A source-grid-load-storage local power grid low-frequency fault coordination control method

Abstract

The present application relates to the field of low-frequency fault coordinated control, and relates to a kind of source network load storage local power grid low-frequency fault coordinated control method.The present application is based on the active power variation of each time point in history when low-frequency fault occurs in local power grid, combined with the dynamic correction of current actual inertia coefficient based on virtual inertia coefficient, and calculates total power shortage, real-time predicts the voltage variation amplitude of corresponding node after each load line is cut off and the load rate of adjacent load line, from which the cuttable line is initially screened, based on the important level of cuttable line, active power telemetry value combined with its historical line stability score, the priority of cuttable line is sorted, and candidate line combination is constructed;The residual power shortage of each combination is obtained, combined with marginal loss function, the pre-output power of each energy storage unit corresponding to each candidate line combination is analyzed;The optimal combination is screened to execute corresponding load shedding operation and energy storage discharge control;The economy and power supply recovery capability of local power grid are improved.

Description

Technical Field

[0001] This invention relates to the field of low-frequency fault coordination control, and specifically to a method for low-frequency fault coordination control in a local power grid with source, grid, load, and storage. Background Technology

[0002] Against the backdrop of energy transition and the construction of new power systems, the local power grid, which integrates multiple entities such as distributed power sources, power loads, and energy storage units, has complex operating characteristics and is highly susceptible to low-frequency faults. Once a low-frequency fault occurs, if power regulation and load control are not carried out in a timely and accurate manner, the grid frequency will continue to drop, seriously threatening the safe and stable operation of the local power grid.

[0003] Existing technologies, such as Chinese Patent Publication No. CN117878952A, disclose a coordinated control method and system based on flexible photovoltaic-storage resources and low-frequency load shedding. By monitoring changes in grid frequency, when the frequency drops to a preset photovoltaic-storage action value, the system provides maximum power support in emergency situations through emergency output control of flexible photovoltaic-storage resources; when the frequency drops to a preset load shedding action value, the system optimizes low-frequency load shedding with the objective function of minimizing control costs until the frequency returns to normal levels, thereby improving the economy and stability of the power system.

[0004] However, the existing technology has the following problems: 1. The existing technology mostly uses a fixed inertia coefficient when calculating the power deficit, without considering the dynamic changes of the virtual inertia of grid-connected new energy sources and the impact of historical disturbances in the power grid. This results in a mismatch between the inertia coefficient and the actual operating state of the power grid, and a large error in the calculation of the power deficit, which poses a hidden danger for the formulation of subsequent control strategies.

[0005] 2. Existing technologies achieve emergency output control of flexible photovoltaic-storage resources by preset photovoltaic-storage action values ​​and preset load shedding action values. However, they do not consider the optimal combination and adjustment based on the coordinated action of the two, which can easily lead to ineffective loss of energy storage resources or unnecessary load shedding. It is difficult to achieve both economic efficiency and high efficiency in fault control at the same time. Summary of the Invention

[0006] This invention aims to address the shortcomings of existing technologies by providing a coordinated control method for low-frequency faults in a local power grid with power generation, grid, load, and storage. Candidate line combinations are constructed by prioritizing total power deficit and switchable lines. The optimal combination is then analyzed based on the marginal loss of each candidate line combination and the expected recovery time of the corresponding candidate line, thus achieving efficient coordinated fault control.

[0007] To achieve the above objectives, the present invention adopts the following technical solution: The present invention provides a method for coordinated control of low-frequency faults in a local power grid, comprising: when a low-frequency fault is detected in the local power grid, dynamically correcting the current actual inertia coefficient based on the changes in active power at historical time points and combining the virtual inertia coefficient of grid-type new energy, and calculating the total power deficit.

[0008] Real-time prediction of voltage change at corresponding nodes after each load line is disconnected and load rate of adjacent load lines are used to initially screen disconnectable lines.

[0009] Based on the importance level of the switchable lines, the active power telemetry values, and the historical line stability scores of the switchable lines, the switchable lines are prioritized and ranked. Candidate line combinations are constructed based on the total power deficit and the priority ranking.

[0010] Obtain the remaining power deficit of candidate line combinations, construct a marginal loss function based on the remaining capacity percentage of each energy storage unit and the current temperature, and analyze the pre-output power of each energy storage unit corresponding to each candidate line combination.

[0011] Based on the marginal loss of each candidate line combination and the expected recovery time of the corresponding candidate line, the optimal combination is analyzed, and the corresponding load shedding operation and energy storage discharge control are executed according to the optimal combination.

[0012] Compared with the prior art, the present invention has the following beneficial effects: (1) The present invention dynamically corrects the current actual inertia coefficient by combining the change of active power at each historical time point with the virtual inertia coefficient of grid-type new energy and calculates the total power deficit, so that the current actual inertia coefficient can reflect the grid operation status in real time, and the total power deficit calculation is more in line with reality, thus eliminating the problem of excessive or insufficient subsequent power compensation from the source.

[0013] (2) This invention predicts the voltage change amplitude of the corresponding node after each load line is cut off and the load rate of the adjacent load line in real time, and initially screens the cut-off lines, accurately predicts the overload risk of the adjacent lines after the selected lines are cut off, avoids screening out lines without actual compensation value, and improves the safety and efficiency of fault control.

