Rapid frequency evaluation method, system and equipment based on power grid feature set and medium

Abstract

The invention relates to the technical field of power system automation, in particular to a rapid frequency evaluation method, system and device based on a power grid feature set and a medium, and the method comprises the steps: obtaining power grid historical data including an operation mode, a source-grid-load-storage commissioning capacity proportion and a new energy access proportion, and constructing the power system feature set; analyzing historical data of a power vacancy event in the feature set, and calculating an equivalent inertia time constant of the power grid when power vacancy occurs; when disturbance occurs, acquiring the characteristic data of the power grid at the moment, matching the characteristic data with the data in the characteristic set, and approximately estimating an equivalent inertia time constant during disturbance; and in combination with the real-time data of the power grid and the matched equivalent inertia time constant, evaluating the frequency change rate and the frequency lowest point in the inertia response time period after disturbance. The frequency change trend of the system can be quickly evaluated based on historical data, real-time state and external disturbance factors of the power grid under the condition of not depending on a complex simulation model.

Description

Technical Field

[0001] This invention relates to the field of power system automation technology, and in particular to a method, system, device and medium for rapid frequency assessment based on power grid feature sets. Background Technology

[0002] With the continuous expansion and increasing complexity of power systems, the stability and security of power grids have become a key research focus in the field of power engineering. Frequency fluctuations in power systems are a crucial indicator of grid stability; abnormal frequency changes can lead to serious consequences such as equipment damage, generator disconnection, and even large-scale power outages. As the proportion of renewable energy sources continues to increase, and power systems become increasingly larger and more complex, the inertia of the power grid decreases significantly, greatly increasing the amplitude of frequency fluctuations and the difficulty of frequency regulation.

[0003] In related research, the evaluation method of power system frequency has attracted widespread attention. With the increasing complexity of power systems, especially the gradual increase in the proportion of new energy sources, traditional frequency evaluation methods are usually based on the dynamic model and numerical simulation of the power grid. They face challenges such as large computational load and poor real-time performance. In the event of power deficit in the actual power grid, they cannot quickly predict the frequency change so that dispatchers can make reasonable responses.

[0004] The information disclosed in this background section is intended only to enhance the understanding of the general background of the invention and should not be construed as an admission or in any way implying that the information constitutes prior art known to those skilled in the art. Summary of the Invention

[0005] This invention provides a method, system, device, and medium for rapid frequency assessment based on power grid feature sets, thereby effectively solving the problems in the background art.

[0006] To achieve the above objectives, the technical solution adopted by this invention is: a rapid frequency assessment method based on power grid feature sets, comprising the following steps:

[0007] Acquire historical power grid data, including operating modes, the proportion of operational capacity of power generation, grid, load, and storage, and the proportion of renewable energy access, and construct a power system feature set;

[0008] Analyze the historical data of power deficit events in the feature set to calculate the equivalent inertial time constant of the power grid when a power deficit occurs;

[0009] When a disturbance occurs, the characteristic data of the power grid at that time is acquired and matched with the data in the feature set. Based on the equivalent inertial time constant of the power grid when a power deficit occurs, the equivalent inertial time constant at the time of the disturbance is approximately estimated.

[0010] By combining real-time power grid data and the matched equivalent inertial time constant, the rate of frequency change and the minimum frequency point within the inertial response time period after the disturbance are evaluated.

[0011] Furthermore, the acquisition of historical power grid data, including operating modes, the proportion of power generation-grid-load-storage capacity in operation, and the proportion of new energy access, and the construction of a power system feature set, includes the following steps:

[0012] Taking into account climate conditions and electricity demand, the characteristics of each season are determined based on typical climate conditions and changes in electricity load levels in different seasons.

[0013] For each season, based on the proportion of renewable energy output, load level, and grid regulation capacity, a typical combination of operating modes is constructed, with different levels of operating modes for renewable energy power generation and load level.

[0014] Considering different combinations of the proportion of power generation, grid, load and storage capacity, between the maximum and minimum output of new energy power generation, combined with different load demand and grid transmission capacity, and with different combinations of energy storage capacity and load demand, several capacity proportion levels are divided.

[0015] Construct a set of typical operating modes based on the maximum load demand ratio of the power system;

[0016] Seasonal load levels, the proportion of operational capacity of source-grid-load-storage, and the maximum load demand ratio were selected as data features, and the K-means algorithm was used to cluster the data.

[0017] Furthermore, the construction of typical operating mode combinations for each season, based on the proportion of renewable energy output, load level, and grid regulation capacity, includes different tiers of operating mode combinations for renewable energy generation and load levels, including:

[0018] Under the maximum operating mode of new energy power generation, the synchronous machine should be started as few times as possible and no rotational standby should be left as much as possible;

[0019] Under the minimum operating mode of new energy power generation, the synchronous generator should be started as many times as possible while maintaining minimum technical output, thereby increasing inertia and frequency regulation capability.

[0020] Furthermore, the construction of typical operating mode combinations of new energy power generation and different load levels also includes:

[0021] After eliminating impossible extreme combinations, several seasonal fluctuations are formed. Based on their operating characteristics, they are recorded as seasonal sequences in order of frequency response characteristics from best to worst.

[0022] Furthermore, considering different combinations of the proportions of power generation, grid, load, and storage capacity, and dividing the power generation between the maximum and minimum output of new energy generation into several capacity proportion levels based on different load demands and grid transmission capabilities, and combining different energy storage capacities and load demands, it also includes:

[0023] Extreme combinations that do not conform to actual operating conditions are eliminated to form multiple typical commissioning modes; based on each working characteristic, the modes are sorted from best to worst according to the system's frequency response capability, scheduling capability, and stability to form a commissioning percentage sequence.

