Power distribution network voltage sag evaluation method and device, computer equipment and program product

Abstract

The application relates to a power distribution network voltage sag evaluation method and device, computer equipment and program product. The method comprises the following steps: generating a photovoltaic output scene by using a trained generative adversarial network, and generating random fault data of a power distribution network; based on the photovoltaic output scene and the random fault data, continuously iterating until the node voltage of a grid-connected point in the power distribution network after a fault occurs meets a constraint condition, determining a first node voltage of each node in the power distribution network under normal conditions and a second node voltage after a fault occurs; calculating the difference between the first node voltage of each node and the second node voltage of the node to obtain a voltage sag amplitude, and the node is a voltage sag node when the voltage sag amplitude is greater than zero; and determining an evaluation result of the voltage sag of the power distribution network based on the voltage sag node and the number of users corresponding to all nodes in the power distribution network. The method can improve the accuracy of the voltage sag evaluation of the power distribution network.

Description

Technical Field

[0001] This application relates to the field of power system distribution network technology, and in particular to a method, apparatus, computer equipment, computer-readable storage medium and computer program product for evaluating voltage sags in distribution networks. Background Technology

[0002] The widespread integration of new energy sources has brought new influencing factors to the analysis of voltage sags in distribution networks and provided support for the proactive management of voltage sags. However, the integration of distributed power sources such as photovoltaics has changed the power supply structure of the distribution network, making it difficult to accurately characterize the risk of voltage sags and thus failing to effectively guide the accurate management of voltage sags.

[0003] The output of distributed photovoltaic (PV) systems is subject to fluctuations due to multiple complex factors, including solar irradiance, temperature variations, and load fluctuations. Furthermore, voltage sag events in the distribution network are random. Therefore, in the context of distributed PV systems connected to the distribution network, the output fluctuations of multiple generating units and the randomness of voltage sag events introduce multiple uncertainties into the distribution network, increasing the difficulty of voltage sag risk assessment.

[0004] In related technologies, existing voltage sag risk assessments often focus on deterministic power sources, assuming that the output power of distributed photovoltaic (PV) power remains constant. This ignores the fact that distributed PV output is affected by the actual environment and has a certain degree of randomness and time-varying nature, resulting in an inaccurate assessment of voltage sags in the distribution network. Summary of the Invention

[0005] Therefore, it is necessary to provide a method, apparatus, computer equipment, computer-readable storage medium, and computer program product for assessing voltage sags in distribution networks that can effectively improve the accuracy of voltage sag assessment in distribution networks, addressing the aforementioned technical problems.

[0006] Firstly, this application provides a method for assessing voltage sags in distribution networks, including:

[0007] A trained generative adversarial network is used to generate photovoltaic power output scenarios.

[0008] Based on historical fault data of the distribution network, random fault data of the distribution network is generated;

[0009] Based on the photovoltaic power output scenario and the random fault data, the process is continuously iterated in the distribution network until the node voltage of the grid connection point in the distribution network after the fault meets the constraint condition. The iteration is then stopped, and the first node voltage of each node in the distribution network under normal conditions and the second node voltage after the fault are determined.

[0010] For each node in the distribution network, the difference between the first node voltage and the second node voltage of the node is calculated to obtain the voltage sag value. If the voltage sag value is greater than zero, the node is a voltage sag node.

[0011] The assessment result of the voltage sag in the distribution network is determined based on the number of users corresponding to the voltage sag node and the number of users corresponding to all nodes in the distribution network.

[0012] In one embodiment, the iterative process based on the photovoltaic output scenario and the random fault data in the distribution network until the node voltage at the grid connection point in the distribution network after a fault meets the constraint conditions includes:

[0013] In the photovoltaic output scenario, obtain the first voltage matrix of the distribution network under normal conditions;

[0014] The random fault data is injected into the distribution network to obtain the second voltage matrix of the distribution network; both the first voltage matrix and the second voltage matrix include the voltage of each node in the distribution network.

[0015] The first voltage matrix and the second voltage matrix are superimposed to obtain the node voltage matrix of the distribution network after a fault occurs.

[0016] Based on the grid-connected node voltage in the node voltage matrix, the output short-circuit current of the grid-connected node is determined. The output short-circuit current is injected into the fault node of the distribution network after replacing the random fault data. The second voltage matrix is ​​updated based on the current, and the node voltage matrix is ​​continuously iterated until the node voltage of the grid-connected node in the distribution network after the fault meets the constraint conditions.

[0017] In one embodiment, the constraints include:

[0018] ;

[0019] in, Let be the node voltage at the grid connection point in the distribution network after a fault occurs, following the k-th iteration. This represents the node voltage at the grid connection point in the distribution network after a fault occurs, following the (k+1)th iteration. This is the preset convergence accuracy.

[0020] In one embodiment, determining the assessment result of the distribution network voltage sag based on the number of users corresponding to the voltage sag node and the number of users corresponding to all nodes in the distribution network includes:

[0021] When the preset voltage threshold is greater than the preset value, for the first voltage sag node whose voltage sag amplitude is greater than the preset voltage threshold, the ratio between the number of users corresponding to the first voltage sag node and the number of users corresponding to all nodes in the distribution network is used as the evaluation result of the voltage sag of the distribution network.

[0022] When the preset voltage threshold is less than the preset value, for the second voltage sag node whose voltage sag amplitude is less than the preset voltage threshold, the ratio between the number of users corresponding to the second voltage sag node and the number of users corresponding to all nodes in the distribution network is used as the evaluation result of the voltage sag of the distribution network; the evaluation result is used to characterize the severity of the voltage sag of the distribution network.

