New energy access stability evaluation method and system under power grid digital twinning

Abstract

The application provides a new energy access stability evaluation method and system under power grid digital twinning, relates to the technical field of smart power grid, and comprises the following steps: taking the geographic coordinates of a new energy power station as a spatial constraint, and searching time sequence environment working condition data; performing power generation environment transition simulation, and outputting simulation time period working condition sequence; performing multi-threaded new energy power generation quantization simulation, and outputting time sequence new energy power generation data; after performing multi-cycle disturbed fluctuation confrontation fitting, calling multiple time sequence operation steady state anomaly records of multiple local power transmission devices; after constructing multiple node instability feature vectors, performing grid connection compensation strategy matching, and outputting multiple grid connection stability compensation strategies. The application solves the technical problem that the prior art is usually based on a simplified power grid model, cannot fully reflect the complex dynamic response of the power grid under different time scales and different disturbance conditions, and reduces the long-term operation stability of the power grid after new energy access.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of smart grids, in particular to a new energy access stability evaluation method and system under a grid digital twin. BACKGROUND

[0002] The proportion of new energy (such as wind energy, solar energy, etc.) in the power system is increasing, which, although providing more options for energy supply, also brings new challenges, especially in terms of grid stability and reliability. The existing grid stability analysis method is usually based on a simplified grid model, which cannot fully reflect the complex dynamic response of the grid under different time scales and different disturbance conditions, especially the insufficient consideration of multiple superimposed effects of factors such as weather fluctuations, load changes, etc. This makes the stability evaluation of the grid usually only reflect some typical problems in the short term, and cannot effectively evaluate the stability of the grid under long-period and multi-disturbance conditions, so that the grid may ignore some long-term potential risks and cannot take timely adjustment measures, thereby reducing the long-term operation stability of the grid after the access of new energy. SUMMARY

[0003] The application provides a new energy access stability evaluation method and system under a grid digital twin, aiming to solve the technical problem that the existing technology is usually based on a simplified grid model, which cannot fully reflect the complex dynamic response of the grid under different time scales and different disturbance conditions, thereby reducing the long-term operation stability of the grid after the access of new energy.

[0004] The first aspect of the application provides a new energy access stability evaluation method under a grid digital twin, the method comprising: retrieving time sequence environmental condition data satisfying a preset time period with the geographical coordinates of a new energy power station as a spatial constraint, wherein the new energy power station is connected to a local power transmission network; performing power generation environment transition simulation based on the time sequence environmental condition data to output a simulation period condition sequence; performing multi-threaded new energy power generation quantization simulation based on the simulation period condition sequence to output time sequence new energy power generation data; injecting the time sequence new energy power generation data into a power transmission grid twin model through a new energy grid connection point to perform multi-cycle disturbed fluctuation confrontation fitting, and then retrieving a plurality of time sequence operation steady state abnormal records of a plurality of local power transmission devices from a plurality of high-fidelity dynamic models; constructing a plurality of node instability feature vectors based on the plurality of time sequence operation steady state abnormal records, and performing grid connection compensation strategy matching based on the plurality of node instability feature vectors to output a plurality of grid connection stability compensation strategies.

[0005] In a second aspect, the application discloses a new energy access stability evaluation system under a power grid digital twin, which is used for the new energy access stability evaluation method under the power grid digital twin, and comprises a working condition data retrieval module, a power generation environment transition simulation module, a power generation quantization simulation module, an abnormal record calling module and a strategy matching module.

[0006] The one or more technical solutions provided in the application have at least the following beneficial effects:

[0007] By taking the geographic coordinates of the new energy power station as the spatial constraint to retrieve the time-series environmental working condition data, and in combination with the access situation of the local power transmission network, the accuracy and relevance of the environmental working condition data are ensured, which means that the basic data for evaluation is more in line with the geographical and environmental conditions of actual power grid operation, thereby enhancing the accuracy of stability evaluation; by performing power generation environment transition simulation based on the time-series environmental working condition data and outputting the simulation period working condition sequence, the transition process of the power generation environment under different conditions can be accurately simulated, the influence of environmental fluctuations on power generation output is reduced, and the reliability of the evaluation result is improved; by performing multi-threaded new energy power generation quantization simulation and outputting the time-series new energy power generation data, the fluctuations and changes of the new energy power generation amount can be quickly and accurately simulated, making the new energy power generation amount prediction for power grid access more reliable; by injecting the time-series new energy power generation data into the power transmission grid twin model and performing multi-cycle disturbed fluctuation confrontation fitting, the actual operation state of the power grid after the access of new energy can be simulated, and the stability of the power grid under different loads and disturbances can be dynamically reflected, this step provides a warning for the state change of the power grid through the high-fidelity dynamic model, ensuring the stable operation of the power grid after the access of new energy; by constructing the node instability feature vector based on the multiple time-series running steady-state abnormal records and performing grid-connected compensation strategy matching, dynamic adaptation to local instability of the power grid can be realized, and through automatic strategy matching, compensation strategies can be accurately selected according to the characteristics of the unstable nodes, avoiding errors and delays caused by manual intervention, and improving the ability of the power grid to respond to emergencies.

[0008] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description

[0009] Figure 1 A schematic diagram of the process for evaluating the stability of new energy access under the digital twin of the power grid provided in this application embodiment.

[0010] Figure 2 A schematic diagram of the structure of a new energy access stability assessment system under a power grid digital twin provided in this application embodiment.

[0011] Figure labeling: 10, Operating condition data retrieval module, 20, power generation environment transition simulation module, 30, power generation quantification simulation module, 40, anomaly record retrieval module, 50, strategy matching module. Detailed Implementation

[0012] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided below.