[0014] (3) The present invention prioritizes the cut-off lines based on the importance level of the cut-off lines, the active power telemetry value, and the historical line stability score of the cut-off lines. Based on the total power deficit and the priority ranking, candidate line combinations are constructed, making the priority ranking and candidate combination construction more practical reference value, avoiding the bias of subjective judgment, and improving the stability of load shedding.

[0015] (4) By obtaining the remaining power deficit of candidate line combinations, constructing a marginal loss function based on the remaining capacity percentage of each energy storage unit and the current temperature, this invention analyzes the pre-output power of each energy storage unit corresponding to each candidate line combination, avoids excessive energy storage loss due to ignoring the operating status, reduces the equipment operation and maintenance cost of grid fault control, and improves the energy storage utilization efficiency.

[0016] (5) Based on the marginal loss of each candidate line combination and the expected recovery time of the corresponding candidate line, the present invention analyzes the optimal combination, executes the corresponding load shedding operation and energy storage discharge control according to the optimal combination, realizes the comprehensive minimization of energy storage life loss and load shedding recovery cost, and improves the economy and power supply recovery capability of the local power grid. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a schematic diagram of the method steps of the present invention;

[0019] Figure 2 This is a schematic diagram illustrating the specific steps of the method for obtaining the actual inertia coefficient in this invention.

[0020] Figure 3 This is a schematic diagram of the steps in the analysis method for the optimal combination in this invention. Detailed Implementation

[0021] Various exemplary embodiments of the present invention will now be described in detail with reference to the accompanying drawings. It should be noted that, unless otherwise specifically stated, the relative arrangement, numerical expressions, and values ​​of the components and steps set forth in these embodiments do not limit the scope of the invention. Furthermore, it should be understood that, for ease of description, the dimensions of the various parts shown in the drawings are not drawn to actual scale.

[0022] The following description of at least one exemplary embodiment is merely illustrative and is in no way intended to limit the invention or its application or use. Techniques, methods, and apparatus known to those skilled in the art may not be discussed in detail, but where appropriate, such techniques, methods, and apparatus should be considered part of the specification.

[0023] In all examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not as limitations. Therefore, other examples of exemplary embodiments may have different values.

[0024] Please see Figure 1 As shown, the present invention provides a method for coordinated control of low-frequency faults in a local power grid, including S1, when a low-frequency fault is detected in the local power grid, dynamically correcting the current actual inertia coefficient based on the changes in active power at historical time points and combining the virtual inertia coefficient of grid-type new energy, and calculating the total power deficit.

[0025] Considering that the grid inertia coefficient is the core parameter for calculating power deficit, and its value is related to the real-time operating status of the grid, while the virtual inertia of grid-connected new energy will change dynamically with grid disturbances, and the active power change data of historical grid disturbances can objectively reflect the actual fluctuation law of the inertia coefficient, if a fixed inertia coefficient is used when calculating the power deficit, the inertia coefficient will be seriously mismatched with the real-time operating status of the grid, resulting in a large error in the calculation of power deficit, which may easily lead to excessive load shedding or insufficient energy storage compensation in the future, and will not be able to effectively suppress frequency drops.

[0026] Meanwhile, considering that the virtual inertia of grid-connected new energy sources is superimposed on the grid's intrinsic inertia, jointly affecting the system's frequency response characteristics, directly using the total inertia coefficient including virtual inertia when recursively estimating the inertia coefficient using historical disturbance data would cause the estimation result to fluctuate drastically with the virtual inertia, failing to reflect the stable trend of the grid's intrinsic inertia. Therefore, during the historical estimation stage, it is necessary to remove the influence of virtual inertia by subtracting the corresponding virtual inertia coefficient from the total inertia of each disturbance to restore the true inertia level of the grid.

[0027] When a low-frequency fault occurs, the grid-type new energy is providing real-time virtual inertia support through control strategies. This support capability, together with the physical inertia of the equipment, affects the system frequency response. Therefore, when calculating the current actual inertia coefficient, the current virtual inertia coefficient should be added, and the total power deficit should be calculated accordingly.

[0028] Based on this, it should be noted that the analysis method for the total power deficit includes: S11, obtaining the change in active power of each load line of the local power grid at each time point within a set historical time window. If the change in active power of a certain load line at multiple consecutive time points exceeds the set normal fluctuation range, then the load line is identified as a disturbance line, and the time period corresponding to the multiple consecutive time points is recorded as a disturbance.

[0029] It should be explained that the change in active power is the difference between the telemetry value of active power at each time point and the previous time point. In a specific embodiment of the present invention, the set historical time window is the past 7 days, and "continuously greater than the set normal fluctuation range" means that the value is greater than the set normal fluctuation range for 3 or more consecutive time points, to avoid false disturbances and misjudgments caused by factors such as data acquisition equipment errors and instantaneous load impacts. The sampling interval for each time point is 1 second. In addition, the normal fluctuation range is set to 1% of the rated active power. Implementers may also set other specific values, but it should not exceed 5% to avoid the fluctuation range being too large and thus missing disturbances.

