A method and device for predicting commutation failure of a high voltage direct current transmission system

Abstract

The application discloses a high-voltage direct current power transmission system commutation failure prediction method and device, and distinguishes short-circuit faults and breaking faults from each other in causing high-voltage direct current power transmission system commutation failure, and utilizes two machine learning models of classification and regression to respectively predict whether different faults cause commutation failure, so that the robustness of a prediction result is ensured, meanwhile, the machine learning models are adopted, manual experience is not relied on, and the calculation time is greatly reduced.

Description

Technical Field

[0001] This invention relates to a method and device for predicting commutation failure in a high-voltage direct current transmission system, belonging to the field of power system and automation technology. Background Technology

[0002] To effectively address the geographically reversed distribution of energy resources and load centers, thyristor-based high-voltage direct current (HVDC) transmission systems have been widely adopted worldwide due to their advantages in large-capacity, long-distance power transmission. However, the dense integration of HVDC into the AC grid has created complex multi-infeed HVDC systems, posing new challenges to the safe and stable operation of the power system. Inverter-side commutation failure is one of the most common faults in HVDC transmission systems. Sustained commutation failures can lead to HVDC blocking, causing significant impact on the receiving-end AC system and potentially resulting in widespread power outages.

[0003] Online safety and stability analysis and control (OSA) possesses the ability to provide risk warnings and control measures recommendations for current and future scenarios, making it an important tool for dispatchers in power grid operation control. Currently, OSA employs time-domain simulation by establishing detailed dynamic simulation models and considering protection and control response characteristics to assess whether anticipated faults will trigger characteristic events such as DC commutation failure. However, the high proportion of renewable energy and power electronic equipment integrated into the grid exacerbates the strong time-varying and uncertain factors of the power system, significantly increasing the size of the anticipated fault set. Completing detailed time-domain simulations of a large-scale anticipated fault set within a 10-minute online computation cycle would consume substantial computational resources.

[0004] To avoid the time-consuming nature of time-domain simulations and improve evaluation speed, some studies have calculated the short-circuit ratio and effective short-circuit ratio of the AC system after a fault, and used thresholds to determine whether the DC system has experienced commutation failure or blocking. However, using short-circuit ratio thresholds relies too heavily on human experience. Other studies have further introduced machine learning methods, using steady-state features related to network structure and fault location as inputs to establish fast judgment models for DC commutation failure and DC blocking based on shallow or deep learning. However, simply using steady-state features as inputs makes it difficult to guarantee the robustness of the prediction results of machine learning algorithms. Summary of the Invention

[0005] This invention provides a method and apparatus for predicting commutation failure in a high-voltage direct current transmission system, which solves the problems disclosed in the background art.

[0006] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows:

[0007] A method for predicting commutation failure in a high-voltage direct current transmission system includes:

[0008] Based on the grid ground-state operation information and short-circuit fault information, the short-circuit voltage and multi-infeed short-circuit ratio of the receiving-end converter station node after the short-circuit fault, and the multi-infeed interaction factor between the receiving-end converter station nodes after the short-circuit fault are obtained.

[0009] The lowest short-circuit voltage, the multi-infeed short-circuit ratio of the lowest voltage receiving-end converter station node, and the maximum value of the multi-infeed interaction factor of the lowest voltage receiving-end converter station node to other receiving-end converter stations are input into the pre-trained first classification model and first regression model to obtain the first prediction result of whether the short-circuit fault will cause commutation failure and the prediction value of commutation failure margin after the short-circuit fault, respectively.

[0010] Based on the grid's base state operation information and fault information, obtain the voltage change value of the receiving-end converter station node before and after the fault, the sum of the active power change of the connected AC lines, the sum of the reactive power change of the connected AC lines, and the multi-infeed short-circuit ratio of the receiving-end converter station node after the fault.

[0011] The maximum voltage change of the receiving-end converter station node before and after the fault, the sum of the maximum active power change of the connected AC lines, the sum of the maximum reactive power change of the connected AC lines, and the maximum change in the multi-infeed short-circuit ratio are input into the pre-trained second classification model and second regression model to obtain the second prediction result of whether the fault will cause commutation failure and the prediction value of commutation failure margin after the fault. Among them, the change in the multi-infeed short-circuit ratio is the absolute value of the difference between the multi-infeed short-circuit ratio of the receiving-end converter station node after the fault and the multi-infeed short-circuit ratio of the receiving-end converter station node before the fault.

[0012] Based on the first prediction result of whether a short-circuit fault will cause commutation failure, the predicted value of commutation failure margin after a short-circuit fault, the second prediction result of whether an interruption fault will cause commutation failure, and the predicted value of commutation failure margin after an interruption fault, the prediction result of commutation failure in the high-voltage direct current transmission system is obtained.

[0013] Based on the grid's ground-state operation information and short-circuit fault information, the short-circuit voltage and multi-infeed short-circuit ratio of the receiving-end converter station nodes after a short-circuit fault, as well as the multi-infeed interaction factors between the receiving-end converter station nodes after a short-circuit fault, are obtained, including:

[0014] Based on the power grid's ground-state operating mode information, the system's ground-state admittance matrix is ​​constructed, and ground-state power flow calculations are performed.

[0015] Based on the location of the node where the short-circuit fault occurred and the location of the receiving-end converter station node in the short-circuit fault information, the system ground state guided matrix is ​​simplified to obtain a multi-port equivalent impedance matrix that includes the short-circuit node and the receiving-end converter station node.

[0016] Based on the solution calculated from the multi-port equivalent impedance matrix and the ground-state power flow, the short-circuit voltage and multi-infeed short-circuit ratio of the receiving-end converter station nodes after a short-circuit fault, as well as the multi-infeed interaction factor between the receiving-end converter station nodes after a short-circuit fault, are obtained.

