Surge voltage suppression system and control method thereof

Abstract

The invention relates to the technical field of power electronics, and discloses a surge voltage suppression system and a control method thereof. The system comprises a surge feature extraction unit, a suppression map generation unit, a strategy deduction and check unit and an instruction synthesis and distribution unit which are connected in sequence. According to the method, surge feature descriptors are extracted and fused with a historical case library and a real-time power grid topology, and a suppression atlas containing multi-path parameter combination is dynamically generated; and performing time sequence deduction and interaction influence calculation on each path in the map, iteratively correcting parameters to output a checked strategy sequence, and finally converting the strategy sequence into an instruction to control the action of the physical suppression equipment. According to the method, the whole process of the suppression strategy from historical experience learning, dynamic generation to global collaborative rehearsal optimization is realized, and the adaptability, the collaboration and the reliability of responding to surges in a complex power grid environment are improved.

Description

Technical Field

[0001] This invention relates to the field of power electronics technology, specifically to a surge voltage suppression system and its control method. Background Technology

[0002] In power systems, surge voltages are common transient disturbances that threaten equipment safety and grid stability. Existing suppression technologies mainly rely on physical suppression devices installed at critical nodes. These devices are typically triggered based on preset local voltage or current thresholds, and their suppression strategies are relatively fixed, depending on the device's own response characteristics. When the grid encounters complex or novel surges, fixed thresholds and a single local response mode are insufficient to achieve optimal global suppression, potentially leading to protection blind spots or overshoot.

[0003] Existing technical solutions typically lack systematic learning and utilization of historical suppression experience, and also lack the ability to globally extrapolate and verify multi-device collaborative strategies before implementation. The strategy library lags behind changes in grid topology and surge characteristics, resulting in insufficient adaptability. Furthermore, when multiple suppression devices respond to the same surge event, their timing and energy sharing may lack coordination, leading to strategy conflicts that weaken suppression effects and even trigger secondary oscillations. Therefore, how to dynamically generate suppression strategies that match the current grid state based on real-time surge characteristics, and how to anticipate and eliminate collaborative conflicts between multiple devices before strategy execution, have become key issues in improving the intelligence and reliability of suppression systems. Summary of the Invention

[0004] The purpose of this invention is to provide a surge voltage suppression system and its control method to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides a surge voltage suppression system, the system comprising: The surge feature extraction unit samples the original voltage waveform of the power grid node, separates the steady-state component and transient disturbance component from the sampled data, identifies the amplitude, steepness and duration of the transient disturbance component, purifies the feature signal with the help of preset noise filtering methods, and forms a feature descriptor of the surge event based on the purified feature signal. The suppression map generation unit receives the feature descriptor of the surge event, performs multi-dimensional matching between the feature descriptor and surge records in the historical case library, filters out similar historical records based on the matching degree, integrates the suppression parameters of the similar historical records with the current power grid topology, and generates a surge suppression dynamic map containing multiple suppression paths and their parameter combinations through a map construction model. The strategy deduction and verification unit analyzes each suppression path in the surge suppression dynamic map, performs time-series deduction of the expected suppression action for each path, calculates the interactive impact of different suppression actions between power grid nodes, marks potential risk conflict points and performance weaknesses based on the deduction results, iteratively corrects the parameters of the suppression path according to the marking results, and outputs the verified suppression strategy sequence. The instruction synthesis and distribution unit converts the verified suppression strategy sequence into an operation instruction set that can be executed by multiple physical suppression devices, binds an execution timestamp and device identifier to each operation instruction, and distributes the operation instruction set to the corresponding physical suppression device through a real-time communication link.

[0006] Preferably, the surge event feature descriptor includes disturbance amplitude level, disturbance rise edge index, disturbance energy integral value, and disturbance event time label; the surge suppression dynamic map includes a topology graph with nodes and suppression devices as vertices, suppression path edges connecting vertices, and a set of device action timing and intensity parameters attached to each edge; the verified suppression strategy sequence includes a set of suppression action steps ordered by time, device identifier and action parameters corresponding to each step, and associated constraints between steps; and the operation instruction set includes device address code, action type code, action intensity value, and precise trigger time.

[0007] Preferably, the specific working steps of the surge feature extraction unit include: The voltage waveforms of the power grid nodes are captured within a set sampling period to obtain the original sampling sequence; An adaptive baseline calibration algorithm is applied to the original sampling sequence to extract the power frequency steady-state voltage waveform, thus obtaining the residual signal sequence. Multi-scale wavelet transform is performed on the residual signal sequence to identify the time-frequency region where the coefficients after the transform exceed the threshold, and to locate the start and end times of the surge disturbance. Extract the peak value of the disturbance waveform, the average slope from the start time to the peak time, and the integral area of ​​the disturbance waveform on the time axis from the located time-frequency region. The extracted peak value, mean slope, and integral area are compared with the pre-stored background noise statistical features to filter out spurious disturbance events caused by measurement noise, and the event information confirmed as a real surge is encapsulated as the feature descriptor of the surge event.

[0008] Preferably, the specific steps for generating a surge suppression dynamic spectrum in the suppression spectrum generation unit include: Receive the feature descriptor of the surge event, and calculate the comprehensive similarity between the feature descriptor and each record in the historical case database in three dimensions: amplitude, waveform, and occurrence location; Select several historical records whose overall similarity exceeds a preset threshold as a set of similar historical records; Read the combination of suppression devices, action sequence and parameter settings that were successfully applied from the set of similar historical records, as the basic suppression scheme; The model is constructed by inputting the real-time topology of the current power grid, the availability status of each suppression device, and the node impedance parameters into the graph. In the surge suppression model, a basic suppression scheme is used as a seed, and adaptive expansion and variation are performed based on the current power grid parameters to simulate and generate multiple electrical paths from the disturbance occurrence node to each available suppression device. Preliminary action timing and intensity parameters are configured for each path, and finally the surge suppression dynamic spectrum is formed.

[0009] Preferably, the specific steps for time-series deduction and iterative correction in the strategy deduction and verification unit include: Select a suppression path from the surge suppression dynamic map, and unfold the actions of all suppression devices on the suppression path according to the timestamp to form an initial action sequence chain; A simplified electromagnetic transient simulation model including key nodes and lines of the power grid is established, and each action in the initial action time sequence chain is used as an excitation input to the simulation model. Run the simulation model to calculate and record the voltage over-limit, equipment overload or resonance phenomena that may occur in other non-target nodes in the power grid under the action of the initial action sequence chain, and mark the location and time of the anomaly as the risk conflict point; Simultaneously assess the degree of attenuation of the target surge by the suppression path. If the residual voltage after attenuation exceeds the safety limit, mark the location where the residual voltage exceeds the limit as a weak point in performance. Based on information on all risk conflict points and performance weaknesses, the timing or intensity parameters of the action of relevant equipment in the initial action sequence chain are adjusted in reverse to generate a corrected action sequence chain. The corrected action timing chain is input into the simulation model again for verification. This process is repeated until no new risk conflict points are generated and the attenuation requirements are met. At this point, the corrected and verified action timing chain is saved as a qualified suppression strategy. Traverse all suppression paths in the surge suppression dynamic graph, and for each path, perform a deduction and correction process from selecting a suppression path from the surge suppression dynamic graph to saving the corrected and verified action sequence chain as a qualified suppression strategy.