[0023] In one embodiment, the process of obtaining the trained generative adversarial network includes:

[0024] The historical photovoltaic power output data of the power distribution network is obtained, and the historical photovoltaic power output data is preprocessed to obtain the processed historical photovoltaic power output data.

[0025] Random sampling noise is used as the input to the generator network, and the processed historical photovoltaic power output data is used as the input to the discriminator network. The network is continuously trained until the output photovoltaic power output data of the generator adversarial network reaches a Nash equilibrium state, thus obtaining the trained generator adversarial network. The generator adversarial network includes a generator network and a discriminator network.

[0026] In one embodiment, the step of acquiring historical photovoltaic power output data of the distribution network and preprocessing it to obtain processed historical photovoltaic power output data includes:

[0027] The missing photovoltaic output data in the historical photovoltaic output data is filled with the mean of the historical photovoltaic output data to obtain the filled historical photovoltaic output data. The filled historical photovoltaic output data is then cleaned to obtain the cleaned historical photovoltaic output data.

[0028] The cleaned historical photovoltaic power output data is converted into a matrix with a preset dimension, and the matrix is ​​normalized and standardized to obtain the processed historical photovoltaic power output data.

[0029] Secondly, this application also provides an assessment device for voltage sag in a distribution network, comprising:

[0030] The generation module is used to generate photovoltaic power output scenarios using a trained generative adversarial network.

[0031] The generation module is also used to generate random fault data of the distribution network based on historical fault data of the distribution network.

[0032] The determination module is used to continuously iterate in the distribution network based on the photovoltaic output scenario and the random fault data until the node voltage of the grid connection point in the distribution network after the fault meets the constraint condition, then stop the iteration, and determine the first node voltage of each node in the distribution network under normal conditions and the second node voltage after the fault.

[0033] The calculation module is used to calculate the difference between the first node voltage and the second node voltage of each node in the distribution network to obtain the voltage sag value. If the voltage sag value is greater than zero, the node is a voltage sag node.

[0034] The determining module is further configured to determine the evaluation result of the voltage sag of the distribution network based on the number of users corresponding to the voltage sag node and the number of users corresponding to all nodes in the distribution network.

[0035] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:

[0036] A trained generative adversarial network is used to generate photovoltaic power output scenarios.

[0037] Based on historical fault data of the distribution network, random fault data of the distribution network is generated;

[0038] Based on the photovoltaic power output scenario and the random fault data, the process is continuously iterated in the distribution network until the node voltage of the grid connection point in the distribution network after the fault meets the constraint condition. The iteration is then stopped, and the first node voltage of each node in the distribution network under normal conditions and the second node voltage after the fault are determined.

[0039] For each node in the distribution network, the difference between the first node voltage and the second node voltage of the node is calculated to obtain the voltage sag value. If the voltage sag value is greater than zero, the node is a voltage sag node.

[0040] The assessment result of the voltage sag in the distribution network is determined based on the number of users corresponding to the voltage sag node and the number of users corresponding to all nodes in the distribution network.

[0041] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, performs the following steps:

[0042] A trained generative adversarial network is used to generate photovoltaic power output scenarios.

[0043] Based on historical fault data of the distribution network, random fault data of the distribution network is generated;

[0044] Based on the photovoltaic power output scenario and the random fault data, the process is continuously iterated in the distribution network until the node voltage of the grid connection point in the distribution network after the fault meets the constraint condition. The iteration is then stopped, and the first node voltage of each node in the distribution network under normal conditions and the second node voltage after the fault are determined.

[0045] For each node in the distribution network, the difference between the first node voltage and the second node voltage of the node is calculated to obtain the voltage sag value. If the voltage sag value is greater than zero, the node is a voltage sag node.

[0046] The assessment result of the voltage sag in the distribution network is determined based on the number of users corresponding to the voltage sag node and the number of users corresponding to all nodes in the distribution network.

[0047] Fifthly, this application also provides a computer program product, including a computer program that, when executed by a processor, performs the following steps:

[0048] A trained generative adversarial network is used to generate photovoltaic power output scenarios.

[0049] Based on historical fault data of the distribution network, random fault data of the distribution network is generated;

[0050] Based on the photovoltaic power output scenario and the random fault data, the process is continuously iterated in the distribution network until the node voltage of the grid connection point in the distribution network after the fault meets the constraint condition. The iteration is then stopped, and the first node voltage of each node in the distribution network under normal conditions and the second node voltage after the fault are determined.

[0051] For each node in the distribution network, the difference between the first node voltage and the second node voltage of the node is calculated to obtain the voltage sag value. If the voltage sag value is greater than zero, the node is a voltage sag node.

[0052] The assessment result of the voltage sag in the distribution network is determined based on the number of users corresponding to the voltage sag node and the number of users corresponding to all nodes in the distribution network.

[0053] The aforementioned method, apparatus, computer equipment, computer-readable storage medium, and computer program product for assessing voltage sag in distribution networks firstly generate photovoltaic (PV) power output scenarios using a trained generative adversarial network (GAN). Based on historical fault data of the distribution network, random fault data is generated. Based on the PV power output scenarios and random fault data, the process iterates continuously in the distribution network until the node voltage at the grid connection point after a fault meets the constraint conditions. The iteration stops, and the first node voltage under normal conditions and the second node voltage after a fault are determined for each node in the distribution network. For each node in the distribution network, the difference between the first node voltage and the second node voltage is calculated to obtain the voltage sag amplitude. If the voltage sag amplitude is greater than zero, the node is considered a voltage sag node. Based on the number of users corresponding to the voltage sag node and the number of users corresponding to all nodes in the distribution network, the assessment result of the distribution network voltage sag is determined. In this way, the impact of the dynamic output of distributed photovoltaic units in the distribution network on voltage sag characteristics, as well as the randomness and correlation of photovoltaic output itself, are fully considered, which improves the accuracy of scenario generation, makes the voltage sag scenario of distributed photovoltaic access more accurate, and improves the accuracy of voltage sag assessment. Attached Figure Description

[0054] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the description of the embodiments of this application or related technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0055] Figure 1 This is a diagram illustrating the application environment of a distribution network voltage sag assessment method in one embodiment.