[0024] Furthermore, the construction of a typical operating mode set based on the maximum load demand ratio of the power system includes:

[0025] The ranges for high-load operation, medium-load operation, and low-load operation are defined based on the maximum load demand ratio.

[0026] A set of typical operating modes is constructed based on the different defined ranges.

[0027] Furthermore, the selection of seasonal load levels, the proportion of source-grid-load-storage operational capacity, and the maximum load-demand ratio as data features, and the use of the K-means algorithm to cluster the data, includes the following steps:

[0028] Each operating mode is represented as a vector feature x. i The vector features include:

[0029] x i = [Seasonal load level, ratio of operating capacity of power generation, grid, load and storage, and maximum load demand ratio];

[0030] Select n cluster centers:

[0031] C = {c j |j=1,2,3...n};

[0032] In the formula, C represents the set of cluster centers, and for each method x i Calculate its distance to all cluster centers c j The Euclidean distance;

[0033]

[0034] In the formula, x i,k and c j,k The operating mode is x respectively i and cluster center c j The value for the k-th feature is repeatedly calculated until the clustering error converges. Each cluster center represents a typical operating feature, and the feature clusters C1, C2, ..., C3 in the output feature set are calculated. nUsed for future fault matching.

[0035] Furthermore, the step of analyzing historical data on power deficit events in the feature set and calculating the equivalent inertial time constant of the power grid when a power deficit occurs includes the following steps:

[0036] Collect historical data including frequency response data, power deficit data, and system fundamental parameters;

[0037] The frequency change rate is extracted from the frequency response data and used as a matching feature to match the data in the feature set.

[0038] The equivalent inertial time constant is calculated based on the power deficit data.

[0039] The obtained equivalent inertial time constant is compared with the historical typical value. If the equivalent inertial time constant deviates from the historical typical value by more than a set threshold, the accuracy of the data is checked.

[0040] If the calculated equivalent inertial time constant deviates from the actual operating characteristics, adjustments should be made based on the system's fundamental parameters.

[0041] Furthermore, among the collected historical data, the frequency response data, power deficit data, and system basic parameters are:

[0042] The frequency response data includes frequency change curves, collects system frequencies f(t) recorded in historical events, and covers the entire process from the start of the event to the gradual recovery of the frequency; the data is output in time series format.

[0043] If the power deficit data can be directly measured, the power change is extracted directly from the faulty equipment; if it cannot be directly measured, it is estimated by the difference between the total power generation or total load power of the system before and after the event; if there are multiple fault points in the system, the total power deficit is comprehensively assessed.

[0044] The system's basic parameters include the rated power f0 and the system's reference power S. base .

[0045] Further, extracting the frequency change rate from the frequency response data includes:

[0046] Calculate the rate of change of frequency

[0047]

[0048] In the formula: f(t) i+1 ) and f(t) i ) represents the frequency values ​​at two adjacent time points; t i+1 -t iIt is the time interval between two points in time;

[0049] Select a set inertial response time period, and from the frequency change rate Extract the initial segment data to calculate the system inertia;

[0050] The frequency change rate was analyzed using a multi-point average smoothing algorithm. Noise reduction:

[0051]

[0052] Further, the calculation of the equivalent inertial time constant based on the power deficit data includes:

[0053] Based on the power-frequency dynamic response equation, the equivalent inertial time constant H at this point is calculated. For a single time point:

[0054]

[0055] In the formula, ΔP is the power deficit detected in real time, and f0 is the rated frequency of the system;

[0056] For multi-point data, calculate the inertial time constant H for each point. i And calculate the average:

[0057]

[0058] Furthermore, when the disturbance occurs, the characteristic data of the power grid at that time is acquired and matched with the data in the feature set. Based on the equivalent inertial time constant of the power grid when a power deficit occurs, the equivalent inertial time constant at the time of the disturbance is approximately estimated, including the following steps:

[0059] When a fault occurs, acquire the characteristic data of the power grid at that time and organize it into a multi-dimensional feature vector;

[0060] The distance between the multidimensional feature vector and each cluster center in the feature set data is calculated using Euclidean distance.

[0061] Traverse all the cluster centers and find the cluster center with the smallest distance to the multidimensional feature vector;

[0062] Obtain the equivalent inertial time constant of the cluster center as an approximate estimate.

[0063] Furthermore, the rate of change of the frequency is:

[0064]

[0065] In the formula, ΔP is the power deficit detected in real time, and H eqLet f0 be the system's equivalent inertial time constant, and f0 be the system's rated frequency.

[0066] Lowest frequency f min for:

[0067]

[0068] In the formula, Δt represents the time required for the system frequency f0 to drop from its initial value to its lowest point; if the system frequency response characteristics are stable, then an empirical value of Δt is directly selected based on historical data of similar disturbances.

[0069] The present invention also includes a fast frequency assessment system based on a power grid feature set, using the method described above, the system comprising:

[0070] The feature set construction unit is used to acquire historical power grid data, including operating mode, the proportion of power generation-grid-load-storage capacity in operation, and the proportion of new energy access, and to construct a power system feature set.

[0071] The analysis unit is used to analyze historical data of power deficit events in the feature set and calculate the equivalent inertial time constant of the power grid when a power deficit occurs.

