Electromagnetic radiation suppression method for flywheel energy storage system of data center

Abstract

The invention relates to the technical field of flywheel energy storage electromagnetic suppression, in particular to an electromagnetic radiation suppression method for a data center flywheel energy storage system, which comprises the following steps: acquiring grounding potential data to generate a response node set, comparing a rotating speed with a power event to construct a synchronous topology, identifying a coupling path to generate a path map, and adjusting a magnetic shielding assembly. And monitoring magnetic flux change to generate an electromagnetic compatibility state report. According to the method, the sudden change potential and the recovery rate of a grounding node are analyzed, response nodes with disturbance characteristics are screened, rotating speed and load power events are fused, multi-source data response mapping in a time domain is constructed, a coupling path is recognized based on a node physical connection relation, and the magnetic saturation trend is recognized by collecting magnetic flux density and hysteresis characteristics. The excitation phase is matched to implement dynamic adjustment, the response adaptability of the shielding assembly is improved, the stability is judged according to the magnetic flux density change rate trend, the reset operation of the shielding structure is controlled, the shielding effectiveness retention is enhanced, and electromagnetic suppression in a dynamic interference scene is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of electromagnetic suppression of flywheel energy storage, and in particular to a method for suppressing electromagnetic radiation of a flywheel energy storage system for a data center. Background Art

[0002] The field of electromagnetic suppression technology for flywheel energy storage includes control and intervention measures for electromagnetic radiation problems generated by flywheel energy storage systems during operation. The core content revolves around the electromagnetic interference generated by high-speed rotating components and their drive systems in flywheel energy storage devices to the external environment during energy conversion and transmission. The electromagnetic emission is constrained through hardware layout, circuit optimization and shielding design. It covers the structural layout of the energy storage system, electromagnetic compatibility design, electrical connection methods and grounding methods, and focuses on solving the needs of identifying and controlling electromagnetic radiation sources caused by the energy storage system under high-speed and high-frequency conditions to ensure its stable operation in complex power systems.

[0003] Among them, the electromagnetic radiation suppression method for the flywheel energy storage system in the data center refers to a method system for limiting or weakening the electromagnetic radiation interference problem generated by the flywheel energy storage device configured in the data center during operation, including the identification of the radiation path in the flywheel motor drive circuit, the construction of the filter circuit of the power input and output ports, the electromagnetic shielding structure design of the metal shell, and the adjustment and configuration of the system grounding method. It mainly reduces the leakage of high-frequency interference signals by designing a low-pass filter network, constructs a closed magnetic field confinement area by deploying conductive shielding materials at key parts of the flywheel system, and optimizes the electrical grounding form to form a low-impedance leakage path, thereby structurally suppressing electromagnetic radiation.

[0004] In existing flywheel energy storage systems, electromagnetic interference source identification relies heavily on structural planning and layout parameters, lacking a real-time mechanism for identifying dynamic disturbance behavior during operation. This makes it difficult to accurately track the actual propagation path of high-frequency disturbances. The lack of a synchronous comparison method based on event time series during processing prevents the establishment of an effective causal chain between different signal sources, making identification results prone to deviations or misjudgments. Topological connectivity is not used as an auxiliary criterion in path construction, resulting in a lack of structural logic constraints for the resulting shielding strategy, which in turn lacks precise spatial positioning capabilities. The shielding structure design lacks a dynamic adjustment mechanism based on magnetic state, making it unable to identify the nonlinear response of magnetic components caused by varying excitation conditions during actual operation. This can easily lead to fluctuations in shielding effectiveness, especially in high-speed, high-frequency power environments, resulting in unstable shielding zones. The reset control mechanism fails to effectively assess the long-term changes in magnetic components, and configuration adjustments remain at the static parameter setting stage. This risks degrading shielding capabilities over long-term operation, weakening the radiation path recovery capability and suscepting the formation of new electromagnetic leakage paths under persistent or intermittent excitation. Summary of the Invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a method for suppressing electromagnetic radiation for a flywheel energy storage system in a data center.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: a method for suppressing electromagnetic radiation of a flywheel energy storage system in a data center, comprising the following steps: S1: Obtain the time series data of the ground potential of nodes in the flywheel energy storage system grounding network, perform potential differentiation and mark the mutation points, select the nodes whose potential change rate exceeds the surge threshold within a single sampling cycle and returns to steady state within the next two cycles, and generate a transient response node set based on the timestamp; S2: Collecting flywheel speed time series data and load power time series data, detecting speed mutation events and power transient events respectively, extracting event times, performing time window comparison with the timestamps of the transient response node set, screening nodes with time differences less than the time synchronization threshold, and obtaining a synchronous response node topology; S3: Based on the synchronous response node topology, the corresponding grounding module identifier, shield connection point and rack grounding point are retrieved, an electromagnetic coupling path is constructed, the number of nodes and trigger characteristics in the path are counted, and an electromagnetic radiation path map is generated; S4: Read the magnetic shielding component number in the electromagnetic radiation path map, collect magnetic flux density data and hysteresis loop characteristics, determine whether there is a magnetic saturation trend, screen the magnetic shielding components currently in the excitation phase, and perform magnetic circuit optimization adjustment to generate a magnetic shielding optimization configuration table.

[0007] As a further solution of the present invention, the transient response node set includes a mutation intensity label, a steady-state recovery identifier, and a timestamp feature; the synchronous response node topology includes an event type correspondence, an inter-node time offset value, and a node physical hierarchy structure; the electromagnetic radiation path map includes the number of path nodes, path excitation trigger characteristics, and a coupling module structure identifier; and the magnetic shielding optimization configuration table includes a magnetic saturation prediction state, a component excitation phase identifier, and optimization adjustment parameters.

[0008] As a further solution of the present invention, the specific steps of obtaining the transient response node set are: S111: Based on the ground potential time series data of each node in the flywheel energy storage system grounding network, the potential change rate of the node in the current sampling period is calculated, and a sudden change detection of the potential difference is performed. The node change rate values ​​whose change rate exceeds the set surge threshold are screened to obtain sudden change potential rate data; S112: Based on the sudden potential change rate data and the ground potential change rate sequence data of the corresponding node in the subsequent two cycles, determine whether the potential has recovered to the steady-state range. If the judgment criteria for recovery to the steady state are met, classify it as a transient disturbance node, record the timestamp corresponding to the moment when the sudden change occurred, and obtain a transient disturbance time label set. S113: According to the transient disturbance time label set, the unique node identification information corresponding to each node at the mutation moment is extracted, and the disturbance response intensity value of the node is calculated by combining the node identification and the time label. The time label and node identification of the node whose disturbance response intensity value is greater than the disturbance response threshold are combined to generate a transient response node set.

