Source load impact and system bearing life evaluation method, system, equipment and medium

Abstract

The invention discloses a source load impact and system bearing life evaluation method and system, equipment and a medium, and relates to the technical field of traction power supply, and the method comprises the steps: collecting data to recognize impact, building a source load model, analyzing an impact diffusion path, evaluating system bearing, fusing multi-dimensional damage, calculating the residual life of the equipment, and automatically updating parameters. And generating a closed-loop regulation and control strategy. The system comprises a data analysis and dynamic modeling module, a bearing evaluation and conduction analysis module, a damage coupling and life prediction module and a parameter correction and strategy generation module. The traction impact time domain-frequency domain-topology integrated quantization is realized, and the problem of impact time-space diffusion assessment is solved. A life model is constructed, and multi-dimensional stress damage is accurately mapped. The micro-simulation and on-line filtering technology is utilized to ensure that model parameters are adaptively updated along with working conditions, and the problem of disjunction between evaluation and actual conditions is solved. The method supports side cloud cooperation and closed-loop pre-control, effectively improves the safety margin of the power grid, and delays the aging of equipment.

Description

Technical Field

[0001] This invention relates to the field of traction power supply technology, and in particular to a method, system, device, and medium for assessing source load impact and system load life. Background Technology

[0002] In recent years, the scale and capacity of railway electrification networks have continued to increase. The traction and regenerative braking processes of electric locomotives exhibit strong power pulsations and phase switching characteristics on a millisecond to minute scale. This, coupled with the single-phase to three-phase coupling between the traction network and the public power grid, causes the "source-grid-load" system to operate under wider frequency domains, faster dynamics, and stronger coupling. Rapid power surges / reversals, phase segmentation, and grid connection / disconnection switching on the traction side can easily trigger voltage surges, frequency and phasor disturbances, increased imbalance, harmonic / interharmonic energy injection, and voltage flicker. These effects, propagated through equivalent source impedance and network topology, alter the power flow distribution and reactive voltage coupling of the public power grid, eroding the stability and load margins of critical sections and causing additional stress on upstream power sources and substation equipment. At the equipment level, wide-spectrum harmonics and high dv / dt and di / dt lead to increased additional losses and hot spot temperature rise, increasing the risk of partial discharge and transient overvoltage. The electromagnetic force pulsation and torsional vibration caused by regenerative power reversal and unbalanced current will generate considerable stress spectrum on winding clamps, conductor supports, core fasteners, circuit breaker mechanisms, etc., forming low-cycle / high-cycle composite fatigue, which, together with thermal stress, accelerates insulation and structural degradation.

[0003] Existing engineering methods often rely on fundamental or quasi-steady-state assumptions, employing static ZIP loads, simplified induction motors, or aggregated load curves for assessment. These methods struggle to characterize the steepness (dv / dt, di / dt), broadband energy distribution, and their coupling with voltage stability margins of traction shocks. At the system level, traditional time-averaged assessment frameworks fail to quantify the spatial diffusion and temporal superposition effects of shocks on the grid structure, unable to answer the questions of "where, when, and by what magnitude" the shocks propagate to downstream critical sections. At the asset level, current life assessment methods often treat thermal and mechanical stresses separately, estimating them based on Arrhenius or S-N curves, lacking a model system that unifies, accumulates, and updates the four stress channels (electrical, thermal, discharge, and mechanical) within the same dimension. Furthermore, while high-frequency recording and edge sampling capabilities are widely available on-site, the data is mostly limited to over-limit alarms and post-event reviews, failing to serve as "first evidence" for constraint mechanism models and adaptive drive parameters, leading to a disconnect between assessment and control and actual operating conditions.

[0004] With the further penetration of power electronic equipment into the power source, grid, and load sides, the coupling between traction loads and new energy fluctuations and flexible interconnection devices is becoming increasingly complex. Single-point, single-frequency, and single-channel indicators are insufficient to support cross-station and cross-regional operation decisions and life-based maintenance. There is an urgent need for a hybrid modeling framework with measured data as the core constraint, frequency response impedance and graph topology diffusion as the framework, and online parameter identification as the link. On the one hand, it optimizes the impact load stability influence model and the whole system load assessment model, simultaneously measuring the erosion of stability and load capacity by steepness energy, impedance cross margin, and broadband energy within the same system. On the other hand, it constructs a joint theoretical model of the electrical and mechanical life of grid-side equipment under source-load impact, mapping hotspot temperature rise, partial discharge equivalent, overvoltage accumulation, negative sequence, and electromagnetic force pulsation into a unified damage metric and RUL, enabling adaptive online updates that adapt to changes in operating conditions, aging, topology, and control strategies. If the above system is realized, it will enable the impact events in the traction-public power grid coupling scenario to achieve an engineering closed loop of "early warning, quantitative assessment, strategic control, and life-determined maintenance", significantly improving the safety margin and economy of the power grid under the background of high power electronics. Summary of the Invention

[0005] Therefore, the technical problem to be solved by the present invention is that existing evaluation methods are difficult to quantify transient impacts, cannot describe the spatiotemporal diffusion of impacts in the power grid, and do not consider multi-physics coupling in equipment life assessment.

[0006] The above-mentioned technical problems are solved by the following technical solutions: A method for assessing source load impact and system load life includes collecting data to identify impacts, establishing a source load model, analyzing the impact diffusion path, assessing system load, integrating multi-dimensional damage, calculating the remaining life of the equipment and automatically updating parameters, and generating a closed-loop control strategy.

[0007] In a preferred embodiment of the source-load impact and system load life assessment method of the present invention: synchronous measurement data of the source side and the load side are acquired; the synchronous measurement data is analyzed to identify impact events; and a dynamic response characteristic model reflecting the interaction between the source and the load is constructed based on the impact event data; a stability assessment index for quantifying the impact intensity is calculated based on the dynamic response characteristic model; a mapping relationship from local injection to global evolution of the impact is established in combination with the power grid topology characteristics; the spatiotemporal transmission characteristics of the impact are analyzed to obtain the load state data of the entire system; frequency domain energy characteristics and current fluctuation characteristics are extracted from the impact event data; a multi-dimensional physical field coupled damage assessment model is constructed; the health status and remaining life of the equipment to be assessed are calculated through the damage assessment model; the parameters of the dynamic response characteristic model and the damage assessment model are updated through an adaptive correction mechanism using the synchronous measurement data; and a system control strategy is generated based on the stability assessment index, the load state data of the entire system, the health status, and the remaining life.