[0056] Figure 2 This is a flowchart illustrating a method for assessing voltage sags in a distribution network in one embodiment;

[0057] Figure 3 This is a flowchart illustrating a method for assessing voltage sags in a distribution network, as described in another embodiment.

[0058] Figure 4 This is a structural block diagram of a distribution network voltage sag assessment device in one embodiment;

[0059] Figure 5 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0060] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0061] It should be noted that the terms "first," "second," etc., used in this application can be used to describe various elements, but these elements are not limited by these terms. These terms are only used to distinguish the first element from the second element. The terms "comprising" and "having," and any variations thereof, used in this application, are intended to cover non-exclusive inclusion. The term "multiple" used in this application refers to two or more. The term "and / or" used in this application refers to one of the embodiments, or any combination of multiple embodiments.

[0062] The method for evaluating voltage sags in distribution networks provided in this application can be applied to, for example... Figure 1 In the application environment shown, terminal 102 communicates with server 104 via a network. A data storage system can store the data that server 104 needs to process. The data storage system can be integrated onto server 104 or located in the cloud or on other network servers. Terminal 102 can be, but is not limited to, various personal computers, laptops, smartphones, tablets, drones, low-altitude aircraft, IoT devices, and portable wearable devices. IoT devices can include smart speakers, smart TVs, smart air conditioners, smart in-vehicle devices, projection devices, etc. Portable wearable devices can include smartwatches, smart bracelets, head-mounted devices, etc. Head-mounted devices can be virtual reality (VR) devices, augmented reality (AR) devices, smart glasses, etc. Server 104 can be a standalone physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing cloud computing services.

[0063] In one exemplary embodiment, such as Figure 2 As shown, a method for evaluating voltage sags in distribution networks is provided, and this method is applied to... Figure 1 Taking terminal 102 as an example, the explanation includes the following steps 202 to 210. Wherein:

[0064] Step 202: Generate photovoltaic power output scenarios using the trained generative adversarial network.

[0065] Among them, photovoltaic output refers to the power output of nodes in the distribution network per unit time; generative adversarial network is a deep learning model in which two neural networks, a generator network and a discriminator network, compete and learn from each other. After training, it can generate photovoltaic output scenarios similar to the actual working scenarios of the distribution network.

[0066] For example, a generative adversarial network (GAN) is trained using historical photovoltaic (PV) output data from a distribution network to obtain a trained GAN, which is then used to generate PV output scenarios. These PV output scenarios represent the actual PV output data of various nodes in the distribution network under real-world conditions.

[0067] Step 204: Generate random fault data of the distribution network based on historical fault data of the distribution network.

[0068] The random fault data includes the fault type, fault location, and fault time.

[0069] Optionally, the historical fault data of the distribution network is statistically analyzed, a probability distribution model is selected, the distribution parameters of the probability distribution model are estimated using the historical fault data, the probability distribution of the historical fault data is obtained, and random fault data of the distribution network is generated using a preset method.

[0070] The preset method can be the Monte Carlo method or other methods that can generate random fault data; this application does not limit this.

[0071] Step 206: Based on the photovoltaic power output scenario and random fault data, continuously iterate in the distribution network until the node voltage of the grid connection point in the distribution network after the fault meets the constraint condition, stop the iteration, and determine the first node voltage of each node in the distribution network under normal conditions and the second node voltage after the fault.

[0072] The grid connection point is the connection point where distributed photovoltaic power sources are connected to the power grid.

[0073] For example, random fault data is applied to a photovoltaic power output scenario, and it is continuously iterated in the distribution network until the node voltage of the grid connection point in the distribution network after the fault meets the constraint condition. The iteration stops, and the first node voltage of each node in the distribution network under normal conditions and the second node voltage after the fault are determined after the iteration is completed.

[0074] Step 208: For each node in the distribution network, calculate the difference between the first node voltage and the second node voltage of the node to obtain the voltage sag value. If the voltage sag value is greater than zero, the node is a voltage sag node.

[0075] Optionally, for each node in the distribution network, the difference between the first node voltage and the second node voltage of the node is calculated to obtain the voltage sag value. If the voltage sag value is greater than zero, that is, there is a node in the distribution network whose node voltage has decreased after the iteration is completed, the node is determined as a voltage sag node.

[0076] Step 210: Determine the assessment result of voltage sag in the distribution network based on the number of users corresponding to the voltage sag node and the number of users corresponding to all nodes in the distribution network.

[0077] For example, the degree of voltage sag at the voltage sag node is determined based on the number of users corresponding to the voltage sag node and the number of users corresponding to all nodes in the distribution network, thus determining the assessment result of the voltage sag in the distribution network.