[0009] As a further solution of the present invention, the specific steps of obtaining the synchronous response node topology are: S211: Obtain flywheel speed time series data and load power time series data within the flywheel energy storage system operation cycle, calculate the speed increment change value and incremental change rate between adjacent sampling points, identify speed mutation events and transient power mutation events, record the corresponding times, and generate a dual-event time label set; S212: Based on the dual-event time tag set and in combination with the corresponding time tags of each node in the transient response node set, determine the time difference between each event and the node timestamp, set a time synchronization threshold, retain node records whose absolute value of the time difference is less than or equal to the time synchronization threshold, and obtain a synchronization matching node index set; S213: According to the synchronous matching node index set, the connection relationships between all marked nodes are identified and numbered one by one, an adjacency matrix is ​​constructed and whether there is a direct connection relationship between the nodes is marked, and a synchronous response node topology is established.

[0010] As a further solution of the present invention, the specific steps of obtaining the electromagnetic radiation path map are: S311: Based on the physical connection relationship of each node in the synchronous response node topology, extract the grounding module identification information, shield connection point number, and rack grounding point number corresponding to each node connection segment, archive and summarize the grounding parameters corresponding to the nodes in the path segment item by item, and establish a node grounding feature index set; S312: Based on the node grounding feature index set, perform structural identification and trigger direction sorting on the node sequence in each path, determine the electromagnetic coupling characteristics within the path, mark the shielding integrity of the path and record the structural information, and generate a coupling path feature set based on the node sequence; S313: Based on the coupling path feature set, the nodes in each path are classified into node types, a path propagation sequence identifier is established, the trigger start order and propagation chain direction of each node are identified, the number of nodes in each path and the proportion of each type of node are counted, and the difference in path structure is represented by different color lines to generate an electromagnetic radiation path map.

[0011] As a further solution of the present invention, the specific steps for obtaining the magnetic shielding optimization configuration table are: S411: reading the magnetic shielding component number information marked on each path in the electromagnetic radiation path map, collecting magnetic flux density data, determining whether a saturation trend occurs, screening all magnetic shielding components that meet the saturation trend determination and recording their numbers, and obtaining a magnetic saturation component number set; S412: Based on the magnetic saturation component number set, the current excitation state of each component is collected. For each magnetic shielding component, the excitation waveform is identified as being in a rising edge time window. The current magnetic flux density, loop magnetic resistance, magnetic path length, number of coil turns, and driving frequency are extracted. The magnetic circuit impact value of the magnetic shielding component is calculated. The magnetic shielding components having a magnetic circuit impact value greater than a reference value are identified. A target list of magnetic permeability structures requiring reconstruction is established, and a magnetic circuit adjustment structure set is established. S413: Based on the magnetic circuit adjustment structure set, locate the corresponding magnetic permeance path for each magnetic shielding component to be adjusted in the structure diagram, judge the adjustment direction based on the magnetic circuit length and magnetic permeability change trend, record the component number, adjustment direction, target magnetic permeability value, and excitation frequency, and construct a magnetic shielding optimization configuration table.

[0012] As a further embodiment of the present invention, the method further comprises: S5: Counting the rate of change of the magnetic flux density of each component in the magnetic shielding optimization configuration table, monitoring the change amplitude within the moving average window, and triggering a magnetic circuit reset instruction if the continuous detection cycle shows a convergence trend, adjusting the magnetic shielding component to the initial configuration state, and generating an electromagnetic compatibility status report; The electromagnetic compatibility status report includes the magnetic flux density change trend, shielding configuration stability index, and reset adjustment trigger record.

[0013] As a further solution of the present invention, the specific steps of obtaining the electromagnetic compatibility status report are: S511: Calculate the magnetic flux density change rate between adjacent sampling periods based on all magnetic shield component numbers recorded in the magnetic shield optimization configuration table, calculate the absolute value of the continuous change rate of each magnetic shield component, and determine whether all of them are less than a preset change amplitude threshold. If so, the magnetic shield of the corresponding component is marked as tending to a stable state, and a magnetic flux convergence trend determination result is obtained. S512: Based on the magnetic shielding components identified in the magnetic flux convergence trend determination result, a sliding average window is set and a sequence of magnetic flux density change values ​​is extracted within each window. The consistency and convergence characteristics of the change amplitude are evaluated. If the change trend is stable and the difference is within the range, the corresponding magnetic shielding component is determined to meet the magnetic circuit recovery condition, and is added to the reset execution candidate set according to the number to obtain a magnetic circuit reset target set. S513: According to the magnetic circuit reset target set, compare the current magnetic permeability configuration of each magnetic shielding component with the initial factory configuration parameters, and perform deviation evaluation on the current parameters and initial values ​​of each magnetic permeability path. If the deviation range is within the specified threshold, the magnetic circuit reset action can be triggered to reset the magnetic permeability path configuration to the initial state. The status and path information of all restored components are summarized in the order of component numbers to generate an electromagnetic compatibility status report.

[0014] Compared with the prior art, the advantages and positive effects of the present invention are: In the present invention, by analyzing the sudden change potential and recovery rate of the grounding node, the response nodes with disturbance characteristics are screened, the electromagnetic disturbance identification accuracy is improved, the speed and load power events are integrated, and a multi-source data response mapping in the time domain is constructed to enhance the correlation of interference events. The coupling path is identified based on the physical connection relationship of the node, and the spatial pertinence of the propagation path construction is enhanced. The magnetic flux density and hysteresis characteristics are collected to identify the magnetic saturation trend, the excitation phase is matched to implement dynamic adjustment, the response adaptability of the shielding component is improved, the stability is judged by the trend of the magnetic flux density change rate, the reset operation of the shielding structure is controlled, the shielding effectiveness retention is enhanced, and a data-driven response control closed loop is constructed to achieve electromagnetic suppression in dynamic interference scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 It is a flow chart of the main steps of the present invention; Figure 2 Obtaining a flow chart for the transient response node set of the present invention; Figure 3 A flowchart of obtaining the synchronous response node topology of the present invention; Figure 4 A flow chart for obtaining the electromagnetic radiation path map of the present invention; Figure 5 A flow chart for obtaining a magnetic shield optimization configuration table of the present invention; Figure 6 The present invention provides a flowchart for obtaining an electromagnetic compatibility status report. DETAILED DESCRIPTION

[0016] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0017] In the description of the present invention, it should be understood that the terms "length," "width," "up," "down," "front," "back," "left," "right," "vertical," "horizontal," "top," "bottom," "inside," "outside," and the like, indicating positions or relationships, are based on the positions or relationships shown in the accompanying drawings and are intended only to facilitate the description of the present invention and simplify the description. They do not indicate or imply that the devices or elements referred to must have a specific orientation, be constructed, or operate in a specific orientation. Therefore, they should not be construed as limiting the present invention. Furthermore, in the description of the present invention, "plurality" means two or more, unless otherwise expressly and specifically defined.