[0008] In a preferred embodiment of the source-load impact and system load life assessment method of the present invention: the steps of acquiring synchronous measurement data of the source side and the load side, analyzing the synchronous measurement data to identify impact events, and constructing a dynamic response characteristic model reflecting the interaction between the source and the load based on the impact event data include: calculating the time-domain rate of change characteristics, spectral distribution characteristics, and unbalance characteristics of voltage and current in the synchronous measurement data, marking the time periods when the characteristic values ​​exceed a preset threshold as impact event windows; extracting voltage transient waveform data and current transient waveform data within the impact event window, decoupling the source-side equivalent impedance parameter and the load-side admittance parameter; introducing physical consistency constraints to correct the decoupling of the source-side equivalent impedance parameter and the load-side admittance parameter, and establishing a dynamic response characteristic model including frequency response characteristics.

[0009] In a preferred embodiment of the source-load impact and system carrying capacity lifetime assessment method of the present invention: the step of calculating the stability assessment index of the quantitative impact intensity based on the dynamic response characteristic model, and establishing the mapping relationship from local injection to global evolution of the impact in combination with the power grid topology characteristics, analyzing the spatiotemporal conduction characteristics of the impact, and obtaining the full system carrying capacity status data includes: extracting the impedance ratio parameter in the dynamic response characteristic model, calculating the impedance interaction margin reflecting the voltage stability boundary; calculating the transient rate of change energy and broadband spectrum energy of the impact event data, and weightedly fusing the impedance interaction margin, the transient rate of change energy, and the broadband spectrum energy to obtain the stability assessment index; constructing a power grid topology model composed of network nodes, mapping the power fluctuation characteristics of the impact event data to the node power injection vector corresponding to the network node; applying the conduction analysis mechanism to simulate the spatiotemporal diffusion process of the node power injection vector in the power grid topology model, and calculating the equivalent load increment of the network node; calibrating the diffusion parameters of the conduction analysis mechanism using real-time power flow data, and calculating the load carrying capacity margin of the monitoring section by combining the calibrated diffusion parameters and the equivalent load increment to obtain the full system carrying capacity status data.

[0010] In a preferred embodiment of the source-load impact and system bearing life assessment method of the present invention: the step of extracting frequency domain energy features and current fluctuation features from the impact event data, constructing a multi-dimensional physical field coupled damage assessment model, and calculating the health status and remaining life of the device to be assessed through the damage assessment model includes: mapping the frequency domain energy features as additional loss heat sources, inputting them into the thermal network model of the device to be assessed, calculating the hot spot temperature history, and obtaining thermal aging damage; extracting partial discharge features and transient overvoltage features from the impact event data, converting the partial discharge features and transient overvoltage features into the cumulative loss of the insulating medium, and obtaining insulation damage; calculating the electromagnetic force pulsation spectrum based on the negative sequence component and power reversal rate in the current fluctuation features, processing the electromagnetic force pulsation spectrum, and obtaining mechanical fatigue damage; combining the thermal aging damage, the insulation damage, and the mechanical fatigue damage to obtain the total damage degree, comparing the total damage degree with the benchmark life threshold, and determining the health status and remaining life of the device to be assessed.

[0011] In a preferred embodiment of the source load impact and system load life assessment method of the present invention: the step of updating the parameters of the dynamic response characteristic model and the damage assessment model using the synchronous measurement data through an adaptive correction mechanism, and generating a system control strategy based on the stability assessment index, the full system load state data, the health status, and the remaining life includes: constructing a joint objective function containing the deviation between the output data of the dynamic response characteristic model and the synchronous measurement data; solving the joint objective function to determine the adjustment amount of the parameters of the dynamic response characteristic model and the parameters of the damage assessment model; dividing the parameters of the dynamic response characteristic model and the parameters of the damage assessment model into fast-changing parameters and slow-changing parameters, wherein the fast-changing parameters characterize electrical impedance characteristics, and the slow-changing parameters characterize thermal and aging characteristics; updating the fast-changing parameters within the time period corresponding to the impact event using a parameter update mechanism, and updating the slow-changing parameters within a preset rolling time window.

[0012] In a preferred embodiment of the source load impact and system load life assessment method of the present invention: the generation of system regulation strategy based on stability assessment index, the full system load state data, the health state, and the remaining life includes: establishing a multi-dimensional optimization objective that includes suppressing the impact of impact stability, improving the full system load margin, and delaying equipment life decay; constructing a system safety constraint set based on the operating constraints of the equipment to be evaluated and the power grid topology characteristics; determining the system controllable variables used for implementation of regulation, and analyzing the sensitivity relationship of the system controllable variables to the stability assessment index, the full system load state data, and the remaining life; and solving the multi-dimensional optimization objective under the premise of satisfying the system safety constraint set to determine the target value of the system controllable variables and obtain the system regulation strategy.

[0013] In a preferred embodiment of the source-load impact and system load life assessment method of the present invention: the system control instructions include: adjusting the switching status of the reactive power compensation device to optimize impedance interaction margin; adjusting the transformer tap position to improve voltage distribution; and modifying the output plan of the power generation equipment to achieve rebalancing of the line power flow distribution.

[0014] A source load impact and system load life assessment system includes a data parsing and dynamic modeling module, a load assessment and conduction analysis module, a damage coupling and life prediction module, and a parameter correction and strategy generation module.

[0015] In a preferred embodiment of the source-load impact and system load-bearing life assessment system of the present invention: a data parsing and dynamic modeling module is used to acquire synchronous measurement data from the source side and the load side, parse the synchronous measurement data to identify impact events, and construct a dynamic response characteristic model reflecting the interaction between the source and the load based on the impact event data; a load-bearing assessment and conduction analysis module is used to calculate a stability assessment index for quantifying the impact intensity based on the dynamic response characteristic model, and establish a mapping relationship from local injection to global evolution of the impact in combination with the power grid topology characteristics, analyze the spatiotemporal conduction characteristics of the impact, and obtain the load-bearing state data of the entire system; a damage coupling and life prediction module is used to extract the frequency domain energy characteristics and current fluctuation characteristics from the impact event data, construct a multi-dimensional physical field coupled damage assessment model, and calculate the health status and remaining life of the equipment to be assessed through the damage assessment model; a parameter correction and strategy generation module is used to update the parameters of the dynamic response characteristic model and the damage assessment model using the synchronous measurement data through an adaptive correction mechanism, and generate a system control strategy based on the stability assessment index, the load-bearing state data of the entire system, the health status, and the remaining life.

[0016] A computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the steps of the source load impact and system load life assessment method as described above.

[0017] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of a source load impact and system load life assessment method as described above.

[0018] The beneficial effects of this invention are as follows: It achieves integrated quantization of traction impact across the time domain, frequency domain, and topology, solving the problem of assessing the spatiotemporal diffusion of impact. By constructing a multi-channel joint life model integrating electrical, thermal, and mechanical aspects, it accurately maps multidimensional stress damage, supporting life-based maintenance. Utilizing differentiable simulation and online filtering technology, it ensures that model parameters are adaptively updated according to operating conditions, resolving the disconnect between assessment and reality. This method supports edge-cloud collaboration and closed-loop pre-control, effectively improving the power grid safety margin and delaying equipment aging. Attached Figure Description

[0019] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings of the embodiments of the present invention will be briefly described below. Obviously, the drawings described below only relate to some embodiments of the present invention and are not intended to limit the present invention.

