A method and system for post-disaster transformer availability assessment

Abstract

The application provides a post-disaster transformer availability evaluation method and system, and the method comprises the following steps: obtaining state monitoring data of a post-disaster transformer; obtaining a transformer fault label value based on the state monitoring data; dividing the health degree of the transformer according to the transformer fault label value, and determining the component availability rate range under each health degree; correcting the current component availability rate of the transformer by taking all the transformer action information after the operation of the protection system as a reference; and determining the transformer availability based on the corrected component availability rate. The application provides a post-disaster transformer availability evaluation method, and the evaluation result fully considers the transformer protection action condition, is more reliable, is helpful for the rapid formulation of a post-disaster recovery strategy of a system, and effectively improves the reliability and safety of a power system.

Description

Technical Field

[0001] This invention relates to the field of transformer performance evaluation technology, and in particular to a method and system for evaluating the availability of transformers after a disaster. Background Technology

[0002] Post-disaster system component status assessment can determine whether system components are still usable or require repair or replacement. This assessment helps power system managers understand the status of components in the power system after a disaster, providing necessary information for restoring power system operation, enabling them to develop reasonable repair and recovery plans, and improving the reliability and security of the power system.

[0003] Power transformers are among the most widely used core devices in power system energy transmission and conversion. Their operational health is closely related to the safe and reliable operation of the system. Therefore, when a system fault occurs, a condition assessment of the power transformer is necessary. Currently, when oil-immersed transformers fail, the decomposition of transformer oil, solid insulation materials, and other factors produces gases such as hydrogen (H2), methane (CH4), ethane (C2H6), and ethylene (C2H4). Dissolved Gases Analysis (DGA), based on the compositional analysis of these gases, is a commonly used method for transformer condition assessment. In recent years, with the rapid development of machine learning algorithms, various transformer fault diagnosis methods based on DGA data have emerged, including support vector machines, neural networks, and deep learning. However, many methods primarily focus on condition assessment and rarely consider the specific availability of components. In actual post-disaster systems, when dealing with transformers to be restored, it is necessary to determine their availability based on the current condition monitoring data.

[0004] Therefore, although current research on power transformer condition assessment based on intelligent algorithms such as machine learning has yielded many results, none of them have considered the transformer protection operation, resulting in unreliable final assessment results. Furthermore, existing research cannot fully provide the availability of the transformer equipment to be restored, affecting the formulation of subsequent restoration strategies and failing to adequately guarantee the reliability and security of the power system. Summary of the Invention

[0005] The present invention aims to provide a method and system for assessing the availability of transformers after a disaster to solve the above-mentioned technical problems. The assessment results fully consider the operation of transformer protection, are more reliable, help to quickly formulate post-disaster recovery strategies, and improve the reliability and security of the power system.

[0006] To address the aforementioned technical problems, this invention provides a method for assessing the availability of transformers after a disaster, comprising the following steps:

[0007] Obtain post-disaster transformer condition monitoring data;

[0008] Obtain transformer fault tag values ​​based on condition monitoring data;

[0009] Based on the transformer fault label values, the transformer is classified into health status categories, and the availability range of components under each health status is determined.

[0010] Based on all transformer operation information after the Baoxin system was put into operation, the current component availability of the transformer was corrected.

[0011] Transformer availability is determined based on the corrected component availability.

[0012] The above-mentioned method for assessing the availability of transformers after a disaster fully considers the operation of transformer protection systems, making it more reliable and facilitating the rapid development of post-disaster recovery strategies for the system, thereby effectively improving the reliability and security of the power system.

[0013] Furthermore, the correction of the current component availability rate of the transformer, based on all transformer operation information after the commissioning of the information guarantee system, is specifically as follows:

[0014] Based on all transformer operation information after the Baoxin system is put into operation, if the average number of operations per year for the current transformer component is less than 0.5, it indicates that the component has high reliability, and its highest availability rate under its health condition is taken as its current availability rate; if the average number of operations per year for the current transformer component is greater than or equal to 0.5, it indicates that the component has low reliability, and its availability rate is corrected by reducing the availability rate by a% for each operation under its health condition and at its highest availability rate.