[0078] In the aforementioned method for assessing voltage sag in distribution networks, a photovoltaic (PV) output scenario is generated using a trained generative adversarial network (GAN). Random fault data is generated based on historical fault data of the distribution network, including fault type, location, and time. The method iterates continuously in the distribution network based on the PV output scenario and the random fault data until the node voltage at the grid connection point after a fault meets the constraint conditions. The iteration stops, and the first node voltage under normal conditions and the second node voltage after a fault are determined for each node. For each node in the distribution network, the difference between the first and second node voltages is calculated to obtain the voltage sag amplitude. If the voltage sag amplitude is greater than zero, the node is considered a voltage sag node. The assessment result of the distribution network voltage sag is determined based on the number of users corresponding to the voltage sag node and the total number of users corresponding to all nodes in the distribution network. In this way, the impact of the dynamic output of distributed photovoltaic units in the distribution network on voltage sag characteristics, as well as the randomness and correlation of photovoltaic output itself, are fully considered, which improves the accuracy of scenario generation, makes the voltage sag scenario of distributed photovoltaic access more accurate, and improves the accuracy of voltage sag assessment.

[0079] In an exemplary embodiment, based on photovoltaic power output scenarios and random fault data, the process iterates continuously in the distribution network until the node voltage of the grid-connected point in the distribution network after a fault meets the constraint conditions. This includes: obtaining a first voltage matrix of the distribution network under normal conditions in the photovoltaic power output scenario; injecting random fault data into the distribution network to obtain a second voltage matrix of the distribution network; superimposing the first voltage matrix and the second voltage matrix to obtain a node voltage matrix of the distribution network after a fault; determining the output short-circuit current of the grid-connected point based on the node voltage of the grid-connected point in the node voltage matrix; updating the second voltage matrix based on the output short-circuit current and the current of the faulted node after replacing the random fault data with the current of the faulted node in the distribution network; and iterating the node voltage matrix continuously until the node voltage of the grid-connected point in the distribution network after a fault meets the constraint conditions.

[0080] In actual implementation, under the photovoltaic power output scenario, the first voltage matrix of the distribution network under normal conditions is obtained, which is the normal component network of the distribution network. The specific calculation formula is shown in (1).

[0081] (1)

[0082] in, Z represents the first voltage matrix of the distribution network under normal conditions, and Z represents the node impedance matrix in the distribution network. This is the current matrix of the distribution network under normal conditions. Let f be the voltage at node f in the distribution network under normal conditions. Let f be the current at node f in the distribution network under normal conditions, and n be the number of nodes in the distribution network.

[0083] Inject random fault data into the distribution network, set the synchronous power supply and photovoltaic units in the distribution network to zero, add a reverse current at the fault node, obtain the second voltage matrix of the distribution network, which is the fault component network of the distribution network. The specific calculation formula is shown in (2).

[0084] (2)

[0085] in, To inject random fault data into the distribution network, the second voltage matrix of the distribution network, To inject random fault data into the distribution network, the current matrix in the distribution network is given, and Z is the node impedance matrix in the distribution network. For the voltage fault component at node f in the distribution network, Let f be the current of the fault node in the distribution network. The specific calculation formula is shown in (3).

[0086] (3)

[0087] in, Let f be the self-impedance of node f in the distribution network. Let f be the fault resistance of node f in the distribution network.

[0088] The first voltage matrix and the second voltage matrix are superimposed according to the fault occurrence type using the appropriate symmetric component method to obtain the node voltage matrix of the distribution network after a fault occurs.

[0089] Determine the grid connection point node voltage in the node voltage matrix, calculate the output short-circuit current of the grid connection point, and the specific calculation formula is shown in (4). Use the output short-circuit current of the grid connection point to replace the current of the fault node f in the above distribution network. The process is iterated continuously until the node voltage at the grid connection point in the distribution network meets the constraint conditions after a fault occurs.

[0090] (4)

[0091] in, The output short-circuit current at the grid connection point, The dynamic reactive current reference value at the grid connection point is given by the specific calculation formula shown in (5). The specific calculation formula for the reference value of dynamic active current at the grid connection point is shown in (6).

[0092] (5)

[0093] (6)

[0094] Where K1 and K2 are reactive power compensation coefficients, with values ​​ranging from K1 ≥ 1.5 to K2 ≥ 1.05. This represents the per-unit value of the positive sequence component of the grid connection point voltage. This is the rated current of the photovoltaic power station. The initial active current at the grid connection point. This refers to the inverter current limit at the grid connection point.

[0095] In practice, as the distribution network iterates, only the voltage and current at the grid connection point change.

[0096] In the above embodiments, the output current of distributed photovoltaics changes when a fault occurs. By continuously iterating the output short-circuit current, the dynamic behavior of distributed photovoltaics under fault conditions can be reflected more accurately, which helps to more accurately assess the impact of the fault on voltage dips.

[0097] In one exemplary embodiment, the constraints are specifically shown in formula (7).

[0098] (7)

[0099] in, Let be the node voltage at the grid connection point in the distribution network after a fault occurs, following the k-th iteration. This represents the node voltage at the grid connection point in the distribution network after a fault occurs, following the (k+1)th iteration. This is the preset convergence accuracy.

[0100] In the above embodiments, by setting constraints, it is ensured that the calculation results stop iterating after reaching a certain accuracy, thus avoiding meaningless continued calculations.

[0101] In an exemplary embodiment, the evaluation result of the distribution network voltage sag is determined based on the number of users corresponding to the voltage sag node and the number of users corresponding to all nodes in the distribution network. This includes: when the preset voltage threshold is greater than a preset value, for the first voltage sag node whose voltage sag amplitude is greater than the preset voltage threshold, the ratio between the number of users corresponding to the first voltage sag node and the number of users corresponding to all nodes in the distribution network is used as the evaluation result of the distribution network voltage sag; when the preset voltage threshold is less than the preset value, for the second voltage sag node whose voltage sag amplitude is less than the preset voltage threshold, the ratio between the number of users corresponding to the second voltage sag node and the number of users corresponding to all nodes in the distribution network is used as the evaluation result of the distribution network voltage sag.