[0018] See also Figure 1 , a method for suppressing electromagnetic radiation for a flywheel energy storage system in a data center, comprising the following steps: S1: Obtain the ground potential time series data of each node in the flywheel energy storage system grounding network, perform potential differentiation on each node sequence and mark the mutation points, select the nodes whose potential change rate exceeds the surge threshold (according to the transient overvoltage threshold set by the IEC61000-4-5 standard) within a single sampling cycle and recover to a steady state within the next two cycles, and generate a transient response node set based on the timestamp; S2: Collect flywheel speed time series data and load power time series data, detect speed mutation events and power transient events respectively through the two sets of sequences, extract the event time, and compare the time window with the timestamp of the transient response node set. Nodes with a time difference less than the time synchronization threshold (the standard time window for determining power system synchronization events (usually ≤10ms)) are selected to obtain the synchronous response node topology; S3: Based on the physical connection relationship of the synchronous response node topology, the corresponding grounding module identifier, shield connection point, and rack grounding point are retrieved, and an electromagnetic coupling path (an electromagnetic radiation propagation model that complies with the IEEE C95.3 standard) is constructed. The number of nodes and trigger characteristics in the path are counted to generate an electromagnetic radiation path map. S4: Read the number of the magnetic shielding component (magnetic conductive material component that complies with the MIL-STD-461G standard) in the electromagnetic radiation path map, collect magnetic flux density data and hysteresis loop characteristics (standard magnetization characteristic curve of ferromagnetic materials), determine whether there is a magnetic saturation trend (magnetic flux density reaches more than 90% of the material's saturation magnetization intensity), select the magnetic shielding component currently in the excitation phase, perform magnetic circuit optimization and adjustment, and generate a magnetic shielding optimization configuration table; S5: Count the rate of change of the magnetic flux density of each component in the magnetic shielding optimization configuration table, monitor the change amplitude within the moving average window (the data smoothing window set according to the EMI test standard CISPR16-1-1), and if the continuous detection cycle shows a convergence trend (the rate of change decreases for three consecutive sampling cycles and the absolute value is <5%), trigger the magnetic circuit reset instruction, adjust the magnetic shielding components to the initial configuration state, and generate an electromagnetic compatibility status report.

[0019] The transient response node set includes mutation intensity labels, steady-state recovery identifiers, and timestamp features. The synchronous response node topology includes event type correspondence, inter-node time offset values, and node physical hierarchy structures. The electromagnetic radiation path map includes the number of path nodes, path excitation trigger characteristics, and coupling module structure identifiers. The magnetic shielding optimization configuration table includes magnetic saturation prediction status, component excitation phase identifiers, and optimization adjustment parameters. The electromagnetic compatibility status report includes magnetic flux density change trends, shielding configuration stability indicators, and reset adjustment trigger records.

[0020] See also Figure 2 , the specific steps of S1 are: S111: Based on the ground potential time series data of each node in the flywheel energy storage system grounding network, the potential change rate of the node in the current sampling period is calculated, and a sudden change detection of the potential difference is performed. The node change rate values ​​whose change rate exceeds the set surge threshold are screened to obtain sudden change potential rate data; Based on the ground potential time series data of each node in the grounding network of the flywheel energy storage system, the ground potential value of each node is extracted at any two consecutive sampling moments. For example, the potential of the first sampling point is 4.2V, the potential of the second sampling point is 5.5V, and the potential change value is 1.3V. Combined with the sampling period Δt=0.01s, the change rate is calculated to be 130V / s. In this way, the change rate of all nodes in the complete sequence is traversed in turn. In the process of potential difference mutation detection, the absolute value of the difference between the potential at the current sampling moment and the potential at the previous sampling moment is used for judgment. If the mutation change rate exceeds the threshold, the node is recorded as a candidate mutation node. The threshold is selected with reference to the transient overvoltage limit set by the IEC61000-4-5 standard and is set to 850V / s. If the rate of change exceeds this value, the node is considered to have a mutation. For example, if the potential of node A jumps from 2.1V to 10.7V within a certain period, and the sampling period is still 0.01s, then its rate of change is 860V / s, which has exceeded the threshold and is included in the candidate. At the same time, nodes with a rate of change less than 600V / s are excluded, that is, those that are below the interference judgment standard. The screening basis clearly defines the potential change threshold range. The rate of change between 600~850V / s is the interference judgment fuzzy area and is not recorded. After completing the node traversal according to this rule, a data set of potential mutation rates of nodes with a change rate exceeding the threshold is obtained. Among them, the mutation change rate value is the most critical numerical result, representing the intensity of the potential disturbance experienced by the node, and further serves as the basis for subsequent identification.

[0021] S112: Based on the sudden potential change rate data and the ground potential change rate sequence data of the corresponding node in the subsequent two cycles, determine whether the potential has recovered to the steady-state range. If the judgment criteria for recovery to the steady state are met, classify it as a transient disturbance node, record the timestamp corresponding to the sudden change time, and obtain the transient disturbance time label set; Based on the sudden potential change rate data, a trend judgment is performed based on the ground potential change rate sequence in two consecutive cycles after the node mutation. The judgment criterion is whether the potential change rate drops to the set steady-state range in two consecutive cycles. The steady-state recovery range is set to a change rate of less than 100V / s. If the potential change rate in the two cycles after the node mutation is 95V / s and 78V / s respectively, it is considered to have recovered to the steady state, meeting the transient disturbance characteristics and is classified as a transient disturbance node. If the potential change rate still maintains a high potential change rate in subsequent cycles, such as 340V / s and 400V / s, the node is eliminated. Whether the node meets the steady-state recovery requirements is determined by comparing the change rate sequence. The timestamp of the node mutation moment is recorded at sampling time t=0.25s. For example, node B has a mutation occurring at the 25th sampling point, and the corresponding timestamp is 25×0.01=0.25s. The node number and mutation time are combined and recorded as the label item "B-0.25". All nodes that meet the transient disturbance recovery criteria and their time labels are sorted in sequence to form a time annotation set as the basic data source for response identification.