[0072] The matching unit is used to acquire the power grid characteristic data when a disturbance occurs, and match it with the data in the feature set. Based on the equivalent inertial time constant of the power grid when the power deficit occurs, the equivalent inertial time constant during the disturbance is approximately estimated.

[0073] The evaluation unit is used to evaluate the rate of frequency change and the minimum frequency point within the inertial response time period after the disturbance by combining real-time power grid data and the matched equivalent inertial time constant.

[0074] The present invention also includes a computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the method as described above.

[0075] The present invention also includes a storage medium having a computer program stored thereon, which, when executed by a processor, implements the method as described above.

[0076] The beneficial effects of this invention are as follows: By comparing the characteristics of the system at the moment of disturbance with a power grid feature set composed of massive power grid data, and using a rapid frequency assessment method based on the power grid feature set, the frequency change trend of the system can be quickly assessed based on historical power grid data, real-time status, and external disturbance factors without relying on complex simulation models. This method not only improves the efficiency and accuracy of the assessment but also provides necessary decision-making basis for emergency response, helps the power system cope with sudden frequency fluctuations and disturbances, further ensures the safe and stable operation of the power grid, and has guiding significance for current research on rapid frequency prediction. Attached Figure Description

[0077] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0078] Figure 1 This is a flowchart of the method in Example 1;

[0079] Figure 2 This is a schematic diagram of the system structure in Example 1;

[0080] Figure 3 This is a flowchart of the method in Example 2;

[0081] Figure 4 This refers to the combination of operating modes for new energy output and load levels under different seasons in Example 2;

[0082] Figure 5 This is a schematic diagram of the structure of the computer device of the present invention. Detailed Implementation

[0083] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0084] Example 1:

[0085] like Figure 1 As shown: A fast frequency assessment method based on power grid feature sets includes the following steps:

[0086] Acquire historical power grid data, including operating modes, the proportion of operational capacity of power generation, grid, load, and storage, and the proportion of renewable energy access, and construct a power system feature set;

[0087] Historical data of power deficit events in the feature set are analyzed to calculate the equivalent inertial time constant of the power grid when a power deficit occurs.

[0088] When a disturbance occurs, the characteristic data of the power grid at that time is acquired and matched with the data in the characteristic set. Based on the equivalent inertial time constant of the power grid when the power deficit occurs, the equivalent inertial time constant at the time of the disturbance is approximately estimated.

[0089] By combining real-time power grid data and the matched equivalent inertial time constant, the rate of frequency change and the minimum frequency point within the inertial response time period after the disturbance are evaluated.

[0090] By comparing the system's characteristics at the moment of disturbance with a power grid feature set composed of massive power grid data, a rapid frequency assessment method based on this feature set can quickly assess the system's frequency change trend without relying on complex simulation models. This method is based on historical power grid data, real-time status, and external disturbance factors. It not only improves the efficiency and accuracy of the assessment but also provides necessary decision-making basis for emergency response, helping the power system cope with sudden frequency fluctuations and disturbances, further ensuring the safe and stable operation of the power grid, and providing guidance for current research on rapid frequency prediction.

[0091] In this embodiment, historical power grid data, including operating modes, the proportion of power generation-grid-load-storage capacity in operation, and the proportion of new energy access, are acquired, and a power system feature set is constructed, including the following steps:

[0092] Taking into account climate conditions and electricity demand, the characteristics of each season are determined based on typical climate conditions and changes in electricity load levels in different seasons.

[0093] For each season, based on the proportion of renewable energy output, load level, and grid regulation capacity, a typical combination of operating modes is constructed, with different levels of operating modes for renewable energy power generation and load level.

[0094] Considering different combinations of the proportions of source, grid, load and storage capacity, between the maximum and minimum output of new energy power generation, combined with different load demands and grid transmission capacity, and with different combinations of energy storage capacity and load demand, several capacity proportion levels are divided.

[0095] Construct a set of typical operating modes based on the maximum load demand ratio of the power system;

[0096] Seasonal load levels, the proportion of operational capacity of source-grid-load-storage, and the maximum load demand ratio were selected as data features, and the K-means algorithm was used to cluster the data.

[0097] For each season, based on the proportion of renewable energy output, load level, and grid regulation capacity, typical operating mode combinations are constructed, with different tiers for renewable energy generation and load levels, including:

[0098] Under the maximum operating mode of new energy power generation, the synchronous machine should be started as few times as possible and no rotational standby should be left as much as possible;

[0099] Under the minimum operating mode of new energy power generation, the synchronous generator should be started as many times as possible while maintaining minimum technical output, thereby increasing inertia and frequency regulation capability.

[0100] Among them, constructing typical operating mode combinations of new energy power generation and different load levels also includes:

[0101] After eliminating impossible extreme combinations, several seasonal fluctuations are formed. Based on their operating characteristics, they are recorded as seasonal sequences in order of frequency response characteristics from best to worst.

[0102] Considering different combinations of power generation, grid, load, and storage capacity ratios, and taking into account different load demands and grid transmission capabilities, as well as different combinations of energy storage capacity and load demand, several capacity ratio levels are defined between the maximum and minimum output of new energy power generation, including:

[0103] Extreme combinations that do not conform to actual operating conditions are eliminated to form multiple typical commissioning modes; based on each working characteristic, the modes are sorted from best to worst according to the system's frequency response capability, scheduling capability, and stability to form a commissioning percentage sequence.