[0013] Example 1, as Figure 1 As shown in the embodiment of this application, a method for evaluating the stability of new energy access under a power grid digital twin is provided. The method includes:

[0014] Using the geographical coordinates of the new energy power station as a spatial constraint, time-series environmental condition data that meets a preset time period are retrieved, wherein the new energy power station is connected to the local power transmission network.

[0015] New energy power plants include wind farms or solar power plants. Their geographical coordinates, including latitude and longitude, provide location constraints, defining the climate and environmental region where the power plant is located, thus influencing the regional characteristics of the subsequently acquired time-series environmental condition data. Using the geographical coordinates of the power plant as input, time-series environmental condition data for the corresponding period is retrieved from databases or meteorological data service platforms. This includes wind speed and direction applicable to wind power, radiance and cloud cover applicable to photovoltaic solar power, as well as other meteorological data such as temperature and humidity. This data is provided in time-series format, which can be hourly, daily, or other time periods, reflecting environmental changes over a specific time period. During data retrieval, an appropriate time range or period is selected, such as a year, a quarter, or a month, to ensure that the selected data covers the actual grid-connected period of the power plant and is compatible with the calculation model of new energy power generation capacity.

[0016] Perform power generation environment transition simulation based on the time-series environment condition data, output simulation period condition sequence.

[0017] Through analysis of time-series environment condition data, simulate transition processes under different environmental conditions, for example, wind energy generation is affected by changes in wind speed and direction, while photovoltaic power generation is affected by factors such as sunlight intensity and cloud coverage. This simulation process establishes mathematical models to transition different environmental data to corresponding power generation states. During simulation, techniques such as environmental transition matrices are used to describe how environmental conditions change over time, for example, when transitioning from one stable wind speed state to another, the system's response and power generation capacity change. The result of the simulation is a simulation period condition sequence composed of time series, which reflects the dynamic response of the power plant under different environmental conditions, including the impact of environmental changes on power generation equipment, covering all transition states from start-up, adjustment to shutdown of power generation equipment.

[0018] Perform multi-threaded new energy power generation quantization simulation based on the simulation period condition sequence, output time-series new energy power generation data.

[0019] Use multi-threading technology to accelerate the simulation process, with each thread responsible for different calculation tasks, such as independent calculation of wind and light power generation prediction, and parallel processing of simulation under multiple environmental conditions. Multi-threading can efficiently utilize computing resources, especially when faced with large-scale time-series data, significantly improving simulation efficiency. Specifically, based on the relevant environmental conditions of wind energy, such as wind speed and direction, a wind power generation prediction model is constructed. The model is trained using historical data, such as regression analysis, neural networks, etc. to predict power generation at a specific wind speed; similarly, based on the relevant environmental conditions of light energy, such as irradiance and cloud coverage, a light power generation prediction model is constructed. This model also predicts power generation under specific radiation conditions, also based on historical data. Under the cooperation of the multi-threading model, through wind and light power generation prediction, the final output of time-series new energy power generation data includes specific power generation values over time.

[0020] Inject the time-series new energy power generation data into the power transmission grid twin model through the new energy grid point, perform multi-cycle disturbed fluctuation confrontation fitting, and retrieve multiple time-series operation steady-state anomaly records of multiple local power transmission devices from multiple high-fidelity dynamic models.

[0021] The time-series new energy power generation data is injected into the digital twin model of the power grid in real time. The digital twin model is a virtual replica of the power grid, which simulates the physical operation and dynamic characteristics of the power grid. By injecting time-series new energy power generation data, the real-time impact of new energy power generation volatility on the power grid is simulated. For example, if it is a wind power station connected to the grid, the wind speed data and the predicted power generation will dynamically update the relevant variables in the power grid model. Similarly, the irradiance and cloud cover of photovoltaic power stations are also injected to reflect the volatility of solar energy.

[0022] After injecting the time-series new energy power generation data, the power grid twin model simulates the impact of power generation fluctuations on the stability of the power grid. Multi-cycle disturbance volatility refers to fitting and evaluating the stability of the power grid response by analyzing disturbances in multiple cycles, such as rapid changes in wind speed or fluctuations in cloud cover. These disturbances include sharp fluctuations in wind speed or rapid changes in sunlight in the short term, which affect power generation and in turn cause voltage fluctuations or frequency fluctuations in the power grid. By fitting these fluctuations, the response of the power grid to these disturbances can be predicted, and the dynamic stability of the power grid can be evaluated.

[0023] In the digital twin model of the power grid, each local power transmission device in the power grid, such as transformers, switches, and feeders, has a corresponding high-fidelity dynamic model to simulate its response and stability characteristics. By interfacing with the power grid twin model, these high-fidelity dynamic models can be accessed to evaluate the behavior of each power transmission device under different operating conditions. The multiple time-series operation steady-state anomaly records retrieved from multiple high-fidelity dynamic models refer to abnormal phenomena that occur in each local power transmission device during a specific period, such as voltage fluctuations, overcurrents, and frequency instability. These records are feedback on the response of the power grid under various disturbances, reflecting the risk of instability that may occur during operation of the power grid.

[0024] After constructing multiple node instability feature vectors based on the multiple time-series operation steady-state anomaly records, a grid-connected compensation strategy matching is performed based on the multiple node instability feature vectors, and multiple grid-connected stability compensation strategies are output.

[0025] Node instability feature vectors are feature data extracted from the time-series operation steady-state anomaly records of each power transmission node. These node instability feature vectors can reflect the stability indicators of each node in the power grid when encountering disturbances. Node instability features include voltage deviation, oscillation frequency, and damping ratio, which indicate whether the nodes in the power grid have a tendency to lose stability and whether they may cause voltage collapse or frequency deviation.