[0022] S113: Extract the unique node identification information corresponding to each node at the mutation moment according to the transient disturbance time label set, and combine the node identification and time label to use the formula: ; Calculate the The disturbance response intensity value of each node , according to the time label and node identifier of the node whose disturbance response intensity value is greater than the disturbance response threshold, a transient response node set is generated, where, Representation node In the cycle The potential change value within Indicates the The time sequence number of the cycle, Representation node The standard deviation of the potential change rate in the two cycles before and after the mutation cycle, Representation node In the The stable potential offset value of the non-mutation cycle, Representation node The mutation peak change rate, Representation node The benchmark value of the steady-state potential change rate; According to the transient disturbance time label set, the node number and the corresponding mutation time in each label item are extracted, and the node time mapping is established item by item to generate the transient disturbance node set. Then, the disturbance response intensity value of each node in the set is calculated. If the change value of node C in the 1st to 5th cycle is 1.2, 2.8, 0.5, 0.9, and 1.4V respectively, then the corresponding , then the item is: ; Assume that the standard deviation of the potential change rate of the node in the period before and after the mutation is =12.5V / s, corresponding to the steady-state periodic offset potential value If they are 0.3V, 0.5V, and 0.7V respectively, then the sum of the squares is 0.09+0.25+0.49=0.83, and the square root is: ; The sudden peak change rate of node C is 870V / s, the steady-state reference change rate is 90V / s, and the absolute value of the difference is 780. The final calculation formula is: ; This value is compared with the disturbance response threshold of 0.035 (the setting is based on the statistical analysis results of the disturbance characteristics of multiple actual nodes in the flywheel energy storage system. 20 representative node samples were selected. Among the disturbance response intensity values ​​measured during the mutation period, about 85% were in the range of 0.02~0.034, and in this interval, the node potential took no more than three cycles to recover to the steady state after the mutation. When the intensity value exceeded 0.035, the disturbance showed the characteristics of high instantaneous peak, fast potential drop and significantly reduced subsequent fluctuation. Therefore, 0.035 was used as the critical value for distinction. This value rises moderately with the increase of the peak value of the mutation potential change rate and the sum of the squares of the steady-state offset. It has dynamic adaptability and is suitable for actual working conditions with wide node distribution and complex electrical ground network structure. It can stably reflect the significant boundary between mutation response and steady-state recovery). If , then node C is included in the transient response node set. This result shows that the severity of the mutation of node C reaches the recording standard and can be included in the analysis scope as a response node.

[0023] Table 1 Transient disturbance sample node parameters

[0024] Table 1 lists the key calculation parameters of node C. Referring to Table 1, the disturbance response intensity value is 0.0403, which meets the response node identification conditions and is added to the transient response node set.

[0025] The disturbance response intensity value is a quantitative indicator used to characterize the degree of transient disturbance experienced by a node in the grounding network of a flywheel energy storage system within a specific time period. It comprehensively reflects the cumulative degree of potential change rate of the node in multiple cycles before and after the mutation moment, the relative position of the mutation position in the time series, the degree of potential fluctuation before and after the mutation, and the degree of steady-state offset. It is normalized and adjusted by the difference between the disturbance peak and the steady-state baseline change rate. It is specifically reflected in the aggregation of the disturbance energy borne by the node in the grounding network. The higher the value, the more drastic and continuous mutation of the node potential in a short period of time, and has higher transient response characteristics. It is suitable for identifying and determining key disturbance sources, sensitive nodes and potential disturbance propagation paths, and is the core criterion for constructing a transient response node set.

[0026] See also Figure 3 , the specific steps of S2 are: S211: Obtain flywheel speed time series data and load power time series data within the flywheel energy storage system operation cycle, calculate the speed increment change value and incremental change rate between adjacent sampling points, identify speed mutation events and transient power mutation events, record the corresponding times, and generate a dual-event time label set; The flywheel speed time series data and load power time series data within the operating cycle of the flywheel energy storage system are obtained. First, the sampling interval is standardized and the sampling period is set to 10ms, that is, 100 data points are recorded per second. For the flywheel speed sequence, the speed difference between two adjacent time points is extracted. For example, at t1=1.0s and t2=1.01s, the speeds are 1500rpm and 1810rpm respectively, then the speed increment is 310rpm, which exceeds the set threshold of 300rpm. t2 is marked as the speed mutation moment. This threshold refers to the typical characteristics of the dynamic response of the rapid change of speed in the high inertia system, and its rationality is determined within the scope of the IEC60034 standard. The load power sequence is sampled. Using the same processing method, the difference between the power values ​​at every two adjacent moments is calculated. If the power increases instantaneously from 5.2 kW to 7.0 kW at t3 = 2.04 s, a change of 1.8 kW, exceeding the set power mutation threshold of 1.5 kW, then t3 is recorded as the power mutation moment. This threshold is set based on the typical instantaneous power increase characteristics in power load disturbance experiments to avoid misjudging power fluctuations. All time points that meet the conditions are marked with event types and stored in the event sequence, where the label "TS" indicates a speed mutation and "PL" indicates a power disturbance, for example, "TS-1.01" and "PL-2.04". This method completes the time label extraction and organizes it into a unified data set, forming a dual-event time label set.

[0027] S212: Based on the dual-event time tag set and the corresponding time tags of each node in the transient response node set, determine the time difference between each event and the node timestamp, set a time synchronization threshold, retain node records with an absolute value of the time difference less than or equal to the time synchronization threshold, and obtain a synchronization matching node index set; According to the dual event time label set, each mutation event time point is traversed and compared with the time label corresponding to the node in the transient response node set obtained in the previous stage. The difference between the node mutation time and the event time is extracted. If node R1 mutates at 1.013 seconds and event TS-1.01, the time difference between the two is 3ms, and the absolute value is less than the set threshold of 10ms, which meets the power system synchronization event judgment condition. The threshold is derived from the maximum allowable drift standard of the synchronization clock error (IEEE1588). The judgment standard is based on whether |t n -t e |≤0.01s for screening. After traversing all node and event combinations in sequence, record the node numbers of all nodes with a time difference within 10ms, such as nodes R1, R3, and R6, and organize them into an array [1, 3, 6], which is recorded as the synchronization matching node index set. Exclude the synchronization failure nodes and mark the event association for each index, such as R1→TS and R3→PL, to ensure the unique mapping between nodes and events, and obtain the synchronization matching node index set.

[0028] S213: Based on the synchronization matching node index set, the connection relationships between all marked nodes are identified and numbered one by one, an adjacency matrix is ​​constructed, and whether there are direct connections between nodes is noted. Nodes that have met the time synchronization conditions are marked as synchronization state 1, and nodes that are not involved are marked as state 0, and a synchronization response node topology is established; According to the synchronous matching node index set, the system connection relationship parameters of each node are extracted, including the bus number, feeder number, upstream and downstream node number and connection path identifier. Combined with the preset power grid topology table, the network structure to which it belongs is searched according to the node number, and the adjacent node relationship matrix is ​​sorted out. For example, node 1 is connected to nodes 2 and 5, and node 3 is connected to nodes 4 and 6. All connection relationships are uniformly marked, and the node status identifier of the node in the synchronous state is set to 1, and the others are set to 0. Finally, a topology table containing the structure number, connection path and synchronization status is constructed to generate the synchronous response node topology.

[0029] Table 2 Synchronous matching node structure and topology table

[0030] Table 2 lists the structural topology parameter information of each node after time synchronization screening. Referring to Table 2, the network propagation path of the node in the synchronous response can be determined based on the node position and connection path, supporting the process of forming the synchronous response topology.