[0030] S12. The corresponding system disturbance power is obtained by vector summing the changes in active power of each load line corresponding to each disturbance at the first time point.

[0031] S13. Collect the bus voltage frequency change rate and system disturbance power of the local power grid for each disturbance within a set historical time window, and calculate the initial inertia coefficient by combining the bus voltage frequency change rate. The bus voltage frequency change rate for each disturbance refers to the voltage change rate corresponding to the first time point of each disturbance. This process refers to the inverse operation of the rotor motion equation; the specific formula is as follows: .

[0032] in The initial inertia coefficient is expressed in seconds. The per-unit value representing the system disturbance power, that is, the ratio of the system disturbance power to the rated power. The frequency change rate of the bus voltage is represented by Hz / s. For the formula constant 2 in the 50Hz power frequency scenario, and The product of . This formula is based on the rotor motion equation. It is derived that, due to Substitute the values ​​into the equation and solve. The total inertia coefficient of the system at that moment is obtained from a single disturbance event.

[0033] S14. Based on the initial inertia coefficients of each disturbance, combined with the system's preset inertia coefficients and the virtual inertia coefficients of the grid-type new energy sources during each disturbance process, the current actual inertia coefficients are obtained through weighted recursive updates. For example... Figure 2 As shown, the specific implementation steps are as follows: S141, obtain the system preset inertia coefficient and the virtual inertia coefficient of the grid-type new energy in each disturbance process, and perform a weighted summation of the initial inertia coefficient and preset inertia coefficient corresponding to the first disturbance within the set historical time window, and use the difference between the result and the corresponding virtual inertia coefficient as the updated body inertia coefficient.

[0034] In a specific embodiment of the present invention, the system preset inertia coefficient is typically stored as a fixed value in the power system database in engineering practice and obtained through offline simulation calculation during the power grid planning and design phase. This is existing technology and will not be described in detail here. The inertia coefficient of this type of source-grid-load-storage local power grid under typical operating conditions is mostly concentrated between 2 and 6 seconds; in this embodiment, 4 seconds is used as an example.

[0035] It should be explained that the virtual inertia coefficient of grid-connected new energy refers to the equivalent inertia constant of grid-connected new energy (such as photovoltaic and wind power) by simulating the inertia response characteristics of traditional synchronous generators through control strategies. The unit is seconds. It is used to characterize the virtual inertia support capability of new energy units to grid frequency changes. It changes dynamically with the operating status of new energy units and control system parameters. Its specific value can be read in real time through the communication interface.

[0036] Furthermore, considering that the initial inertia coefficient better reflects the actual measured inertia level, while the system preset inertia coefficient is a fixed value set based on power grid design parameters and theoretical operating conditions, which deviates from the real-time operating state of the power grid, assigning a higher weight to the initial inertia coefficient allows the updated physical inertia coefficient to better reflect the actual power grid situation. Therefore, in this embodiment, the weight of the initial inertia coefficient is set to 70%, and the weight of the system preset inertia coefficient is set to 30%. Implementers can also set other specific weight values, and the sum of the two is 1. By subtracting the weighted result from the corresponding virtual inertia coefficient, the superimposed influence of the virtual inertia of grid-connected new energy on the actual physical inertia of the power grid is accurately removed, restoring the true inertia level of the power grid itself.

[0037] S142. Record the difference between the initial inertia coefficient and the corresponding virtual inertia coefficient for each remaining disturbance within the set historical time window as the initial body inertia coefficient for each disturbance.

[0038] S143. The initial body inertia coefficients of each iteration are weighted and calculated together with the previous updated body inertia coefficients to obtain the updated body inertia coefficients of each disturbance.

[0039] Considering that the previous updated inertia coefficient already matches the actual operating state of the power grid at the time of the previous disturbance, although it can reflect the instantaneous inertia characteristics of the current disturbance of the power grid, it is easily affected by equipment acquisition errors and instantaneous extreme fluctuations of the power grid, resulting in local deviations. Therefore, the previous updated inertia coefficient is given a higher weight. In this embodiment, the initial inertia coefficient corresponding to each disturbance is set to 30%, and the weight of the previous updated inertia coefficient is set to 70%.

[0040] S144. The sum of the updated ontological inertia coefficient corresponding to the last disturbance within the set historical time window and the current virtual inertia coefficient is recorded as the current actual inertia coefficient.

[0041] Since the virtual inertia provided by new energy sources and the physical inertia both affect the system's frequency response, they must be added together to obtain the actual system inertia. The current actual inertia coefficient is used for subsequent calculations of the total power deficit. Real-time updates of the actual inertia coefficient are implemented for each low-frequency fault occurrence.

[0042] S15. Calculate the total power deficit based on the current actual inertia coefficient and the current bus voltage frequency change rate.

[0043] .

[0044] in This represents the total power deficit. Represents the current actual inertia coefficient. Represents the current rate of change of the bus voltage frequency. The rated power reference value represents the local power grid of source, grid, load and storage. The rated power reference value is determined artificially during the system design stage. It is usually taken as the total installed capacity of the local power grid and stored as a fixed parameter in the power system database. This formula is existing technology and will not be described in detail in this invention.