[0026] According to different node instability characteristic vectors, corresponding grid-connected compensation strategies are found and matched. The grid-connected compensation strategies are generated by historical data analysis, expert knowledge base and algorithm reasoning to generate control strategies suitable for the current instability characteristics. For example, for areas with large frequency fluctuations, more frequency control devices are enabled; for areas with large voltage fluctuations, voltage regulation capability is strengthened. After strategy matching, multiple grid stability compensation strategies are output, which will be applied to grid operation to ensure that the grid can still operate stably after new energy access.

[0027] Further, it also includes:

[0028] The physical topology of the local power transmission network and the grid device information are retrieved; according to the grid device information, a plurality of high-fidelity dynamic models of a plurality of local power transmission devices are retrieved and called from a grid equipment model; the plurality of high-fidelity dynamic models are connected according to the physical topology to construct a power transmission grid twin model, wherein the power transmission grid twin model is provided with the new energy grid connection point.

[0029] The physical topology of the local power transmission network describes how each component in the grid is interconnected, for example, how transmission lines connect different substations, how they are distributed to each area, and how each component is connected. These information provides geographical and structural constraints for the grid model. The grid device information includes detailed specifications and technical parameters of various equipment in the grid, such as the capacity and voltage regulation range of the transformer, the setting parameters of the voltage transformer and protection device, the operation mode of the switch device, and the dynamic characteristics of the power equipment, such as response speed, load bearing capacity, etc. These grid device information defines the performance and limitations of the grid components, which affects the way the grid responds.

[0030] According to the retrieved grid device information, retrieve the high-fidelity dynamic model corresponding to each local power transmission device. High-fidelity dynamic model is a model that can accurately simulate the dynamic response of grid equipment. It takes into account the behavior of grid equipment under different working conditions, such as load fluctuations, voltage changes, short-circuit faults, etc. High-fidelity means that the model has very high accuracy and can reflect the details of actual grid operation. Dynamic model means that the model can capture the change process of the grid on the time axis, especially the response when a disturbance occurs, such as how load fluctuations, power generation changes or faults affect voltage, frequency and equipment response.

[0031] According to the retrieved physical topology, the high-fidelity dynamic model is connected according to the actual power grid structure, and the mutual connection mode of each node and how the current and voltage are transmitted in the power grid are determined, which is equivalent to building a virtual replica of the power grid, i.e. digital twin. In the power transmission grid twin model, a new energy grid connection point is set, and the new energy grid connection point is particularly concerned about the following contents: power generation characteristic simulation, i.e. connecting the new energy power generation data with the power station input port in the digital twin model; grid connection influence analysis, i.e. simulating the influence of new energy power generation on the power grid, especially how to affect the stability of the power grid under the condition of power grid load fluctuation and new energy volatility.

[0032] Further, based on the time sequence environment working condition data, a power generation environment transition simulation is performed to output a simulation period working condition sequence, and the method comprises:

[0033] By aggregating the time sequence environment working condition data, a plurality of extreme environment working conditions are obtained; based on the plurality of extreme environment working conditions, an environment working condition transition matrix is constructed; the transition probability of the environment working condition transition matrix is calibrated online using the time sequence environment working condition data to output an environment working condition transition model; the environment working condition transition model is driven to perform new energy power generation environment transition simulation to output the simulation period working condition sequence.

[0034] The time sequence environment working condition data includes meteorological data closely related to new energy power generation. By aggregating the time sequence environment working condition data, samples of extreme environment working conditions are screened out. For example, extreme working conditions of wind energy include wind speed greater than a certain critical value or less than a certain low threshold; extreme working conditions of light energy include strong sunlight or complete absence of sunlight. These extreme environment working conditions represent conditions that may cause new energy power generation to exhibit significant changes, such as dramatic fluctuations in power generation power. Through these data, the behavior of new energy power generation under extreme conditions can be simulated and evaluated.

[0035] The environmental condition transition matrix describes the probability of the environmental condition transitioning from one state to another within a certain time period. For example, an element in the matrix represents the probability of transitioning from a state of "wind speed higher than 10 m / s" to a state of "wind speed lower than 3 m / s". This matrix is based on historical time-series environmental data. Each row of the transition matrix represents the current environmental state, and each column represents the next state. The value of each matrix element represents the probability of transitioning from the current environmental state to the next state. Specifically, extreme environmental conditions are divided into different state intervals. For example, wind speed can be divided into low wind speed, medium wind speed, and high wind speed intervals. Illumination intensity can be divided into weak light, normal light, and strong light intervals. The transition frequency between each environmental condition is calculated, i.e., the frequency of transitioning from a certain state to another state. For example, the number of times of transitioning from a low wind speed state to a medium wind speed state is counted. Based on the state transition frequency, the transition probability between each pair of states is calculated. These probability values constitute the environmental condition transition matrix.

[0036] Online calibration of transition probability refers to dynamically updating the transition probability in the environmental condition transition matrix when new data arrives. Specifically, based on time-series environmental condition data, the current state transition is counted. By comparing the actual state change with the predicted transition probability in the historical data, the transition probability in the environmental condition transition matrix is adjusted. With the continuous arrival of new data, the environmental condition transition matrix is continuously calibrated to improve its prediction accuracy for future environmental condition changes. The calibrated environmental condition transition matrix is the updated environmental condition transition model. This model can more accurately describe the dynamic rules of environmental state changes and provide higher accuracy for subsequent transition simulation of power generation environments.

[0037] Using the calibrated environmental condition transition model for transition simulation, the model predicts the possible environmental state at the next time based on the current environmental state and the transition probability in the transition matrix. The appearance and change of extreme environmental conditions are considered in the simulation process, so as to simulate how the new energy power generation system responds under different environmental conditions.