[0020] Figure 1 A detailed flowchart of the source load impact and system load life assessment method is shown; Figure 2A schematic diagram of the layered structure of the edge-cloud collaboration used in the source load impact and system load life assessment system is shown. Figure 3 A simplified flowchart of the source load impact and system load life assessment method is shown. Detailed Implementation

[0021] To enable those skilled in the art to better understand the present invention, the present invention will be further described in detail below with reference to specific embodiments and accompanying drawings.

[0022] The terminology used in this invention is that which is currently widely used in the art in consideration of the function of the invention; however, these terms may vary according to the intent of those skilled in the art, precedent, or new technology in the art. Furthermore, specific terms may be chosen by the applicant, and in such cases, their detailed meanings will be described in the detailed description of the invention. Therefore, the terms used in this specification should not be construed as simple names, but rather based on their meanings and the overall description of the invention.

[0023] Example 1, referring to Figure 1 This embodiment provides a method for assessing source load impact and system load-bearing life, including: S1: Collect data to identify impacts and establish a source-load model.

[0024] Specifically, synchronous measurement data from the source and load sides are acquired, the synchronous measurement data is analyzed to identify impact events, and a dynamic response characteristic model reflecting the interaction between the source and load is constructed based on the impact event data.

[0025] In this embodiment, high-precision synchronous sampling is first performed using edge computing units to capture transient waveforms of voltage and current. By calculating indicators such as steepness, spectral energy, and imbalance, typical impact events such as traction, regeneration, or phase-by-phase passage are automatically identified and extracted. Then, the source-side equivalent impedance and load-side mixed admittance are inverted using the regularized least squares method to construct an equivalent Thevenin or Norton model that includes frequency response characteristics.

[0026] By using high-frequency synchronous recording and event-driven mechanisms, fleeting transient impacts and broadband disturbances can be accurately extracted from massive amounts of data, solving the problem that traditional steady-state monitoring is unable to capture millisecond-level power pulsations and phase switching characteristics. At the same time, the establishment of a frequency response impedance model provides a mathematical basis that conforms to physical measurements for subsequent analysis of source-network coupled oscillations and voltage stability.

[0027] S2: Analyze the impact diffusion path and assess the system load.

[0028] Specifically, a stability assessment index for quantifying impact intensity is calculated based on a dynamic response characteristic model. A mapping relationship between impact from local injection to global evolution is established by combining the characteristics of the power grid topology. The spatiotemporal transmission characteristics of the impact are analyzed to obtain the load-bearing state data of the entire system.

[0029] In this implementation, a unified impulse stability impact index (ISI) is first defined, which includes transient steepness, frequency domain energy, and impedance margin deficit. Then, a propagation model of impulse energy in the power grid topology is constructed using the Graph Laplace diffusion operator, mapping locally injected power disturbances to the equivalent load increments of all nodes in the network, thereby calculating the spatiotemporal carrying capacity margin (ACM) of critical sections.

[0030] This approach breaks away from the limitations of traditional methods that only focus on single-point exceedances or time-period averages, and can comprehensively consider the intensity and scope of the impact under the same dimension. It can not only reflect the degree of impact on voltage stability, but also intuitively quantify the spatial diffusion path and temporal superposition effect of impact energy in the power grid, thereby accurately identifying the weak links and potential risk sections of the system's carrying capacity.

[0031] S3: Integrate multidimensional damage to calculate the remaining lifespan of the device.

[0032] Frequency domain energy characteristics and current fluctuation characteristics are extracted from impact event data to construct a multidimensional physical field coupled damage assessment model. The health status and remaining life of the equipment to be assessed are calculated through the damage assessment model.

[0033] In this embodiment, the process establishes a four-channel joint model of electricity, heat, discharge, and force, inputs harmonic and interharmonic losses into the thermal network to calculate the hot spot temperature rise, converts transient overvoltage into partial discharge accumulation, and transforms the electromagnetic force pulsation caused by unbalanced current and power reversal into mechanical fatigue stress.

[0034] Ultimately, by employing rainflow counting and cumulative damage theory, the overall health and remaining life (RUL) of the equipment are output. This overcomes the shortcomings of existing technologies that separately assess thermal aging and mechanical wear, achieving unified quantification of multi-physical field composite damage to equipment. It can also accurately reveal the accelerated aging mechanism of broadband harmonics and frequent impacts on transformer insulation and mechanical structures, providing a scientific basis for the shift from periodic maintenance to life-based maintenance, and significantly improving the level of precision in asset management.

[0035] S4: Automatically update parameters and generate closed-loop control strategies.

[0036] The parameters of the dynamic response characteristic model and the damage assessment model are updated through an adaptive correction mechanism using synchronous measurement data, and system control strategies are generated based on stability assessment indicators, full system load status data, health status and remaining life.

[0037] In this embodiment, a differentiable simulation model is constructed, and fast-changing and slow-changing parameters are separated and updated online based on measured data using adjoint automatic differentiation and unscented Kalman filtering techniques. Based on the updated model, a multi-objective optimization problem aimed at suppressing impacts, releasing load margins, and delaying aging is solved, generating control commands such as reactive power compensation and tap changer adjustment.

[0038] It can solve the problem of offline model parameters gradually becoming distorted as equipment ages and operating conditions change, ensuring that the evaluation results are always consistent with the actual situation on site; through a closed-loop feedback mechanism, it can maximize the carrying capacity of the power grid and extend the service life of key equipment while ensuring the safety and stability of the system, achieving a qualitative leap from passive monitoring to proactive prevention and control.

[0039] It should be noted that existing technical solutions mostly assume fundamental or quasi-steady-state conditions and use static load models to evaluate systems. These methods struggle to characterize the steepness and broadband energy distribution of traction shocks, and fail to reflect their coupling relationship with voltage stability margins. Furthermore, traditional time-average-based evaluation frameworks struggle to quantify the spatial diffusion of shocks on the grid structure and their superposition effects over time. Asset life assessments also often calculate thermal and mechanical stresses separately, lacking a unified quantitative system.

[0040] Therefore, to address the aforementioned problems, this invention constructs a hybrid modeling framework with measured data as the core constraint, simultaneously measuring kurtosis energy and impedance cross-margin within the same system. It utilizes graph topology diffusion operators to analyze the spatiotemporal transmission patterns of shocks and uniformly maps multi-channel electrothermal stresses into equipment damage metrics. This method enables early warning and quantifiable assessment of shock events, and ensures the model's adaptability to real-world operating conditions through online parameter updates, thereby effectively improving the safety margin and operational economy of the power grid.