[0104] In this embodiment, a set of typical operating modes is constructed based on the maximum load demand ratio of the power system, including:

[0105] The classification of high-load, medium-load, and low-load operating modes is defined based on the maximum load demand ratio.

[0106] Construct a set of typical operating modes based on different division ranges.

[0107] The data features selected include seasonal load levels, the proportion of operational capacity of the power generation, grid, load, and storage systems, and the maximum load-demand ratio. The K-means algorithm is then used to cluster the data, including the following steps:

[0108] Each operating mode is represented as a vector feature x. i Vector features include:

[0109] x i = [Seasonal load level, ratio of operating capacity of power generation, grid, load and storage, and maximum load demand ratio];

[0110] Select n cluster centers:

[0111] C = {c j |j=1,2,3...n};

[0112] In the formula, C represents the set of cluster centers, and for each method x i Calculate its distance to all cluster centers c j The Euclidean distance;

[0113]

[0114] In the formula, x i,k and c j,k The operating mode is x respectively i and cluster center c j The value for the k-th feature is repeatedly calculated until the clustering error converges. Each cluster center represents a typical operating feature, and the feature clusters C1, C2, ..., C3 in the output feature set are calculated. n Used for future fault matching.

[0115] Analyzing historical data of power deficit events in the feature set to calculate the equivalent inertial time constant of the power grid at the time of power deficit includes the following steps:

[0116] Collect historical data including frequency response data, power deficit data, and system fundamental parameters;

[0117] Extract the rate of change of frequency from the frequency response data and use it as a matching feature to match the data in the feature set;

[0118] Calculate the equivalent inertial time constant based on the power deficit data;

[0119] The obtained equivalent inertial time constant is compared with the historical typical value. If the equivalent inertial time constant deviates from the historical typical value by more than a set threshold, the accuracy of the data is checked.

[0120] If the calculated equivalent inertial time constant deviates from the actual operating characteristics, adjustments should be made based on the system's fundamental parameters.

[0121] Among the collected historical data, frequency response data, power deficit data, and system fundamental parameters are:

[0122] Frequency response data includes frequency change curves, collected system frequencies f(t) recorded from historical events, and data covering the entire process from the start of the event to the gradual recovery of the frequency; the data is output in time series format.

[0123] If the power deficit data can be directly measured, the power change is extracted directly from the faulty equipment; if it cannot be directly measured, it is estimated by the difference between the total power generation or total load power of the system before and after the event. If there are multiple fault points in the system, the total power deficit is comprehensively assessed.

[0124] The system's basic parameters include the rated power f0 and the system's reference power S. base .

[0125] As a preferred embodiment of the above, extracting the frequency change rate from the frequency response data includes:

[0126] Calculate the rate of change of frequency

[0127]

[0128] In the formula: f(t) i+1 ) and f(t) i ) represents the frequency values ​​at two adjacent time points; t i+1 -t i It is the time interval between two points in time;

[0129] Select the inertial response time period and analyze the frequency change rate. Extract the initial segment data to calculate the system inertia;

[0130] The rate of change of frequency was analyzed using a multi-point average smoothing algorithm. Noise reduction:

[0131]

[0132] The equivalent inertial time constant is calculated based on the power deficit data, including:

[0133] Based on the power-frequency dynamic response equation, the equivalent inertial time constant H at this point is calculated. For a single time point:

[0134]

[0135] In the formula, ΔP is the power deficit detected in real time, and f0 is the rated frequency of the system;

[0136] For multi-point data, calculate the inertial time constant H for each point. i And calculate the average:

[0137]

[0138] In this embodiment, when a disturbance occurs, the characteristic data of the power grid at that time is acquired and matched with the data in the feature set. Based on the equivalent inertial time constant of the power grid when a power deficit occurs, the equivalent inertial time constant at the time of the disturbance is approximately estimated, including the following steps:

[0139] When a fault occurs, acquire the characteristic data of the power grid at that time and organize it into a multi-dimensional feature vector;

[0140] Calculate the distance between a multidimensional feature vector and each cluster center in the feature set data using Euclidean distance;

[0141] Traverse all cluster centers and find the cluster center with the smallest distance to the multidimensional feature vector;

[0142] Obtain the equivalent inertial time constant of the cluster center as an approximate estimate.

[0143] The rate of change of frequency is:

[0144]

[0145] In the formula, ΔP is the power deficit detected in real time, and H eq Let f0 be the system's equivalent inertial time constant, and f0 be the system's rated frequency.

[0146] Lowest frequency f min for:

[0147]

[0148] In the formula, Δt represents the time required for the system frequency f0 to drop from its initial value to its lowest point; if the system frequency response characteristics are stable, then an empirical value of Δt is directly selected based on historical data of similar disturbances.

[0149] like Figure 2 As shown, this embodiment also includes a fast frequency assessment system based on a power grid feature set, using the method described above. The system includes:

[0150] The feature set construction unit is used to acquire historical power grid data, including operating mode, the proportion of power generation-grid-load-storage capacity in operation, and the proportion of new energy access, and to construct a power system feature set.

[0151] The analysis unit is used to analyze historical data of power deficit events in the feature set and calculate the equivalent inertial time constant of the power grid when a power deficit occurs.

[0152] The matching unit is used to acquire the characteristic data of the power grid when a disturbance occurs, and match it with the data in the feature set. Based on the equivalent inertial time constant of the power grid when the power deficit occurs, the equivalent inertial time constant during the disturbance is approximately estimated.