[0017] The short-circuit voltage at the receiving-end converter station node after a short-circuit fault is calculated using the following formula:

[0018]

[0019] Among them, U k This represents the short-circuit voltage at node k of the receiving-end converter station after a short-circuit fault. The voltage amplitude at node k of the receiving-end converter station is obtained from the ground-state power flow calculation. Z represents the voltage magnitude at node j obtained from the ground-state power flow calculation. Node j is the node where the short-circuit fault occurs. kj Z represents the mutual impedance between nodes k and j at the receiving end of the converter station in the multi-port equivalent impedance matrix. jj Let be the self-impedance of node j in the multi-port equivalent impedance matrix;

[0020] The formula for calculating the multi-infeed short-circuit ratio of the receiving-end converter station node after a short-circuit fault is as follows:

[0021]

[0022] Among them, MISCR k Z represents the multi-infeed short-circuit ratio at node k of the receiving-end converter station after a short-circuit fault. kk,短路 P represents the self-impedance of node k at the receiving end of the converter station in the multi-port equivalent impedance matrix before disconnection. d,k P represents the DC power at node k of the receiving-end converter station. d,c Z represents the DC power at node c of the receiving-end converter station. kc,短路 is the mutual impedance of receiving-end converter station node k and receiving-end converter station node c in the multi-port equivalent impedance matrix before disconnection, and h is the number of receiving-end converter station nodes.

[0023] The multi-feed interaction factor between nodes of the receiving-end converter station after a short-circuit fault is calculated using the following formula:

[0024]

[0025] Among them, MIIF ck Z is the multi-feed interaction factor of receiving-end converter station node k to receiving-end converter station node c after a short-circuit fault. ck Let be the mutual impedance between receiving-end converter station node c and receiving-end converter station node k in the multi-port equivalent impedance matrix.

[0026] Based on the grid's base-state operation information and fault interruption information, obtain the voltage change values ​​of the receiving-end converter station nodes before and after the fault interruption, the sum of the active power changes of the connected AC lines, the sum of the reactive power changes of the connected AC lines, and the multi-infeed short-circuit ratio of the receiving-end converter station nodes after the fault interruption, including:

[0027] Based on the power grid's ground-state operating mode information, the system's ground-state admittance matrix is ​​constructed, and ground-state power flow calculations are performed.

[0028] Modify the system ground state admittance matrix based on the fault information to obtain the system admittance matrix after the fault, and perform power flow calculation after the fault.

[0029] Based on the location of the receiving-end converter station node, the admittance matrix after system interruption is simplified to obtain the multi-port equivalent impedance matrix including the receiving-end converter station node.

[0030] Based on the multi-port equivalent impedance matrix, the solution of the ground-state power flow calculation, and the solution of the power flow calculation after the fault is interrupted, the voltage change value of the receiving-end converter station node before and after the fault is interrupted, the sum of the active power change of the connected AC lines, the sum of the reactive power change of the connected AC lines, and the multi-infeed short-circuit ratio of the receiving-end converter station node after the fault is interrupted are obtained.

[0031] The voltage change at the receiving-end converter station node before and after the fault is interrupted is calculated using the following formula:

[0032]

[0033] Wherein, ΔU k The voltage change at node k of the receiving-end converter station before and after the fault is interrupted is given. The voltage amplitude at node k of the receiving-end converter station is obtained from the ground-state power flow calculation. The voltage amplitude at node k of the receiving-end converter station is obtained from the power flow calculation after the fault is interrupted.

[0034] The sum of the changes in active power of the AC lines connected to the receiving-end converter station nodes before and after the fault is interrupted is calculated using the following formula:

[0035]

[0036] Wherein, ΔP k R is the sum of the changes in active power of the AC lines connected to node k of the receiving-end converter station before and after the fault is interrupted. k This represents the number of AC lines connected to node k of the receiving-end converter station. The active power at node k of the receiving-end converter station is obtained from the ground-state power flow calculation. The active power of node k in the receiving-end converter station is obtained from the power flow calculation after the fault is interrupted.

[0037] The sum of the reactive power changes of the AC lines connected to the receiving-end converter station nodes before and after the fault is interrupted is calculated using the following formula:

[0038]

[0039] Where, ΔQ k It is the sum of the zero-power changes of the AC lines connected to node k of the receiving-end converter station before and after the fault is interrupted. The zero power of node k at the receiving-end converter station, obtained from the ground-state power flow calculation. The zero power of node k in the receiving-end converter station obtained from the power flow calculation after the fault is interrupted;

[0040] The multi-infeed short-circuit ratio of the receiving-end converter station node after the fault is interrupted is calculated using the following formula:

[0041]

[0042] Among them, MISCR k,开断 Z represents the multi-infeed short-circuit ratio of node k at the receiving-end converter station after a fault is interrupted. kk,开断 P represents the self-impedance of node k at the receiving end of the converter station in the multi-port equivalent impedance matrix after a fault is interrupted. d,k P represents the DC power at node k of the receiving-end converter station. d,c Z represents the DC power at node c of the receiving-end converter station. kc,开断 is the mutual impedance between node k and node c of the receiving-end converter station in the multi-port equivalent impedance matrix after the fault is interrupted, and h is the number of nodes in the receiving-end converter station.

[0043] Based on the first prediction result of whether a short-circuit fault will cause commutation failure, the predicted value of commutation failure margin after a short-circuit fault, the second prediction result of whether an interruption fault will cause commutation failure, and the predicted value of commutation failure margin after an interruption fault, the prediction results of commutation failure in the high-voltage direct current transmission system are obtained, including:

[0044] If the first prediction result is yes and the predicted value of the commutation failure margin after the short-circuit fault is lower than the first threshold, or if the second prediction result is yes and the predicted value of the commutation failure margin after the opening fault is lower than the second threshold, then it is determined that the fault caused the commutation failure of the high-voltage direct current transmission system.

[0045] A commutation failure prediction system for a high-voltage direct current transmission system includes:

[0046] The short-circuit fault characteristic module obtains the short-circuit voltage and multi-infeed short-circuit ratio of the receiving-end converter station node after a short-circuit fault, as well as the multi-infeed interaction factor between the receiving-end converter station nodes after a short-circuit fault, based on the grid ground-state operation mode information and short-circuit fault information.

[0047] The short-circuit fault prediction module inputs the lowest short-circuit voltage, the multi-infeed short-circuit ratio of the lowest voltage receiving-end converter station node, and the maximum value of the multi-infeed interaction factor of the lowest voltage receiving-end converter station node to other receiving-end converter stations into the pre-trained first classification model and first regression model, respectively, to obtain the prediction results of whether the short-circuit fault will cause commutation failure and the prediction results of the commutation failure margin value after the short-circuit fault.

[0048] The fault interruption feature module obtains the voltage change value of the receiving-end converter station node before and after the fault interruption, the sum of the active power change of the connected AC lines, the sum of the reactive power change of the connected AC lines, and the multi-infeed short-circuit ratio of the receiving-end converter station node after the fault interruption, based on the grid base state operation mode information and fault interruption information.

