, update the optimal ratio parameters after every 5 feedbacks, build a gradient prediction model, calculate the differential in real time through a 1μs sliding window, detect the rate of change of the threat index, and use the long short-term memory network (LSTM) training history Sequence, predict the threat index trend of the next 5μs, preset 3 typical topologies, when When the corresponding threshold is exceeded or the LSTM-predicted threat level increases, the existing link is immediately cut off, and the preset topology is automatically selected according to the interference type to form a new ring topology. The optimal parameters of previous reconstructions are recorded, and the FPGA configuration file is automatically updated after every 10 valid reconstructions are completed.

[0054] Compared with the existing design, by deploying a 5μm thick graphene conductive film, its high conductivity is used to achieve surface reflection and skin effect loss of high-frequency electromagnetic waves, effectively reducing electromagnetic leakage. For low-frequency interference below 1GHz, ferrite magnetic sheets are deployed at edge points and dense seams to absorb magnetic loss, thereby improving the protection capability of the low-frequency band. Plasma discharge units are embedded in the standing wave antinode area, and the ionized gas responds instantaneously to transient pulses such as EMP, thereby enhancing the protection against transient strong interference. The field strength, phase, and pulse parameters of each node area are collected in real time through a 16-channel detection array, and the reflection coefficients of graphene, ferrite, and plasma nodes and the connection order of nodes in the ring topology are automatically adjusted according to real-time data, realizing intelligent dynamic optimization and utilizing phase The difference matrix and PID algorithm fix the reflection phase difference between adjacent nodes at 180°, forming standing wave nodes and enhancing the multipath reflection attenuation effect. Based on the real-time collected data and prediction model, the coverage area ratio of different nodes is automatically adjusted to optimize material utilization, increase the density of graphene nodes in high-risk grids, and reduce the proportion of plasma nodes in low-risk grids, thereby improving material utilization and saving costs. By constructing a gradient prediction model and a long short-term memory network (LSTM), differentials are calculated in real time and future threat index trends are predicted. When it is detected that the corresponding threshold is exceeded or the LSTM predicts an increase in the threat level, the reconstruction mechanism is immediately activated, and the preset topology configuration is automatically selected according to the interference type to form a new ring topology, which improves the response speed and anti-interference ability of the system.

[0055] Example 5: The present invention provides an electromagnetic shielding system for power distribution, which includes a modeling and positioning module, a data acquisition and signal generation module, a data processing and algorithm module, a shielding reinforcement and dynamic control module, and an intelligent optimization and feedback module. The modeling and positioning module establishes a three-dimensional rectangular coordinate system with the corner of the shielding room as the origin and generates a shielding model in proportion; marks simulation points, determines surface points, edge points, and vertex points through geometric operations, locates key positions of the shielding room structure, and refines spatial discretization to support high-precision analysis of local areas. The initial division is into 1m³ cubes, and the octree algorithm is used to segment layer by layer to a resolution of 0.25m³; and the structural parameters of each grid are associated; The data acquisition and signal generation module uses a 16-channel synchronous sampling electromagnetic wave detection array to collect field strength and phase in real time. A new pulse parameter sensor is added to obtain pulse width and rise time. Standard test signals and transient pulses are generated to simulate real electromagnetic environments. The data processing and algorithm module is used to purify raw data, unify the data format, perform sliding window denoising, and remove outliers; perform frequency band normalization to map field strength to the [0,1] interval; and perform fast Fourier transform (FFT) to convert time-domain phase distribution to frequency-domain distribution, extract phase periodicity, identify key electromagnetic leakage features and abnormal areas, predict field strength in unmeasured areas, and quantify three-dimensional spatial risks. The shield reinforcement and dynamic control module implements differentiated reinforcement according to risk level, suppresses leakage and standing waves, dynamically adjusts the shield structure, and enhances multipath interference attenuation; The intelligent optimization and feedback module iteratively optimizes risk assessment parameters through historical data, verifies the reinforcement effect, drives system self-optimization, synchronizes unit group numbers with timestamps, and tracks the spatiotemporal distribution of leakage events.

[0056] The above description of the present invention and its embodiments is non-limiting. The drawings show only one embodiment of the present invention, and the actual structure is not limited thereto. In short, if a person skilled in the art is inspired by the above and, without departing from the purpose of the present invention, designs structures and embodiments similar to the technical solution without creatively designing, they shall fall within the scope of protection of the present invention.

Claims

1. A method for electromagnetic shielding for power distribution, characterized in that: This includes establishing a shielding model corresponding to the shielding room based on its volume data, arranging a 16-channel synchronous sampling electromagnetic wave detection array within the shielding model, simulating and locating the electromagnetic wave conditions within the shielding room, using a density clustering algorithm to identify abnormal areas, selecting key detection points, constructing a phase difference matrix to perform a fast Fourier transform on the time-domain phase data, converting it into a frequency-domain distribution, extracting the periodic phase variation characteristics at specific frequencies, calculating the standing wave ratio to establish a frequency-distance-attenuation three-dimensional surface model, dividing the reinforcement area into three levels, calculating the deviation between the actual attenuation value and the expected value, and establishing a reinforcement effect feedback loop. The 16-channel detection array is used to collect pulse parameters and construct time domain waveform factors, quantify the time domain characteristics of the pulse, distinguish the shielding requirements of different waveforms in the same frequency band, and combine the pulse waveform characteristics with the frequency band risk. By calculating the modified impact index, standing wave ratio and leakage path positioning based on the waveguide equation, an electromagnetic risk heat map is generated, and the high-risk areas of electromagnetic leakage are identified. By calculating the joint density and comprehensive risk score, the core leakage area and auxiliary protection area are divided.