[0031] See also Figure 4 , the specific steps of S3 are: S311: Based on the physical connection relationship of each node in the synchronous response node topology, extract the grounding module identification information, shield connection point number, and rack grounding point number corresponding to each node connection segment, archive and summarize the grounding parameters corresponding to the nodes in the path segment item by item, and establish a node grounding feature index set; Based on the physical connection relationship of the synchronous response node topology, each connection path composed of wires between nodes is structurally deconstructed and identification information is extracted. During the execution process, the topological connection matrix obtained in the previous step is first called to retrieve the node pairs in the connection path one by one. For example, the connection between node numbers N1 and N4 is a path segment. For this path segment, the electrical structure ports connected to its starting point and end point are extracted, and the connected grounding module code is obtained through structural port association. For example, N1 is connected to the GND-01 module, which represents the connection to the bus, and N4 is connected to the GND-02 module, which represents the connection to the shielding shell. The grounding module identification needs to be numbered and confirmed according to the factory electrical installation drawing; next, the interface parameters corresponding to the cables at both ends of the path are compared. For example, the N1 connection point is the shielding connector model SPX-M16, and it is judged to be the shielding point number SP-003 according to the numbering rule. Then, combined with the cabinet and rack structure diagram in the equipment installation diagram, its positioning coordinate value on the structural layout diagram is extracted (for example, N1 is installed on the third layer of rack K7, and the positioning point is X=1.2 m, Y=0.3 m), by matching the positioning coordinates to the grounding reference point list, the corresponding rack grounding point number is obtained as J-K7-3. After all parameters are extracted, they are merged to form a grounding parameter index entry containing the path number, start and end nodes, grounding module number, shield connection point number, and rack grounding point number. To improve structural clarity, the parameters of each path need to be recorded separately and numbered and summarized into a table, such as the path segment number "P-001", the associated nodes "N1-N4", the corresponding grounding module "GND-01 to GND-02", the shield number "SP-003", and the grounding point "J-K7-3". This process is carried out in sequence throughout the topological network, and finally a node grounding feature index set is obtained.

[0032] S312: Based on the node grounding feature index set, perform structural identification and trigger direction sorting on the node sequence in each path, determine the electromagnetic coupling characteristics within the path, mark the shielding integrity of the path and record the structural information, and generate a coupling path feature set based on the node sequence; Based on the node grounding feature index set, the structural features and grounding relationship analysis are performed for each node pair in each connection path. First, based on the known grounding module type, whether there is a cross-module connection in the path is determined. If the path starts with the grounding module GND-01 and ends with GND-03, and there is a relay node in between, it is marked as a multi-module path segment. The differences in grounding point locations in this structure may significantly affect the coupling path and require priority identification. Next, the node spacing is measured. Based on the physical coordinates of the nodes in the process wiring diagram, the Euclidean distance between each two adjacent nodes is calculated. If the path "P-002" consists of nodes N5→N6→N8, with node spacing of 0.6 meters and 0.9 meters, respectively, the total path length is 1.5 meters, which is recorded as the length parameter of the path segment. If the grounding point spacing between the node pairs exceeds 300 mm, it is marked as a high-spacing path segment. The shield connection point number sequence and the corresponding cable number are further recorded. For example, N5 uses shielded cable SP-009 and N6 uses shielded cable SP-010. The structural balance parameters within the path are established item by item. For each type of structural indicator, such as path length, grounding point spacing, and shielding number difference, a basic characteristic parameter table is established to evaluate the structural level of the path in the overall topology. For example, path "P-002" is connected within the same rack and contains multiple grounding modules and multiple shielding contacts. Its structural level is defined as "composite structure" and recorded in the characteristic field. Finally, a complete record table of parameter items such as structure type, length distribution, grounding point density, and shielding sequence is formed as a multi-dimensional structure archive of the path to generate a coupled path feature set.

[0033] S313: Based on the coupling path feature set, nodes in each path are classified into node types, path propagation sequence identifiers are established, the triggering start order and propagation chain direction of each node are identified, the number of nodes in each path and the proportion of each type of node are counted, and different color lines are used to represent path structure differences to generate an electromagnetic radiation path map; According to the coupling path feature set, the node number sequence involved in the path segment is traversed one by one, the path source node is determined according to the trigger timestamp of the starting node, and the propagation chain is constructed in ascending order of the node numbers in the topology. If the node sequence in the path segment "P-003" is N10→N11→N13→N15, and the triggering order is N10=0.04s, N11=0.06s, N13=0.08s, N15=0.1s, then it is defined as the source, relay, relay, and terminal node in turn, and the propagation direction is marked with arrows on the path structure diagram, from left to right as the start to terminal path direction; for the nodes in the path structure The total number of nodes is counted. Paths containing four nodes with a complete, uninterrupted structural sequence are labeled "continuous paths." If the path contains a jump node (e.g., N10→N13) and the intermediate node timestamp is missing, it is recorded as an "intermittent path." Next, the distribution of grounding modules at each node is counted. If multiple nodes share a grounding module (e.g., N10 and N11 both use GND-01), the ground sharing rate for that path is recorded as 50%. A propagation sequence table is compiled based on the node spacing and grounding point numbers. This table is then summarized by segment number, path type is distinguished by color, and the trigger direction is annotated with a line type. Finally, a complete path diagram is drawn. This path diagram should include information such as node number, propagation direction, node trigger time, and structural classification. It integrates topological structure and electromagnetic propagation characteristics to generate an electromagnetic radiation path map.

[0034] Table 3 Grounding path structure parameters

[0035] Table 3 lists the grounding module combination, path length, number of nodes, and sharing rate data of typical path segments. Table 3 can clearly show the differences in structural parameters of different paths, providing support for subsequent map generation.

[0036] See also Figure 5 , the specific steps of S4 are: S411: Read the magnetic shielding component number information marked on each path in the electromagnetic radiation path map, collect magnetic flux density data, determine whether a saturation trend occurs, screen all magnetic shielding components that meet the saturation trend judgment and record their numbers to obtain a magnetic saturation component number set; The magnetic shielding component number information in the electromagnetic radiation path map is read, and a list of components that meet the magnetic material requirements of the MIL-STD-461G standard is extracted. When performing the screening operation, all component numbers with the "M" prefix, such as M001 and M002, are extracted based on the map identifier. The material field in the component technical specification table is called to determine whether it is a standard magnetic material such as iron-based alloy, permalloy, or pure nickel alloy. The numbers that meet the standards are retained and recorded in the valid component list. The magnetic flux density is then collected using the embedded magnetic flux sensor module. Measurement points are set in the core and edge areas of the component, and the sampling time is set to 2ms. The reading value is: the magnetic flux of component M001 at t=2.00s is 0.79T, and at t=2. At 01s, the magnetic flux is 0.81T, and the collected average value is 0.80T. This value is compared with its saturation magnetic flux density. If the component material is 1J85 alloy, its saturation magnetic flux density is 0.89T, then the current magnetic flux is 89.9% of the saturation magnetization intensity. Combined with the magnetic saturation judgment reference value of 0.9×Bs=0.801T, because 0.80T is lower than the judgment threshold, the saturation mark is not triggered; then measure component M002, read the magnetic flux value of 0.84T, and its saturation magnetic flux density is 0.88T, corresponding to 90.9%, which is greater than the saturation judgment threshold of 0.792T. M002 is recorded as a magnetic saturation component. Such data are archived and organized according to component number, magnetic flux value, and judgment result, and finally a set of component numbers that contains all component numbers that meet the magnetic saturation trend conditions is formed, which is the magnetic saturation component number set.