[0153] The evaluation unit is used to combine real-time power grid data and the matched equivalent inertial time constant to evaluate the rate of frequency change and the minimum frequency point within the inertial response time period after a disturbance.

[0154] Example 2:

[0155] like Figure 3 As shown, the method in this embodiment specifically includes the following steps:

[0156] Step 1: Construct a power system feature set based on historical grid data, according to the operating mode, the proportion of power generation-grid-load-storage capacity in operation, and the proportion of new energy access.

[0157] Step 2: Analyze the historical data of power deficit events in the feature set and calculate the equivalent inertial time constant of the power grid when a power deficit event occurs;

[0158] Step 3: Based on real-time data, detect the occurrence of disturbances. When a disturbance occurs, match the power grid characteristic data at that time with the data in the power grid characteristic set to approximately estimate the equivalent inertial time constant of the system during the disturbance.

[0159] Step 4: Combining real-time power grid data and the matched system equivalent inertial time constant, quickly assess the rate of frequency change RoCoF and the minimum frequency point within the inertial response time period after the disturbance.

[0160] In step 1, a power system feature set based on historical grid data is constructed according to the operating mode, the proportion of power generation-grid-load-storage capacity in operation, and the proportion of renewable energy access. Specifically:

[0161] Step 1-1, as follows Figure 4 As shown, the power system dataset is constructed based on seasonal load fluctuations. This requires considering climate conditions and electricity demand to build a feature set. The year is divided into four seasons: spring, summer, autumn, and winter. Within each season, the characteristics of each season are determined based on typical climate conditions and changes in power load levels. For each season, based on the proportion of renewable energy output, load levels (peak, off-peak, and valley), and grid regulation capacity, typical operating mode combinations are constructed, with different tiers for renewable energy generation and load levels. Under the maximum renewable energy generation operating mode, synchronous generators are operated as few times as possible and rotating reserves are minimized to make the frequency response characteristics as harsh as possible. Under the minimum renewable energy generation operating mode, synchronous generators are operated as many times as possible while maintaining minimum technical output, increasing inertia and frequency regulation capabilities to make the frequency regulation response characteristics as optimistic as possible. Several tiers are divided between the maximum and minimum renewable energy generation outputs and combined with different load levels. Extreme combinations that are impossible are eliminated, forming several seasonal fluctuations. Based on their operating characteristics, these are denoted as seasonal sequences {F1, F2, ..., F...} in descending order of frequency response characteristics. n};

[0162] Steps 1-2: Constructing a Power System Dataset Based on the Source-Grid-Load-Storage Capacity Ratio. The requirements for constructing a power system dataset based on the source-grid-load-storage capacity ratio include considering different combinations of source, grid, load, and storage capacity ratios. Under the maximum operating mode of renewable energy generation, the system should minimize the activation of traditional energy generation, prioritizing the joint dispatch of renewable energy generation and energy storage systems to ensure stable operation of the power system under high renewable energy generation. Under the maximum operating mode of traditional energy generation, the system should maximize the activation of traditional energy generation, reduce the proportion of renewable energy generation, and increase the participation of energy storage systems to enhance the system's regulation capacity, ensuring the system's reliability and stability during peak load periods. Between the maximum and minimum output of renewable energy generation, combining different load demands and grid transmission capacity with different combinations of energy storage capacity and load demand, several capacity ratio levels are defined, eliminating extreme combinations that do not meet actual operating conditions, forming multiple typical commissioning modes. Based on each operating characteristic, according to the system's frequency response capability, dispatch capability, and stability, the data is sorted from best to worst, forming a commissioning ratio sequence {T1, T2, ..., T...}. n};

[0163] Steps 1-3: The maximum load demand ratio reflects the ratio between the actual load and the maximum load capacity of the system. It has a significant impact on the regulation capability, inertial response capability, and frequency stability of the power grid. Constructing a typical operating mode set based on the maximum load demand ratio of the power system requires first defining the high-load operating mode R. load >80%; Medium load operation mode: 50% <R load ≤80%; Low load operation mode: R load ≤50%; of which R load The maximum load demand ratio is as follows:

[0164]

[0165] Among them, P load It is the actual load demand, P max The maximum load capacity of the year;

[0166] Steps 1-4: Select seasonal load levels, the proportion of source-grid-load-storage operational capacity, and the maximum load-demand ratio as data features. Use the K-means algorithm to cluster the data. Each operating mode can be represented as a vector feature x. i Vectors contain

[0167] x i =[Seasonal load level, ratio of source-grid-load-storage operational capacity, maximum load demand ratio](2)

[0168] Select n cluster centers;

[0169] C = {c j |j=1,2,3...n} (3)

[0170] Where C represents the set of cluster centers, for each method x i Calculate its distance to all cluster centers c j The Euclidean distance;

[0171]

[0172] Where, x i,k and c j,k The operating mode is x respectively i and cluster center c j The value of the k-th feature is repeatedly calculated until the clustering error converges. Each cluster center represents a typical operating feature, and the feature clusters C1, C2, ..., C3 in the output feature set are calculated. n For future fault matching;

[0173] Step 2: Analyze the historical data of power deficit events in the feature set, and calculate the equivalent inertial time constant of the power grid at the time of the power deficit event. Specifically:

[0174] Step 2-1: Collect historical frequency response data, power deficit data, and basic system parameters;

[0175] The frequency response data includes frequency change curves, collected system frequencies f(t) recorded in historical events, which can be obtained through a PMU or frequency recorder. The data needs to cover the entire process from the start of the event to the gradual recovery of the frequency. The frequency sampling rate should be as high as possible; this paper recommends 10Hz, i.e., 10 samples per second, to ensure the capture of fast inertial response characteristics. The data format can be output as a time series, for example, time (s): [0.0, 0.1, 0.2, 0.3, 0.4, ...], frequency (Hz): [50.0, 49.95, 49.90, 49.85, ...].