[0049] The fault prediction module inputs the maximum voltage change of the receiving-end converter station node before and after the fault, the sum of the maximum active power change of the connected AC lines, the sum of the maximum reactive power change of the connected AC lines, and the maximum change in the multi-infeed short-circuit ratio into the pre-trained second classification model and second regression model to obtain the prediction results of whether the fault will cause commutation failure and the prediction results of the commutation failure margin after the fault. Among them, the change in the multi-infeed short-circuit ratio is the absolute value of the difference between the multi-infeed short-circuit ratio of the receiving-end converter station node after the fault and the multi-infeed short-circuit ratio of the receiving-end converter station node before the fault.

[0050] The results module obtains the prediction results of commutation failure in the high-voltage direct current transmission system based on the prediction results of whether a short-circuit fault will cause commutation failure, the prediction results of the commutation failure margin after a short-circuit fault, the prediction results of whether an opening fault will cause commutation failure, and the prediction results of the commutation failure margin after an opening fault.

[0051] The short-circuit fault characteristic module obtains the short-circuit voltage and multi-infeed short-circuit ratio of the receiving-end converter station nodes after a short-circuit fault, as well as the multi-infeed interaction factor between the receiving-end converter station nodes after a short-circuit fault. The process includes:

[0052] Based on the power grid's ground-state operating mode information, the system's ground-state admittance matrix is ​​constructed, and ground-state power flow calculations are performed.

[0053] Based on the location of the node where the short-circuit fault occurred and the location of the receiving-end converter station node in the short-circuit fault information, the system ground state guided matrix is ​​simplified to obtain a multi-port equivalent impedance matrix that includes the short-circuit node and the receiving-end converter station node.

[0054] Based on the solution calculated from the multi-port equivalent impedance matrix and the ground-state power flow, the short-circuit voltage and multi-infeed short-circuit ratio of the receiving-end converter station nodes after a short-circuit fault, as well as the multi-infeed interaction factor between the receiving-end converter station nodes after a short-circuit fault, are obtained.

[0055] The fault interruption feature module acquires the voltage change value of the receiving-end converter station node before and after the fault interruption, the sum of the active power change of the connected AC lines, the sum of the reactive power change of the connected AC lines, and the multi-infeed short-circuit ratio of the receiving-end converter station node after the fault interruption. The process includes:

[0056] Based on the power grid's ground-state operating mode information, the system's ground-state admittance matrix is ​​constructed, and ground-state power flow calculations are performed.

[0057] Modify the system ground state admittance matrix based on the fault information to obtain the system admittance matrix after the fault, and perform power flow calculation after the fault.

[0058] Based on the location of the receiving-end converter station node, the admittance matrix after system interruption is simplified to obtain the multi-port equivalent impedance matrix including the receiving-end converter station node.

[0059] Based on the multi-port equivalent impedance matrix, the solution of the ground-state power flow calculation, and the solution of the power flow calculation after the fault is interrupted, the voltage change value of the receiving-end converter station node before and after the fault is interrupted, the sum of the active power change of the connected AC lines, the sum of the reactive power change of the connected AC lines, and the multi-infeed short-circuit ratio of the receiving-end converter station node after the fault is interrupted are obtained.

[0060] A computer-readable storage medium storing one or more programs, the one or more programs including instructions that, when executed by a computing device, cause the computing device to perform a method for predicting commutation failures in a high-voltage direct current transmission system.

[0061] The beneficial effects achieved by this invention are as follows: This invention distinguishes between commutation failures caused by short-circuit faults and open-circuit faults in high-voltage direct current transmission systems. It uses two machine learning models, classification and regression, to predict whether different faults will cause commutation failures. This not only ensures the robustness of the prediction results, but also significantly reduces the computation time because it does not rely on human experience due to the use of machine learning models. Attached Figure Description

[0062] Figure 1 A flowchart of a method for predicting commutation failure in a high-voltage direct current transmission system;

[0063] Figure 2 A flowchart for model training for short-circuit faults;

[0064] Figure 3 This is a flowchart of the model training process for open / closed faults. Detailed Implementation

[0065] The present invention will be further described below with reference to the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and should not be used to limit the scope of protection of the present invention.

[0066] like Figure 1 As shown, a method for predicting commutation failure in a high-voltage direct current transmission system includes the following steps:

[0067] Step 1: Based on the grid ground-state operation information and short-circuit fault information, obtain the short-circuit voltage and multi-infeed short-circuit ratio of the receiving-end converter station node after the short-circuit fault, and the multi-infeed interaction factor between the receiving-end converter station nodes after the short-circuit fault.

[0068] Step 2: Input the lowest short-circuit voltage, the multi-infeed short-circuit ratio of the lowest voltage receiving-end converter station node, and the maximum value of the multi-infeed interaction factor of the lowest voltage receiving-end converter station node to other receiving-end converter stations into the pre-trained first classification model and first regression model to obtain the first prediction result of whether the short-circuit fault will cause commutation failure and the prediction value of commutation failure margin after the short-circuit fault.

[0069] Step 3: Based on the grid base state operation information and fault information, obtain the voltage change value of the receiving-end converter station node before and after the fault, the sum of the active power change of the connected AC lines, the sum of the reactive power change of the connected AC lines, and the multi-infeed short-circuit ratio of the receiving-end converter station node after the fault.

[0070] Step 4: Input the maximum voltage change value of the receiving-end converter station node before and after the fault, the sum of the maximum active power change of the connected AC lines, the sum of the maximum reactive power change of the connected AC lines, and the maximum change in the multi-infeed short-circuit ratio into the pre-trained second classification model and second regression model to obtain the second prediction result of whether the fault will cause commutation failure and the prediction value of commutation failure margin after the fault. Among them, the change in the multi-infeed short-circuit ratio is the absolute value of the difference between the multi-infeed short-circuit ratio of the receiving-end converter station node after the fault and the multi-infeed short-circuit ratio of the receiving-end converter station node before the fault.

[0071] Step 5: Based on the first prediction result of whether a short-circuit fault will cause commutation failure, the predicted value of commutation failure margin after a short-circuit fault, the second prediction result of whether an opening fault will cause commutation failure, and the predicted value of commutation failure margin after an opening fault, obtain the prediction result of commutation failure of the high-voltage direct current transmission system.