[0041] Example 2, refer to Figure 1 This embodiment provides a method for assessing source load impact and system load-bearing life, including: S1: Data Acquisition for Impact Identification and Source-Load Model Establishment: Acquire synchronous measurement data from the source and load sides, analyze the synchronous measurement data to identify impact events, and construct a dynamic response characteristic model reflecting the interaction between the source and load based on the impact event data.

[0042] S2: Analyze the impact diffusion path and assess the system load: Calculate the stability assessment index of the quantitative impact intensity based on the dynamic response characteristic model, and establish the mapping relationship from local injection to global evolution of the impact in combination with the power grid topology characteristics. Analyze the spatiotemporal transmission characteristics of the impact and obtain the load status data of the entire system.

[0043] S3: Integrating multidimensional damage to calculate the remaining life of the equipment: Extracting frequency domain energy characteristics and current fluctuation characteristics from impact event data, constructing a multidimensional physical field coupled damage assessment model, and calculating the health status and remaining life of the equipment to be assessed through the damage assessment model.

[0044] S4: Automatic parameter update and closed-loop control strategy generation: Utilize synchronous measurement data to update the parameters of the dynamic response characteristic model and damage assessment model through an adaptive correction mechanism, and generate system control strategies based on stability assessment indicators, full system load status data, health status, and remaining lifespan.

[0045] By establishing a complete closed loop of "sensing-assessment-prediction-control," the problem of disconnect between monitoring data and mechanistic models in existing technologies is overcome. Its core advantage lies in the fusion of multi-source heterogeneous data such as "waveform, phasor, and topology," realizing full-chain quantification of the impact of traction shock, and providing a solid data foundation and logical support for subsequent precise control.

[0046] Specifically, the process involves acquiring synchronous measurement data from the source and load sides, analyzing this data to identify impact events, and constructing a dynamic response characteristic model reflecting the interaction between the source and load based on the impact event data. This includes: calculating the time-domain rate of change characteristics, spectral distribution characteristics, and imbalance characteristics of voltage and current in the synchronous measurement data; marking periods where the characteristic values ​​exceed preset thresholds as impact event windows; extracting transient voltage and current waveform data within the impact event windows; decoupling the equivalent impedance parameters of the source side and the admittance parameters of the load side; introducing physical consistency constraints to correct the decoupling of the equivalent impedance parameters of the source side and the admittance parameters of the load side; and establishing a dynamic response characteristic model that includes frequency response characteristics.

[0047] After accurately capturing fleeting impact characteristics, a mathematical model conforming to physical laws is established. Through joint detection of multiple indicators in the time-frequency domain, background noise is effectively distinguished from actual traction / regenerative impacts. The introduction of KK consistency constraints forces the identified impedance parameters to meet causal requirements, avoiding the loss of physical meaning caused by pure mathematical fitting, and significantly improving the model's generalization ability and reliability under complex operating conditions. Specifically, based on the dynamic response characteristic model, a stability assessment index for quantifying the impact intensity is calculated. A mapping relationship from local injection to global evolution of the impact is established in conjunction with the power grid topology characteristics. The spatiotemporal conduction characteristics of the impact are analyzed to obtain the system-wide load-bearing status data, including: extracting the impedance ratio parameter from the dynamic response characteristic model and calculating the impedance cross-margin reflecting the voltage stability boundary; calculating the transient rate of change energy and broadband spectral energy of the impact event data, and weightedly fusing the impedance cross-margin, transient rate of change energy, and broadband spectral energy to obtain the stability assessment index; constructing a power grid topology model composed of network nodes, mapping the power fluctuation characteristics of the impact event data to the node power injection vector of the corresponding network node; applying a conduction analysis mechanism to simulate the spatiotemporal diffusion process of the node power injection vector in the power grid topology model, and calculating the equivalent load increment of the network node; calibrating the diffusion parameters of the conduction analysis mechanism using real-time power flow data, and combining the calibrated diffusion parameters with the equivalent load increment to calculate the load-bearing margin of the monitoring section, thus obtaining the system-wide load-bearing status data.

[0048] The challenges of quantifying the impact and tracing the diffusion path are solved in this step: the unified index of source-load hybrid impedance and impact stability impact unifies transient steepness and steady-state impedance margin in the same dimension, comprehensively measuring voltage stability; and the introduction of the Graph Laplace diffusion operator allows the evaluation system to clearly analyze the flow trajectory of impact energy in complex grid structures, and then intuitively present the dynamic changes in the bearing capacity of key sections, providing dispatchers with visualized spatiotemporal early warning.

[0049] Specifically, frequency domain energy features and current fluctuation features are extracted from impact event data to construct a multi-dimensional physical field coupled damage assessment model. The health status and remaining life of the equipment under assessment are calculated using this model, including: mapping frequency domain energy features to additional loss heat sources and inputting them into the thermal network model of the equipment under assessment to calculate the hotspot temperature history and obtain thermal aging damage; extracting partial discharge features and transient overvoltage features from the impact event data and converting them into cumulative losses of the insulating medium to obtain insulation damage; calculating the electromagnetic force pulsation spectrum based on the negative sequence component and power reversal rate in the current fluctuation features, processing the electromagnetic force pulsation spectrum, and obtaining mechanical fatigue damage; combining thermal aging damage, insulation damage, and mechanical fatigue damage to obtain the total damage degree, and comparing the total damage degree with the baseline life threshold to determine the health status and remaining life of the equipment under assessment.

[0050] Traditional life assessment focuses only on single thermal aging, which has significant limitations. Our invention has constructed a damage assessment system that integrates multiple physical fields such as electricity, heat, and force. By uniformly converting broadband harmonic heating, overvoltage partial discharge, and mechanical vibration caused by power surges into life loss, it can more realistically reflect the combined damage of electric locomotive impacts on key equipment such as transformers. It has accurate quantitative basis, enabling the implementation of the life-based maintenance strategy and effectively avoiding over-maintenance or neglect.

[0051] Specifically, the parameters of the dynamic response characteristic model and the damage assessment model are updated using an adaptive correction mechanism based on synchronous measurement data. The system control strategy is generated based on stability assessment indicators, overall system load-bearing status data, health status, and remaining lifespan. This includes: constructing a joint objective function that includes the deviation between the output data of the dynamic response characteristic model and the synchronous measurement data; solving the joint objective function to determine the adjustment amounts of the parameters of the dynamic response characteristic model and the damage assessment model; classifying the parameters of the dynamic response characteristic model and the damage assessment model into fast-changing parameters and slow-changing parameters, where fast-changing parameters characterize electrical impedance characteristics and slow-changing parameters characterize thermal and aging characteristics; and using a parameter update mechanism to update the fast-changing parameters within the corresponding impact event period and update the slow-changing parameters within a preset rolling time window.