[0010] Preferably, the specific working steps of the instruction synthesis and distribution unit include: Read the first suppression strategy in the verified suppression strategy sequence, and translate each suppression action step in the suppression strategy into a standard instruction frame containing device address code, function code, parameter register and write value according to the communication protocol specification of the device manufacturer. Each standard instruction frame is assigned a precise absolute timestamp, which is calculated by adding the relative delay time specified in the suppression strategy to the system base time; Standard instruction frames with absolute timestamps are classified and aggregated according to device address codes to form instruction data packets sent to each physical suppression device. Through a high-speed parallel communication interface, under the coordination of a unified synchronous clock signal, each instruction data packet is distributed to the buffer area of ​​the corresponding physical suppression device; The physical suppression device performs the corresponding action when the local clock reaches the time indicated by the absolute timestamp in the instruction.

[0011] Preferably, the system further includes a disturbance backtracking analysis unit, comprising: After each surge suppression action is completed, collect voltage waveform data of all network nodes and action response logs of all physical suppression devices; The actual recorded waveform data is compared with the simulated waveform generated during the strategy deduction stage to identify the deviation between the suppression effect and the expectation. The causes of the deviation are analyzed, and the analysis results, along with the feature descriptor of this surge event, the final suppression strategy, and the actual effect evaluation, are stored as a new case record in the historical case library for use in optimizing subsequent matching and graph generation processes.

[0012] Preferably, the system further includes a resource efficiency assessment unit, comprising: Periodically calculate the cumulative number of actions, total energy absorption, and response delay time of each physical suppression device; Calculate the current health score and remaining service life prediction for each physical suppression device based on the statistical results; The current health score and the predicted remaining service life are fed back to the suppression map generation unit, serving as an important basis for screening available suppression devices and allocating suppression intensity when constructing a surge suppression dynamic map.

[0013] Preferably, the system further includes a system synchronization and timekeeping unit, comprising: It receives high-precision satellite timing signals to provide a unified microsecond-level precision clock source for all units in the system that require a time reference. The communication link delay between the instruction synthesis and distribution unit and each physical suppression device is monitored, and the communication link delay is dynamically compensated to the timestamp of the distributed instruction.

[0014] Preferably, the method includes: Step 1: Sample the original voltage waveform of the power grid node, separate the steady-state component and transient disturbance component from the sampled data, identify the amplitude, steepness and duration of the transient disturbance component, purify the characteristic signal with the help of preset noise filtering methods, and form a characteristic descriptor of the surge event based on the purified characteristic signal. Step 2: Receive the feature descriptor of the surge event, perform multi-dimensional matching between the feature descriptor and surge records in the historical case library, filter out similar historical records based on the matching degree, integrate the suppression parameters of the similar historical records with the current power grid topology, and generate a surge suppression dynamic map containing multiple suppression paths and their parameter combinations through a graph construction model. Step 3: Analyze each suppression path in the surge suppression dynamic map, perform time-series simulation of the expected suppression action for each path, calculate the interactive impact of different suppression actions between power grid nodes, mark potential risk conflict points and performance weaknesses based on the simulation results, iteratively correct the parameters of the suppression path according to the marking results, and output the verified suppression strategy sequence. Step 4: The verified suppression strategy sequence is converted into an operation instruction set that can be executed by multiple physical suppression devices. Each operation instruction is bound with an execution timestamp and a device identifier. The operation instruction set is then distributed to the corresponding physical suppression device through a real-time communication link.

[0015] Compared with the prior art, the beneficial effects of the present invention are: By performing multi-dimensional matching between real-time extracted surge feature descriptors and a historical case library, and integrating this with the current power grid topology to generate a dynamic surge suppression map, the generation of suppression strategies no longer relies on a fixed rule base. This method can autonomously learn from past successful or failed cases, providing historically validated parameter combinations suitable for the current network connectivity for current surge events. The dynamic map presents multiple feasible suppression paths and their parameters, avoiding the limitations of a single strategy, enhancing the system's adaptability to complex and variable surge conditions and the flexibility of strategy selection, and improving the targeting and reliability of suppression measures.

[0016] By independently performing time-series simulations of each suppression path in the dynamic graph and calculating the interactive effects of different suppression actions between grid nodes, the global effect of the strategy can be simulated before the actual actions are executed. This process can accurately locate potential conflicts in the timing of different equipment action commands, energy dissipation paths, and weak links in the network structure regarding suppression effectiveness. Annotations based on the simulation results drive iterative corrections to the suppression parameters, ensuring that the final output strategy sequence preemptively avoids collaborative risks and optimizes action timing and parameter coordination. This achieves a shift from "trigger-action" to "prediction-verification-optimization-execution," improving the overall efficiency and safety margin of collaborative suppression in complex grid environments. Attached Figure Description

[0017] Figure 1 This is a schematic diagram illustrating the working principle of the surge voltage suppression system described in this invention. Figure 2 A flowchart illustrating the operation of the surge feature extraction unit; Figure 3 A flowchart illustrating the operation of the instruction synthesis and distribution unit; Figure 4 A pie chart showing the percentage of causes of surge suppression deviation; Figure 5 A bar chart comparing the communication link delay of physically suppressed devices. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] Please see Figure 1This invention provides a surge voltage suppression system, comprising: a surge feature extraction unit that samples the original voltage waveform of a power grid node, separates steady-state components and transient disturbance components from the sampled data, identifies the amplitude, steepness, and duration of the transient disturbance components, purifies the feature signal using preset noise filtering methods, and forms a feature descriptor for the surge event based on the purified feature signal; a suppression map generation unit receives the feature descriptor of the surge event, performs multi-dimensional matching with surge records in a historical case library, filters similar historical records based on the matching degree, and integrates the suppression parameters of the similar historical records with the current power grid topology to generate a surge suppression dynamic map containing multiple suppression paths and their parameter combinations through a map construction model; and a strategy deduction and verification unit analyzes each suppression path in the surge suppression dynamic map, performs time-series deduction of the expected suppression action for each path, calculates the interaction effects of different suppression actions between power grid nodes, marks potential risk conflict points and effectiveness weaknesses based on the deduction results, iteratively corrects the parameters of the suppression path based on the marking results, and outputs a verified suppression strategy sequence. The instruction synthesis and distribution unit transforms the verified suppression strategy sequence into an operation instruction set that can be executed by multiple physical suppression devices, binds an execution timestamp and device identifier to each operation instruction, and distributes the operation instruction set to the corresponding physical suppression device through a real-time communication link.