[0102] In practice, the evaluation result of voltage sag in the distribution network is determined based on the number of users corresponding to the voltage sag node and the number of users corresponding to all nodes in the distribution network. The specific calculation formula is shown in (8).

[0103] (8)

[0104] in, The evaluation result for voltage sag in the distribution network is given, where X is the preset voltage threshold. This represents the number of users corresponding to either the first or second transient node in the i-th failure. This represents the number of users corresponding to all nodes in the distribution network.

[0105] In one embodiment, X can be 100; when X is greater than 100... Let X be the number of users corresponding to the first voltage sag node in the i-th fault. The first voltage sag node is a voltage sag node with a voltage sag amplitude greater than 100. When X is less than 100... The number of users corresponding to the second voltage sag node in the i-th fault is denoted as , where the second voltage sag node is a voltage sag node with a voltage sag amplitude of less than 100.

[0106] In practical implementation, the assessment results of voltage sag in the distribution network The larger the value, the more severe the voltage dip in the distribution network.

[0107] In the above embodiments, the voltage sag frequency index provides a clear and quantifiable evaluation standard by statistically analyzing the frequency of voltage sag events within a specific time period. This allows grid operators to more intuitively understand the severity of voltage sags.

[0108] In an exemplary embodiment, the process of acquiring the trained generative adversarial network includes: acquiring historical photovoltaic power output data of the distribution network, preprocessing the historical photovoltaic power output data to obtain processed historical photovoltaic power output data; using random sampling noise as input to the generator network, using the processed historical photovoltaic power output data as input to the discriminator network, and continuously training until the photovoltaic power output data output by the generative adversarial network reaches a Nash equilibrium state, thereby obtaining the trained generative adversarial network; the generative adversarial network includes a generator network and a discriminator network.

[0109] In practice, historical photovoltaic power output data of the distribution network is obtained and preprocessed to obtain processed historical photovoltaic power output data.

[0110] Randomly sampled noise is used as input to the generator network, and processed historical photovoltaic power output data is used as input to the discriminator network. The network is continuously trained. During training, the parameters of the discriminator network are fixed while the parameters of the generator network are updated until the loss function of the generator network is minimized. This process continues until the generative adversarial network reaches Nash equilibrium, meaning that the photovoltaic power output scenario generated by the generator network closely approximates the real historical photovoltaic power output data, and the discriminator network has difficulty distinguishing between real and fake data. This results in the trained generative adversarial network.

[0111] In one embodiment, the discriminator network feeds back its judgments on the input data to the generator network, enabling the generator network to continuously adjust its parameters to generate more realistic samples. The discriminator network adjusts its parameters based on the continuously generated samples to better detect differences in the input data. The generator network continuously improves the quality of the generated samples, while the discriminator network continuously improves its discrimination ability.

[0112] In the above embodiments, through adversarial training, the generative adversarial network can learn the distribution characteristics of real data and generate false data that is difficult to distinguish from real data, making the assessment of voltage sag in the distribution network more accurate.

[0113] In an exemplary embodiment, historical photovoltaic (PV) output data of the distribution network is acquired and preprocessed to obtain processed historical PV output data. This includes: filling missing PV output data in the historical PV output data with the mean of the historical PV output data to obtain filled historical PV output data; cleaning the filled historical PV output data to obtain cleaned historical PV output data; converting the cleaned historical PV output data into a matrix under a preset dimension; and normalizing and standardizing the matrix to obtain processed historical PV output data.

[0114] In practice, historical photovoltaic power output data is repaired and cleaned. The average value of historical photovoltaic power output data is used to fill in the missing photovoltaic power output data in the historical photovoltaic power output data, resulting in cleaned historical photovoltaic power output data.

[0115] For example, based on the period and time sequence of historical photovoltaic power output data, nodes with a data missing rate of more than 1% are selected for data repair and cleaning, and the repaired and cleaned data is processed into historical photovoltaic power output data for each time period.

[0116] The cleaned historical photovoltaic power output data is converted into a matrix with preset dimensions, and the matrix is ​​normalized and standardized to obtain the processed historical photovoltaic power output data.

[0117] For example, the input data used to generate photovoltaic power output data scenarios is the historical photovoltaic power output data of N nodes for one year. One year is 365 days. The historical photovoltaic power output data is converted into a three-dimensional matrix, and the data size is represented as N*T*D, where N is the data volume of the distribution network node, T is 24 hours, and D is 365 days. In addition, the historical photovoltaic power output data is divided into rainy season and dry season according to the preset time period. For example, May to September is the rainy season, and other months are the dry season. The photovoltaic power output data information of all nodes for each day is specifically shown in formula (9).

[0118] (9)

[0119] Where n is the number of nodes in the distribution network, and t is the time point. This represents the photovoltaic output data of node n at time t.

[0120] The historical photovoltaic power output data is normalized and standardized, and the data is mapped to a preset range. The specific formula is shown in (10).

[0121] (10)

[0122] in, For the photovoltaic power output data of node i, This is the minimum photovoltaic output data. This represents the maximum photovoltaic output data.

[0123] In the above embodiments, by preprocessing the historical photovoltaic power output data, noise and outliers in the data can be removed, and standards can be unified to improve the accuracy and reliability of the historical photovoltaic power output data.

[0124] To illustrate the method for assessing voltage sags in the distribution network in this application, an embodiment is provided below, with a detailed flowchart as follows. Figure 3 As illustrated, this application provides an example of a method for assessing voltage sags in a distribution network under a specific scenario.