[0052] To address the challenge of model parameters drifting due to equipment aging and environmental changes, we developed a differentiable simulation + dual-track filtering mechanism. Furthermore, our fast and slow variable separation update strategy ensures rapid tracking of electrical transients while also considering the long-term statistical regularities of thermal and aging parameters, guaranteeing that the model maintains consistency with the real physical system throughout its entire lifecycle, achieving digital twin-level online self-adaptation.

[0053] Specifically, based on stability assessment indicators, system load-bearing state data, health status, and remaining lifetime, the system regulation strategy is generated as follows: A multi-dimensional optimization objective is established, encompassing suppressing the impact of shocks on stability, improving the system load-bearing margin, and delaying equipment lifespan degradation; a system safety constraint set is constructed based on the operating constraints of the equipment to be evaluated and the grid topology characteristics; the system controllable variables used for implementation are determined, and the sensitivity relationship of these variables to stability assessment indicators, system load-bearing state data, and remaining lifetime is analyzed; under the premise of satisfying the system safety constraint set, the multi-dimensional optimization objective is solved, the target values ​​of the system controllable variables are determined, and the system regulation strategy is obtained.

[0054] Our invention transforms the assessment results into actual control actions, achieving a closed-loop technology. The multi-objective optimization algorithm can find the optimal balance between system stability, power supply capacity, and equipment lifespan. The generated strategy is no longer a single-dimensional emergency measure, but a globally optimal solution that takes into account both short-term safety and long-term asset health, significantly improving the overall efficiency of power grid operation.

[0055] Specifically, system control commands include: adjusting the switching status of reactive power compensation devices to optimize impedance cross-margin; adjusting transformer tap positions to improve voltage distribution; and revising the output plans of power generation equipment to achieve rebalancing of line power flow distribution.

[0056] Only by translating the theoretical results of multi-objective optimization into specific execution instructions for physical devices can direct intervention in the power grid's operating status be achieved. Furthermore, by switching reactive power compensation devices on and off, node impedance characteristics can be directly corrected to improve voltage stability margins; voltage deviations can be eliminated by adjusting transformer tap changes; and power flow distribution can be optimized by coordinating with generator output adjustments.

[0057] This multi-device collaborative control ensures that the system can maintain a safe operating boundary when facing impact loads, and implements the assessed load margin release and life extension strategies at the physical level, realizing a complete technical closed loop from monitoring and assessment to execution control.

[0058] Example 3, referring to Figure 1 This embodiment provides a source load impact and system load life assessment system, including a data parsing and dynamic modeling module, a load assessment and conduction analysis module, a damage coupling and life prediction module, and a parameter correction and strategy generation module.

[0059] Specifically, the data parsing and dynamic modeling module is used to acquire synchronous measurement data from the source and load sides, parse the synchronous measurement data to identify impact events, and construct a dynamic response characteristic model reflecting the interaction between the source and load based on the impact event data.

[0060] The data parsing and dynamic modeling module addresses the shortcomings of traditional monitoring methods, such as insufficient sampling rates and a lack of event-driven mechanisms, enabling precise capture of millisecond-level broadband impacts. Through real-time decoupling and modeling at the edge, massive amounts of raw waveforms are transformed into physically meaningful impedance and source parameters, significantly reducing data transmission bandwidth requirements and providing a high-fidelity model foundation for subsequent system-level stability analysis.

[0061] Specifically, the load-bearing assessment and transmission analysis module is used to calculate the stability assessment index of the quantitative impact intensity based on the dynamic response characteristic model, and to establish the mapping relationship from local injection to global evolution of the impact in combination with the power grid topology characteristics, analyze the spatiotemporal transmission characteristics of the impact, and obtain the load-bearing status data of the entire system.

[0062] The load-bearing capacity assessment and conduction analysis module can overcome the limitations of single-point assessment and provide a global view of the spatiotemporal distribution of impact energy in the power grid. By quantifying the diffusion and superposition effects of impacts in the network, it can intuitively identify weak links in the system's load-bearing capacity and potential voltage exceedance risk areas, enabling dispatchers to grasp the actual extent to which impacts erode the stability margin of the entire network.

[0063] Specifically, the damage coupling and life prediction module is used to extract frequency domain energy characteristics and current fluctuation characteristics from impact event data, construct a multi-dimensional physical field coupled damage assessment model, and calculate the health status and remaining life of the equipment to be assessed through the damage assessment model.

[0064] The damage coupling and life prediction module establishes a life evaluation system that integrates electro-thermal-mechanical multi-physics fields, changing the previous reliance on single thermal aging or offline tests to estimate life. By accumulating damage caused by multi-dimensional stress in real time, it can accurately reflect the accelerating effect of operating condition fluctuations on equipment aging, providing scientific data support for the shift from periodic maintenance to condition-based maintenance and life-based maintenance.

[0065] Specifically, the parameter correction and strategy generation module is used to update the parameters of the dynamic response characteristic model and the damage assessment model through an adaptive correction mechanism using synchronous measurement data, and to generate system control strategies based on stability assessment indicators, full system load status data, health status and remaining life.

[0066] The parameter calibration and strategy generation module constructs a closed loop for evaluation and control, ensuring that model parameters always keep pace with equipment aging and changes in the operating environment, thus solving the problem of offline models becoming distorted over time. By generating comprehensive control strategies that take into account system stability, carrying capacity, and equipment lifespan, it enables proactive intervention and optimization of the power grid's operating status, effectively improving the system's safety margin and asset utilization efficiency.

[0067] Example 4, refer to Figure 1 This embodiment provides a method for assessing source load impact and system load life.

[0068] First, the overall technical solution and variable definition. This method proposes a three-layer architecture of "mechanism-data fusion": synchronous sampling and event recognition are performed at the edge, system-level evaluation and strategy generation are performed at the cloud, and parameters are updated online through a dual-track mechanism of "differentiable simulation + statistical filtering".

[0069] Time-domain signals are mainly composed of bus phasors and high-sampling waveforms: voltage Current Sampling frequency fs, event window Frequency domain quantities are obtained from short-time Fourier transform or wavelet transform, specifically time-frequency energy. , The source-charge frequency response is modeled using the equivalent Thevenin / Norton model: source impedance Mixed load admittance: The parameter vector θ needs to be identified. The network is an undirected graph. Modeling, Laplace matrix: L=D−W Where D is the degree matrix and W is the weighted adjacency matrix.

[0070] Equipment set D includes generators, main transformers, voltage regulators and compensation devices, circuit breakers, busbars and cables, etc. In the lifespan measurement, temperature T(t), stress amplitude Δσ(t), and partial discharge equivalent qpd(t) are all obtained through a combination of actual measurement and simulation.