[0038] Further, based on the simulation period condition sequence, multi-threaded new energy power generation quantization simulation is performed, and time-series new energy power generation data is output. Previously, the method includes:

[0039] The wind energy time series power generation data, the light energy time series power generation data and the historical time series environmental working condition data of the new energy power station are called; the historical time series environmental working condition data is decomposed according to the power generation correlation attribute, and wind energy correlation environmental working condition data and light energy correlation environmental working condition data are obtained; a wind energy power generation prediction model is constructed based on the wind energy time series power generation data and the wind energy correlation environmental working condition data; a light energy power generation prediction model is constructed based on the light energy time series power generation data and the light energy correlation environmental working condition data; the wind energy power generation prediction model and the light energy power generation prediction model are isolated and connected in parallel, and a power addition engine is configured at the output end, so that the construction of a double-thread power generation prediction network is completed.

[0040] The wind energy time series power generation data is a data set reflecting the actual power generation of a wind power station in each time period, and this data is usually collected by hours or minutes, showing the power generation of a wind turbine at different times; the light energy time series power generation data is a similar time series data set, recording the actual power generation of a photovoltaic power station in each time period; the historical time series environmental working condition data includes wind speed, solar radiation, temperature, humidity, cloud cover and other environmental data, which reflects the changes of external meteorological conditions and directly affects the power generation capacity of wind energy and light energy.

[0041] Each type of power generation is closely related to specific environmental working conditions. In order to construct an accurate prediction model, the environmental working condition attributes related to power generation are first identified. For wind power generation, wind speed and wind direction are the most important environmental factors, and for light power generation, solar radiation intensity and cloud cover are the main influencing factors. Through data analysis methods such as correlation analysis and regression analysis, the influence of different environmental working condition attributes on wind and light power generation is quantified. According to this power generation correlation attribute, the historical time series environmental working condition data is decomposed. Specifically, for wind energy, environmental working condition data related to wind energy generation, such as wind speed and wind direction, is extracted to form wind energy correlation environmental working condition data; for light energy, environmental working condition data related to light energy generation, such as solar radiation and cloud cover, is extracted to form light energy correlation environmental working condition data.

[0042] Through the correlation of wind energy time series power generation data and wind energy correlation environmental working condition data, the relationship between environmental working conditions and wind energy power generation can be identified, and a prediction model can be constructed. Specifically, a suitable regression model, such as linear regression, random forest regression, support vector regression, etc., is selected for wind energy power generation prediction. The wind energy time series power generation data and the wind energy correlation environmental working condition data are used as features to train the model. The prediction performance of the model is evaluated through cross-validation, and the accuracy of the model can be evaluated using mean square error and mean absolute error. The constructed wind energy power generation prediction model can predict the wind energy power generation in a future period of time under given future environmental working conditions.

[0043] Through the association of the light energy time series power generation data and the light energy associated environmental working condition data, the relationship between the environmental working condition and the light energy power generation can be identified, and then a light energy power generation prediction model is constructed. This process is similar to the construction process of the wind energy power generation prediction model, and thus will not be described herein for the sake of brevity of the description.

[0044] The wind energy power generation prediction model and the light energy power generation prediction model are isolated and in parallel, which means that the wind energy power generation prediction model predicts the wind energy power generation based on the data of the wind speed and other environmental working conditions, while the light energy power generation prediction model predicts the light energy power generation based on the environmental data such as solar radiation, and the two are independent, and the prediction of the wind energy and the light energy is processed in parallel. After the output of the results of the two independent prediction models, they are combined, and the power summation engine will collect the predicted wind energy and light energy power generation from the two models and sum them up to obtain the total new energy power generation in the time period. Finally, the output of the dual-thread prediction network is a time series new energy total power generation prediction, which integrates the wind energy and light energy power generation and can provide a reference for power grid dispatching.

[0045] Further, the method for performing multi-thread new energy power generation simulation based on the simulated time period working condition sequence and outputting time series new energy power generation data comprises the following steps.

[0046] The simulated time period working condition sequence is decomposed according to the power generation associated attributes to obtain a light energy associated simulated working condition sequence and a wind energy associated simulated working condition sequence; the light energy associated simulated working condition sequence and the wind energy associated simulated working condition sequence are input into the dual-thread power generation prediction network to perform multi-thread new energy power generation prediction, and the time series new energy power generation data is output.

[0047] The simulated time period working condition data is decomposed according to the power generation associated attributes, specifically, the environmental data related to the wind energy power generation is extracted from the simulated time period working condition sequence, and the main environmental variables are the wind speed and the wind direction. Through the screening and extraction of these factors, a working condition sequence for wind energy power generation prediction, i.e., a wind energy associated simulated working condition sequence, is obtained. The environmental data related to the light energy power generation is extracted from the simulated time period working condition sequence, and the main environmental variables are the solar radiation intensity and the cloud coverage. The extraction of these data can obtain a working condition sequence for light energy power generation prediction, i.e., a light energy associated simulated working condition sequence.

[0048] The light energy associated simulation working condition sequence and the wind energy associated simulation working condition sequence are respectively taken as inputs and transmitted to the previously constructed double-thread power generation prediction network, which includes two sub-models running in parallel, one for predicting wind power generation and the other for predicting light power generation. The two prediction tasks are run in parallel to calculate the predicted values of wind power generation and light power generation, respectively. The predicted values of wind power generation and light power generation are added by a power addition engine to obtain comprehensive time-series new energy power generation data.

[0049] Further, the wind energy associated environmental working condition data includes time-series wind speed records and time-series wind direction records, and the light energy associated environmental working condition data includes time-series irradiance records and time-series cloud coverage records.