[0020] In one embodiment of the present invention, the surge event descriptor includes the disturbance amplitude level, the disturbance rise edge exponent, the disturbance energy integral value, and the time label of the disturbance event. The surge suppression dynamic graph includes a topology graph with nodes and suppression devices as vertices, suppression path edges connecting the vertices, and a set of device action timing and intensity parameters attached to each edge. The verified suppression strategy sequence includes a set of suppression action steps ordered chronologically, the device identifier and action parameters corresponding to each step, and the associated constraints between steps. The operation instruction set includes the device address code, action type code, action intensity value, and precise trigger time.

[0021] In practical implementation, the surge event descriptor consists of multiple quantized fields used to accurately characterize the core electrical features of a surge disturbance. The disturbance amplitude level is a discrete classification identifier, categorized based on the ratio of the surge voltage peak to the system nominal voltage; for example, a ratio between 1.5 and 2.0 is classified as level one. The disturbance rise edge exponent is a dimensionless value used to characterize the intensity of the voltage rise from the disturbance's inception to its peak; its calculation relies on statistical processing of the waveform slope during the rise phase. The disturbance energy integral value is a scalar obtained by integrating the purified disturbance voltage waveform over time, reflecting the total energy carried by the surge event. The time stamp records the absolute moment the surge event was accurately identified, with microsecond-level precision. The surge suppression dynamic map is a structured data model containing topological and action information. Its vertex set consists of monitoring nodes in the power grid and various physical suppression devices, while the edge set represents the possible electrical connections and signal transmission paths from the disturbance source node to each suppression device. Each edge is accompanied by a detailed set of device action timing and intensity parameters. The action timing specifies the order and interval of activation of multiple suppression devices on the same path, while the intensity parameter set includes specific values ​​such as the voltage clamping value or current absorption level that each device needs to apply.

[0022] In some embodiments, the verified suppression strategy sequence is represented as a list of instructions arranged in strict chronological order. Each entry in the list is a suppression action step, explicitly indicating the unique identifier of the physical suppression device executing the step, as well as the type of action the device needs to perform and the corresponding parameter value, such as setting the compensation voltage of a series voltage compensator to -230 volts. Inter-step constraints define the logical or temporal dependencies between different steps; for example, step two must be initiated within five milliseconds after step one is completed. The operation instruction set is the final hardware-level mapping of the suppression strategy sequence, and each operation instruction contains four fixed fields: the device address code corresponds to the unique physical address of the physical suppression device in the network; the action type code indicates the specific operation the device needs to perform; the action intensity value is a specific parameter value matching the action type code; and the precise trigger time is an absolute point in time based on the system's unified clock, at which the instruction will be executed by the device.

[0023] It is understandable that the criteria for classifying disturbance amplitude levels can be pre-configured in the system's parameter database, and the threshold values ​​can be adjusted according to different voltage levels of the power grid. The calculation of the disturbance rise edge exponent can be achieved by performing piecewise linear fitting on the rise phase waveform and then obtaining the average slope. The calculation of the disturbance energy integral value involves processing discrete sampled data, and its mathematical expression is as follows:

[0024] Where: characters Represents the integral value of the perturbation energy, character For sampling point sequence number, character With characters These correspond to the sampling point numbers at the start and end times of the surge disturbance, respectively. Representing the The purified instantaneous value of the perturbation voltage at each sampling point, character This represents the fixed sampling time interval of the system. The generation of time stamps relies on a high-precision clock source provided by the system synchronization and timekeeping unit.

[0025] Optionally, the construction of the surge suppression dynamic graph can be based on graph theory algorithms, abstracting the power grid topology and suppression devices as vertices of the graph, and electrical connections and signal paths as edges. The device action timing and intensity parameter sets are attached to the corresponding edges in key-value pairs, allowing the strategy derivation unit to directly read and modify them. The associated constraints in the suppression strategy sequence can be internally represented using a directed acyclic graph to formally describe the sequential relationship between steps, ensuring the logical correctness of action execution. During synthesis, the precise trigger times in the operation instruction set are superimposed with communication link delay compensation values, ensuring that the action is triggered precisely when the device's local clock reaches the specified moment after the instruction arrives in the device buffer, achieving synchronous coordination among multiple devices.

[0026] In some embodiments, the disturbance rising edge exponent in the feature descriptor can be extracted from waveform data using a function model that takes the voltage sequence of the rising phase as input and outputs a scalar value reflecting the average rate of change. The device action timing sequence in the surge suppression dynamic spectrum can be further subdivided into multiple phases, each corresponding to a specific angular interval of the power grid frequency cycle to achieve synchronization with the fundamental voltage of the power grid. The storage format of the suppression strategy sequence uses an extensible markup language to facilitate parsing and transmission between different software modules. The distribution of the operation instruction set follows strict industrial communication protocols, such as a high-speed Ethernet-based precise clock synchronization protocol, to ensure that instructions are delivered on time.

[0027] In one embodiment of the present invention, see [reference] Figure 2The surge feature extraction unit operates by capturing voltage waveforms of power grid nodes within a set sampling period to obtain the original sampling sequence. An adaptive baseline calibration algorithm is applied to the original sampling sequence to extract the power frequency steady-state voltage waveform, resulting in a residual signal sequence. A multi-scale wavelet transform is performed on the residual signal sequence to identify time-frequency regions where the transformed coefficients exceed a threshold, thus locating the start and end times of the surge disturbance. From the located time-frequency regions, the peak value of the disturbance waveform, the mean slope from the start time to the peak time, and the integral area of ​​the disturbance waveform on the time axis are extracted. The extracted peak value, mean slope, and integral area are compared with pre-stored background noise statistical features to filter out spurious disturbance events caused by measurement noise. Events confirmed as genuine surges are encapsulated as feature descriptors for the surge events.

[0028] The process of generating a surge suppression dynamic spectrum in the suppression spectrum generation unit includes receiving the feature descriptor of the surge event and calculating the comprehensive similarity between the feature descriptor and each record in the historical case library in three dimensions: amplitude, waveform, and occurrence location. Several historical records with a comprehensive similarity exceeding a preset threshold are selected as a set of similar historical records. The successfully applied suppression device combinations, action sequences, and parameter settings are read from the set of similar historical records as the basic suppression scheme. The real-time topology of the current power grid, the availability status of each suppression device, and the node impedance parameters are input into the spectrum construction model. In the spectrum construction model, using the basic suppression scheme as a seed, adaptive expansion and variation are performed based on the current power grid parameters to simulate and generate multiple electrical paths from the disturbance occurrence node to each available suppression device. Preliminary action timing and intensity parameters are configured for each path, ultimately forming the surge suppression dynamic spectrum.