[0071] Secondly, impact event detection and parametric characterization: To map traction impacts from different sources and operating conditions to a standardized feature space, this invention proposes a joint detection and decomposition of four types of indicators: steepness, spectrum, imbalance, and energy.

[0072] 1. Steepness index: Robust estimation using moving difference and Savitzky-Golay filtering is employed at high sampling rates.

[0073] 2. Spectral Concentration and Broadband Energy make To focus on frequency bands containing harmonics, interharmonics, and high-frequency sidebands, the following definitions are provided: in This is a template for the frequency band sensitivity of devices / systems.

[0074] Definition of spectral concentration: Used to distinguish between "narrowband harmonics" and "broadband energy".

[0075] 3 Imbalance / Negative Order Indicators: The negative sequence current ratio introduced on the three-phase public grid side is... Compared with the negative sequence voltage It is calculated in real time by the synchronous phasor.

[0076] 4. Event energy and power reversal: Event active / active energy: Power reversal rate: Through the above indicators, this invention unifies event location, classification traction / regeneration / phase segmentation / grid connection / disconnection / control switching and intensity quantization into a standardized vector. .

[0077] Then, the source-charge hybrid impedance and the impact stability effect are unified indices, namely SIS: To characterize the three types of influences—transient kurtosis, frequency domain energy, and stability margin—under the same dimension, this invention defines an impact stability influence index: Among them, the margin loss item Defined by frequency response impedance ratio: Weighting function Used to emphasize frequency bands sensitive to voltage stability. To ensure physical consistency, Z... S Z L Apply Kramers–Kronig consistency with low / high frequency asymptotic regularization: in [∙] represents the Hilbert transform. The parameters α, β, γ, δ of the ISI are obtained through multi-point calibration to maximize the correlation between the ISI and the voltage over-limit amplitude, duration, and subsequent temperature rise peak.

[0078] Next, the system-wide "Impact Diffusion-Bearing Margin ACM" assessment model: To analyze the transmission mechanism of traction impact in a space frame, this invention uses a graphical Laplace diffusion kernel to characterize the spatiotemporal propagation of event injection. Let the nodal impact injection vector be... System impact power density satisfy: Will Mapped to node equivalent load increment The equivalent loads at key nodes / sections are obtained: Define the spatiotemporal carrying capacity margin: and with The valley value and valley width are used as operational alarm parameters. To mitigate parameter uncertainties, this invention combines synchronous phasor power flow to perform short-term calibration of W and α: in λ is the a priori edge weight, and λ is the sparsity regularity, ensuring that the diffusion kernel converges under measurement consistency.

[0079] Then, the combined damage and RUL of the electrical lifetime-mechanical lifetime "four-channel": To convert multi-domain stress into uniform damage, this invention defines total equipment damage as follows: in The baseline lifetime is determined by the superposition of three branches: thermal, slope, and partial discharge. Where T(t) is obtained by a gradient RC thermal network driven by additional losses. In terms of frequency domain loss increments: As input to the thermal network, the temperature field's response over time is obtained. The mechanical life channel employs rainflow counting and the S–N / Coffin–Manson model: Stress amplitude Calculated from electromagnetic force and vibrational energy spectrum: For frequency bands related to mechanical coupling, This represents the amplitude of the structure transfer function. The above four channels accumulate under a unified dimension, forming a parallel lifetime update mechanism that combines event-driven short windows and scrolling long windows.

[0080] Then, parameter identification, differentiable simulation, and online updates are performed. To ensure the model continuously fits the real-world conditions during operation, this invention constructs a multi-objective, cross-domain, differentiable joint loss function: The gradient is calculated using adjoint-automatic differentiation to obtain offline initial values; in the online stage, unscented Kalman filtering / ensemble Kalman filtering is used for "fast-slow variable" separation and updating: Fast parameters are updated according to the event window. Slow parameters are updated via a scrolling window.

[0081] State transition: in, Step size, , This addresses process and measurement noise. The dual-track update ensures that model parameters remain interpretable and consistent despite fluctuations in data quality and gradual changes in equipment.

[0082] Then, the multi-objective optimization of the evaluation-prevention closed loop: In obtaining , and Subsequently, this invention generates a pre-control strategy with the objectives of "suppressing impact, releasing load, and delaying aging." Let the control vector... Represent reactive power compensation, transformer tap change, regenerative shaping settings, and power flow transfer quantities, respectively, to construct a multi-objective problem: The sensitivity of ISI and ACM to control variables is explicitly expressed in linearized or quadratic approximation form, and solved using sequential quadratic programming or mixed-integer quadratic programming. Heuristic fast solutions are used for short-cycle operations, while robust optimization is employed for long-cycle operations to handle uncertainty sets. : The final strategy is issued in accordance with IEC61850 / GOOSE and gRPCAPI, forming a pre-control closed loop that can be executed at the minute level.

[0083] Specifically, such as Figure 1 As shown, this innovative device adopts a hierarchical "edge-cloud collaborative" structure. The edge side is deployed at traction substations and key public network nodes, including: a synchronous sampling and timing module (GPS / IEEE1588PTP), a high-frequency acquisition and anti-aliasing filtering module, an event detection and decomposition module, a local impedance inversion and fast ISI estimation module, and a buffering and uploading module. The cloud side is deployed at the dispatch / master station, including: a graph topology and Laplace diffusion engine, a power flow consistency calibrator, an ACM evaluator, a four-channel lifetime engine for thermal network solving / slope-partial discharge reduction / rainflow counting, differentiable simulation and dual-track filtering, and a multi-objective strategy optimizer.

[0084] In this embodiment, the edge measurement unit (EMU) employs a 24-bit Σ-Δ ADC with a sampling rate ≥6.4kHz, configurable up to 12.8kHz. Synchronization is achieved via GPS 1PPS+PTP dual-redundant timing, with a timescale error ≤1μs. Voltage measurements are taken from the secondary side of the 27.5kV PT at the traction substation, or a μPMU is installed at the 110 / 220kV bus. Current measurements utilize a zero-flux transformer or a 0.2S-class CT secondary, with optional current clamping at the locomotive's current-receiving side. The analog front-end features a band-stop-low-pass cascaded anti-aliasing filter with a cutoff frequency of 0.4f_s and inter-channel time offset ≤0.5μs. The edge unit possesses ≥4 cores of ARM+DSP or x86 small industrial control computing power, ≥8GB RAM, and local temporary storage of ≥48h of raw waveforms and feature summaries, supporting OPCUA / IEC61850MMS / GOOSE and gRPC uploads.

[0085] Specifically, S1: Collecting data to identify impacts and establishing a source-load model includes: Data acquisition, calibration and preprocessing; event detection and decomposition; source-load frequency response identification and ISI calculation.

[0086] Data acquisition, calibration, and preprocessing: S1.1: Sampling and Alignment. Acquiring three-phase voltage and current. Discrete quantities such as circuit breaker position and AT / BT segmentation status are also included; linear interpolation is used to align discrete events to waveform time scales.