[0045] This invention dynamically corrects the current actual inertia coefficient by combining the changes in active power at various historical time points with the virtual inertia coefficient of grid-connected new energy sources, and calculates the total power deficit. This allows the current actual inertia coefficient to reflect the grid operation status in real time, and the total power deficit calculation to be more realistic, thus preventing the problem of excessive or insufficient power compensation in the future from the source.

[0046] S2. Real-time prediction of the voltage change amplitude of the corresponding node after each load line is cut off and the load rate of adjacent load lines, and initial screening of cut-off lines.

[0047] Considering that disconnecting load lines will trigger a redistribution of node power, directly affecting the voltage change of the corresponding node, and that adjacent load lines of the same node will experience load rate changes due to power transfer, and that the matching degree between the active power telemetry value of the candidate line and the total power deficit determines the actual compensation value of load shedding, blindly shedding loads without anticipating the above-mentioned related effects can easily cause the node voltage to drop below the safety threshold, cause secondary faults due to overload of adjacent lines, or waste power supply resources by disconnecting lines with no actual compensation value. Therefore, it is necessary to first predict the node voltage change after disconnecting each load line and the load rate of adjacent load lines in real time based on the line operating parameters, and then initially screen out the lines with actual load shedding value.

[0048] Based on this, in a preferred embodiment of the present invention, the specific method for screening switchable lines includes: S21, extracting active power telemetry values, current node voltage and current from the status information of each load line, and screening out each candidate line whose active power telemetry value is not zero.

[0049] S22. Based on the active power telemetry values ​​of each candidate line and the voltage-power sensitivity coefficient, analyze the voltage change amplitude, and combine this with the current node voltage to obtain the predicted node voltage after the candidate line is disconnected. Specifically, the product of the active power telemetry value of each candidate line and the voltage-power sensitivity coefficient is recorded as the voltage change amplitude, and the sum of the voltage amplitude and the current node voltage is the predicted node voltage after the candidate line is disconnected.

[0050] The voltage-power sensitivity coefficient reflects the impact of changes in the active power of a load line on its connected node voltage, i.e., the voltage change caused by a unit change in active power. The voltage-power sensitivity coefficient is obtained as follows: In offline mode, the local power grid can be simulated using power flow calculation software such as PSASP. Specifically, a small disturbance is superimposed on the active power of the candidate line. Calculate the change in voltage at the corresponding node. Then the voltage-power sensitivity coefficient is approximately: The simulation using the power flow calculation software PSASP is a well-known technique in the field, and will not be described in detail here.

[0051] S23. Determine whether adjacent load lines will experience overload based on the current changes of adjacent lines corresponding to the historical disturbance records of each candidate line. The specific implementation steps include: S231. Record other load lines sharing nodes with each candidate line as adjacent load lines, and calculate the current sensitivity factor based on the current changes of each adjacent load line during the historical disturbances of each candidate line.

[0052] The current sensitivity factor is used to characterize the impact of changes in the active power of the candidate line on the current of adjacent load lines. It is obtained as follows: the ratio of the current change of each adjacent load line to the corresponding active power change of the candidate line during each historical disturbance is obtained, and the current sensitivity factor of each candidate line is calculated by averaging the values.

[0053] S232. The product of the current sensitivity factor of each adjacent load line and the telemetry value of the active power of the corresponding candidate line is recorded as the current change.

[0054] S233. The sum of the current change of each adjacent load line and the current is recorded as the predicted current, and the ratio of the predicted current to the corresponding rated current is recorded as the load rate of each adjacent load line.

[0055] S234. If the load rate of an adjacent load line of a candidate line is greater than the set load rate threshold, it is determined that the adjacent load line will be overloaded.

[0056] S24. If the telemetry value of the active power of a candidate line is less than or equal to the total power deficit, and the predicted voltage of the corresponding node after the candidate line is cut off is not lower than the preset safety threshold and the adjacent lines are not overloaded, then the line is marked as a cut-off line.

[0057] This invention uses real-time prediction of the voltage change amplitude of the corresponding node after each load line is cut off and the load rate of adjacent load lines to initially screen cut-off lines, accurately predict the overload risk of adjacent lines after the candidate lines are cut off, avoids screening out lines with no actual compensation value, and improves the safety and efficiency of fault control.

[0058] S3. Prioritize the cut-off lines based on their importance level, active power telemetry values, and historical line stability scores. Construct candidate line combinations based on total power deficit and priority ranking.

[0059] Considering that the load shedding priority of switchable lines is related to the load importance level, active power telemetry value, and historical line stability rating, the load importance level determines the social and economic impact of load shedding, the active power telemetry value directly affects the compensation effect of the total power deficit, and the historical line stability rating reflects the operational reliability of the power grid after the line is switched on or off, all three are indispensable. If the ranking is based on only a single indicator or the priority is determined by subjective judgment, it is easy to cause the load shedding scheme to be out of touch with the actual compensation needs of the power grid, or even cause serious losses due to prioritizing the shedding of lines with high importance levels. Therefore, it is necessary to combine the importance level of switchable lines, active power telemetry value, and historical line stability rating to conduct a comprehensive priority ranking, and construct candidate line combinations based on the compensation needs of the total power deficit.