[0125] First, the missing photovoltaic output data in the historical photovoltaic output data is filled with the mean of the historical photovoltaic output data, and then the data is cleaned to obtain the cleaned historical photovoltaic output data. The cleaned historical photovoltaic output data is then converted into a matrix of preset dimensions and normalized and standardized to obtain the processed historical photovoltaic output data.

[0126] Randomly sampled noise is used as input to the generator network, and processed historical photovoltaic power output data is used as input to the discriminator network. The network is continuously trained. During training, the parameters of the discriminator network are fixed while the parameters of the generator network are updated until the loss function of the generator network is minimized. This process continues until the generative adversarial network reaches Nash equilibrium, meaning that the photovoltaic power output scenario generated by the generator network closely approximates the real historical photovoltaic power output data, and the discriminator network can hardly distinguish between real and fake data. This results in the trained generative adversarial network, which is then used to generate photovoltaic power output scenarios.

[0127] Statistical analysis is performed on historical fault data of the distribution network, a probability distribution model is selected, the distribution parameters of the probability distribution model are estimated using historical fault data, the probability distribution of historical fault data is obtained, and random fault data of the distribution network is generated using a preset method.

[0128] In the photovoltaic power output scenario, the first voltage matrix of the distribution network under normal conditions is obtained, which is the normal component network of the distribution network. The specific calculation formula is shown in (1). Random fault data is injected into the distribution network, the synchronous power supply and photovoltaic units in the distribution network are set to zero, and a reverse current is added at the fault node to obtain the second voltage matrix of the distribution network, which is the fault component network of the distribution network. The specific calculation formula is shown in (2).

[0129] The first voltage matrix and the second voltage matrix are superimposed according to the fault occurrence type using the appropriate symmetric component method to obtain the node voltage matrix of the distribution network after a fault occurs.

[0130] Determine the grid connection point node voltage in the node voltage matrix, calculate the output short-circuit current of the grid connection point, and the specific calculation formula is shown in (4). Use the output short-circuit current of the grid connection point to replace the current of the fault node f in the above distribution network. The process is iterated continuously until the node voltage at the grid connection point in the distribution network meets the constraint conditions after a fault occurs.

[0131] When the preset voltage threshold is greater than the preset value, for the first voltage sag node whose voltage sag amplitude is greater than the preset voltage threshold, the ratio of the number of users corresponding to the first voltage sag node to the number of users corresponding to all nodes in the distribution network is used as the evaluation result of the distribution network voltage sag; when the preset voltage threshold is less than the preset value, for the second voltage sag node whose voltage sag amplitude is less than the preset voltage threshold, the ratio of the number of users corresponding to the second voltage sag node to the number of users corresponding to all nodes in the distribution network is used as the evaluation result of the distribution network voltage sag.

[0132] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages in other steps. It is understood that the steps in different embodiments can be freely combined as needed, and all non-contradictory solutions formed by such combinations are within the scope of protection of this application.

[0133] Based on the same inventive concept, this application also provides a distribution network voltage sag assessment device for implementing the above-described distribution network voltage sag assessment method. The solution provided by this device is similar to the implementation described in the above-described method. Therefore, the specific limitations of one or more distribution network voltage sag assessment device embodiments provided below can be found in the limitations of the distribution network voltage sag assessment method described above, and will not be repeated here.

[0134] In one exemplary embodiment, such as Figure 4 As shown, an evaluation device for voltage sag in a distribution network is provided, comprising: a generation module 401, a determination module 402, and a calculation module 403, wherein:

[0135] The generation module is used to generate photovoltaic power output scenarios using a trained generative adversarial network.

[0136] The generation module is also used to generate random fault data of the distribution network based on historical fault data of the distribution network; the random fault data includes fault type, fault location and fault time.

[0137] The determination module is used to continuously iterate in the distribution network based on the photovoltaic output scenario and the random fault data until the node voltage of the grid connection point in the distribution network after the fault meets the constraint condition, then stop the iteration, and determine the first node voltage of each node in the distribution network under normal conditions and the second node voltage after the fault.

[0138] The calculation module is used to calculate the difference between the first node voltage and the second node voltage of each node in the distribution network to obtain the voltage sag value. If the voltage sag value is greater than zero, the node is a voltage sag node.

[0139] The determining module is further configured to determine the evaluation result of the voltage sag of the distribution network based on the number of users corresponding to the voltage sag node and the number of users corresponding to all nodes in the distribution network.

[0140] In one exemplary embodiment, the device further includes an iteration module for obtaining a first voltage matrix of the distribution network under normal conditions in the photovoltaic output scenario;

[0141] The random fault data is injected into the distribution network to obtain the second voltage matrix of the distribution network; the voltage matrix includes the voltage of each node in the distribution network;

[0142] The first voltage matrix and the second voltage matrix are superimposed to obtain the node voltage matrix of the distribution network after a fault occurs.

[0143] Based on the node voltages of the grid-connected points in the node voltage matrix, the node voltage matrix is ​​iterated continuously until the node voltages of the grid-connected points in the distribution network after a fault meet the constraint conditions.

[0144] In one exemplary embodiment, the constraints include:

[0145] ;

[0146] in, Let be the node voltage at the grid connection point in the distribution network after a fault occurs, following the k-th iteration. This represents the node voltage at the grid connection point in the distribution network after a fault occurs, following the (k+1)th iteration. This is the preset convergence accuracy.