[0029] In practical implementation, the surge feature extraction unit's workflow begins with continuous monitoring of the grid node voltage. Within a set sampling period, the unit's high-speed analog-to-digital converter captures the real-time voltage waveform of the grid node, obtaining an original sampling sequence composed of instantaneous voltage values ​​at equal time intervals. An adaptive baseline calibration algorithm is applied to the original sampling sequence. This algorithm uses the power frequency period as a reference, dynamically fitting and extracting the power frequency steady-state voltage waveform that changes slowly with the load. Subtracting this steady-state waveform from the original sampling sequence yields the residual signal sequence. A multi-scale wavelet transform is performed on the residual signal sequence, calculating the detail coefficients of the residual signal at different decomposition scales. By identifying time-frequency regions where the magnitude of the detail coefficients exceeds a preset threshold, the precise start and end times of the surge disturbance can be located. From the original residual signal segment corresponding to the located time-frequency region, the absolute peak value of the disturbance waveform is extracted, the arithmetic mean of the waveform slope from the start time to the peak time is calculated as the slope mean, and the absolute value of this waveform segment is numerically integrated on the time axis to obtain the integration area. The extracted peak value, mean slope, and integral area are compared with the background noise statistical features pre-stored in the unit memory. The background noise statistical features include the root mean square value of the noise voltage and the peak distribution range. False disturbance events caused by measurement noise are filtered out by threshold comparison. Event information that meets the surge discrimination conditions is encapsulated into the feature descriptor of the surge event.

[0030] In some embodiments, the adaptive baseline calibration algorithm is implemented through a sliding time window with a width of two power frequency cycles. Within each window, the algorithm uses the least squares method to fit a sinusoidal fundamental wave, thereby updating the steady-state waveform reference in real time. The number of decomposition levels for the multi-scale wavelet transform is pre-set based on the system's highest sampling frequency and the surge frequency band of interest, for example, a five-level decomposition. The threshold for detail coefficients is related to the decomposition scale; the lower the scale, the higher the corresponding threshold, to distinguish high-frequency noise from real fast surges. Background noise statistical characteristics are obtained through long-term monitoring and learning during undisturbed periods and are updated periodically.

[0031] The process of generating a surge suppression dynamic spectrum in the suppression spectrum generation unit begins with receiving feature descriptors from the surge feature extraction unit. The processor within the unit calculates the comprehensive similarity between the feature descriptors and each record in the historical case library across three dimensions: amplitude, waveform, and location. The comprehensive similarity is calculated using a weighted Euclidean distance model. Several historical records with a comprehensive similarity exceeding a preset threshold are selected to form a set of similar historical records. From this set, the combination of suppression devices, the order of actions, and parameter settings successfully applied to each historical record are read one by one. This information is then fused and normalized to form a set of basic suppression schemes. The real-time topology of the current power grid, the online availability status of each physical suppression device, and the node impedance parameters obtained from the state estimation system are all input into the spectrum to construct the model. In the surge suppression model, a basic suppression scheme is used as a seed scheme. Based on the real-time parameters of the current power grid, the model is adaptively expanded and mutated. The expansion process includes finding more electrical connection paths from the disturbance occurrence node to each available physical suppression device in the current topology. The mutation process includes randomly perturbing the action timing and intensity parameters in the seed scheme to generate a new scheme. The multiple feasible paths generated by simulation and the configuration parameters together constitute the surge suppression dynamic spectrum.

[0032] It is understandable that the weighted Euclidean distance model that combines similarity can be specifically expressed as:

[0033] Where: characters Represents the calculated overall similarity, characters ,character and characters These are preset weighting coefficients for three dimensions: amplitude, waveform shape, and location of occurrence. Represents the normalized difference between the current surge amplitude and the amplitude of historical cases, character The value representing the difference in waveform shape characteristics between the two characters. The electrical distance represents the location of the occurrence. A preset threshold is a constant between zero and one, used to filter out sufficiently similar historical cases. The fusion of basic suppression schemes can employ a voting mechanism or a weighted average mechanism.

[0034] Optionally, the real-time topology can be derived from the real-time wiring diagram information provided by the power grid energy management system. The availability status of physical suppression devices is obtained through reports from the devices' own status monitoring units, with statuses including ready, busy, faulty, or offline. Node impedance parameters can be the latest results output by the status estimation module. During adaptive expansion and mutation, for each newly generated path, a preliminary set of action timing and intensity parameters is configured. The initial values ​​of the intensity parameters are calculated by querying the device parameter library and combining it with the path's electrical parameters. The final surge suppression dynamic map is stored in memory as a graph data structure, where vertex attributes contain node and device information, and edge attributes contain path impedance, action timing chains, and intensity parameter sets.

[0035] In one embodiment of the present invention, see [reference] Figure 3 The process of timing deduction and iterative correction in the strategy deduction and verification unit includes selecting a suppression path from the surge suppression dynamic spectrum, unfolding the actions of all suppression devices on the suppression path according to timestamps to form an initial action timing chain. A simplified electromagnetic transient simulation model including key nodes and lines of the power grid is established, and each action in the initial action timing chain is used as an excitation input to the simulation model. The simulation model is run to calculate and record the voltage overruns, equipment overloads, or resonance phenomena that may occur at other non-target nodes in the power grid under the action of the initial action timing chain, and the location and time of the anomaly are marked as risk conflict points. At the same time, the attenuation degree of the suppression path on the target surge is evaluated. If the residual voltage after attenuation exceeds the safety limit, the location where the residual voltage exceeds the limit is marked as a performance weakness. Based on the information of all risk conflict points and performance weaknesses, the action time or intensity parameters of relevant devices in the initial action timing chain are adjusted in reverse to generate a corrected action timing chain. The corrected action timing chain is input into the simulation model again for verification. This process is repeated until no new risk conflict points are generated and the attenuation requirements are met. At this point, the corrected and verified action timing chain is saved as a qualified suppression strategy. All suppression paths in the surge suppression dynamic graph are traversed, and for each path, a deduction and correction process is performed from selecting a suppression path from the surge suppression dynamic graph to saving the corrected and verified action timing chain as a qualified suppression strategy.