[0060] Based on this, in a specific embodiment of the present invention, the method for constructing the candidate line combination includes: S31, classifying the load types of each switchable line into different importance levels according to preset rules, and sorting the switchable lines in reverse order of importance level. For example, in this embodiment, the load types corresponding to the first importance level include hospitals, fire protection, data centers, etc.; the load types corresponding to the second importance level include large factories, commercial centers, etc.; and the load types corresponding to the third importance level include residential electricity. The lower the importance level, the higher the ranking.

[0061] S32. Obtain the difference between the active power telemetry value and the maximum active power telemetry value of each cut-away line of the same importance level, and record the ratio of the difference to the maximum active power telemetry value as the optimal difference of active power telemetry value.

[0062] S33. Analyze the line stability score based on the historical switching records of each switchable line, and obtain the optimal line stability gap based on the line stability score of each switchable line and the maximum line stability score. The specific implementation steps are as follows: S331. Obtain the historical switching records of each switchable line, and extract the initial voltage before each switching and the lowest voltage value within a set time period after the switching occurs. The set time period is within five minutes after the switching occurs. The implementer can also set other specific values, but it should not be too short so that the voltage does not reach the actual lowest value within the set time period.

[0063] S332. Obtain the difference between the starting voltage and the minimum voltage value, and record the ratio of this difference to the starting voltage as the voltage change rate.

[0064] S333. Calculate the average voltage change rate of each historical switching record of each switchable line and record it as the historical average voltage change rate.

[0065] S334. The difference between the maximum and minimum values ​​of the historical average voltage change rate in all cut-off lines is recorded as the relative voltage change rate.

[0066] S335. Obtain the absolute difference between the historical average voltage change rate of each switchable line and the maximum value of the historical average voltage change rate among all switchable lines. The ratio of this difference to the relative voltage change rate is recorded as the line stability coefficient. It should be noted that a relative voltage change rate of zero means that the historical average voltage change rate of all switchable lines is relative, in which case the line stability coefficient of each switchable line would be recorded as 1. However, the probability of this event occurring is extremely small. To prevent the denominator from being zero and the fraction from being meaningless due to a zero relative voltage change rate, this is explained here.

[0067] S336. The product of the line stability coefficient and the total score of the set line stability score is recorded as the line stability score. In this embodiment, the total score of the line stability score is set to 100 points.

[0068] S34. The sum of the optimal difference between the active power telemetry values ​​and the optimal difference between the line stability values ​​is recorded as the comprehensive optimal difference. The priority of each switchable line is determined by sorting the switchable lines in ascending order according to the comprehensive optimal difference. This ensures that the smaller the comprehensive optimal difference, the larger the active power telemetry value, and the more stable the line.

[0069] S35. Accumulate the active power telemetry values ​​of each cuttable line in unit step size according to the priority of each cuttable line. When the total active power telemetry value corresponding to a certain accumulation is greater than the total power deficit for the first time, obtain the line set corresponding to each accumulation before that accumulation and record them as candidate line combinations.

[0070] The unit step size refers to a step size based on one line. For example, the active power telemetry value of load line 1 is 100kW, the active power telemetry value of load line 2 is 100kW, the active power telemetry value of load line 3 is 90kW, and the active power telemetry value of load line 4 is 70kW. The total power deficit is 350kW. Load line 1 + load line 2 + load line 3 = 290kW < 350kW, and load line 1 + load line 2 + load line 3 + load line 4 = 360kW > 350kW. Therefore, the candidate line combinations are {load line 1}, {load line 1 + load line 2}, and {load line 1 + load line 2 + load line 3}. It should be noted that when the total active power telemetry value corresponding to a certain accumulation is exactly equal to the total power deficit, the set of candidate lines corresponding to that accumulation and each previous accumulation is recorded as a candidate line combination; in addition, if the active power telemetry value of the highest priority switchable line is greater than the total power deficit, the empty set of no-load lines is taken as one of the candidate line combinations, and the set containing only the highest priority switchable lines is taken as the second candidate line combination.

[0071] S36. Compare and analyze the total remaining output power of the energy storage unit with the total power deficit, and optimize the candidate line combination.

[0072] It should be further explained that if the total remaining output power of the energy storage unit is greater than or equal to the total power deficit, an empty set of no-load lines is added as one of the candidate line combinations, participating in the subsequent selection along with the non-empty candidate line combinations. If the total remaining output power of the energy storage unit is less than the total power deficit, the difference between the total power deficit and the total remaining output power of the energy storage unit is calculated and recorded as the load shedding requirement. Combinations with a total power telemetry value less than the load shedding requirement are eliminated.

[0073] This invention prioritizes switchable lines based on their importance level, active power telemetry values, and historical line stability scores. It constructs candidate line combinations based on total power deficit and priority ranking, making the priority ranking and candidate combination construction more practically valuable, avoiding bias in subjective judgment, and improving the stability of load shedding.

[0074] S4. Obtain the remaining power deficit of the candidate line combination, construct the marginal loss function based on the remaining capacity percentage of each energy storage unit and the current temperature, and analyze the pre-output power of each energy storage unit corresponding to each candidate line combination.