[0087] S1.2: Calibration and Denoising. Amplitude calibration is performed based on the PT / CT ratio; Variational Mode Decomposition (VMD) or Singular Spectrum Analysis (SSA) is used to suppress random noise and power frequency drift; voltage phase is corrected at the 0.1° level.

[0088] S1.3: Phasor Supplementation. Simultaneous phasor calculation for 30–240 fps. Negative / zero order components This provides input for subsequent imbalance indicators and power flow calibration.

[0089] S1.4: Spectrum estimation. Time-frequency energy is generated using an STFT Hanning window, 50% overlap, or an S-transform. The frequency band covers 2–2.5kHz.

[0090] Event detection and decomposition S1.5: Steepness threshold. Calculation: And with robust threshold Select 3–5 to trigger the candidate window.

[0091] S1.6: Spectral Concentration and Energy. Calculate broadband energy within the event window: And calculate the spectral concentration ξ\xiξ to distinguish between narrowband harmonic clusters and broadband disturbances.

[0092] S1.7: Imbalance Criterion. Synchronous Reading , If the value exceeds the configured threshold, such as 2%, it will be marked as an unbalanced event.

[0093] S1.8: Event Classification. (The following text appears to be a separate, unrelated section: "will...") Input a lightweight classifier such as logistic regression / small tree to determine the type of "traction, regeneration, phase segmentation, grid connection / disconnection, control switching", and generate a tagged event list.

[0094] Source-charge frequency response identification and ISI calculation: S1.9: Local equivalent impedance. Calculated within the event window: Using adjacent events and background segments as references, regularized least squares decomposition is performed. and : in Includes K–K uniformity and asymptotic regularization.

[0095] S1.10: Fast stability margin. Calculation: S1.11: ISI Indicator. Summary: Parameters α, β, γ, and δ are calibrated through regression analysis using historical events and records of exceeding limits / temperature rises. The edge-end retains approximate H and w values ​​from a fast ISI estimation, while the cloud performs refined calculations and regression updates.

[0096] Specifically, S2: Analyze the impact diffusion path and assess the system load, including the impact diffusion of the entire system and ACM assessment.

[0097] Whole-system impact diffusion and ACM assessment: S2.1: Network Graph Modeling. Stations / sections / buses are used as nodes, and edge weights Wij are given comprehensively based on equivalent admittance |Yij|, power flow direction, reactive power support capacity, and line reactance ratio; a Laplace matrix L=D−W is constructed.

[0098] S2.2: Injection Mapping. The power perturbation of an event at a physical location is mapped to a node injection u(t), which can be weighted according to proximity and PT / CT location.

[0099] S2.3: Diffusion Solution and Calibration. Solution: Short-term calibration of α and W using synchronous phasor power flow: S2.4: Load Margin. Calculates the equivalent load increment at each node. ,have to: The monitoring interface displays the ACM heatmap and the time of the lowest value, and provides graded alarms in "red / orange / yellow / green".

[0100] S2.5: Additional losses and thermal network. Obtained from time-frequency energy and current spectra: As input to the heat network, solve for the temperature. The thermal network can be modeled using a multi-node RC equivalent or simplified PTH model; parameters are derived from nameplate / testing / in-service identification. S602: Electrical life three-branch. Cumulative: Localized Equivalent It can be calculated using ultrasonic / radio frequency / transient overvoltage proxy variables. S603: Mechanical life branch. The stress amplitude spectrum Δσ is obtained from electromagnetic force-vibration coupling, and accumulated by rainflow counting and S–N / Coffin–Manson curves: S2.6: RUL Output and Grading. Total Damage Output device level: Green: D<0.3; Yellow: 0.3–0.5; Orange: 0.5–0.7; Red: ≥0.7, and enter the life-cycle maintenance list.

[0101] Specifically, S3: Integrates multidimensional damage and calculates the remaining lifespan of the device, including a joint assessment of electrical and mechanical lifespan.

[0102] Joint assessment of electrical and mechanical lifespan: S3.1: Additional losses and thermal network. Obtained from time-frequency energy and current spectrum. As input to the heat network, solve for the temperature. The thermal network can be modeled using a multi-node RC equivalent or simplified PTH model; parameters are derived from nameplate / testing / in-service identification. S602: Electrical life three-branch. Cumulative: Localized Equivalent It can be calculated using ultrasonic / radio frequency / transient overvoltage proxy variables. S603: Mechanical life branch. The stress amplitude spectrum Δσ is obtained from electromagnetic force-vibration coupling, and accumulated by rainflow counting and S–N / Coffin–Manson curves: S3.2: RUL Output and Classification. Total Damage Output device level: Green: D<0.3; Yellow: 0.3–0.5; Orange: 0.5–0.7; Red: ≥0.7, and enter the life-cycle maintenance list.

[0103] Specifically, S4: Automatic parameter update and closed-loop control strategy generation includes: Differentiable simulation and dual-track online updates; evaluation-pre-control closed loop.

[0104] Microsimulation and Dual-track Online Updates: S4.1: Joint Loss Function: Offline initial values ​​are obtained using adjoint-automatic differentiation; online initial values ​​are obtained using UKF / EnKF for fast-slow variable separation. Status Update: Fast variables are updated according to the event window, while slow variables are updated smoothly according to a scrolling window, such as 5–15 minutes, ensuring that the parameters are traceable and interpretable for real-time and aging conditions.

[0105] Assessment-Prevention Closed Loop: S4.2: Establish control vectors: Constructing multi-objective optimization: The constraints are: Short-cycle solutions use SQP / MIQP for fast computation, while intraday solutions utilize robust optimization. The strategy is issued through dual channels of IEC61850 / GOOSE and gRPC, retaining rollback and limiting mechanisms to ensure safety.

[0106] Finally, it should be noted that the methods and devices described in detail above are merely embodiments, and those skilled in the art can modify these embodiments in different ways as long as they do not depart from the scope of the present invention.

Claims

1. A method for source and load impact and system carrying life evaluation, characterized in that, Comprise: Obtain synchronous measurement data of source side and load side, analyze the synchronous measurement data to identify impact events, and construct a dynamic response characteristic model reflecting source-load interaction based on the data of the impact events; Based on the dynamic response characteristic model, calculate the stability evaluation index of the impact intensity, and establish the mapping relationship of the impact from local injection to global evolution combined with the grid topology characteristics, analyze the spatio-temporal conduction characteristics of the impact, and obtain the whole system bearing state data; Extract the frequency energy characteristics and current fluctuation characteristics in the impact event data, construct a multi-dimensional physical field coupled damage assessment model, and calculate the health status and remaining life of the equipment to be evaluated through the damage assessment model; Use the synchronous measurement data to update the parameters of the dynamic response characteristic model and the damage assessment model through an adaptive correction mechanism, and generate system control strategies according to the stability evaluation index, the whole system bearing state data, the health status and the remaining life.