[0036] The instruction synthesis and distribution unit's operation includes reading the first suppression strategy from the verified suppression strategy sequence, translating each suppression action step in the suppression strategy into a standard instruction frame containing device address code, function code, parameter register, and written value, according to the device manufacturer's communication protocol specifications. A precise absolute timestamp is assigned to each standard instruction frame, calculated by adding the relative delay time specified in the suppression strategy to the system base time. The standard instruction frames with absolute timestamps are then categorized and aggregated according to their device address codes to form instruction data packets sent to each physical suppression device. Through a high-speed parallel communication interface, coordinated by a unified synchronization clock signal, each instruction data packet is distributed to the buffer of the corresponding physical suppression device. The physical suppression device executes the corresponding action when its local clock reaches the time indicated by the absolute timestamp in the instruction.

[0037] In practical implementation, the process of timing deduction and iterative correction by the strategy deduction and verification unit begins with the analysis of the surge suppression dynamic spectrum. The unit selects a suppression path from the surge suppression dynamic spectrum, sorts and expands the preset actions of all physical suppression devices on the suppression path according to their timestamps, forming an initial action timing chain that includes action type, intensity parameters, and execution time. A simplified electromagnetic transient simulation model including key nodes and lines of the power grid is established. This simulation model integrates line parameters, transformer models, and load characteristics. Each action in the initial action timing chain is input into the simulation model as a time-varying excitation source, and the simulation model is run within a set simulation time window. After running the simulation model, the voltage overrun, equipment overload, or resonance phenomena that may occur at other non-target nodes in the power grid, excluding the surge occurrence node, under the action of the initial action timing chain are calculated and recorded. Voltage overrun refers to the node voltage exceeding the upper or lower limit of steady-state operation; equipment overload refers to the current flowing through the equipment exceeding its rated capacity; resonance phenomenon refers to the abnormal amplification of voltage or current amplitude at a specific frequency. The location of the power grid where the above abnormal phenomena occur and the time of occurrence are marked as risk conflict points. Simultaneously, the attenuation degree of the suppression path on the target surge is evaluated, and the residual voltage peak of the critical node after the suppression action is compared with the system safety limit. If the residual voltage exceeds the safety limit, the location where the residual voltage exceeds the limit is marked as a weak point. Based on the information of all identified risk conflict points and weak points, the action time or intensity parameters of the relevant physical suppression devices in the initial action sequence chain are adjusted in reverse to generate a corrected action sequence chain. The corrected action sequence chain is used as the excitation input to the simulation model for verification again. This process of simulation, evaluation, marking, and adjustment is repeated until no new risk conflict points are generated in the new simulation results and the surge attenuation meets the preset safety limit requirements. At this point, the finally corrected and verified action sequence chain is saved as a qualified suppression strategy. All suppression paths in the surge suppression dynamic map are traversed, and the above deduction and correction process from selecting the path to saving the qualified suppression strategy is fully executed for each suppression path to form a verified suppression strategy sequence containing multiple optional strategies.

[0038] In some embodiments, a simplified electromagnetic transient simulation model can be achieved through a combination of a frequency domain solver based on the admittance matrix and time domain interpolation. The marking of risk conflict points records not only their location and time, but also the type and severity level of the anomaly. The process of correcting the initial action timing chain can be based on a rule base; for example, when an overvoltage is detected at a node, the rule base suggests delaying the action time of a downstream physical suppression device or increasing its energy absorption level. The generation of the corrected action timing chain can be formally expressed as an adjustment function of the original parameters; for example, the adjustment formula for the action time is:

[0039] Where: characters Represents the new action moment after the adjustment, character Represents the original action moment before adjustment, character This represents the time shift calculated based on risk conflict information; the time shift can be positive or negative. A maximum number of iterations is set in the iteration process to prevent infinite loops.

[0040] The instruction synthesis and distribution unit's workflow begins after the strategy deduction and verification unit outputs the verified suppression strategy sequence. The unit reads the first suppression strategy from the verified sequence and translates each suppression action step in that strategy into a standard instruction frame containing device address code, function code, parameter register, and written values, according to the communication protocol specifications provided by the physical suppression device manufacturer. Each standard instruction frame is assigned a precise absolute timestamp, calculated by adding the relative delay time specified for that step in the suppression strategy to the system reference time provided by the system synchronization and timekeeping unit. The standard instruction frames with absolute timestamps are then categorized and aggregated according to their device address codes, forming instruction data packets sent to each physical suppression device. All instructions from the same physical suppression device are ordered according to their absolute timestamps. Through a high-speed parallel communication interface, coordinated by a unified synchronization clock signal issued by the system synchronization and timekeeping unit, each instruction data packet is distributed to the buffer of the corresponding physical suppression device. The physical suppression device executes the corresponding action when its local clock reaches the time indicated by the absolute timestamp in the instruction, achieving distributed precise coordination.

[0041] In one embodiment of the present invention, the disturbance backtracking analysis unit's operation includes collecting voltage waveform data from all network nodes and action response logs of all physical suppression devices after each surge suppression action is executed. The actual waveform data is compared with the simulated waveforms generated during the strategy deduction phase to identify the deviation between the suppression effect and the expectation. The causes of the deviation are analyzed, and the analysis results, along with the feature descriptor of this surge event, the final suppression strategy executed, and the actual effect evaluation, are stored as a new case record in the historical case library for optimizing subsequent matching and graph generation processes.

[0042] The resource efficiency assessment unit's workflow includes periodically calculating the cumulative number of actions, total energy absorption, and response delay time of each physical suppression device. Based on the statistical results, it calculates the current health score and predicted remaining service life for each physical suppression device. These scores are then fed back to the suppression map generation unit, serving as a crucial basis for selecting available suppression devices and allocating suppression intensity when constructing a dynamic surge suppression map.

[0043] In practical implementation, the disturbance backtracking analysis unit initiates its workflow after each surge suppression action triggered by the command synthesis and distribution unit is completed. The unit collects voltage waveform data from key nodes across the entire network via the data acquisition network, and simultaneously receives action response logs uploaded from each physical suppression device. These logs include the time the device receives the command, the time of execution, the actual action parameters, and the completion status. The disturbance backtracking analysis unit compares the actual waveform data with the simulation waveforms generated by the strategy deduction and verification unit during the deduction phase. The comparison includes the surge voltage decay curve, the maximum overvoltage value at each node, and the transient oscillation frequency and amplitude caused by the suppression action. Differential calculations are used to identify the deviation between the actual suppression effect and the simulation expectations. The potential causes of the deviation are analyzed. The cause analysis may involve subtle differences between simulation model parameters and reality, drift of physical suppression device response characteristics, or fluctuations in communication delay. The analysis results, together with the feature descriptor of this surge event, the final suppression strategy sequence, and the actual effect evaluation report, are stored as a new case record in the historical case library. This is used to optimize the matching and spectrum generation process of subsequent surge feature extraction units and suppression spectrum generation units.