[0176] If the power deficit data can be directly measured, the power change can be directly extracted from the faulty equipment (such as a tripped generator or a faulty line). If it cannot be directly measured, it can be estimated by the difference between the total generating power or total load power of the system before and after the event. If there are multiple fault points in the system (such as multiple units tripping or regional load interruption), the total power deficit needs to be comprehensively assessed.

[0177] The system's basic parameters include the rated power f0 and the system's reference power S. base ;

[0178] Step 2-2: Extract the rate of change from the frequency response data; f(t) is the frequency that changes with time, calculate the rate of change of the frequency.

[0179]

[0180] Where: f(t) i+1 ) and f(t) i ) represents the frequency values ​​at two adjacent time points; t i+1 -t i It is the time interval between two points in time;

[0181] The inertial response time period is typically selected as 0 to 1 second after the event occurs, as the inertial response primarily takes effect during this period. This method uses 0.5 seconds, based on the rate of change of frequency. Extract the initial segment data to calculate the system inertia;

[0182] Since frequency data may contain noise, it is recommended to use multi-point average smoothing.

[0183]

[0184] Steps 2-3: Calculate the equivalent inertial time constant based on the actual historical power deficit events, using the power-frequency dynamic response equation;

[0185]

[0186] The formula is rearranged to deduce the equivalent inertial time constant at this point; for a single time point, it is derived from:

[0187]

[0188] If multi-point data is used, the inertial time constant H at each point can be calculated. i Then calculate the average:

[0189]

[0190] Steps 2-4: Verify and adjust the obtained inertial constant. Compare the calculated inertial time constant H with the historical typical value of the system to check its rationality. If the result deviates significantly from the historical value, the accuracy of the input data needs to be checked again. If it is found that the calculated inertial time constant deviates from the actual operating characteristics, it can be adjusted in combination with the system's operating mode and conditions. Finally, the equivalent inertial time constant H of the system when the power deficit event occurs is obtained. This parameter will serve as a key input for quickly evaluating the rate of frequency change and the lowest frequency in subsequent steps, providing a basis for predicting the frequency trajectory after the disturbance.

[0191] In step 3, based on real-time data, disturbances are detected. When a disturbance occurs, the power grid characteristic data at that moment is matched with the data in the power grid characteristic set to approximately estimate the equivalent inertial time constant of the system at the time of the disturbance. Specifically:

[0192] Step 3-1: Store the operation mode feature clusters obtained from the clustering in Step 1 as a database or model file for easy real-time access. The feature storage format includes cluster center feature values, seasonal load levels, the proportion of source-grid-load-storage operational capacity, maximum load demand ratio, corresponding equivalent inertia H, and typical frequency trajectory. When a new fault occurs, extract the power loss ΔP and the real-time operation mode from the PMU and other real-time monitoring equipment.

[0193] The above features are organized into a multidimensional feature vector X. new ;

[0194] Calculate the new fault feature vector X using Euclidean distance. new With the feature set, each cluster center C i Distance:

[0195]

[0196] Where x k C is the k-th eigenvalue of the new fault feature vector. ik It is cluster center C i The kth eigenvalue;

[0197] Traverse all cluster centers and find the cluster with respect to X. new The cluster center with the smallest distance C best :

[0198]

[0199] Using the matched cluster centers, the equivalent inertia H represented by each cluster is obtained as an approximate estimate. Since this scheme is designed to assess frequency-insecure scenarios and faults, speed is the primary objective. Under the premise of meeting speed requirements, the predicted equivalent inertia H should be as close as possible to the actual response, while being slightly conservative.

[0200] In step 4, by combining real-time power grid data and the matched system equivalent inertial time constant, the frequency change rate RoCoF and the frequency minimum point after the disturbance are quickly evaluated.

[0201] Step 4-1: The rate of frequency change characterizes the initial characteristics of the system's frequency response, and its mathematical expression is:

[0202]

[0203] Where ΔP is the power deficit detected in real time, and Heq Let f0 be the system's equivalent inertial time constant, and f0 be the system's rated frequency. This formula reflects the influence of system inertia on frequency stability; the larger the inertia, the smaller the RoCoF, and the slower the system's frequency decreases. After a disturbance occurs, the current system's equivalent inertial time constant H can be quickly obtained by matching it with the power grid's historical characteristic database. eq Thus, the initial frequency change rate of the system when the disturbance occurs is obtained;

[0204] Step 4-2, Lowest frequency f min It is also one of the key indicators of the frequency response of a power system when power disturbances occur, and its calculation formula is as follows:

[0205]

[0206] Where Δt represents the time required for the system frequency f0 to decrease from its initial value to its minimum point. If the system frequency response characteristics are stable, an empirical value of Δt can be directly selected based on historical data of similar disturbances. For conventional power grids, the main stage of inertial response is usually completed within 0.5-2 seconds. When the system inertia is large, i.e., when traditional thermal power is the main source, Δt≈1.5s; when the system inertia is small, i.e., when a high proportion of new energy sources is present, Δt≈0.5-1s. Based on the actual time for the frequency to decrease from its initial value to its minimum point under similar disturbance scenarios, the historical average value of 1.2s is selected as Δt. Combined with the RoCoF calculated in step 4-1, the minimum frequency during the disturbance can be quickly assessed. Based on the RoCoF and f0 during the disturbance... min The ability to quickly plot frequency response curves is crucial for understanding the dynamic behavior of the system after disturbances and for developing countermeasures. When evaluating f... min When the system approaches the minimum frequency safety threshold set by the system, emergency measures can be triggered immediately; if the RoCoF value is too high, it indicates that the system frequency is dropping too fast, and it is advisable to increase energy storage or spin-back in advance to provide rapid response support.