[0147] In an exemplary embodiment, the determining module is further configured to, when the preset voltage threshold is greater than the preset value, for the first voltage sag node whose voltage sag amplitude is greater than the preset voltage threshold, use the ratio between the number of users corresponding to the first voltage sag node and the number of users corresponding to all nodes in the distribution network as the evaluation result of the voltage sag of the distribution network.

[0148] If the preset voltage threshold is less than the preset value, for the second voltage sag node whose voltage sag amplitude is less than the preset voltage threshold, the ratio between the number of users corresponding to the second voltage sag node and the number of users corresponding to all nodes in the distribution network is used as the evaluation result of the voltage sag of the distribution network.

[0149] In one exemplary embodiment, the device further includes an acquisition module for acquiring historical photovoltaic power output data of the distribution network and preprocessing it to obtain processed historical photovoltaic power output data;

[0150] Random sampling noise is used as the input to the generator network, and the processed historical photovoltaic power output data is used as the input to the discriminator network. The network is continuously trained until the generative adversarial network reaches Nash equilibrium, resulting in a trained generative adversarial network. The generative adversarial network includes a generator network and a discriminator network.

[0151] In one exemplary embodiment, the device further includes a processing module for filling missing photovoltaic output data in the historical photovoltaic output data with the mean of the historical photovoltaic output data and cleaning it to obtain cleaned historical photovoltaic output data.

[0152] The cleaned historical photovoltaic power output data is converted into a matrix of preset dimensions and then normalized and standardized to obtain the processed historical photovoltaic power output data.

[0153] Each module in the aforementioned power distribution network voltage sag assessment device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module.

[0154] In one exemplary embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 5As shown, the computer device includes a processor, memory, input / output interfaces (I / O), a communication interface, a display unit, and input devices. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface, display unit, and input devices are also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operation of the operating system and computer programs in the non-volatile storage media. The database stores photovoltaic power output data in the power distribution network. The I / O interfaces are used for information exchange between the processor and external devices. The communication interface is used for communication with external terminals via a network connection. When executed by the processor, the computer program implements a method for evaluating voltage sags in the power distribution network.

[0155] The display unit of this computer device is used to form a visually visible image and can be a display screen, a projection device, or a virtual reality imaging device. The display screen can be an LCD screen or an e-ink screen. The input device of this computer device can be a touch layer covering the display screen, or buttons, a trackball, or a touchpad set on the casing of the computer device, or an external keyboard, touchpad, or mouse, etc.

[0156] Those skilled in the art will understand that Figure 5 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0157] In one exemplary embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:

[0158] A photovoltaic power output scenario is generated using a trained generative adversarial network.

[0159] Based on historical fault data of the distribution network, random fault data of the distribution network is generated; the random fault data includes fault type, fault location and fault time.

[0160] Based on the photovoltaic power output scenario and the random fault data, the process is continuously iterated in the distribution network until the node voltage of the grid connection point in the distribution network after the fault meets the constraint condition. The iteration is then stopped, and the first node voltage of each node in the distribution network under normal conditions and the second node voltage after the fault are determined.

[0161] For each node in the distribution network, the difference between the first node voltage and the second node voltage of the node is calculated to obtain the voltage sag value. If the voltage sag value is greater than zero, the node is a voltage sag node.

[0162] The assessment result of the voltage sag in the distribution network is determined based on the number of users corresponding to the voltage sag node and the number of users corresponding to all nodes in the distribution network.

[0163] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, the computer program performing the following steps when executed by a processor:

[0164] A photovoltaic power output scenario is generated using a trained generative adversarial network.

[0165] Based on historical fault data of the distribution network, random fault data of the distribution network is generated; the random fault data includes fault type, fault location and fault time.

[0166] Based on the photovoltaic power output scenario and the random fault data, the process is continuously iterated in the distribution network until the node voltage of the grid connection point in the distribution network after the fault meets the constraint condition. The iteration is then stopped, and the first node voltage of each node in the distribution network under normal conditions and the second node voltage after the fault are determined.

[0167] For each node in the distribution network, the difference between the first node voltage and the second node voltage of the node is calculated to obtain the voltage sag value. If the voltage sag value is greater than zero, the node is a voltage sag node.

[0168] The assessment result of the voltage sag in the distribution network is determined based on the number of users corresponding to the voltage sag node and the number of users corresponding to all nodes in the distribution network.

[0169] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, performs the following steps:

[0170] A photovoltaic power output scenario is generated using a trained generative adversarial network.

[0171] Based on historical fault data of the distribution network, random fault data of the distribution network is generated; the random fault data includes fault type, fault location and fault time.

[0172] Based on the photovoltaic power output scenario and the random fault data, the process is continuously iterated in the distribution network until the node voltage of the grid connection point in the distribution network after the fault meets the constraint condition. The iteration is then stopped, and the first node voltage of each node in the distribution network under normal conditions and the second node voltage after the fault are determined.

[0173] For each node in the distribution network, the difference between the first node voltage and the second node voltage of the node is calculated to obtain the voltage sag value. If the voltage sag value is greater than zero, the node is a voltage sag node.

[0174] The assessment result of the voltage sag in the distribution network is determined based on the number of users corresponding to the voltage sag node and the number of users corresponding to all nodes in the distribution network.

[0175] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0176] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, artificial intelligence (AI) processors, etc., and are not limited to these.