[0072] The above method distinguishes between commutation failures caused by short-circuit faults and open-circuit faults in high-voltage direct current transmission systems. It uses two machine learning models, classification and regression, to predict whether different faults will cause commutation failures. This not only ensures the robustness of the prediction results, but also significantly reduces computation time by using machine learning models and eliminating the need for human experience.

[0073] Before implementing the above method, two machine learning models can be constructed, namely a classification model and a regression model. The two models predict whether a fault will cause commutation failure and the commutation failure margin after a fault, respectively. Before using the model, it is necessary to sample and train it in advance using machine learning methods. There are various machine learning methods, such as neural networks and support vector machines. The method is not limited in this invention.

[0074] This invention distinguishes between commutation failures in high-voltage direct current transmission systems caused by short-circuit faults and open-circuit faults. Therefore, it is necessary to train different models for different faults. For short-circuit faults, a first classification model and a first regression model are trained, and for open-circuit faults, a first classification model and a first regression model are trained.

[0075] like Figure 2 As shown, the training process for the first classification model and the first regression model is as follows:

[0076] A1) Construct a large set of three-phase instantaneous short-circuit faults;

[0077] A2) Transient time-domain simulation is used to calculate the fault set. During the transient time-domain simulation, it is marked whether a system commutation failure has occurred, and the commutation safety margin of the system after a short-circuit fault is calculated. The formula for calculating the commutation safety margin σ is as follows:

[0078]

[0079] Where, γ c,set γ is the set value for the critical shut-off angle. min γ is the minimum shut-off angle in the simulation. ref The setpoint for the inverter operating at a given turn-off angle, where λ is the conversion factor, and k is the setpoint. c The turn-off angle trajectory at γ c,set The first-order sensitivity at that point.

[0080] A3) Calculate the short-circuit voltage and multi-infeed short-circuit ratio of the receiving-end converter station node after the short-circuit fault, and the multi-infeed interaction factor between the receiving-end converter station nodes after the short-circuit fault, using the method in step 1.

[0081] A4) The lowest short-circuit voltage, the multi-infeed short-circuit ratio of the lowest voltage receiving-end converter station node, and the maximum value of the multi-infeed interaction factor of the lowest voltage receiving-end converter station node to other receiving-end converter stations are used as machine learning input features. Combined with the label of whether commutation failure has occurred as the total input, machine learning methods are used for training to obtain the first classification model.

[0082] The lowest short-circuit voltage, the multi-infeed short-circuit ratio of the lowest voltage receiving-end converter station node, and the maximum value of the multi-infeed interaction factor of the lowest voltage receiving-end converter station node to other receiving-end converter stations are used as machine learning input features. Combined with the commutation safety margin as the total input, machine learning methods are used for training to obtain the first regression model.

[0083] like Figure 3 As shown, the training process for the second classification model and the second regression model is as follows:

[0084] B1) Construct a large set of open faults;

[0085] B2) Transient time-domain simulation calculation is used for the fault set. During the transient time-domain simulation calculation, the system commutation failure is marked and the commutation safety margin of the system after the fault is interrupted is calculated.

[0086] B3) Calculate the voltage change value of the receiving-end converter station node before and after the fault is broken, the sum of the active power change of the connected AC lines, the sum of the reactive power change of the connected AC lines, and the multi-infeed short-circuit ratio of the receiving-end converter station node after the fault is broken, using the method in step 3.

[0087] B4) The maximum voltage change of the receiving-end converter station node before and after the fault is interrupted, the sum of the maximum active power change of the connected AC lines, the sum of the maximum reactive power change of the connected AC lines, and the maximum change of the multi-infeed short-circuit ratio are used as machine learning input features. Combined with the label of whether commutation failure has occurred as the total input, machine learning methods are used for training to obtain the second classification model.

[0088] The maximum voltage change at the receiving-end converter station node before and after the fault is interrupted, the sum of the maximum active power change of the connected AC lines, the sum of the maximum reactive power change of the connected AC lines, and the maximum change of the multi-infeed short-circuit ratio are used as machine learning input features. Combined with the commutation safety margin as the total input, the machine learning method is used for training to obtain the second regression model.

[0089] In the above method, the grid's basic operating mode information and short-circuit fault information can be obtained first. Based on the grid's basic operating mode information, the system's basic admittance matrix can be constructed, and basic power flow calculation can be performed. Based on the grid's basic operating mode information and short-circuit fault information, the short-circuit voltage, multi-infeed short-circuit ratio, and multi-infeed interaction factor of the receiving-end converter station node after the short-circuit fault can be obtained. The specific process is as follows:

[0090] 11) The power flow solution obtained from the ground state power flow calculation includes the node voltage magnitude and phase angle, the active and reactive power of the line, and the active and reactive power of the transformer, etc.

[0091] 12) Based on the location of the node where the short circuit fault occurred and the location of the receiving-end converter station node in the short circuit fault information, the system ground state guided moment nanoarray is simplified to obtain a multi-port equivalent impedance matrix that includes the short circuit node and the receiving-end converter station node.

[0092] 13) Based on the solution calculated from the multi-port equivalent impedance matrix and the ground-state power flow, obtain the short-circuit voltage and multi-infeed short-circuit ratio of the receiving-end converter station node after the short-circuit fault, and the multi-infeed interaction factor between the receiving-end converter station nodes after the short-circuit fault.

[0093] The calculation formulas for each parameter in step 13) are as follows:

[0094] The short-circuit voltage at the receiving-end converter station node after a short-circuit fault is calculated using the following formula:

[0095]

[0096] Among them, U k This represents the short-circuit voltage at node k of the receiving-end converter station after a short-circuit fault. The voltage amplitude at node k of the receiving-end converter station is obtained from the ground-state power flow calculation. Z represents the voltage magnitude at node j obtained from the ground-state power flow calculation. Node j is the node where the short-circuit fault occurs. kj Z represents the mutual impedance between nodes k and j at the receiving end of the converter station in the multi-port equivalent impedance matrix. jj Let be the self-impedance of node j in the multi-port equivalent impedance matrix;

[0097] The formula for calculating the multi-infeed short-circuit ratio of the receiving-end converter station node after a short-circuit fault is as follows:

[0098]

[0099] Among them, MISCR k Z represents the multi-infeed short-circuit ratio at node k of the receiving-end converter station after a short-circuit fault. kk,短路 P represents the self-impedance of node k at the receiving end of the converter station in the multi-port equivalent impedance matrix before disconnection. d,k P represents the DC power at node k of the receiving-end converter station. d,c Z represents the DC power at node c of the receiving-end converter station. kc,短路 is the mutual impedance of receiving-end converter station node k and receiving-end converter station node c in the multi-port equivalent impedance matrix before disconnection, and h is the number of receiving-end converter station nodes.