[0037] S412: Based on the magnetic saturation component number set, the current excitation state of each component is collected. For the excitation waveform of each magnetic shielding component, whether it is in the rising edge time window is identified. The current magnetic flux density, loop magnetic resistance, magnetic path length, coil turns and driving frequency are extracted using the formula: ; Calculate the Magnetic circuit impact value of a magnetic shielding component , determine the magnetic shielding components whose magnetic circuit impact value is greater than the reference value, establish a target list of magnetic permeability structure reconstruction, and establish a magnetic circuit adjustment structure set, where, Indicates the current magnetic flux density of the component, is the saturation flux density of its magnetic material, Indicates the current driving frequency of the component, in Hz. Indicates the excitation circuit resistance in ohms. represents the number of turns of the excitation winding, Indicates the length of the magnetic permeability path in meters; Based on the magnetic saturation component number set, we extract whether each component is currently in the excitation phase state. The excitation state is determined based on the drive signal frequency and the current waveform phase. The excitation frequency range is set to 1kHz to 20kHz, and the sampling frequency is 50kHz. The power-on cycle of each component is analyzed. If component M005 shows a continuous current rise and a synchronous voltage rise segment in the interval t=0.300s~0.301s, it is determined to be in the excitation rising edge, marked as the excitation state, and recorded in the excitation component set; its magnetic flux density value, number of winding turns, path length, resistance, and frequency parameters are extracted. Taking M005 as an example, its magnetic flux density is 0.83T, saturation magnetic flux is 0.91T, frequency is 15kHz, loop resistance is 2.1Ω, coil turns are 20, and magnetic path length is 0.9m. Substituting the above values ​​into the formula, the calculation process is as follows: ; The result Zg is 0.0367, which is lower than the benchmark value of 0.05 (the setting is based on comparing the numerical distribution characteristics of the excess flux increase term and the magnetic circuit structure impedance term. When the magnetic flux density reaches more than 90% of the saturation intensity, the frequency exceeds 10kHz, the loop resistance is less than 3Ω, and the number of winding turns is not more than 25, the magnetic circuit impact value is concentrated between 0.04 and 0.06. Among them, those above 0.05 are generally accompanied by abnormal fluctuations in the magnetic permeability at the excitation end or non-uniform magnetic field distribution. Therefore, the value of 0.05 is taken as the critical reaction mark value of the magnetic circuit structure. This value increases with increasing frequency and decreasing winding density, and remains stable when the path length is less than 0.8m). Therefore, the magnetic permeability structure of this component does not need to be adjusted. If the result of a component such as M007 is 0.0643, which exceeds the benchmark value, it is included in the list of components that need to be adjusted. After traversing all components, an adjustment number and target direction structure set are formed. This is used as the execution object of the magnetic permeability path optimization, and a magnetic circuit adjustment structure set is established.

[0038] The magnetic circuit impact value is used to measure the degree of deviation between the electromagnetic response intensity of the magnetic shielding component under the excitation state and the bearing limit of the magnetic circuit structure. It specifically reflects the transient magnetic response load formed by the combined action of multiple factors such as the magnetic flux density, excitation frequency, resistance limit, coil structure and magnetic guide path length when the component approaches the magnetic saturation boundary. The larger the value, the more likely the current magnetic circuit system of the component is to produce nonlinear magnetization behavior or approach the saturation state due to excitation coupling, thereby resulting in a decrease in magnetic shielding effectiveness or abnormal magnetic energy accumulation. Therefore, the magnetic circuit impact value not only comprehensively characterizes the degree of coordination and matching between the current magnetic state of the component and the structural configuration, but also provides a quantitative basis for the optimization and adjustment of the magnetic guide path.

[0039] The formula is obtained by the square term of the excess magnetic flux density Reflects the difference between the current component and its saturation critical point, and compares the difference term with the excitation frequency Multiplication reflects the enhanced effect of flux response on magnetic circuit impact at different excitation rates. The loop resistance in the denominator is and winding turns The electromagnetic impedance factor represents the restrictive effect of the circuit on the flow of the excitation current. The greater the impedance, the weaker the response. Therefore, the amplitude is attenuated by using this term as a divisor in the formula, forming a modulation structure that is inversely related to the magnetic saturation trend. The right side is multiplied by The term combines the influence of the length of the magnetic permeability path and the winding structure density on the regulation of the entire magnetic field distribution path, where represents the length of the magnetic circuit body, and As the inverse of the coil density, it is used to weaken the regulating effect of the high-turn winding on the length. Therefore, the integral logic is reflected as follows: the critical difference of magnetic saturation is amplified by the excitation frequency and then weakened according to the electromagnetic impedance, and then the direction is corrected according to the change of the structural path, thereby fully expressing the magnetic circuit impact strength of the magnetic shielding component under the excitation state when it is close to saturation.

[0040] S413: Based on the magnetic circuit adjustment structure set, locate the corresponding magnetic permeance path for each magnetic shielding component to be adjusted in the structure diagram, determine the adjustment direction based on the magnetic circuit length and magnetic permeability change trend, record the component number, adjustment direction, target magnetic permeability value, and excitation frequency, and construct a magnetic shielding optimization configuration table; Based on the magnetic circuit adjustment structure set, the magnetic circuit is optimized and reconstructed for the identified magnetic shielding component numbers that need adjustment. First, the magnetic permeability path segment corresponding to the component is located in the structure diagram, and the starting and ending node numbers are recorded. For example, the M007 connection path segment is N14→N16→N19, with a total path length of 1.2m. An evenly spaced node configuration is used, with a spacing of 0.6m between each segment. The original magnetic permeability structure is judged to be a single path configuration, with an initial permeability value μ of 4200H / m. The direction is determined by combining the magnetic permeability change rule and the current excitation state of the component. If the magnetic permeability distribution along the N14→N19 direction is 4200→3900→3600H / m, the adjustment direction is set to positive permeability increase, and the target value is set to an increase of 5%. The target μ is recorded as 4410, 4095, and 3780H / m respectively. After adjustment, the magnetic circuit configuration needs to replace the magnetic permeability of each path segment and record the corresponding segment number, magnetic permeability before and after adjustment, target change amplitude, and excitation frequency in the configuration table. Finally, a table configuration structure is constructed based on the component number, path direction, target permeability value, and corresponding frequency to generate a magnetic shielding optimization configuration table.