[0050] The time-series wind speed record represents the speed of the wind at a certain moment or time period, usually in meters per second. Wind speed directly affects the output power of a wind turbine. Generally, the greater the wind speed, the greater the output power of the wind turbine. The time-series wind direction record represents the direction of the wind, usually in degrees, indicating the direction from which the wind blows. Wind direction affects the efficiency of a wind turbine, as the orientation of the turbine blades towards the wind direction affects power generation. If the direction of the turbine blades is not adjusted to the optimal wind direction, the power generation efficiency may be reduced even if the wind speed is high.

[0051] The time-series irradiance record represents the solar radiation power received per unit area, usually in watts per square meter. Irradiance directly determines the power generation capacity of a photovoltaic panel. As irradiance increases, the power generation of a photovoltaic cell also increases. The time-series cloud coverage record represents the percentage of the sky covered by clouds, usually represented as a percentage from 0 to 100. The presence of clouds blocks solar radiation, reducing the intensity of light on photovoltaic panels and affecting power generation. When cloud coverage is high, solar radiation is usually reduced, and power generation is correspondingly reduced.

[0052] Further, the time-series new energy power generation data is dynamically injected into the power transmission grid twin model through the new energy grid-connected point, and after performing multi-cycle disturbed fluctuation confrontation fitting, a plurality of time-series operation steady-state anomaly records of a plurality of local power transmission devices are retrieved from a plurality of high-fidelity dynamic models. The method comprises:

[0053] The time-series new energy power generation data is converted into time-series standard power quality signals, wherein the time-series standard power quality signals include a voltage amplitude fluctuation sequence and an active power fluctuation sequence; after the time-series standard power quality signals are synchronized to the new energy grid-connected point through a real-time simulation interface device, the power transmission grid twin model is driven to perform time-series simulation operation of the multiple high-fidelity dynamic models, and multiple time-series operation steady-state characteristic parameters are output; the multiple structured abnormality criteria of the multiple local power transmission devices are used to map and traverse the multiple time-series operation steady-state characteristic parameters, and multiple time-series operation steady-state abnormality records are output.

[0054] Power quality is an important factor affecting the stability of the power grid. In order to further analyze the stability of the power grid, the time-series new energy power generation data is converted into time-series standard power quality signals. The power fluctuation of new energy will cause the voltage fluctuation of the power grid, and the power fluctuation of wind energy and light energy may lead to the instability of the voltage of the power grid. The voltage amplitude fluctuation sequence reflects the change of the voltage amplitude in the power grid. The active power fluctuation is the power fluctuation of the power grid caused by the change of the power generation of new energy. The change of wind speed and solar irradiance will lead to the fluctuation of power generation, and further affect the active power of the power grid.

[0055] The time-series standard power quality signals are synchronized to the actual operating environment of the power grid through a real-time simulation interface device. The real-time simulation interface device is connected to the power grid system in real time and interacts with the state of the new energy grid-connected point. The power transmission grid twin model is a digital and dynamic simulation model that accurately reflects the physical and electrical characteristics of the power grid. By inputting the time-series standard power quality signals into the power transmission grid twin model, the real-time operation of the power grid can be simulated in the simulation environment. In the power transmission grid twin model, each high-fidelity dynamic model can simulate the response of different components in the power grid and the influence of the new energy grid-connected point on the entire power grid. By simulating the power grid through multiple high-fidelity dynamic models, it means simulating the voltage and power fluctuations at different time points and analyzing how these fluctuations affect the stability of the power grid. The simulation results are multiple time-series operation steady-state characteristic parameters, including voltage amplitude, active power, power grid frequency, current fluctuation, etc. These parameters reflect the operating state of the power grid in a specific period.

[0056] The structured abnormality criterion is a set of rules or standards for determining whether an abnormality occurs in the power grid operation. Each criterion is defined based on the physical characteristics of the power grid operation and the power quality requirements. For example, an abnormality is determined when the voltage variation amplitude exceeds a certain threshold, an abnormality is determined when the active power fluctuation is too large or exceeds the normal range, and an abnormality is determined when the power grid frequency or current deviation exceeds a certain range. By comparing multiple time-series operation steady-state characteristic parameters with the structured abnormality criterion one by one, the abnormal conditions that may exist at different time points of the power grid can be identified. Finally, multiple time-series operation steady-state abnormality records are output, which describe the abnormal conditions that occur during the operation of the power grid in detail. Each abnormality record contains information such as the time point of the abnormality occurrence, the type of abnormality, and the severity of the abnormality.

[0057] Further, after constructing multiple node instability feature vectors based on the multiple time-series operation steady-state abnormality records, performing grid compensation strategy matching based on the multiple node instability feature vectors, and outputting multiple grid stability compensation strategies, the method comprises:

[0058] The first voltage maximum deviation rate, the first voltage change rate extreme value, the first dominant oscillation frequency, and the first damping ratio of the first time-series operation steady-state abnormality record are calculated. The first voltage maximum deviation rate, the first voltage change rate extreme value, the first dominant oscillation frequency, and the first damping ratio are structurally combined to generate a first node instability feature vector. The device type of the first local power transmission device is used as a matching constraint, and the first node instability feature vector is used to traverse the strategy knowledge base for historical case matching to obtain a first candidate strategy set. Grid compensation strategy matching is performed based on the multiple node instability feature vectors, and multiple candidate strategy sets are output. After combining and enumerating the multiple candidate strategy sets, the power grid stability is verified cooperatively in the power transmission grid twin model, and the multiple grid stability compensation strategies are selected and output.

[0059] The first voltage maximum deviation rate refers to the maximum degree of voltage change within a certain period of time, which represents the amplitude of voltage change. A larger deviation means that the power grid is at risk of instability. Through the first time-series operation steady-state abnormality record, the maximum amplitude of voltage fluctuation is found, and then divided by the reference value or average value of the voltage to obtain the first voltage maximum deviation rate. The first voltage change rate extreme value refers to the maximum change speed of voltage per unit time within a certain period of time, i.e., the maximum rate of voltage fluctuation. According to the time-series voltage change data, the rate of voltage change is calculated, and the maximum value is found. Rapid voltage change may cause damage to power grid equipment or affect power quality.