[0100] The multi-feed interaction factor of the receiving-end converter station node after a short-circuit fault is calculated using the following formula:

[0101]

[0102] Among them, MIIF ckZ is the multi-feed interaction factor of receiving-end converter station node k to receiving-end converter station node c after a short-circuit fault. ck Let be the mutual impedance between receiving-end converter station node c and receiving-end converter station node k in the multi-port equivalent impedance matrix.

[0103] Furthermore, the lowest short-circuit voltage, the multi-infeed short-circuit ratio of the lowest voltage receiving-end converter station node, and the multi-infeed interaction factor of the lowest voltage receiving-end converter station node can be obtained. These can be used as inputs to the first classification model and the first regression model to obtain the first prediction result of whether a short-circuit fault will cause commutation failure and the prediction value of commutation failure margin after a short-circuit fault.

[0104] Similar to the short-circuit faults mentioned above, obtaining fault information, based on the grid's base-state operating mode information and the fault information, allows us to obtain the voltage change values ​​of the receiving-end converter station nodes before and after the fault, the sum of the active power changes of the connected AC lines, the sum of the reactive power changes of the connected AC lines, and the multi-infeed short-circuit ratio of the receiving-end converter station nodes after the fault. The specific process is as follows:

[0105] 21) The processing procedure for the power grid ground state operation mode information is as described in 11). Since it has already been processed in 11), the system ground state admittance matrix can be directly modified based on the fault information to obtain the admittance matrix after the fault and to perform power flow calculation after the fault. The power flow solution here includes the node voltage magnitude and phase angle after the fault, the active and reactive power of the line, and the active and reactive power of the transformer.

[0106] 22) Based on the location of the receiving-end converter station node, the admittance matrix after system interruption is simplified to obtain the multi-port equivalent impedance matrix including the receiving-end converter station node;

[0107] 23) Based on the multi-port equivalent impedance matrix, the solution of the ground state power flow calculation, and the solution of the power flow calculation after the fault is interrupted, obtain the voltage change value of the receiving-end converter station node before and after the fault is interrupted, the sum of the active power change of the connected AC lines, the sum of the reactive power change of the connected AC lines, and the multi-infeed short-circuit ratio of the receiving-end converter station node after the fault is interrupted.

[0108] The calculation formulas for each parameter in step 23) are as follows:

[0109] The voltage change at the receiving-end converter station node before and after the fault is interrupted is calculated using the following formula:

[0110]

[0111] Wherein, ΔU k The voltage change at node k of the receiving-end converter station before and after the fault is interrupted is given. The voltage amplitude at node k of the receiving-end converter station is obtained from the ground-state power flow calculation. The voltage amplitude at node k of the receiving-end converter station is obtained from the power flow calculation after the fault is interrupted.

[0112] The sum of the changes in active power of the AC lines connected to the receiving-end converter station nodes before and after the fault is interrupted is calculated using the following formula:

[0113]

[0114] Wherein, ΔP k R is the sum of the changes in active power of the AC lines connected to node k of the receiving-end converter station before and after the fault is interrupted. k This represents the number of AC lines connected to node k of the receiving-end converter station. The active power at node k of the receiving-end converter station is obtained from the ground-state power flow calculation. The active power of node k in the receiving-end converter station is obtained from the power flow calculation after the fault is interrupted.

[0115] The sum of the reactive power changes of the AC lines connected to the receiving-end converter station nodes before and after the fault is interrupted is calculated using the following formula:

[0116]

[0117] Where, ΔQ k It is the sum of the zero-power changes of the AC lines connected to node k of the receiving-end converter station before and after the fault is interrupted. The zero power of node k at the receiving-end converter station, obtained from the ground-state power flow calculation. The zero power of node k in the receiving-end converter station obtained from the power flow calculation after the fault is interrupted;

[0118] The multi-infeed short-circuit ratio of the receiving-end converter station node after the fault is interrupted is calculated using the following formula:

[0119]

[0120] Among them, MISCR k,开断 Z represents the multi-infeed short-circuit ratio of node k at the receiving-end converter station after a fault is interrupted. kk,开断 P represents the self-impedance of node k at the receiving end of the converter station in the multi-port equivalent impedance matrix after a fault is interrupted. d,k P represents the DC power at node k of the receiving-end converter station. d,c Z represents the DC power at node c of the receiving-end converter station. kc,开断 is the mutual impedance between node k and node c of the receiving-end converter station in the multi-port equivalent impedance matrix after the fault is interrupted, and h is the number of nodes in the receiving-end converter station.

[0121] The change in the multi-infeed short-circuit ratio of the receiving-end converter station node before and after the fault is interrupted is calculated using the following formula:

[0122] ΔMISCR k=|MISCR k,开断 -MISCR k,短路 |

[0123] Among them, ΔMISCR k This represents the change in the multi-infeed short-circuit ratio of node k in the receiving-end converter station before and after the fault is interrupted.

[0124] Furthermore, the maximum voltage change value of the receiving-end converter station node before and after the fault is interrupted, the sum of the maximum active power change of the connected AC lines, the sum of the maximum reactive power change of the connected AC lines, and the maximum change of the multi-infeed short-circuit ratio can be obtained. These can be used as inputs to the second classification model and the second regression model to obtain the second prediction result of whether the fault interruption will cause commutation failure and the prediction value of commutation failure margin after the fault interruption.

[0125] If the first prediction result is yes and the predicted value of the commutation failure margin after the short-circuit fault is lower than the first threshold, or if the second prediction result is yes and the predicted value of the commutation failure margin after the opening fault is lower than the second threshold, then it is determined that the fault caused the commutation failure of the high-voltage direct current transmission system.

[0126] The above method combines model-driven static computation with data-driven machine learning, significantly reducing computation time and improving the efficiency of traditional time-domain simulation-based commutation failure prediction. It distinguishes between short-circuit faults and open-circuit faults that cause system commutation failures, and then uses classification and regression machine learning models to coordinate with each other, ensuring the robustness of the prediction results. It can quickly scan whether massive NK three-phase short-circuit parallel lines or transformer open-circuit faults across the entire network cause system commutation failures, providing technical support for online safety and stability analysis of power systems to quickly grasp the characteristics of the power grid and take corresponding preventive and control measures.