[0044] In some embodiments, voltage waveform data comes from fault recording devices or synchronous phasor measurement units deployed at various nodes of the power grid, and the data is stored in the standard COMTRADE format. Action response logs are uploaded in the manufacturing message specification format via smart terminals embedded in each physical suppression device. The comparison process employs a specialized waveform comparison algorithm to calculate the root mean square error between the actual waveform and the simulated waveform within a specific time window. Deviation cause analysis can utilize a pre-built knowledge base, which associates different deviation patterns with possible causes. New case records are saved as structured data objects, containing multiple fields such as the original feature descriptor, suppression strategy, actual waveform data segment, deviation analysis report, and effect score.

[0045] The resource efficiency assessment unit operates independently and periodically. By querying the equipment management database and action event logs, the unit calculates the cumulative number of actions, the total energy absorbed, and the response delay from command issuance to actual execution for each physical suppression device since the last assessment cycle. Based on the statistical results, it calculates the current health score and predicted remaining service life for each physical suppression device. The current health score is a comprehensive value reflecting the recent performance status of the device, while the predicted remaining service life is an estimate based on the device's cumulative electrical stress and aging model. The calculated current health score and predicted remaining service life are fed back to the suppression map generation unit in real time. This serves as a crucial basis for the suppression map generation unit to select available suppression devices and allocate suppression intensity when constructing a surge suppression dynamic map. For example, devices with a health score below a threshold will be marked as unavailable, and devices with a shorter predicted remaining service life will have their energy absorption tasks assigned to them in the map reduced accordingly. See Table 1.

[0046] Table 1: Statistics on Resource Efficiency of Physical Suppression Equipment

[0047] It is understandable that the cumulative number of actions is directly accumulated from the device's action counter or event log. The total cumulative energy absorption is calculated by integrating the product of the voltage and current absorbed by the device during each action and then summing the results. The response delay time is the difference between the action execution time recorded in the local log of the physical suppression device and the command issuance time recorded by the instruction synthesis and distribution unit. The current health score can be calculated based on a comprehensive evaluation model, for example:

[0048] Where: characters Represents the current health score, characters Represents the cumulative number of actions after normalization, character Represents the total cumulative energy absorption after normalization, character Represents the normalized average response time, character ,character and characters These are the weighting coefficients for each item, characters. ,character and characters These are scoring functions for frequency, energy, and delay, designed so that the score decreases as losses increase. The remaining service life prediction is based on the equipment's stress-life model, estimated by combining cumulative electrical stress (related to total energy absorption) and years of operation.

[0049] Optionally, the operating cycle of the resource efficiency assessment unit can be set to 24 hours or one week. The equipment management database stores the basic parameters and historical operating records of all physical suppression devices. Data fed back to the suppression map generation unit can be transmitted in real time via shared memory or a message queue. After receiving equipment health and lifespan information, the suppression map generation unit updates these dynamic parameters in its internal equipment status table and calls these parameters for decision-making in subsequent map generation logic. The design of the health scoring function needs to consider differences in equipment type; for example, different weighting coefficients should be used for series-type compensation devices and parallel-type absorption devices. , and .

[0050] In some embodiments, calculating the total cumulative energy absorption requires the physical suppression device to record instantaneous sampled values ​​of voltage and current at each action. Statistics on the average response delay time can exclude extreme timeout data caused by communication anomalies. The current health score output range can be set from 0 to 100, and the score is obtained by querying a preset health score table, which uses the cumulative number of actions, total cumulative energy absorption, and average response delay time as a joint index. The estimation model for the remaining service life prediction value can employ a linear degradation model or a reliability model based on the Weibull distribution. When the suppression map generation unit filters available devices based on device health, if the health of all candidate devices at a critical location is poor, a preventative maintenance alarm for that device may be triggered.

[0051] See Figure 4 This is a pie chart showing the percentage of causes for surge suppression deviations. The chart reflects the core factors contributing to the discrepancy between actual performance and simulation expectations in surge suppression systems. "Differences in simulation model parameters" account for over half of the reasons, indicating that optimizing the match between the model and the actual power grid is crucial for improving suppression accuracy. "Drift in equipment response characteristics" is the second most common cause, suggesting the need to strengthen equipment condition monitoring and calibration. Identifying "differences in simulation model parameters" as the primary source of deviation guides the team to prioritize resource allocation to optimizing the parameter match between the simulation model and the actual power grid, thereby improving the predictive accuracy of suppression strategies. The high percentage of "drift in equipment response characteristics" indicates the need to strengthen condition monitoring throughout the equipment's lifecycle and incorporate equipment drift data into a historical case database to optimize the adaptability of subsequent suppression strategies.

[0052] In one embodiment of the present invention, the core function of the system synchronization and timekeeping unit is to establish and maintain a unified time reference within the entire surge voltage suppression system. The system synchronization and timekeeping unit continuously receives standard time and frequency signals from the Global Positioning System or the BeiDou Navigation Satellite System through an external high-precision satellite timing signal receiving antenna and decoding module. After decoding, filtering, and verifying the integrity of the received original timing signals, the system synchronization and timekeeping unit uses phase-locked loop technology and a disciplined clock algorithm to synchronize the high-stability crystal oscillator clock source inside the unit to the satellite time reference, thereby generating and outputting a unified clock signal with microsecond-level accuracy. This unified clock signal is distributed to the surge feature extraction unit, the suppression map generation unit, the strategy deduction and verification unit, and the instruction synthesis and distribution unit through a dedicated clock distribution network or a precision clock synchronization protocol, providing a consistent time reference for all these units that require precise timestamps. The system synchronization and timing unit continuously monitors the communication link delay between the instruction synthesis and distribution unit and each physical suppression device. The monitoring method includes periodically sending timestamped query messages to each physical suppression device and receiving its timestamped response messages. The one-way communication link delay is estimated by calculating the round-trip time of the message and excluding processing delay. The system synchronization and timing unit provides the dynamically measured communication link delay value to the instruction synthesis and distribution unit in real time.

[0053] In some embodiments, the high-precision satellite timing signal receiving module employs multi-frequency reception technology to mitigate ionospheric delay errors. Phase-locked loop (PLL) technology and a disciplined clock algorithm can smoothly handle clock fluctuations caused by brief loss or interference of satellite signals. The unified clock signal can be output as a pulse-per-second signal combined with serial time messages, or it can be distributed in software via a precision time protocol based on Ethernet. Communication link delay is monitored at fixed time intervals, such as once per second or every five seconds, to accommodate slight variations in network load. When estimating unidirectional communication link delay, it is assumed that the network path is symmetrical, i.e., the uplink and downlink delays are equal, using the formula:

[0054] Where: characters Represents the estimated one-way communication link delay, character The character represents the local time at which the system synchronization and timekeeping unit receives the response message. The local time when the system synchronization and timekeeping unit sends a query message represents the character... This represents the fixed processing delay within the physical suppression device for processing query messages and generating response messages. The fixed processing delay is a constant that is pre-calibrated by the device manufacturer and provided to the system synchronization and timing unit.