[0060] The first dominant oscillation frequency refers to the frequency of the main oscillation mode in the power grid, which is usually determined by the electrical characteristics of the power grid. The connection of new energy such as wind energy and solar energy will affect the oscillation characteristics of the power grid, which may cause oscillation at a specific frequency. Through time series voltage or current data, the dominant frequency of the power grid is extracted using frequency spectrum analysis method. The first damping ratio is a parameter to measure the damping of power grid oscillation. The smaller the damping ratio, the slower the oscillation damping, and the more likely the power grid will be unstable. A lower damping ratio means that the power grid will oscillate for a long time when disturbed. By analyzing the oscillation mode, the speed of oscillation damping is calculated to obtain the damping ratio.

[0061] The structured combination means that these features are combined into a vector according to certain rules to form a unified first node instability feature vector. This vector contains multiple key stability indicators of the power grid at the node.

[0062] In the power grid, different types of local power transmission devices have different effects on the stability of the power grid, and each device has different response characteristics. Therefore, when matching strategies, the type of the first local power transmission device currently analyzed is used as a matching constraint. The strategy knowledge base is a database that stores historical operation data and compensation strategies of the power grid, including the feature vectors and compensation strategies taken when the power grid was unstable in the past. These historical cases are helpful for providing reference for the current power grid instability. By traversing the strategy knowledge base, the current first node instability feature vector is matched with historical cases according to the current first node instability feature vector, and historical cases with similar feature vectors are found. The compensation strategy taken at that time is analyzed, and the matching result is a first candidate strategy set, which contains multiple possible compensation strategies. These strategies can be different types of power grid stability compensation schemes, such as adjusting voltage, power factor, damping control, etc.

[0063] By analogy, the above analysis process is extended to other nodes of the power grid. For each node instability feature vector, historical case matching is performed according to the existing strategy knowledge base to find a compensation strategy suitable for the node, and multiple candidate strategy sets are output.

[0064] The multiple candidate strategy sets are combined and enumerated to form a strategy set containing all possible strategy combinations. During the combination process, various possible coordination modes need to be considered, for example, some strategies may only be effective under certain grid operating conditions, so through combination and enumeration, it is ensured that no effective compensation mode is missed. After the combination of the candidate strategy set, these combined strategies are injected into the power grid twin model for simulation. The power grid twin model simulates the real operating environment of the grid and can perform high-fidelity simulation testing on different strategy combinations. During the simulation process, grid stability coordination verification is performed, that is, the influence of each strategy combination on various aspects of the grid (such as voltage, frequency, power flow, etc.) is tested. The purpose of verification is to ensure that the selected strategy combination can coordinately improve the overall stability of the grid and avoid local instability or cause other unnecessary effects. Based on the simulation results, the most effective multiple grid-connected stability compensation strategies are selected. These strategies perform best in simulation and can effectively solve the stability problem of the grid.

[0065] Further, the multiple grid-connected stability compensation strategies are cooperatively deployed to the multiple local power transmission devices to perform dynamic stability compensation of the local power transmission network.

[0066] The selected multiple grid-connected stability compensation strategies are cooperatively deployed to the corresponding multiple local power transmission devices, including substations, transmission lines, circuit breakers, regulating devices, etc., which are key components of grid stability. Cooperative deployment means that different grid-connected stability compensation strategies are implemented simultaneously at different grid nodes to ensure comprehensive management of the stability of the entire grid. After deployment, these grid-connected stability compensation strategies will adjust the dynamic response of the grid in real time, for example, by adjusting the voltage of the transformer, switching different loads or adjusting the power factor, optimizing the operating state of the grid. Dynamic stability compensation ensures that the grid can quickly recover to a stable state when facing new energy fluctuations, load changes or external disturbances.

[0067] Embodiment two, based on the same inventive concept as the new energy access stability evaluation method under the grid digital twin in the preceding embodiments, as shown in Figure 2 The present application provides a new energy access stability evaluation system under the grid digital twin, which comprises:

[0068] The working condition data retrieval module 10 is configured to retrieve time-series environmental working condition data satisfying a preset time period with the geographical coordinates of the new energy power station as a spatial constraint, wherein the new energy power station is connected to the local power transmission network; the power generation environment transition simulation module 20 is configured to perform power generation environment transition simulation based on the time-series environmental working condition data, and output a simulation period working condition sequence; the power generation quantization simulation module 30 is configured to perform multi-threaded new energy power generation quantization simulation based on the simulation period working condition sequence, and output time-series new energy power generation data; the abnormal record calling module 40 is configured to inject the time-series new energy power generation data into a power transmission grid twin model through a new energy grid connection point, perform multi-cycle disturbed fluctuation confrontation fitting, and then call time-series operation steady-state abnormal records of a plurality of local power transmission devices from a plurality of high-fidelity dynamic models; the strategy matching module 50 is configured to construct a plurality of node instability feature vectors based on the plurality of time-series operation steady-state abnormal records, perform grid compensation strategy matching based on the plurality of node instability feature vectors, and output a plurality of grid stability compensation strategies.

[0069] Further, the abnormal record calling module 40 is configured to perform the following operation steps:

[0070] Call the physical topology and grid device information of the local power transmission network; retrieve and call a plurality of high-fidelity dynamic models of a plurality of local power transmission devices from a grid equipment model according to the grid device information; connect the plurality of high-fidelity dynamic models according to the physical topology, and construct a power transmission grid twin model, wherein the power transmission grid twin model is provided with the new energy grid connection point.