[0075] Considering that the remaining capacity percentage of the energy storage unit determines the depth of discharge, and the greater the depth of discharge, the higher the loss per unit power, the current temperature will affect the temperature rise during the energy storage discharge process. Excessive temperature rise will exacerbate equipment losses. Furthermore, the remaining power deficit of the candidate line combination needs to be compensated by the energy storage units. The rationality of energy storage power allocation determines the economy of grid fault control and the service life of the energy storage units. If the output power is not allocated in conjunction with the real-time operating status of the energy storage, it is impossible to achieve coordinated optimization of load shedding and energy storage discharge, which will easily lead to ineffective loss of energy storage resources. Therefore, a marginal loss function is constructed based on the remaining capacity percentage of each energy storage unit and the current temperature. Combined with the remaining power deficit of the candidate line combination, the pre-output power of each energy storage unit is analyzed.

[0076] Based on this, in a specific embodiment of the present invention, the method for constructing the marginal loss function is as follows: W1, the reciprocal of the rated capacity of each energy storage unit is recorded as the discharge depth per unit output power, and the ratio of the discharge depth to the current remaining capacity percentage is recorded as the discharge loss coefficient.

[0077] W2. Based on the historical discharge records of each energy storage unit, obtain the temperature rise per unit output power at the current temperature and the optimal temperature corresponding to the minimum temperature rise.

[0078] W3. The ratio of the temperature rise per unit output power to the optimal temperature is denoted as the temperature influence coefficient. The marginal loss function is constructed based on the exponential model in combination with the discharge loss coefficient.

[0079] The specific formula for the marginal loss function described in this embodiment is as follows: .

[0080] in Represents marginal loss. Represents the temperature influence coefficient. Represents the discharge loss coefficient. This represents the pre-output power of each energy storage unit. Considering that the losses of the energy storage unit increase exponentially with the increase of output power, an exponential model is adopted, which will not be elaborated further.

[0081] In a preferred embodiment of the present invention, the method for allocating the pre-output power of each energy storage unit corresponding to each candidate line combination includes: firstly, recording the difference between the total power deficit and the telemetry value of the total active power of each line in each candidate line combination as the corresponding remaining power deficit.

[0082] Then, based on the marginal loss function of each energy storage unit, and according to the principle that the marginal loss of each energy storage unit is equal, the remaining power deficit corresponding to each candidate line combination is allocated to each energy storage unit, so as to obtain the pre-output power and corresponding marginal loss of each energy storage unit under each candidate line combination.

[0083] Specifically, by simultaneously solving the marginal loss functions of each energy storage unit, for example... = =M; This yields the pre-output power expression for each energy storage unit, including the unknown M. , By making , sum + Equal to the remaining power deficit, solve for Substituting these values ​​into the above pre-output power expression yields the pre-output power corresponding to each energy storage unit.

[0084] It should be noted that when allocating power according to the principle of equal marginal loss, if the calculated pre-output power of a certain energy storage unit exceeds its rated maximum output power, then its rated maximum output power is taken as the actual allocation value, and the remaining power deficit is redistributed among the other energy storage units according to the principle of equal marginal loss, and so on.

[0085] This invention obtains the remaining power deficit of candidate line combinations, constructs a marginal loss function based on the remaining capacity percentage of each energy storage unit and the current temperature, analyzes the pre-output power of each energy storage unit corresponding to each candidate line combination, avoids excessive energy storage loss due to ignoring the operating status, reduces the equipment operation and maintenance cost of grid fault control, and improves energy storage utilization efficiency.

[0086] S5. Analyze the optimal combination based on the marginal loss of each candidate line combination and the expected recovery time of the corresponding candidate line, and execute the corresponding load shedding operation and energy storage discharge control according to the optimal combination.

[0087] Considering that the implementation effect of candidate line combinations is related to the marginal loss of energy storage, the expected recovery time of the fault, and the scale of load shedding, the marginal loss reflects the economy of energy storage discharge, the expected recovery time reflects the efficiency of fault control, and the scale of load shedding is related to the economy and stability of power grid supply. The three factors together determine the comprehensive effect of fault coordination control. If the optimal combination is selected based on only a single indicator, it is impossible to take into account the economy, efficiency and stability of power grid fault control. Therefore, it is necessary to construct a comprehensive cost evaluation system based on the three factors, analyze the comprehensive cost of each candidate line combination, select the optimal combination with the minimum comprehensive cost, and execute the corresponding load shedding operation and energy storage discharge control accordingly.

[0088] Based on this, such as Figure 3 As shown, the analysis method for the optimal combination includes: S51, obtaining the difference between the marginal loss corresponding to each candidate line combination and the minimum marginal loss among all candidate line combinations, and recording the ratio of this difference to the maximum marginal loss as the marginal loss cost.

[0089] S52. Extract the historical average voltage recovery time of each candidate line in each candidate line combination from the historical switching records of each switchable line.

[0090] S53. Obtain the maximum historical average voltage recovery time of each candidate line in each candidate line combination, and record it as the expected recovery time of the corresponding candidate line combination. Calculate the recovery time cost using the same method as the marginal loss cost. The specific steps are as follows: Obtain the difference between the expected recovery time of each candidate line combination and the minimum expected recovery time among all candidate line combinations, and record the ratio of this difference to the maximum expected recovery time as the recovery time cost.