[0055] After receiving the real-time communication link delay value provided by the system synchronization and timekeeping unit, the instruction synthesis and distribution unit dynamically compensates the latest communication link delay value into the timestamp of the distributed instruction during the process of allocating an absolute timestamp for each standard instruction frame when synthesizing the operation instruction set. Specifically, the instruction synthesis and distribution unit first calculates the original absolute timestamp at which a certain instruction is scheduled to be executed at the physical suppression device based on the suppression strategy sequence, and then subtracts the estimated current communication link delay value provided by the system synchronization and timekeeping unit, pointing to the target physical suppression device, from this original absolute timestamp. This results in an earlier, corrected timestamp for the actual distribution of the instruction. The instruction synthesis and distribution unit sends the instruction at the local time indicated by the corrected timestamp, ensuring that the instruction receives the instruction even after experiencing communication link delays. Then, it arrives at the physical suppression device at the exact time indicated by the original absolute timestamp and is ready to execute. The physical suppression device triggers the action locally according to the original absolute timestamp carried in the instruction, thereby achieving precise coordination of multiple devices under a unified time base and offsetting the asynchronous error caused by variable communication delay.

[0056] It is understandable that system synchronization and timekeeping units are the root of the entire system's time consistency. Dynamic compensation mechanisms for communication link latency are crucial for ensuring strict synchronization of distributed actions, especially when instructions need to traverse multi-hop network devices or network segments with varying loads, where latency fluctuations may occur. The physical suppression device also needs to maintain a local clock synchronized with the system synchronization and timekeeping units, typically achieved by receiving a synchronization clock signal or periodically performing network time synchronization. The original absolute timestamp carried within the instruction is the sole basis for triggering the physical suppression device's actions.

[0057] Optionally, the system synchronization and timing unit can be equipped with a rubidium atomic clock or a temperature-controlled crystal oscillator as a local clock, capable of maintaining high-precision timekeeping for short periods when satellite signals are temporarily unavailable. The clock distribution network can utilize fiber optic transmission to reduce jitter and improve accuracy. Monitoring messages for communication link delays can employ high-priority time synchronization protocol messages to minimize the impact of queuing delays on measurement results. The command synthesis and distribution unit can perform timestamp compensation calculations in real-time when generating each command frame, or it can batch process all command frames within a single command data packet. With a stable communication network architecture, the system synchronization and timing unit can learn and maintain a delay baseline value for each fixed communication path, dynamically compensating only for fluctuations relative to the baseline.

[0058] In some embodiments, the satellite timing signal is decoded to obtain Coordinated Universal Time (UTC) information, which the system synchronization and timekeeping unit converts into the epoch time format used internally by the system. The microsecond-level precision of the unified clock signal means that the time deviation between units is controlled within ±1 microsecond. Fixed processing delay for communication link delay monitoring. This requires precise data obtained through laboratory calibration or from the manufacturer, and can be reconfigured by the system synchronization and timing unit after a device firmware upgrade. When performing latency compensation, the instruction synthesis and distribution unit will consider the current latency value detected. If the time is abnormally high or low, a smoothing filtering algorithm can be enabled to compensate using the moving average of historical delay data, thus preventing commands from being issued too early or too late due to a single measurement anomaly. Upon receiving a command, the physical suppression device will again compare the original absolute timestamp in the command with its own local clock. If a minor deviation is detected and the device is capable, it can perform a final fine-tuning of the local time within a very small preset range.

[0059] See Figure 5 This is a comparative bar chart showing the impact of physical suppression of communication link latency. The "compensated latency" for all devices is slightly lower than the "original one-way latency," demonstrating the value of dynamic latency compensation. Fixed processing latency is much lower than link latency, resulting in a weaker impact on overall synchronization. This chart visually illustrates the differences in link latency and compensation effects among different devices, prioritizing optimization of communication links for high-latency devices; verifying the effectiveness of the latency compensation function of the system synchronization and timekeeping units; and providing device-level latency data support for the instruction timestamp compensation strategy. The fact that the "compensated latency" for all devices is slightly lower than the "original one-way latency" directly demonstrates that the latency compensation mechanism of the system synchronization and timekeeping units can effectively offset some link fluctuations, providing data support for the time synchronization of multiple device actions.

[0060] It should be noted that, in this document, relational terms such as "first" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0061] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A surge voltage suppression system, characterized by, Comprise: The surge feature extraction unit samples the original voltage waveform of the power grid node, separates the steady-state component and the transient disturbance component from the sampling data, identifies the amplitude, steepness and duration of the transient disturbance component, purifies the feature signal by means of pre-set noise filtering means, and forms the characteristic descriptor of the surge event according to the purified feature signal; The suppression atlas generation unit receives the characteristic descriptor of the surge event, performs multidimensional matching between the characteristic descriptor and the surge records in the historical case library, filters out similar historical records according to the matching degree, fuses the suppression parameters of the similar historical records and the current power grid topology, and generates a surge suppression dynamic atlas containing multiple suppression paths and parameter combinations through an atlas construction model; The strategy deduction and checking unit analyzes each suppression path in the surge suppression dynamic atlas, time-series deduces the expected suppression actions of each path, calculates the interaction influence between different suppression actions in the power grid nodes, labels potential risk conflict points and performance weak points based on the deduction results, iteratively modifies the parameters of the suppression path according to the labeling results, and outputs the checked suppression strategy sequence; The instruction synthesis and distribution unit converts the checked suppression strategy sequence into an operation instruction set that can be executed by multiple physical suppression devices, binds an execution timestamp and a device identifier to each operation instruction, and distributes the operation instruction set to the corresponding physical suppression device through a real-time communication link.

2. A surge voltage suppression system according to claim 1, wherein, The characteristic descriptor of the surge event includes disturbance amplitude level, disturbance rising edge index, disturbance energy integral value, and time label of the disturbance event, the surge suppression dynamic atlas includes a topological graph with nodes and suppression devices as vertices, suppression path edges connecting the vertices, and device action time sequence and intensity parameter set attached to each edge, the checked suppression strategy sequence includes a set of suppression action steps sorted by time, device identifier and action parameters corresponding to each step, and associated constraint conditions between steps, and the operation instruction set includes device address code, action type code, action strength value, and accurate trigger time.