[0071] Further, the power generation environment transition simulation module 20 is configured to perform the following operation steps:

[0072] Obtain a plurality of extreme environmental working conditions by aggregating the time-series environmental working condition data; construct an environmental working condition transition matrix based on the plurality of extreme environmental working conditions; perform online calibration of transition probabilities of the environmental working condition transition matrix using the time-series environmental working condition data, and output an environmental working condition transition model; drive the environmental working condition transition model to perform new energy power generation environment transition simulation, and output the simulation period working condition sequence.

[0073] Further, the power generation quantization simulation module 30 is configured to perform the following operation steps:

[0074] retrieve wind energy time series power generation data, light energy time series power generation data and historical time series environmental working condition data of the new energy power station; decompose the historical time series environmental working condition data according to power generation correlation attributes to obtain wind energy correlation environmental working condition data and light energy correlation environmental working condition data; construct a wind energy power generation prediction model based on the wind energy time series power generation data and the wind energy correlation environmental working condition data; construct a light energy power generation prediction model based on the light energy time series power generation data and the light energy correlation environmental working condition data; isolate and parallel the wind energy power generation prediction model and the light energy power generation prediction model, and configure a power addition engine at an output end to complete construction of a double-thread power generation prediction network.

[0075] Further, the power generation simulation module 30 is configured to perform the following operation steps:

[0076] decompose the simulated period working condition sequence according to power generation correlation attributes to obtain light energy correlation simulated working condition sequence and wind energy correlation simulated working condition sequence; input the light energy correlation simulated working condition sequence and the wind energy correlation simulated working condition sequence into the double-thread power generation prediction network to perform multi-thread new energy power generation prediction and output the time series new energy power generation data.

[0077] Further, the wind energy correlation environmental working condition data includes time series wind speed records and time series wind direction records, and the light energy correlation environmental working condition data includes time series irradiance records and time series cloud cover records.

[0078] Further, the abnormal record retrieval module 40 is configured to perform the following operation steps:

[0079] convert the time series new energy power generation data into time series standard power quality signals, wherein the time series standard power quality signals include voltage amplitude fluctuation sequence and active power fluctuation sequence; after synchronizing the time series standard power quality signals to the new energy grid-connected point through a real-time simulation interface device, drive the power transmission grid twin model to perform time series simulation operation of the multiple high-fidelity dynamic models and output multiple time series operation steady-state characteristic parameters; and map and traverse the multiple time series operation steady-state characteristic parameters using multiple structured abnormality criteria of the multiple local power transmission devices to output the multiple time series operation steady-state abnormal records.

[0080] Further, the strategy matching module 50 is configured to perform the following operation steps:

[0081] The first voltage maximum deviation rate, the first voltage change rate extreme value, the first dominant oscillation frequency and the first damping ratio of the first time sequence operation steady-state abnormal record are calculated; the first voltage maximum deviation rate, the first voltage change rate extreme value, the first dominant oscillation frequency and the first damping ratio are structured and combined to generate a first node instability feature vector; the first node instability feature vector is used to traverse a strategy knowledge base to perform historical case matching with a device type of the first local power transmission device as a matching constraint, and a first candidate strategy set is obtained; grid-connected compensation strategy matching is performed based on the plurality of node instability feature vectors, and a plurality of candidate strategy sets are output; after the plurality of candidate strategy sets are combined and enumerated, the power transmission grid twin model is injected to perform power grid stability cooperative verification, and the plurality of grid-connected stability compensation strategies are screened and output.

[0082] Further, the plurality of grid-connected stability compensation strategies are cooperatively deployed to the plurality of local power transmission devices to perform dynamic stability compensation of the local power transmission network.

[0083] The above is only a preferred embodiment of the present application, and does not limit the present application in any form. Although the present application has been disclosed as above with a preferred embodiment, it is not intended to limit the present application. Any person skilled in the art can make some changes or modifications to the above disclosed technical content to obtain equivalent embodiments with equivalent changes, without departing from the technical solution of the present application. Any modification, equivalent change and modification of the above embodiments based on the technical essence of the present application are still within the scope of the technical solution of the present application.

Claims

1. A method for assessing the stability of renewable energy access under a power grid digital twin, characterized in that, The method includes: Using the geographical coordinates of the new energy power station as a spatial constraint, time-series environmental condition data that meet a preset time period are retrieved, wherein the new energy power station is connected to the local power transmission network. Based on the aforementioned time-series environmental operating condition data, a power generation environment transition simulation is performed, and a simulation period operating condition sequence is output. Based on the simulated time period operating condition sequence, perform multi-threaded quantitative simulation of new energy power generation and output time-series new energy power generation data; By dynamically injecting the time-series new energy power generation data into the power grid twin model through the new energy grid connection point, and performing multi-cycle disturbance fluctuation countermeasure fitting, multiple time-series steady-state anomaly records of multiple local power transmission devices are retrieved from multiple high-fidelity dynamic models. After constructing multiple node instability feature vectors based on the multiple time-series steady-state anomaly records, grid connection compensation strategy matching is performed based on the multiple node instability feature vectors to output multiple grid connection stability compensation strategies. Based on the simulated time period operating condition sequence, a multi-threaded quantitative simulation of new energy power generation is performed, and time-series new energy power generation data is output. Prior to this, the method includes: Retrieve the wind power time-series power generation data, solar power time-series power generation data, and historical time-series environmental condition data of the new energy power station; Based on the power generation correlation attributes, the historical time series environmental condition data are decomposed to obtain wind energy-related environmental condition data and solar energy-related environmental condition data. Based on the wind energy time-series power generation data and wind energy-related environmental condition data, a wind energy power generation prediction model is constructed. Based on the aforementioned solar energy time-series power generation data and solar energy-related environmental condition data, a solar energy power generation prediction model is constructed. By isolating and paralleling the wind power generation prediction model and the solar power generation prediction model, and configuring a power summing engine at the output end, the construction of a dual-thread power generation prediction network is completed. The method involves performing multi-threaded quantitative simulations of renewable energy generation based on the simulated time period operating condition sequence, and outputting time-series renewable energy generation data. Based on the power generation correlation attributes, the simulated time period operating condition sequence is decomposed to obtain the solar energy correlation simulated operating condition sequence and the wind energy correlation simulated operating condition sequence; The solar energy-related simulation operating condition sequence and the wind energy-related simulation operating condition sequence are input into the dual-thread power generation prediction network to perform multi-threaded new energy power generation prediction and output the time-series new energy power generation data. The wind energy-related environmental condition data includes time-series wind speed records and time-series wind direction records, and the solar energy-related environmental condition data includes time-series irradiance records and time-series cloud coverage records.