[0127] Based on the same technical solution, this invention also discloses a software system for the above-mentioned method, a commutation failure prediction system for a high-voltage direct current transmission system, comprising:

[0128] The short-circuit fault characteristic module obtains the short-circuit voltage and multi-infeed short-circuit ratio of the receiving-end converter station node after a short-circuit fault, as well as the multi-infeed interaction factor between the receiving-end converter station nodes after a short-circuit fault, based on the grid ground-state operation mode information and short-circuit fault information.

[0129] The short-circuit fault characteristic module obtains the short-circuit voltage and multi-infeed short-circuit ratio of the receiving-end converter station nodes after a short-circuit fault, as well as the multi-infeed interaction factor between the receiving-end converter station nodes after a short-circuit fault. The process includes:

[0130] 11) Based on the power grid's ground-state operating mode information, construct the system's ground-state admittance matrix and perform ground-state power flow calculations;

[0131] 12) Based on the location of the node where the short circuit fault occurred and the location of the receiving-end converter station node in the short circuit fault information, the system ground state guided moment nanoarray is simplified to obtain a multi-port equivalent impedance matrix that includes the short circuit node and the receiving-end converter station node.

[0132] 13) Based on the solution calculated from the multi-port equivalent impedance matrix and the ground-state power flow, obtain the short-circuit voltage and multi-infeed short-circuit ratio of the receiving-end converter station node after the short-circuit fault, and the multi-infeed interaction factor between the receiving-end converter station nodes after the short-circuit fault.

[0133] The short-circuit fault prediction module inputs the lowest short-circuit voltage, the multi-infeed short-circuit ratio of the lowest voltage receiving-end converter station node, and the maximum value of the multi-infeed interaction factor of the lowest voltage receiving-end converter station node to other receiving-end converter stations into the pre-trained first classification model and first regression model, respectively, to obtain the prediction results of whether the short-circuit fault will cause commutation failure and the prediction results of the commutation failure margin value after the short-circuit fault.

[0134] The fault interruption feature module obtains the voltage change value of the receiving-end converter station node before and after the fault interruption, the sum of the active power change of the connected AC lines, the sum of the reactive power change of the connected AC lines, and the multi-infeed short-circuit ratio of the receiving-end converter station node after the fault interruption, based on the grid base state operation mode information and fault interruption information.

[0135] The fault interruption feature module acquires the voltage change value of the receiving-end converter station node before and after the fault interruption, the sum of the active power change of the connected AC lines, the sum of the reactive power change of the connected AC lines, and the multi-infeed short-circuit ratio of the receiving-end converter station node after the fault interruption. The process includes:

[0136] 21) Based on the power grid's ground-state operating mode information, construct the system's ground-state admittance matrix and perform ground-state power flow calculations;

[0137] 22) Modify the system ground state admittance matrix according to the fault information to obtain the system admittance matrix after the fault, and perform power flow calculation after the fault.

[0138] 23) Based on the location of the receiving-end converter station node, the admittance matrix after system interruption is simplified to obtain the multi-port equivalent impedance matrix including the receiving-end converter station node;

[0139] 24) Based on the multi-port equivalent impedance matrix, the solution of the ground state power flow calculation, and the solution of the power flow calculation after the fault is interrupted, obtain the voltage change value of the receiving-end converter station node before and after the fault is interrupted, the sum of the active power change of the connected AC lines, the sum of the reactive power change of the connected AC lines, and the multi-infeed short-circuit ratio of the receiving-end converter station node after the fault is interrupted.

[0140] The fault prediction module inputs the maximum voltage change of the receiving-end converter station node before and after the fault, the sum of the maximum active power changes of the connected AC lines, the sum of the maximum reactive power changes of the connected AC lines, and the maximum change in the multi-infeed short-circuit ratio into the pre-trained classification and regression models. It then obtains the prediction results of whether the fault will cause commutation failure and the prediction results of the commutation failure margin after the fault. The change in the multi-infeed short-circuit ratio is the absolute value of the difference between the multi-infeed short-circuit ratio of the receiving-end converter station node after the fault and the multi-infeed short-circuit ratio of the receiving-end converter station node before the fault.

[0141] The results module obtains the prediction results of commutation failure in the high-voltage direct current transmission system based on the prediction results of whether a short-circuit fault will cause commutation failure, the prediction results of the commutation failure margin after a short-circuit fault, the prediction results of whether an opening fault will cause commutation failure, and the prediction results of the commutation failure margin after an opening fault.

[0142] In the above system, the data processing flow and methods of each module have the same steps, which will not be described again here.

[0143] Based on the same technical solution, the present invention also discloses a computer-readable storage medium storing one or more programs, the one or more programs including instructions that, when executed by a computing device, cause the computing device to perform a method for predicting commutation failure in a high-voltage direct current transmission system.

[0144] Based on the same technical solution, the present invention also discloses a computing device, including one or more processors, one or more memories, and one or more programs, wherein the one or more programs are stored in the one or more memories and configured to be executed by the one or more processors, and the one or more programs include instructions for executing a method for predicting commutation failure in a high-voltage direct current transmission system.

[0145] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0146] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0147] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0148] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0149] The above are merely embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention are included within the scope of the claims of the present invention pending approval.