[0091] S54. Obtain the ratio of the number of candidate routes to the maximum number of candidate routes in each candidate route combination, and denote it as the scale cost.

[0092] S55. The candidate line combination that minimizes the sum of marginal loss cost, recovery time cost, and scale cost is designated as the optimal combination. Obtain the pre-output power of each energy storage unit corresponding to the optimal combination and the corresponding candidate lines, control each energy storage unit to output at the pre-output power, and disconnect the corresponding candidate lines.

[0093] This invention analyzes the optimal combination based on the marginal loss of each candidate line combination and the expected recovery time of the corresponding candidate line, and executes the corresponding load shedding operation and energy storage discharge control according to the optimal combination. This achieves the comprehensive minimization of energy storage life loss and load shedding recovery cost, and improves the economy and power supply recovery capability of the local power grid.

[0094] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.

[0095] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0096] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0097] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0098] Finally, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for coordinated control of low-frequency faults in a local power grid with source, grid, load, and storage, characterized in that, include: When a low-frequency fault is detected in the local power grid, the current actual inertia coefficient is dynamically corrected based on the changes in active power at various historical time points, combined with the virtual inertia coefficient of grid-type new energy, and the total power deficit is calculated. Real-time prediction of voltage change amplitude at corresponding nodes after each load line is disconnected and load rate of adjacent load lines, from which initial screening of disconnectable lines is performed. Based on the importance level of the switchable lines, the active power telemetry value, and the historical line stability score of the switchable lines, the switchable lines are prioritized and ranked. Candidate line combinations are constructed based on the total power deficit and priority ranking. Obtain the remaining power deficit of candidate line combinations, construct a marginal loss function based on the remaining capacity percentage of each energy storage unit and the current temperature, and analyze the pre-output power of each energy storage unit corresponding to each candidate line combination. Based on the marginal loss of each candidate line combination and the expected recovery time of the corresponding candidate line, the optimal combination is analyzed, and the corresponding load shedding operation and energy storage discharge control are executed according to the optimal combination.

2. The method for coordinated control of low-frequency faults in a local power grid based on source, grid, load, and storage as described in claim 1, is characterized in that... The analysis method for the total power deficit includes: The active power change of each load line in the local power grid is obtained at each time point within a set historical time window. If the active power change of a certain load line exceeds the set normal fluctuation range at multiple consecutive time points, the load line is identified as a disturbance line, and the time period corresponding to the multiple consecutive time points is recorded as a disturbance. The corresponding system disturbance power is obtained by vector summing the changes in active power of each load line corresponding to each disturbance at the first time point; Collect the bus voltage frequency change rate and system disturbance power of the local power grid during each disturbance within a set historical time window, and calculate the initial inertia coefficient by combining the bus voltage frequency change rate. Based on the initial inertia coefficients of each disturbance, combined with the system's preset inertia coefficients and the virtual inertia coefficients of grid-type new energy during each disturbance process, the current actual inertia coefficients are obtained through weighted recursive updates. The total power deficit is calculated based on the current actual inertia coefficient and the current bus voltage frequency change rate.

3. The method for coordinated control of low-frequency faults in a local power grid based on source, grid, load, and storage as described in claim 2, is characterized in that... The method for obtaining the current actual inertia coefficient includes: Obtain the system's preset inertia coefficient and the virtual inertia coefficient of the grid-type new energy during each disturbance process. Then, perform a weighted summation of the initial inertia coefficient and the preset inertia coefficient corresponding to the first disturbance within the set historical time window. Use the difference between the result and the corresponding virtual inertia coefficient as the updated body inertia coefficient. The difference between the initial inertia coefficient and the corresponding virtual inertia coefficient for each remaining disturbance within the set historical time window is recorded as the initial body inertia coefficient for each disturbance. The initial body inertia coefficients of each iteration are weighted and calculated together with the previous updated body inertia coefficients to obtain the updated body inertia coefficients of each disturbance. The sum of the updated ontological inertia coefficient corresponding to the last disturbance within the set historical time window and the current virtual inertia coefficient is recorded as the current actual inertia coefficient.

4. The method for coordinated control of low-frequency faults in a local power grid based on source, grid, load, and storage as described in claim 1, characterized in that, The specific method for screening cuttable lines includes: Extract active power telemetry values, current node voltage, and current current from the status information of each load line, and filter out candidate lines whose active power telemetry values ​​are not zero. Based on the active power telemetry values ​​of each candidate line and the voltage-power sensitivity coefficient, the voltage change amplitude is analyzed, and the predicted node voltage after the candidate line is disconnected is obtained by combining the current node voltage. Based on the historical disturbance records of each candidate line, determine whether the adjacent load line will be overloaded. If the telemetry value of the active power of a candidate line is less than or equal to the total power deficit, and the predicted voltage of its corresponding node is not lower than the preset safety threshold and the adjacent lines are not overloaded after the candidate line is cut off, then the line is marked as a cut-off line.