[0041] Table 4 Optimized configuration of magnetic shielding components

[0042] Table 4 lists the permeability settings and adjustment ratios before and after adjustment for the optimization of the magnetic guide path structure of a typical excitation component. See Table 4, which serves as a configuration reference source for executing the magnetic guide path adjustment task.

[0043] See also Figure 6 , the specific steps of S5 are: S511: Calculate the magnetic flux density change rate between adjacent sampling periods based on the numbers of all magnetic shield components recorded in the magnetic shield optimization configuration table. Calculate the absolute value of the continuous change rate of each magnetic shield component and determine whether it is less than a preset change amplitude threshold. The magnetic shield of the corresponding component is marked as tending to a stable state, and the magnetic flux convergence trend determination result is obtained. The magnetic flux density change rate of the magnetic shielding components listed in the magnetic shielding optimization configuration table is statistically analyzed. Three consecutive sampling cycles are set for each component to acquire data. The sampling cycle is 10ms. Time series data is constructed and the magnetic flux density difference operation is performed point by point. The change rate is obtained by dividing the change between adjacent points by the flux value of the previous cycle. The magnetic flux density of component M012 at t=0.100s, 0.110s, and 0.120s is 0.81T, 0.79T, and 0.77T, respectively. The change rates are (0.79–0.81) / 0.81=–2.47% and (0.77–0.79) / 0.79=–2.53%, respectively. They are all negative values ​​and the absolute values ​​are If the value is lower than 5%, the direction is consistent and the amplitude is within the convergence judgment interval; then the component M015 is tested, the sampling values ​​are 0.84T, 0.82T, and 0.83T, and the corresponding change rates are –2.38% and –1.22%. The third segment of change is positive and does not meet the continuous decline condition, so it is not included; in this way, all components in the magnetic shielding optimization configuration table are processed one by one, and the components that meet the three-segment continuous decline and the absolute value of the change rate is less than 5% are numbered, a magnetic flux convergence trend sequence is generated, and a data list is established by number. Finally, 8 numbers that meet the trend criteria, such as M012, M018, and M021, are extracted from all components and recorded to obtain a magnetic flux convergence trend judgment set.

[0044] S512: Based on the magnetic shielding components identified in the magnetic flux convergence trend determination result, a sliding average window is set and a sequence of magnetic flux density change values ​​is extracted within each window. The consistency and convergence characteristics of the change amplitude are evaluated. If the change trend is stable and the difference is within the range, the corresponding magnetic shielding component is determined to meet the magnetic circuit recovery conditions and is added to the reset execution candidate set according to the number to obtain the magnetic circuit reset target set. Based on the component numbers in the flux convergence trend judgment set, the change amplitude in the sliding average window is evaluated, and the window width is set to three sampling periods. The flux density sequence of M012 from t=0.100s to t=0.120s is [0.81T, 0.79T, 0.77T], and the average value is 0.79T. Each difference in the window is normalized and combined with the frequency, resistance, and permeability parameters to construct a convergence trend evaluation item. The frequency of M012 is 15kHz, the resistance is 2.1Ω, and the permeability is 4100H / m. Compared with the standard frequency of 10kHz and the standard permeability of 4000H / m, the change is obtained. The amplitude is within the preset deviation tolerance and is marked as a stable change. The measured window of M018 is [0.84T, 0.82T, 0.80T], the average value is 0.82T, the frequency is 16kHz, the resistance is 1.9Ω, and the magnetic permeability is 4300H / m. It is judged that its amplitude is also in the controllable range and is also included. If it is detected that M019 has a certain section of change exceeding 4% in the sequence [0.87T, 0.84T, 0.81T], it will be eliminated. Finally, the component set that meets the reset judgment conditions includes 5 components numbered M012, M018, M021, M022, and M025, and the magnetic circuit reset target set is established.

[0045] S513: Based on the magnetic circuit reset target set, the current magnetic permeability configuration of each magnetic shielding component is compared with the initial factory configuration parameters. Deviation evaluation is performed on the current parameters of each magnetic permeability path and the initial values. If the deviation range is within the specified threshold, the magnetic circuit reset action is triggered to reset the magnetic permeability path configuration to the initial state. The status and path information of all restored components are summarized in component number order to generate an electromagnetic compatibility status report. According to the magnetic circuit reset target set, taking components M012 and M018 as examples, the magnetic permeability of their current magnetic shielding configuration is compared with the factory initial configuration. The current magnetic permeability of path P23 of component M012 is 4050H / m, the initial value is 4000H / m, and the deviation is +1.25%, which is less than the maximum allowable offset of 5%. It is determined that it can be reset directly, corresponding to a magnetic permeability path length of 1.1m and 25 turns. The current configuration of component M018 is 3850H / m, the initial value is 3800H / m, the deviation is +1.32%, the resistance is 2.3Ω, and the frequency is 14.5kHz, which is also within the allowable range. The reset is confirmed. After each reset, the component number, path number, original and current magnetic permeability, change amplitude and status identification are recorded, and finally structured data in tabular form is generated to form a unified record format and generate an electromagnetic compatibility status report.

[0046] Table 5 Magnetic circuit reset component status record

[0047] As shown in Table 5, the reset components M012 and M018 have recovered their magnetic permeability within the adjustment range. The table summarizes the reset status data by number. See Table 5 for complete reset task archive information.

[0048] The above are merely preferred embodiments of the present invention and do not limit the present invention in any other form. Any technician familiar with the profession may use the technical content disclosed above to change or modify it into an equivalent embodiment with equivalent changes and apply it to other fields. However, any simple modification, equivalent change and modification made to the above embodiment based on the technical essence of the present invention without departing from the content of the technical solution of the present invention shall still fall within the scope of protection of the technical solution of the present invention.

Claims

1. A method for suppressing electromagnetic radiation of a flywheel energy storage system in a data center, characterized in that: The following steps are involved: S1: Obtain the time series data of the ground potential of nodes in the flywheel energy storage system grounding network, perform potential differentiation and mark the mutation points, select the nodes whose potential change rate exceeds the surge threshold within a single sampling cycle and returns to steady state within the next two cycles, and generate a transient response node set based on the timestamp; S2: Collecting flywheel speed time series data and load power time series data, detecting speed mutation events and power transient events respectively, extracting event times, performing time window comparison with the timestamps of the transient response node set, screening nodes with time differences less than the time synchronization threshold, and obtaining a synchronous response node topology; S3: Based on the synchronous response node topology, the corresponding grounding module identifier, shield connection point and rack grounding point are retrieved, an electromagnetic coupling path is constructed, the number of nodes and trigger characteristics in the path are counted, and an electromagnetic radiation path map is generated; S4: Read the magnetic shielding component number in the electromagnetic radiation path map, collect magnetic flux density data and hysteresis loop characteristics, determine whether there is a magnetic saturation trend, screen the magnetic shielding components currently in the excitation phase, and perform magnetic circuit optimization adjustment to generate a magnetic shielding optimization configuration table.