Claims

1. A method for predicting commutation failure in a high-voltage direct current transmission system, characterized in that, include: Based on the grid ground-state operation information and short-circuit fault information, the short-circuit voltage and multi-infeed short-circuit ratio of the receiving-end converter station node after the short-circuit fault, and the multi-infeed interaction factor between the receiving-end converter station nodes after the short-circuit fault are obtained. The lowest short-circuit voltage, the multi-infeed short-circuit ratio of the lowest voltage receiving-end converter station node, and the maximum value of the multi-infeed interaction factor of the lowest voltage receiving-end converter station node to other receiving-end converter stations are input into the pre-trained first classification model and first regression model to obtain the first prediction result of whether the short-circuit fault will cause commutation failure and the prediction value of commutation failure margin after the short-circuit fault, respectively. Based on the grid's base state operation information and fault information, obtain the voltage change value of the receiving-end converter station node before and after the fault, the sum of the active power change of the connected AC lines, the sum of the reactive power change of the connected AC lines, and the multi-infeed short-circuit ratio of the receiving-end converter station node after the fault. The maximum voltage change of the receiving-end converter station node before and after the fault, the sum of the maximum active power change of the connected AC lines, the sum of the maximum reactive power change of the connected AC lines, and the maximum change in the multi-infeed short-circuit ratio are input into the pre-trained second classification model and second regression model to obtain the second prediction result of whether the fault will cause commutation failure and the prediction value of commutation failure margin after the fault. Among them, the change in the multi-infeed short-circuit ratio is the absolute value of the difference between the multi-infeed short-circuit ratio of the receiving-end converter station node after the fault and the multi-infeed short-circuit ratio of the receiving-end converter station node before the fault. Based on the first prediction result of whether a short-circuit fault will cause commutation failure, the predicted value of commutation failure margin after a short-circuit fault, the second prediction result of whether an interruption fault will cause commutation failure, and the predicted value of commutation failure margin after an interruption fault, the prediction result of commutation failure in the high-voltage direct current transmission system is obtained.

2. The method for predicting commutation failure in a high-voltage direct current transmission system according to claim 1, characterized in that, Based on the grid's ground-state operation information and short-circuit fault information, the short-circuit voltage and multi-infeed short-circuit ratio of the receiving-end converter station nodes after a short-circuit fault, as well as the multi-infeed interaction factors between the receiving-end converter station nodes after a short-circuit fault, are obtained, including: Based on the power grid's ground-state operating mode information, the system's ground-state admittance matrix is ​​constructed, and ground-state power flow calculations are performed. Based on the location of the node where the short circuit fault occurred and the location of the receiving-end converter station node in the short circuit fault information, the system ground state admittance matrix is ​​simplified to obtain a multi-port equivalent impedance matrix that includes the short circuit node and the receiving-end converter station node. Based on the solution calculated from the multi-port equivalent impedance matrix and the ground-state power flow, the short-circuit voltage and multi-infeed short-circuit ratio of the receiving-end converter station nodes after a short-circuit fault, as well as the multi-infeed interaction factor between the receiving-end converter station nodes after a short-circuit fault, are obtained.

3. The method for predicting commutation failure in a high-voltage direct current transmission system according to claim 2, characterized in that, The short-circuit voltage at the receiving-end converter station node after a short-circuit fault is calculated using the following formula: Among them, U k This represents the short-circuit voltage at node k of the receiving-end converter station after a short-circuit fault. The voltage amplitude at node k of the receiving-end converter station is obtained from the ground-state power flow calculation. Z represents the voltage magnitude at node j obtained from the ground-state power flow calculation. Node j is the node where the short-circuit fault occurs. kj Z represents the mutual impedance between nodes k and j at the receiving end of the converter station in the multi-port equivalent impedance matrix. jj Let be the self-impedance of node j in the multi-port equivalent impedance matrix; The formula for calculating the multi-infeed short-circuit ratio of the receiving-end converter station node after a short-circuit fault is as follows: Among them, MISCR k Z represents the multi-infeed short-circuit ratio at node k of the receiving-end converter station after a short-circuit fault. kk,短路 P represents the self-impedance of node k at the receiving end of the converter station in the multi-port equivalent impedance matrix before disconnection. d,k P represents the DC power at node k of the receiving-end converter station. d,c Z represents the DC power at node c of the receiving-end converter station. kc,短路 is the mutual impedance of receiving-end converter station node k and receiving-end converter station node c in the multi-port equivalent impedance matrix before disconnection, and h is the number of receiving-end converter station nodes. The multi-feed interaction factor between nodes of the receiving-end converter station after a short-circuit fault is calculated using the following formula: Among them, MIIF ck Z is the multi-feed interaction factor of receiving-end converter station node k to receiving-end converter station node c after a short-circuit fault. ck Let be the mutual impedance between receiving-end converter station node c and receiving-end converter station node k in the multi-port equivalent impedance matrix.

4. The method for predicting commutation failure in a high-voltage direct current transmission system according to claim 1, characterized in that, Based on the grid's base-state operation information and fault interruption information, obtain the voltage change values ​​of the receiving-end converter station nodes before and after the fault interruption, the sum of the active power changes of the connected AC lines, the sum of the reactive power changes of the connected AC lines, and the multi-infeed short-circuit ratio of the receiving-end converter station nodes after the fault interruption, including: Based on the power grid's ground-state operating mode information, the system's ground-state admittance matrix is ​​constructed, and ground-state power flow calculations are performed. Modify the system ground state admittance matrix based on the fault information to obtain the system admittance matrix after the fault, and perform power flow calculation after the fault. Based on the location of the receiving-end converter station node, the admittance matrix after system interruption is simplified to obtain the multi-port equivalent impedance matrix including the receiving-end converter station node. Based on the multi-port equivalent impedance matrix, the solution of the ground-state power flow calculation, and the solution of the power flow calculation after the fault is interrupted, the voltage change value of the receiving-end converter station node before and after the fault is interrupted, the sum of the active power change of the connected AC lines, the sum of the reactive power change of the connected AC lines, and the multi-infeed short-circuit ratio of the receiving-end converter station node after the fault is interrupted are obtained.

5. The method for predicting commutation failure in a high-voltage direct current transmission system according to claim 4, characterized in that, The voltage change at the receiving-end converter station node before and after the fault is interrupted is calculated using the following formula: Wherein, ΔU k The voltage change at node k of the receiving-end converter station before and after the fault is interrupted is given. The voltage amplitude at node k of the receiving-end converter station is obtained from the ground-state power flow calculation. The voltage amplitude at node k of the receiving-end converter station is obtained from the power flow calculation after the fault is interrupted. The sum of the changes in active power of the AC lines connected to the receiving-end converter station nodes before and after the fault is interrupted is calculated using the following formula: Wherein, ΔP k R is the sum of the changes in active power of the AC lines connected to node k of the receiving-end converter station before and after the fault is interrupted. k This represents the number of AC lines connected to node k of the receiving-end converter station. The active power at node k of the receiving-end converter station is obtained from the ground-state power flow calculation. The active power of node k in the receiving-end converter station is obtained from the power flow calculation after the fault is interrupted. The sum of the reactive power changes of the AC lines connected to the receiving-end converter station nodes before and after the fault is interrupted is calculated using the following formula: Where, ΔQ k It is the sum of the zero-power changes of the AC lines connected to node k of the receiving-end converter station before and after the fault is interrupted. The zero power of node k at the receiving-end converter station, obtained from the ground-state power flow calculation. The zero power of node k in the receiving-end converter station obtained from the power flow calculation after the fault is interrupted; The multi-infeed short-circuit ratio of the receiving-end converter station node after the fault is interrupted is calculated using the following formula: Among them, MISCR k,开断 Z represents the multi-infeed short-circuit ratio of node k at the receiving-end converter station after a fault is interrupted. kk,开断 P represents the self-impedance of node k at the receiving end of the converter station in the multi-port equivalent impedance matrix after a fault is interrupted. d,k P represents the DC power at node k of the receiving-end converter station. d,c Z represents the DC power at node c of the receiving-end converter station. kc,开断 is the mutual impedance between node k and node c of the receiving-end converter station in the multi-port equivalent impedance matrix after the fault is interrupted, and h is the number of nodes in the receiving-end converter station.