2. The source-load impact and system bearing life assessment method of claim 1, wherein: The synchronous measurement data of the source side and the load side is obtained, the synchronous measurement data is analyzed to identify impact events, and a dynamic response characteristic model reflecting source-load interaction is constructed based on the data of the impact events, comprising: Calculate the time-domain variation rate characteristics, frequency spectrum distribution characteristics and unbalance degree characteristics of voltage and current in the synchronous measurement data, and mark the time period with characteristic values exceeding the preset threshold as an impact event window; Extract the voltage transient waveform data and current transient waveform data in the impact event window, decouple the source side equivalent impedance parameters and the load side admittance parameters; Introduce a physical consistency constraint condition to modify the decoupled source side equivalent impedance parameters and load side admittance parameters, and establish a dynamic response characteristic model containing frequency response characteristics.

3. The source-load impact and system bearing life assessment method of claim 1, wherein: Based on the dynamic response characteristic model, the stability evaluation index of the impact intensity is calculated, and the mapping relationship of the impact from local injection to global evolution is established combined with the grid topology characteristics, the spatio-temporal conduction characteristics of the impact are analyzed, and the whole system bearing state data is obtained, comprising: Extract the impedance ratio parameters in the dynamic response characteristic model, and calculate the impedance interaction margin reflecting the voltage stability boundary; Calculate the transient variation rate energy and wideband spectrum energy of the impact event data, and weight and fuse the impedance interaction margin, the transient variation rate energy and the wideband spectrum energy to obtain the stability evaluation index; Construct a power grid topology structure model composed of network nodes, and map the power fluctuation characteristics of the impact event data to the node power injection vector corresponding to the network nodes; Apply a conduction analysis mechanism to simulate the spatio-temporal diffusion process of the node power injection vector in the power grid topology structure model, and calculate the equivalent load increment of the network nodes; Calibrate the diffusion parameters of the conduction analysis mechanism using real-time power flow data, and calculate the load bearing margin of the monitoring section combined with the calibrated diffusion parameters and the equivalent load increment to obtain the whole system bearing state data. 4.The source-load impact and system carrying life assessment method of claim 1, wherein: the extracting the frequency domain energy features and the current fluctuation features in the impact event data, constructing a multi-dimensional physical field coupling damage assessment model, and calculating the health state and the remaining life of the to-be-assessed equipment through the damage assessment model comprises: mapping the frequency domain energy features to additional loss heat sources, inputting the additional loss heat sources into a thermal network model of the to-be-assessed equipment, calculating a thermal point temperature history, and obtaining thermal aging damage; extracting partial discharge features and transient overvoltage features in the impact event data, converting the partial discharge features and the transient overvoltage features into cumulative loss of insulating medium, and obtaining insulation damage; calculating an electromagnetic force pulsation spectrum according to negative sequence components and power reversal rates in the current fluctuation features, processing the electromagnetic force pulsation spectrum, and obtaining mechanical fatigue damage; and combining the thermal aging damage, the insulation damage, and the mechanical fatigue damage to obtain total damage degree, comparing the total damage degree with a reference life threshold, and determining the health state and the remaining life of the to-be-assessed equipment. 5.The source-load impact and system carrying life assessment method of claim 1, wherein: the updating the parameters of the dynamic response feature model and the damage assessment model through an adaptive correction mechanism and generating a system control strategy according to the stability assessment index, the full-system carrying state data, the health state, and the remaining life using the synchronous measurement data comprises: constructing a joint objective function containing a deviation between output data of the dynamic response feature model and the synchronous measurement data; solving the joint objective function to determine adjustment amounts of the parameters of the dynamic response feature model and the parameters of the damage assessment model; dividing the parameters of the dynamic response feature model and the parameters of the damage assessment model into fast-varying parameters and slow-varying parameters, wherein the fast-varying parameters represent electrical impedance characteristics, and the slow-varying parameters represent thermal and aging characteristics; and using a parameter updating mechanism to update the fast-varying parameters within a period corresponding to the impact event, and updating the slow-varying parameters within a preset rolling time window. 6.The source-load impact and system carrying life assessment method of claim 1, wherein: the generating a system control strategy according to the stability assessment index, the full-system carrying state data, the health state, and the remaining life comprises: establishing a multi-dimensional optimization objective containing suppression of impact stability influence, improvement of full-system carrying margin, and delay of equipment life attenuation; constructing a system safety constraint set based on operating constraints of the to-be-assessed equipment and the power grid topology features; determining system controllable variables for implementing control, and analyzing sensitivity relationships of the system controllable variables to the stability assessment index, the full-system carrying state data, and the remaining life; and solving the multi-dimensional optimization objective under the premise of meeting the system safety constraint set, determining target values of the system controllable variables, and obtaining a system control strategy. 7.The source-load impact and system carrying life assessment method of claim 6, wherein: ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ The system regulation instruction comprises: Adjusting the switching state of the reactive power compensation device to optimize the impedance interaction margin; Adjusting the transformer tap position to improve the voltage distribution; Correcting the power generation equipment output plan to achieve rebalancing of the line power flow distribution.

8. A source and load impact and system on life evaluation system, characterized by, Comprise: A data analysis and dynamic modeling module for obtaining source-side and load-side synchronous measurement data, analyzing the synchronous measurement data to identify impact events, and constructing a dynamic response characteristic model reflecting source-load interaction based on the data of the impact events; A bearing evaluation and conduction analysis module for calculating a stability evaluation index quantifying impact strength based on the dynamic response characteristic model, and establishing a mapping relationship of impact from local injection to global evolution in combination with the grid topology characteristics, analyzing the time-space conduction characteristics of the impact, and obtaining full-system bearing state data; A damage coupling and life prediction module for extracting frequency energy features and current fluctuation features in the impact event data, constructing a multi-dimensional physical field coupling damage evaluation model, and calculating the health status and remaining life of the equipment to be evaluated through the damage evaluation model; A parameter correction and strategy generation module for updating the parameters of the dynamic response characteristic model and the damage evaluation model through an adaptive correction mechanism using the synchronous measurement data, and generating system regulation strategies according to the stability evaluation index, the full-system bearing state data, the health status, and the remaining life.

9. An electronic device, comprising: Comprise: A memory and a processor; The memory is used to store computer executable instructions, and the processor is used to execute the computer executable instructions, which realize the steps of the source-load impact and system bearing life evaluation method in any one of claims 1 to 7 when executed by the processor.

10. A computer-readable storage medium, characterized in that, It stores computer executable instructions, which realize the steps of the source-load impact and system bearing life evaluation method in any one of claims 1 to 7 when executed by the processor.

Images