[0207] Please see Figure 5 The diagram shows a structural schematic of a computer device provided in an embodiment of this application. An embodiment of this application provides a computer device 400, including a processor 410 and a memory 420. The memory 420 stores a computer program executable by the processor 410. When the computer program is executed by the processor 410, it performs the method described above.

[0208] This application embodiment also provides a storage medium 430, on which a computer program is stored, and the computer program is executed by a processor 410 to perform the above method.

[0209] The storage medium 430 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0210] In the description of this invention, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. "A plurality of" means two or more, unless otherwise explicitly specified.

[0211] In this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," "linking," and "fixing," etc., should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0212] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Furthermore, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0213] Any process or method description in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing a particular logical function or process, and the scope of the preferred embodiments of the invention includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as will be understood by those skilled in the art to which embodiments of the invention pertain.

[0214] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include: an electrical connection having one or more wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Alternatively, the computer-readable medium may be paper or other suitable media on which the program can be printed, since the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in a computer memory.

[0215] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0216] Those skilled in the art will understand that all or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.

[0217] The storage medium mentioned above can be a read-only memory, a disk, or an optical disk, etc. Although embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present invention.

Claims

1. A method for rapid frequency assessment based on power grid feature sets, characterized in that, Includes the following steps: Acquire historical power grid data, including operating modes, the proportion of operational capacity of power generation, grid, load, and storage, and the proportion of renewable energy access, and construct a power system feature set; Analyze the historical data of power deficit events in the feature set to calculate the equivalent inertial time constant of the power grid when a power deficit occurs; When a disturbance occurs, the characteristic data of the power grid at that time is acquired and matched with the data in the feature set. Based on the equivalent inertial time constant of the power grid when a power deficit occurs, the equivalent inertial time constant at the time of the disturbance is approximately estimated. By combining real-time power grid data and the matched equivalent inertial time constant, the rate of frequency change and the minimum frequency point within the inertial response time period after the disturbance are evaluated.

2. The method for rapid frequency assessment based on power grid feature sets according to claim 1, characterized in that, The acquisition of historical power grid data, including operating modes, the proportion of power generation-grid-load-storage capacity in operation, and the proportion of new energy access, and the construction of a power system feature set, includes the following steps: Taking into account climate conditions and electricity demand, the characteristics of each season are determined based on typical climate conditions and changes in electricity load levels in different seasons. For each season, based on the proportion of renewable energy output, load level, and grid regulation capacity, a typical combination of operating modes is constructed, with different levels of operating modes for renewable energy power generation and load level. Considering different combinations of the proportion of power generation, grid, load and storage capacity, between the maximum and minimum output of new energy power generation, combined with different load demand and grid transmission capacity, and with different combinations of energy storage capacity and load demand, several capacity proportion levels are divided. Construct a set of typical operating modes based on the maximum load demand ratio of the power system; Seasonal load levels, the proportion of operational capacity of source-grid-load-storage, and the maximum load demand ratio were selected as data features, and the K-means algorithm was used to cluster the data.

3. The method for rapid frequency assessment based on power grid feature sets according to claim 2, characterized in that, The aforementioned method constructs typical operating mode combinations for each season based on the proportion of renewable energy output, load level, and grid regulation capacity, with different tiers for renewable energy generation and load levels, including: Under the maximum operating mode of new energy power generation, the synchronous machine should be started as few times as possible and no rotational standby should be left as much as possible; Under the minimum operating mode of new energy power generation, the synchronous generator should be started as many times as possible while maintaining minimum technical output, thereby increasing inertia and frequency regulation capability.

4. The method for rapid frequency assessment based on power grid feature sets according to claim 2, characterized in that, The construction of typical operating mode combinations, which categorizes new energy power generation and load levels into different operating mode combinations, also includes: After eliminating impossible extreme combinations, several seasonal fluctuations are formed. Based on their operating characteristics, they are recorded as seasonal sequences in order of frequency response characteristics from best to worst.

5. The method for rapid frequency assessment based on power grid feature sets according to claim 2, characterized in that, The consideration of different combinations of the proportions of power generation, grid, load, and storage capacity, within the range of maximum and minimum output of new energy power generation, combined with different load demands and grid transmission capabilities, and with different combinations of energy storage capacity and load demand, further includes: Extreme combinations that do not conform to actual operating conditions are eliminated to form multiple typical commissioning modes; based on each working characteristic, the modes are sorted from best to worst according to the system's frequency response capability, scheduling capability, and stability to form a commissioning percentage sequence.

6. The method for rapid frequency assessment based on power grid feature sets according to claim 2, characterized in that, The construction of a typical operating mode set based on the maximum load demand ratio of the power system includes: The ranges for high-load operation, medium-load operation, and low-load operation are defined based on the maximum load demand ratio. A set of typical operating modes is constructed based on the different defined ranges.