2. The electromagnetic shielding method for power distribution according to claim 1, characterized in that: A coordinate system is established with a corner of the shielding room as the origin. According to the three-dimensional rectangular coordinate system, a three-dimensional shielding model corresponding to the shielding room is established in proportion. A 16-channel synchronous sampling electromagnetic wave detection array is arranged, and a signal generation link for transmitting standard electromagnetic waves is built.

3. The electromagnetic shielding method for power distribution according to claim 1, wherein: A sliding window denoising method is used for the 16-channel synchronous sampling electromagnetic wave detection array data. The median of the five adjacent sampling points is calculated, and the median is used to replace the outliers to remove noise interference. All field strength data are normalized by frequency band and the core points are marked.

4. The electromagnetic shielding method for power distribution according to claim 1, wherein: Starting from the core point, the density-connected points are merged into the same cluster, the average field strength of each cluster is calculated, the abnormal clusters are marked, the key detection points in the cluster are selected, the phase difference between each detection point is calculated, and the phase difference matrix is constructed.

5. The electromagnetic shielding method for power distribution according to claim 1, characterized in that: The phase difference between each key detection point is calculated, and a phase difference matrix is constructed. Combined with the three-dimensional coordinates of the detection points, the detection point pairs that meet the node and antinode conditions are mapped into space, and the node areas and antinode areas of the standing wave are marked. The standing wave pattern is classified using a clustering algorithm, and the phase difference of each segment is calculated segmentally. The complete leakage path is located through the phase accumulation effect. In the three-dimensional model of the shielded room, the confirmed waveguide leakage path is marked, and the phase difference of each segment is calculated segmentally. The complete leakage path is located through the phase accumulation effect. In the three-dimensional model of the shielded room, the confirmed waveguide leakage path is marked.

6. The electromagnetic shielding method for power distribution according to claim 1, characterized in that: Quantify the time domain characteristics of the pulse, distinguish the shielding requirements of different waveforms in the same frequency band, set the basic frequency band weight, preset the basic risk level according to the electromagnetic wave frequency band, and integrate the basic frequency band weight with the time domain waveform factor.

7. The electromagnetic shielding method for power distribution according to claim 1, characterized in that: The shielded room space is refined into a grid, and parameters such as seam density and comprehensive risk score are calculated. The core leakage area and auxiliary protection area are divided. The reflection phase is adjusted in real time using an adjustable phase metasurface. The synthetic field strength is optimized through a three-dimensional electromagnetic coupling model. The GPS clock source is used to ensure the synchronization of detection data of each unit group. The scenario parameters, optimal reinforcement parameters and weight coefficients of previous leakage incidents are recorded.

8. The electromagnetic shielding method for power distribution according to claim 1, characterized in that: Three types of heterogeneous attenuation nodes are deployed in the core leakage area and auxiliary protection area of the three-dimensional model of the shielded room. Through the standing wave interference effect, the attenuation increases exponentially with the number of nodes, so that the reflection phase difference between adjacent nodes in the ring topology is fixed at 180°, forming standing wave nodes.

9. The electromagnetic shielding method for power distribution according to claim 8, characterized in that: A 16-channel detection array is used to collect the field strength, phase, and pulse parameters of each node area in real time. The connection order of the nodes in the ring topology is adjusted. The phase difference matrix and PID algorithm are used to fix the reflection phase difference of adjacent nodes to 180°. A gradient prediction model and a long-short-term memory network are constructed. The differential is calculated in real time and the trend of the future threat index is predicted. The preset topology configuration is automatically selected according to the interference type to form a new ring topology.

10. An electromagnetic shielding system for power distribution, used to implement the electromagnetic shielding method for power distribution according to any one of claims 1 to 9, characterized in that: It includes a modeling and positioning module, a data acquisition and signal generation module, a data processing and algorithm module, a shielding reinforcement and dynamic control module, and an intelligent optimization and feedback module. The modeling and positioning module establishes a three-dimensional rectangular coordinate system with the corner of the shielding room as the origin, generates a shielding model in proportion, locates the key positions of the shielding room structure, and refines the spatial discretization. The data acquisition and signal generation module uses a 16-channel synchronous sampling electromagnetic wave detection array to collect field strength and phase in real time, generate standard test signals and transient pulses, and simulate the real electromagnetic environment; The data processing and algorithm module is used to purify the raw data, unify the data format, perform sliding window denoising, and remove outliers; The shield reinforcement and dynamic control module implements differentiated reinforcement according to risk level, suppresses leakage and standing waves, dynamically adjusts the shield structure, and enhances multipath interference attenuation; The intelligent optimization and feedback module iteratively optimizes risk assessment parameters through historical data, verifies the reinforcement effect, drives system self-optimization, synchronizes unit group numbers with timestamps, and tracks the spatiotemporal distribution of leakage events.

Images