2. The electromagnetic radiation suppression method for a flywheel energy storage system in a data center according to claim 1, characterized in that: The transient response node set includes a mutation intensity label, a steady-state recovery identifier, and a timestamp feature; the synchronous response node topology includes an event type correspondence, an inter-node time offset value, and a node physical hierarchy structure; the electromagnetic radiation path map includes the number of path nodes, path excitation trigger characteristics, and a coupling module structure identifier; the magnetic shielding optimization configuration table includes a magnetic saturation prediction state, a component excitation phase identifier, and optimization adjustment parameters.

3. The electromagnetic radiation suppression method for a flywheel energy storage system in a data center according to claim 1, characterized in that: The specific steps of obtaining the transient response node set are: S111: Based on the ground potential time series data of each node in the flywheel energy storage system grounding network, the potential change rate of the node in the current sampling period is calculated, and a sudden change detection of the potential difference is performed. The node change rate values ​​whose change rate exceeds the set surge threshold are screened to obtain sudden change potential rate data; S112: Based on the sudden potential change rate data and the ground potential change rate sequence data of the corresponding node in the subsequent two cycles, determine whether the potential has recovered to the steady-state range. If the judgment criteria for recovery to the steady state are met, classify it as a transient disturbance node, record the timestamp corresponding to the moment when the sudden change occurred, and obtain a transient disturbance time label set. S113: According to the transient disturbance time label set, the unique node identification information corresponding to each node at the mutation moment is extracted, and the disturbance response intensity value of the node is calculated by combining the node identification and the time label. The time label and node identification of the node whose disturbance response intensity value is greater than the disturbance response threshold are combined to generate a transient response node set.

4. The electromagnetic radiation suppression method for a flywheel energy storage system in a data center according to claim 1, characterized in that: The specific steps for obtaining the synchronous response node topology are: S211: Obtain flywheel speed time series data and load power time series data within the flywheel energy storage system operation cycle, calculate the speed increment change value and incremental change rate between adjacent sampling points, identify speed mutation events and transient power mutation events, record the corresponding times, and generate a dual-event time label set; S212: Based on the dual-event time tag set and in combination with the corresponding time tags of each node in the transient response node set, determine the time difference between each event and the node timestamp, set a time synchronization threshold, retain node records whose absolute value of the time difference is less than or equal to the time synchronization threshold, and obtain a synchronization matching node index set; S213: According to the synchronous matching node index set, the connection relationships between all marked nodes are identified and numbered one by one, an adjacency matrix is ​​constructed and whether there is a direct connection relationship between the nodes is marked, and a synchronous response node topology is established.

5. The electromagnetic radiation suppression method for a flywheel energy storage system in a data center according to claim 1, characterized in that: The specific steps for obtaining the electromagnetic radiation path map are: S311: Based on the physical connection relationship of each node in the synchronous response node topology, extract the grounding module identification information, shield connection point number, and rack grounding point number corresponding to each node connection segment, archive and summarize the grounding parameters corresponding to the nodes in the path segment item by item, and establish a node grounding feature index set; S312: Based on the node grounding feature index set, perform structural identification and trigger direction sorting on the node sequence in each path, determine the electromagnetic coupling characteristics within the path, mark the shielding integrity of the path and record the structural information, and generate a coupling path feature set based on the node sequence; S313: Based on the coupling path feature set, the nodes in each path are classified into node types, a path propagation sequence identifier is established, the trigger start order and propagation chain direction of each node are identified, the number of nodes in each path and the proportion of each type of node are counted, and the difference in path structure is represented by different color lines to generate an electromagnetic radiation path map.

6. The electromagnetic radiation suppression method for a flywheel energy storage system in a data center according to claim 1, characterized in that: The specific steps for obtaining the magnetic shielding optimization configuration table are: S411: reading the magnetic shielding component number information marked on each path in the electromagnetic radiation path map, collecting magnetic flux density data, determining whether a saturation trend occurs, screening all magnetic shielding components that meet the saturation trend determination and recording their numbers, and obtaining a magnetic saturation component number set; S412: Based on the magnetic saturation component number set, the current excitation state of each component is collected. For each magnetic shielding component, the excitation waveform is identified as being in a rising edge time window. The current magnetic flux density, loop magnetic resistance, magnetic path length, number of coil turns, and driving frequency are extracted. The magnetic circuit impact value of the magnetic shielding component is calculated. The magnetic shielding components having a magnetic circuit impact value greater than a reference value are identified. A target list of magnetic permeability structures requiring reconstruction is established, and a magnetic circuit adjustment structure set is established. S413: Based on the magnetic circuit adjustment structure set, locate the corresponding magnetic permeance path for each magnetic shielding component to be adjusted in the structure diagram, judge the adjustment direction based on the magnetic circuit length and magnetic permeability change trend, record the component number, adjustment direction, target magnetic permeability value, and excitation frequency, and construct a magnetic shielding optimization configuration table.

7. The electromagnetic radiation suppression method for a flywheel energy storage system in a data center according to claim 1, characterized in that: The method further comprises: S5: Counting the rate of change of the magnetic flux density of each component in the magnetic shielding optimization configuration table, monitoring the change amplitude within the moving average window, and triggering a magnetic circuit reset instruction if the continuous detection cycle shows a convergence trend, adjusting the magnetic shielding component to the initial configuration state, and generating an electromagnetic compatibility status report; The electromagnetic compatibility status report includes the magnetic flux density change trend, shielding configuration stability index, and reset adjustment trigger record.

8. The electromagnetic radiation suppression method for a flywheel energy storage system in a data center according to claim 7, characterized in that: The specific steps for obtaining the electromagnetic compatibility status report are: S511: Calculate the magnetic flux density change rate between adjacent sampling periods based on all magnetic shield component numbers recorded in the magnetic shield optimization configuration table, calculate the absolute value of the continuous change rate of each magnetic shield component, and determine whether all of them are less than a preset change amplitude threshold. If so, the magnetic shield of the corresponding component is marked as tending to a stable state, and a magnetic flux convergence trend determination result is obtained. S512: Based on the magnetic shielding components identified in the magnetic flux convergence trend determination result, a sliding average window is set and a sequence of magnetic flux density change values ​​is extracted within each window. The consistency and convergence characteristics of the change amplitude are evaluated. If the change trend is stable and the difference is within the range, the corresponding magnetic shielding component is determined to meet the magnetic circuit recovery condition, and is added to the reset execution candidate set according to the number to obtain a magnetic circuit reset target set. S513: According to the magnetic circuit reset target set, compare the current magnetic permeability configuration of each magnetic shielding component with the initial factory configuration parameters, and perform deviation evaluation on the current parameters and initial values ​​of each magnetic permeability path. If the deviation range is within the specified threshold, the magnetic circuit reset action can be triggered to reset the magnetic permeability path configuration to the initial state. The status and path information of all restored components are summarized in the order of component numbers to generate an electromagnetic compatibility status report.

Images