6. The method for predicting commutation failure in a high-voltage direct current transmission system according to claim 1, characterized in that, Based on the first prediction result of whether a short-circuit fault will cause commutation failure, the predicted value of commutation failure margin after a short-circuit fault, the second prediction result of whether an interruption fault will cause commutation failure, and the predicted value of commutation failure margin after an interruption fault, the prediction results of commutation failure in the high-voltage direct current transmission system are obtained, including: If the first prediction result is yes and the predicted value of the commutation failure margin after the short-circuit fault is lower than the first threshold, or if the second prediction result is yes and the predicted value of the commutation failure margin after the opening fault is lower than the second threshold, then it is determined that the fault caused the commutation failure of the high-voltage direct current transmission system.

7. A commutation failure prediction system for a high-voltage direct current transmission system, characterized in that, include: The short-circuit fault characteristic module obtains the short-circuit voltage and multi-infeed short-circuit ratio of the receiving-end converter station node after a short-circuit fault, as well as the multi-infeed interaction factor between the receiving-end converter station nodes after a short-circuit fault, based on the grid ground-state operation mode information and short-circuit fault information. The short-circuit fault prediction module inputs the lowest short-circuit voltage, the multi-infeed short-circuit ratio of the lowest voltage receiving-end converter station node, and the maximum value of the multi-infeed interaction factor of the lowest voltage receiving-end converter station node to other receiving-end converter stations into the pre-trained first classification model and first regression model, respectively, to obtain the prediction results of whether the short-circuit fault will cause commutation failure and the prediction results of the commutation failure margin value after the short-circuit fault. The fault interruption feature module obtains the voltage change value of the receiving-end converter station node before and after the fault interruption, the sum of the active power change of the connected AC lines, the sum of the reactive power change of the connected AC lines, and the multi-infeed short-circuit ratio of the receiving-end converter station node after the fault interruption, based on the grid base state operation mode information and fault interruption information. The fault prediction module inputs the maximum voltage change of the receiving-end converter station node before and after the fault, the sum of the maximum active power change of the connected AC lines, the sum of the maximum reactive power change of the connected AC lines, and the maximum change in the multi-infeed short-circuit ratio into the pre-trained second classification model and second regression model to obtain the prediction results of whether the fault will cause commutation failure and the prediction results of the commutation failure margin after the fault. Among them, the change in the multi-infeed short-circuit ratio is the absolute value of the difference between the multi-infeed short-circuit ratio of the receiving-end converter station node after the fault and the multi-infeed short-circuit ratio of the receiving-end converter station node before the fault. The results module obtains the prediction results of commutation failure in the high-voltage direct current transmission system based on the prediction results of whether a short-circuit fault will cause commutation failure, the prediction results of the commutation failure margin after a short-circuit fault, the prediction results of whether an opening fault will cause commutation failure, and the prediction results of the commutation failure margin after an opening fault.

8. The high-voltage direct current transmission system commutation failure prediction system according to claim 7, characterized in that, The short-circuit fault characteristic module obtains the short-circuit voltage and multi-infeed short-circuit ratio of the receiving-end converter station nodes after a short-circuit fault, as well as the multi-infeed interaction factor between the receiving-end converter station nodes after a short-circuit fault. The process includes: Based on the power grid's ground-state operating mode information, the system's ground-state admittance matrix is ​​constructed, and ground-state power flow calculations are performed. Based on the location of the node where the short circuit fault occurred and the location of the receiving-end converter station node in the short circuit fault information, the system ground state admittance matrix is ​​simplified to obtain a multi-port equivalent impedance matrix that includes the short circuit node and the receiving-end converter station node. Based on the solution calculated from the multi-port equivalent impedance matrix and the ground-state power flow, the short-circuit voltage and multi-infeed short-circuit ratio of the receiving-end converter station nodes after a short-circuit fault, as well as the multi-infeed interaction factor between the receiving-end converter station nodes after a short-circuit fault, are obtained.

9. A commutation failure prediction system for a high-voltage direct current transmission system according to claim 7, characterized in that, The fault interruption feature module acquires the voltage change value of the receiving-end converter station node before and after the fault interruption, the sum of the active power change of the connected AC lines, the sum of the reactive power change of the connected AC lines, and the multi-infeed short-circuit ratio of the receiving-end converter station node after the fault interruption. The process includes: Based on the power grid's ground-state operating mode information, the system's ground-state admittance matrix is ​​constructed, and ground-state power flow calculations are performed. Modify the system ground state admittance matrix based on the fault information to obtain the system admittance matrix after the fault, and perform power flow calculation after the fault. Based on the location of the receiving-end converter station node, the admittance matrix after system interruption is simplified to obtain the multi-port equivalent impedance matrix including the receiving-end converter station node. Based on the multi-port equivalent impedance matrix, the solution of the ground-state power flow calculation, and the solution of the power flow calculation after the fault is interrupted, the voltage change value of the receiving-end converter station node before and after the fault is interrupted, the sum of the active power change of the connected AC lines, the sum of the reactive power change of the connected AC lines, and the multi-infeed short-circuit ratio of the receiving-end converter station node after the fault is interrupted are obtained.

10. A computer-readable storage medium for storing one or more programs, characterized in that, The one or more programs include instructions that, when executed by a computing device, cause the computing device to perform any of the methods according to claims 1 to 6.

Images