[0177] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0178] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for assessing voltage sags in a distribution network, characterized in that, The method includes: A photovoltaic power output scenario is generated using a trained generative adversarial network. Historical fault data of the distribution network are statistically analyzed, and random fault data of the distribution network is generated based on a probability distribution model; the random fault data includes fault type, fault location, and fault time. Based on the photovoltaic power output scenario and the random fault data, the process is continuously iterated in the distribution network until the node voltage of the grid connection point in the distribution network after the fault meets the constraint condition. The iteration is then stopped, and the first node voltage of each node in the distribution network under normal conditions and the second node voltage after the fault are determined. For each node in the distribution network, the difference between the first node voltage and the second node voltage of the node is calculated to obtain the voltage sag value. If the voltage sag value is greater than zero, the node is a voltage sag node. The evaluation result of the voltage sag in the distribution network is determined based on the number of users corresponding to the voltage sag node and the number of users corresponding to all nodes in the distribution network. The process of iterating continuously in the distribution network based on the photovoltaic output scenario and the random fault data until the node voltage at the grid connection point in the distribution network after a fault meets the constraint conditions includes: In the photovoltaic output scenario, obtain the first voltage matrix of the distribution network under normal conditions; The random fault data is injected into the distribution network to obtain the second voltage matrix of the distribution network; both the first voltage matrix and the second voltage matrix include the voltage of each node in the distribution network. The first voltage matrix and the second voltage matrix are superimposed to obtain the node voltage matrix of the distribution network after a fault occurs. Based on the grid-connected node voltage in the node voltage matrix, the output short-circuit current of the grid-connected node is determined. The output short-circuit current is injected into the fault node of the distribution network after replacing the random fault data. The second voltage matrix is ​​updated based on the current, and the node voltage matrix is ​​continuously iterated until the node voltage of the grid-connected node in the distribution network after the fault meets the constraint conditions.

2. The method according to claim 1, characterized in that, The constraints include: ； in, Let be the node voltage at the grid connection point in the distribution network after a fault occurs, following the k-th iteration. This represents the node voltage at the grid connection point in the distribution network after a fault occurs, following the (k+1)th iteration. This is the preset convergence accuracy.

3. The method according to claim 1, characterized in that, The assessment result of the voltage sag in the distribution network, based on the number of users corresponding to the voltage sag node and the number of users corresponding to all nodes in the distribution network, includes: When the preset voltage threshold is greater than the preset value, for the first voltage sag node whose voltage sag amplitude is greater than the preset voltage threshold, the ratio between the number of users corresponding to the first voltage sag node and the number of users corresponding to all nodes in the distribution network is used as the evaluation result of the voltage sag of the distribution network. When the preset voltage threshold is less than the preset value, for the second voltage sag node whose voltage sag amplitude is less than the preset voltage threshold, the ratio between the number of users corresponding to the second voltage sag node and the number of users corresponding to all nodes in the distribution network is used as the evaluation result of the voltage sag of the distribution network; the evaluation result is used to characterize the severity of the voltage sag of the distribution network.

4. The method according to claim 1, characterized in that, The process of obtaining the trained generative adversarial network includes: The historical photovoltaic power output data of the power distribution network is obtained, and the historical photovoltaic power output data is preprocessed to obtain the processed historical photovoltaic power output data. Random sampling noise is used as the input to the generator network, and the processed historical photovoltaic power output data is used as the input to the discriminator network. The network is continuously trained until the photovoltaic power output data output by the generative adversarial network reaches a Nash equilibrium state, thus obtaining the trained generative adversarial network. The generative adversarial network includes a generator network and a discriminator network.

5. The method according to claim 4, characterized in that, The process of acquiring historical photovoltaic power output data of the distribution network and preprocessing it to obtain processed historical photovoltaic power output data includes: The missing photovoltaic output data in the historical photovoltaic output data is filled with the mean of the historical photovoltaic output data to obtain the filled historical photovoltaic output data. The filled historical photovoltaic output data is then cleaned to obtain the cleaned historical photovoltaic output data. The cleaned historical photovoltaic power output data is converted into a matrix with a preset dimension, and the matrix is ​​normalized and standardized to obtain the processed historical photovoltaic power output data.

6. A device for evaluating voltage sag in a distribution network, characterized in that, The device includes: The generation module is used to generate photovoltaic power output scenarios using a trained generative adversarial network. The generation module is also used to statistically analyze historical fault data of the distribution network and generate random fault data of the distribution network based on a probability distribution model; the random fault data includes fault type, fault location, and fault time. The determination module is used to continuously iterate in the distribution network based on the photovoltaic output scenario and the random fault data until the node voltage of the grid connection point in the distribution network after the fault meets the constraint condition, then stop the iteration, and determine the first node voltage of each node in the distribution network under normal conditions and the second node voltage after the fault. The calculation module is used to calculate the difference between the first node voltage and the second node voltage of each node in the distribution network to obtain the voltage sag value. If the voltage sag value is greater than zero, the node is a voltage sag node. The determining module is further configured to determine the evaluation result of the voltage sag of the distribution network based on the number of users corresponding to the voltage sag node and the number of users corresponding to all nodes in the distribution network. An iterative module is used to: obtain a first voltage matrix of the distribution network under normal conditions in the photovoltaic power output scenario; inject the random fault data into the distribution network to obtain a second voltage matrix of the distribution network; both the first and second voltage matrices include the voltage of each node in the distribution network; superimpose the first and second voltage matrices to obtain a node voltage matrix of the distribution network after a fault occurs; determine the output short-circuit current of the grid-connected node based on the node voltage of the grid-connected node in the node voltage matrix; replace the current of the faulted node with the output short-circuit current after injecting the random fault data into the distribution network; update the second voltage matrix based on the current; and continuously iterate the node voltage matrix until the node voltage of the grid-connected node in the distribution network after a fault meets the constraint conditions.

7. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 5.

9. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 5.

Images