3. A surge voltage suppression system as defined in claim 2, wherein, The specific working steps of the surge feature extraction unit include: Capture the voltage waveform of the power grid node within the set sampling period to obtain the original sampling sequence; Apply an adaptive baseline calibration algorithm to the original sampling sequence to strip out the power frequency steady-state voltage waveform and obtain a residual signal sequence; Perform multi-scale wavelet transform on the residual signal sequence, identify the time-frequency region where the transformed coefficients exceed the threshold, and locate the start time and end time of the surge disturbance; Extract the peak value of the disturbance waveform, the average slope from the start time to the peak time, and the integral area of the disturbance waveform on the time axis from the located time-frequency region; Compare the extracted peak value, average slope and integral area with the pre-stored background noise statistical characteristics, filter out the pseudo-disturbance events caused by measurement noise, and encapsulate the event information confirmed as a real surge as the characteristic descriptor of the surge event.

4. A surge voltage suppression system as claimed in claim 3, wherein, The specific steps of generating the surge suppression dynamic atlas in the suppression atlas generation unit include: receiving a feature descriptor of the surge event, calculating a comprehensive similarity of the feature descriptor with each record in a historical case library in three dimensions of amplitude, waveform, and occurrence location; selecting a number of historical records with a comprehensive similarity exceeding a preset threshold as a similar historical record set; reading a combination of suppression devices, an action sequence, and parameter settings successfully applied from the similar historical record set as a basic suppression scheme; inputting real-time topological structure of a current power grid, available states of each suppression device, and node impedance parameters into a graph construction model; in the graph construction model, expanding and mutating adaptively according to current power grid parameters with the basic suppression scheme as a seed, simulating to generate a plurality of electrical paths from a disturbance occurrence node to each available suppression device, and configuring an initial action timing and intensity parameter for each path, and finally forming the surge suppression dynamic graph.

5. A surge voltage suppression system as claimed in claim 4, wherein, The specific steps of time sequence deduction and iterative correction in the strategy deduction and checking unit include: selecting a suppression path from the surge suppression dynamic graph, expanding all suppression devices on the suppression path according to a time stamp to form an initial action timing chain; establishing a simplified electromagnetic transient simulation model containing key nodes and lines of the power grid, and inputting each action in the initial action timing chain into the simulation model as an excitation; running the simulation model, calculating and recording possible voltage over-limit, device overload or resonance phenomena at other non-target nodes in the power grid under the action of the initial action timing chain, and marking the position and time of the abnormal occurrence as a risk conflict point; simultaneously evaluating the attenuation degree of the suppression path to the target surge, and if the residual voltage after attenuation exceeds the safety limit, marking the position of the residual voltage over-limit as a weak point of efficiency; according to the information of all risk conflict points and weak points of efficiency, reversely adjusting the action time or intensity parameter of the related devices in the initial action timing chain to generate a corrected action timing chain; inputting the corrected action timing chain into the simulation model again for verification, and repeating the process until no new risk conflict point is generated and the attenuation requirement is met, at which time the corrected and verified action timing chain is saved as a qualified suppression strategy; iterating through all suppression paths in the surge suppression dynamic graph, and executing the deduction and correction process from selecting a suppression path from the surge suppression dynamic graph to saving the corrected and verified action timing chain as a qualified suppression strategy for each path.

6. A surge voltage suppression system as defined in claim 5, wherein, The specific working steps of the instruction synthesis and distribution unit include: reading the first suppression strategy in the checked suppression strategy sequence, and translating each suppression action step in the suppression strategy into a standard instruction frame containing a device address code, a function code, a parameter register, and a written value according to the communication protocol specification of the device manufacturer; allocating an accurate absolute time stamp to each standard instruction frame, which is calculated by adding a relative delay time specified in the suppression strategy to a system reference time; classifying and gathering the standard instruction frames with absolute time stamps according to the device address code to form an instruction data packet sent to each physical suppression device; Through a high-speed parallel communication interface, under the coordination of a unified synchronous clock signal, each instruction data packet is distributed to the corresponding buffer area of the physical suppression device; The physical suppression device executes the corresponding action according to the absolute timestamp in the instruction when the local clock reaches the time indicated by the absolute timestamp.

7. The surge voltage suppression system of claim 1, wherein, It also includes a disturbance backtracking analysis unit, comprising: After each surge suppression action is executed, the voltage recording data of all nodes in the network and the action response logs of all physical suppression devices are collected; The actual recording data is compared with the simulation waveform generated in the strategy deduction stage to identify the deviation between the suppression effect and the expectation; The causes of the deviation are analyzed, and the analysis results, together with the characteristic descriptors of the current surge event, the final executed suppression strategy and the actual effect evaluation, are stored as a new case record in the historical case library for optimizing the subsequent matching and atlas generation process.

8. A surge voltage suppression system as claimed in claim 7, wherein, It also includes a resource performance evaluation unit, comprising: Periodically, the cumulative action times, the total energy absorption and the response delay time of each physical suppression device are counted; According to the statistical results, the current health score and the remaining service life prediction value of each physical suppression device are calculated; The current health score and the remaining service life prediction value are fed back to the suppression atlas generation unit as an important basis for screening available suppression devices and allocating suppression intensity when constructing the surge suppression dynamic atlas.

9. The surge voltage suppression system of claim 1, wherein, It also includes a system synchronization and timekeeping unit, comprising: Receiving high-precision satellite time signal, providing a unified microsecond-level precision clock source for all units in the system that need time reference; Monitoring the communication link delay between the instruction synthesis and distribution unit and each physical suppression device, and dynamically compensating the communication link delay to the timestamp of the distributed instruction.

10. A method for surge voltage suppression control, applied to a surge voltage suppression system according to any one of claims 1 to 9, characterized in that, The method comprises: Step one: sampling the original voltage waveform of the power grid node, separating the steady-state component and the transient disturbance component from the sampling data, identifying the amplitude, steepness and duration of the transient disturbance component, purifying the characteristic signal by means of pre-set noise removal means, and forming the characteristic descriptor of the surge event according to the purified characteristic signal; Step two: receiving the characteristic descriptor of the surge event, performing multi-dimensional matching of the characteristic descriptor with the surge records in the historical case library, screening similar historical records according to the matching degree, fusing the suppression parameters of the similar historical records and the current power grid topology, and generating a surge suppression dynamic atlas containing multiple suppression paths and their parameter combinations through a graph construction model; Step three: analyzing each suppression path in the surge suppression dynamic atlas, timing the expected suppression action of each path, calculating the interaction between different suppression actions among the nodes in the power grid, labeling potential risk conflict points and performance weak points based on the analysis results, iteratively correcting the parameters of the suppression path according to the labeling results, and outputting the checked suppression strategy sequence; Step four: converting the checked suppression strategy sequence into an operation instruction set that can be executed by multiple physical suppression devices, binding an execution timestamp and a device identifier to each operation instruction, and distributing the operation instruction set to the corresponding physical suppression devices through a real-time communication link.

Images