5. The method for coordinated control of low-frequency faults in a local power grid based on source, grid, load, and storage as described in claim 4, is characterized in that... The method for determining whether an adjacent load line will experience an overload includes: Other load lines sharing nodes with each candidate line are denoted as adjacent load lines. Based on the current change of each adjacent load line during each historical disturbance of each candidate line, the current sensitivity factor is calculated. The product of the current sensitivity factor of each adjacent load line and the telemetry value of the active power of the corresponding candidate line is recorded as the current change. The sum of the current change of each adjacent load line and the current is recorded as the predicted current, and the ratio of the predicted current to the corresponding rated current is recorded as the load rate of each adjacent load line. If the load rate of an adjacent load line of a candidate line is greater than the set load rate threshold, it is determined that the adjacent load line will be overloaded.

6. The method for coordinated control of low-frequency faults in a local power grid based on source, grid, load, and storage as described in claim 1, is characterized in that... The method for constructing the candidate route combination includes: Based on the load type of each switchable line, they are divided into different importance levels according to preset rules, and the switchable lines are sorted in reverse order according to their importance level. Obtain the difference between the active power telemetry value and the maximum active power telemetry value of each cut-away line of the same importance level, and record the ratio of the difference to the maximum active power telemetry value as the optimal difference of active power telemetry value. The line stability score is analyzed based on the historical switching records of each switchable line, and the optimal line stability gap is obtained based on the line stability score of each switchable line and the maximum line stability score. The sum of the optimal difference between the active power telemetry values ​​and the optimal difference between the line stability values ​​is recorded as the comprehensive optimal difference. The priority of each severable line is determined by sorting the severable lines of the same importance level in ascending order according to the comprehensive optimal difference. Based on the priority of each switchable line, the active power telemetry values ​​of each switchable line are accumulated sequentially in unit steps. When the total active power telemetry value corresponding to a certain accumulation is greater than the total power deficit for the first time, the line set corresponding to each accumulation before that accumulation is obtained and recorded as candidate line combinations respectively. The total remaining output power of the energy storage unit is compared with the total power deficit, and the candidate line combination is optimized.

7. The method for coordinated control of low-frequency faults in a local power grid based on source, grid, load, and storage as described in claim 6, is characterized in that... The method for obtaining the line stability score includes: Obtain historical switching records for each switchable line, and extract the starting voltage before each switching and the lowest voltage value within a set time period after the switching occurs. Obtain the difference between the starting voltage and the minimum voltage value, and record the ratio of this difference to the starting voltage as the voltage change rate. The average voltage change rate of each switchable line in each historical switching record is calculated and recorded as the historical average voltage change rate. The difference between the maximum and minimum historical average voltage change rates of all switchable lines is denoted as the relative voltage change rate. Obtain the absolute difference between the historical average voltage change rate of each switchable line and the maximum historical average voltage change rate among all switchable lines, and record the ratio of this difference to the relative voltage change rate as the line stability coefficient. The line stability score is the product of the line stability coefficient and the total score of the set line stability rating.

8. The method for coordinated control of low-frequency faults in a local power grid based on source, grid, load, and storage as described in claim 1, is characterized in that... The marginal loss function is constructed as follows: The reciprocal of the rated capacity of each energy storage unit is denoted as the depth of discharge per unit output power, and the ratio of the depth of discharge to the current percentage of remaining capacity is denoted as the discharge loss coefficient. Based on the historical discharge records of each energy storage unit, the temperature rise per unit output power at the current temperature and the optimal temperature corresponding to the minimum temperature rise are obtained respectively. The ratio of the temperature rise per unit output power to the optimal temperature is denoted as the temperature influence coefficient. A marginal loss function is constructed based on an exponential model, in conjunction with the discharge loss coefficient.

9. The method for coordinated control of low-frequency faults in a local power grid based on source, grid, load, and storage as described in claim 8, is characterized in that... The method for allocating the pre-output power of each energy storage unit corresponding to each candidate circuit combination includes: The difference between the total power deficit and the telemetry value of the total active power of each line in each candidate line combination is recorded as the corresponding remaining power deficit. Based on the marginal loss function of each energy storage unit, and in accordance with the principle that the marginal loss of each energy storage unit is equal, the remaining power deficit corresponding to each candidate line combination is allocated to each energy storage unit, so as to obtain the pre-output power and corresponding marginal loss of each energy storage unit under each candidate line combination.

10. The method for coordinated control of low-frequency faults in a local power grid based on source, grid, load, and storage as described in claim 9, is characterized in that... The analysis method for the optimal combination includes: The difference between the marginal loss corresponding to each candidate route combination and the minimum marginal loss among all candidate route combinations is obtained, and the ratio of this difference to the maximum marginal loss is recorded as the marginal loss cost. Extract the historical average voltage recovery time of each candidate line in each candidate line combination from the historical switching records of each switchable line. The maximum value of the historical average voltage recovery time of each candidate line in each candidate line combination is obtained and recorded as the expected recovery time of the corresponding candidate line combination. The recovery time cost is obtained by using the same calculation method as the marginal loss cost. The ratio of the number of candidate routes to the maximum number of candidate routes in each candidate route combination is recorded as the scale cost. The candidate route combination that minimizes the sum of marginal loss cost, recovery time cost, and scale cost is denoted as the optimal combination.

Images