2. The method for assessing the stability of new energy access under a power grid digital twin as described in claim 1, characterized in that, The method further includes: Retrieve the physical topology and grid device information of the local power transmission network; Based on the power grid device information, multiple high-fidelity dynamic models of multiple local power transmission devices are retrieved and called from the power grid equipment model; Based on the physical topology, the multiple high-fidelity dynamic models are connected to construct a power transmission grid twin model, wherein the power transmission grid twin model includes the new energy grid connection point.

3. The method for assessing the stability of new energy access under a power grid digital twin as described in claim 2, characterized in that, The method includes performing a power generation environment transition simulation based on the aforementioned time-series environmental operating condition data, and outputting a simulation period operating condition sequence. By aggregating the time-series environmental operating conditions data, various extreme environmental operating conditions are obtained; An environmental condition transition matrix is ​​constructed based on the aforementioned multiple extreme environmental conditions; The transition probability of the environmental condition transition matrix is ​​calibrated online using the time-series environmental condition data, and the environmental condition transition model is output. The environmental condition transition model is driven to simulate the environmental transition of new energy power generation, and the operating condition sequence of the simulation period is output.

4. The method for assessing the stability of new energy access under a power grid digital twin as described in claim 2, characterized in that, The method involves dynamically injecting the time-series renewable energy generation data into the power grid twin model through renewable energy grid connection points, performing multi-cycle disturbance fluctuation adversarial fitting, and then retrieving multiple time-series steady-state anomaly records of multiple local power transmission devices from multiple high-fidelity dynamic models. The time-series renewable energy generation data is converted into a time-series standard power quality signal, wherein the time-series standard power quality signal includes a voltage amplitude fluctuation sequence and an active power fluctuation sequence. After synchronizing the time-series standard power quality signal to the new energy grid connection point through the real-time simulation interface device, the power transmission grid twin model is driven to perform time-series simulation of the multiple high-fidelity dynamic models, and outputs multiple time-series steady-state characteristic parameters. Using multiple structured anomaly criteria of the multiple local power transmission devices, the multiple time-series steady-state characteristic parameters are mapped and traversed to output the multiple time-series steady-state anomaly records.

5. The method for assessing the stability of new energy access under a power grid digital twin as described in claim 4, characterized in that, After constructing multiple node instability feature vectors based on the multiple time-series steady-state anomaly records, grid connection compensation strategy matching is performed based on the multiple node instability feature vectors to output multiple grid connection stability compensation strategies. The method includes: Calculate the first maximum voltage deviation rate, the first extreme value of the first voltage change rate, the first dominant oscillation frequency, and the first damping ratio of the first time-series steady-state anomaly record; The first node instability feature vector is generated by structurally combining the first voltage maximum deviation rate, the first voltage change rate extreme value, the first dominant oscillation frequency, and the first damping ratio. Using the device type of the first local power transmission device as a matching constraint, historical case matching is performed by traversing the strategy knowledge base using the instability feature vector of the first node to obtain the first candidate strategy set. Based on the instability feature vectors of the multiple nodes, grid connection compensation strategy matching is performed by analogy, and multiple candidate strategy sets are output; After combining and enumerating the multiple candidate strategy sets, they are injected into the power grid twin model for grid stability collaborative verification, and the multiple grid connection stability compensation strategies are screened and output.

6. The method for assessing the stability of new energy access under a power grid digital twin as described in claim 5, characterized in that, The multiple grid-connected stability compensation strategies are deployed collaboratively to the multiple local power transmission devices to perform dynamic stability compensation for the local power transmission network.

7. A new energy access stability assessment system under a power grid digital twin, characterized in that, The system is used to implement the method for assessing the stability of new energy access under a power grid digital twin as described in any one of claims 1-6, the system comprising: The operating condition data retrieval module is used to retrieve time-series environmental operating condition data that meets a preset time period, with the geographical coordinates of the new energy power station as the spatial constraint. The new energy power station is connected to the local power transmission network. The power generation environment transition simulation module is used to perform power generation environment transition simulation based on the time-series environmental condition data and output the operating condition sequence of the simulation period. The power generation quantification simulation module is used to perform multi-threaded new energy power generation quantification simulation based on the operating condition sequence of the simulation period, and output time-series new energy power generation data; The abnormal record retrieval module is used to dynamically inject the time-series new energy power generation data into the power grid twin model through the new energy grid connection point, perform multi-cycle disturbance fluctuation adversarial fitting, and retrieve multiple time-series steady-state abnormal records of multiple local power transmission devices from multiple high-fidelity dynamic models. The strategy matching module is used to construct multiple node instability feature vectors based on the multiple time-series steady-state anomaly records, and then perform grid connection compensation strategy matching based on the multiple node instability feature vectors to output multiple grid connection stability compensation strategies.

Images