7. The method for rapid frequency assessment based on power grid feature sets according to claim 2, characterized in that, The process of selecting seasonal load levels, the proportion of source-grid-load-storage operational capacity, and the maximum load-demand ratio as data features, and using the K-means algorithm to cluster the data, includes the following steps: Each operating mode is represented as a vector feature x. i The vector features include: x i = [Seasonal load level, ratio of operating capacity of power generation, grid, load and storage, and maximum load demand ratio]; Select n cluster centers: C={c j |j=1,2,3...n}; In the formula, C represents the set of cluster centers, and for each method x i Calculate its distance to all cluster centers c j The Euclidean distance; In the formula, x i,k and c j,k The operating mode is x respectively i and cluster center c j The value for the k-th feature is repeatedly calculated until the clustering error converges. Each cluster center represents a typical operating feature, and the feature clusters C1, C2, ..., C3 in the output feature set are calculated. n Used for future fault matching.

8. The method for rapid frequency assessment based on power grid feature sets according to claim 1, characterized in that, The step of analyzing historical data of power deficit events in the feature set and calculating the equivalent inertial time constant of the power grid at the time of power deficit includes the following steps: Collect historical data including frequency response data, power deficit data, and system fundamental parameters; The frequency change rate is extracted from the frequency response data and used as a matching feature to match the data in the feature set. The equivalent inertial time constant is calculated based on the power deficit data. The obtained equivalent inertial time constant is compared with the historical typical value. If the equivalent inertial time constant deviates from the historical typical value by more than a set threshold, the accuracy of the data is checked. If the calculated equivalent inertial time constant deviates from the actual operating characteristics, adjustments should be made based on the system's fundamental parameters.

9. The method for rapid frequency assessment based on power grid feature sets according to claim 8, characterized in that, Among the collected historical data, including frequency response data, power deficit data, and system basic parameters: The frequency response data includes frequency change curves, collects system frequencies f(t) recorded in historical events, and covers the entire process from the start of the event to the gradual recovery of the frequency; the data is output in time series format. If the power deficit data can be directly measured, the power change is extracted directly from the faulty equipment; if it cannot be directly measured, it is estimated by the difference between the total power generation or total load power of the system before and after the event; if there are multiple fault points in the system, the total power deficit is comprehensively assessed. The system's basic parameters include the rated power f0 and the system's reference power S. base .

10. The method for rapid frequency assessment based on power grid feature sets according to claim 9, characterized in that, Extracting the frequency change rate from the frequency response data includes: Calculate the rate of change of frequency In the formula: f(t) i+1 ) and f(t) i ) represents the frequency values ​​of two adjacent time points; t i+1 -t i It is the time interval between two points in time; Select a set inertial response time period, and from the frequency change rate Extract the initial segment data to calculate the system inertia; The frequency change rate was analyzed using a multi-point average smoothing algorithm. Noise reduction:

11. The method for rapid frequency assessment based on power grid feature sets according to claim 8, characterized in that, The calculation of the equivalent inertial time constant based on the power deficit data includes: Based on the power-frequency dynamic response equation, the equivalent inertial time constant H at this point is calculated. For a single time point: In the formula, ΔP is the power deficit detected in real time, and f0 is the rated frequency of the system; For multi-point data, calculate the inertial time constant H for each point. i And calculate the average:

12. The method for rapid frequency assessment based on power grid feature sets according to claim 2, characterized in that, When the disturbance occurs, the characteristic data of the power grid at that time is acquired and matched with the data in the feature set. Based on the equivalent inertial time constant of the power grid when the power deficit occurs, the equivalent inertial time constant at the time of the disturbance is approximately estimated, including the following steps: When a fault occurs, acquire the characteristic data of the power grid at that time and organize it into a multi-dimensional feature vector; The distance between the multidimensional feature vector and each cluster center in the feature set data is calculated using Euclidean distance. Traverse all the cluster centers and find the cluster center with the smallest distance to the multidimensional feature vector; Obtain the equivalent inertial time constant of the cluster center as an approximate estimate.

13. The method for rapid frequency assessment based on power grid feature sets according to claim 1, characterized in that, The frequency change rate is: In the formula, ΔP is the power deficit detected in real time, and H eq Let f0 be the system's equivalent inertial time constant, and f0 be the system's rated frequency. Lowest frequency f min for: In the formula, Δt represents the time required for the system frequency f0 to decrease from its initial value to its minimum point; If the system frequency response characteristics are stable, then the empirical value of Δt can be directly selected based on historical data of similar disturbances.

14. A rapid frequency assessment system based on power grid feature sets, characterized in that, The system, using the method as described in any one of claims 1 to 13, comprises: The feature set construction unit is used to acquire historical power grid data, including operating mode, the proportion of power generation-grid-load-storage capacity in operation, and the proportion of new energy access, and to construct a power system feature set. The analysis unit is used to analyze historical data of power deficit events in the feature set and calculate the equivalent inertial time constant of the power grid when a power deficit occurs. The matching unit is used to acquire the power grid characteristic data when a disturbance occurs, and match it with the data in the feature set. Based on the equivalent inertial time constant of the power grid when the power deficit occurs, the equivalent inertial time constant during the disturbance is approximately estimated. The evaluation unit is used to evaluate the rate of frequency change and the minimum frequency point within the inertial response time period after the disturbance by combining real-time power grid data and the matched equivalent inertial time constant.

15. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method as described in any one of claims 1-13.

16. A storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1-13.

Images