[0015] Furthermore, the acquisition of transformer fault tag values ​​based on condition monitoring data specifically includes:

[0016] The condition monitoring data is used as the input to a pre-set deep neural network, which is then trained and used for prediction and verification based on the condition monitoring data to obtain transformer fault label values.

[0017] Furthermore, the process of classifying the transformer's health status based on its fault tag value and determining the component availability range for each health status is specifically expressed as follows:

[0018]

[0019] In the formula, R represents the availability range; y represents the transformer fault label value, where 0 represents normal, 1 represents medium and low temperature overheating, 2 represents high temperature overheating, 3 represents low energy discharge, 4 represents high energy discharge, and 5 represents partial discharge.

[0020] Furthermore, the determination of transformer availability based on the corrected component availability rate is specifically expressed as follows:

[0021]

[0022] In the formula: P represents the transformer availability, where 1 indicates the transformer is available and 0 indicates it is unavailable; r represents the corrected component availability rate, expressed as:

[0023] r = R max -a%*n

[0024] In the formula: R max This represents the highest availability rate at the current health level, and n represents the average number of transformer protection operations per year (rounded up).

[0025] This invention also provides a post-disaster transformer availability assessment system for implementing a post-disaster transformer availability assessment method, comprising:

[0026] The data acquisition module is used to acquire post-disaster transformer status monitoring data.

[0027] The fault tag value acquisition module is used to acquire transformer fault tag values ​​based on condition monitoring data;

[0028] The availability determination module is used to classify the health status of transformers based on transformer fault tag values ​​and determine the availability range of components under each health status.

[0029] The availability correction module is used to correct the current availability of transformer components by referencing all transformer operation information after the Baoxin system was put into operation.

[0030] The availability determination module is used to determine the transformer availability based on the corrected component availability.

[0031] The system provided by the above solution can quickly and accurately assess the availability of transformers after a disaster. The system has a simple architecture and is easy to implement. Its assessment results fully consider the operation of transformer protection and are more reliable. It helps to quickly formulate post-disaster recovery strategies and effectively improves the reliability and security of the power system.

[0032] Furthermore, the availability correction module is used to correct the current component availability of the transformer by referring to all transformer operation information after the commissioning of the information guarantee system, specifically as follows:

[0033] Based on all transformer operation information after the Baoxin system is put into operation, if the average number of operations per year for the current transformer component is less than 0.5, it indicates that the component has high reliability, and its highest availability rate under its health condition is taken as its current availability rate; if the average number of operations per year for the current transformer component is greater than or equal to 0.5, it indicates that the component has low reliability, and its availability rate is corrected by reducing the availability rate by a% for each operation under its health condition and at its highest availability rate.

[0034] Furthermore, the fault tag value acquisition module is used to acquire transformer fault tag values ​​based on condition monitoring data, specifically as follows:

[0035] The condition monitoring data is used as the input to a pre-set deep neural network, which is then trained and used for prediction and verification based on the condition monitoring data to obtain transformer fault label values.

[0036] Furthermore, the availability determination module is used to classify the transformer's health status based on the transformer fault tag value, and determine the component availability range under each health status, specifically expressed as follows:

[0037]

[0038] In the formula, R represents the availability range; y represents the transformer fault label value, where 0 represents normal, 1 represents medium and low temperature overheating, 2 represents high temperature overheating, 3 represents low energy discharge, 4 represents high energy discharge, and 5 represents partial discharge.

[0039] Furthermore, the availability determination module is used to determine the transformer availability based on the corrected component availability rate, specifically as follows:

[0040]

[0041] In the formula: P represents the transformer availability, where 1 indicates the transformer is available and 0 indicates it is unavailable; r represents the corrected component availability rate, expressed as:

[0042] r = R max -a%*n

[0043] In the formula: R max This represents the highest availability rate at the current health level, and n represents the average number of transformer protection operations per year (rounded up). Attached Figure Description

[0044] Figure 1 This is a schematic diagram of a post-disaster transformer availability assessment method according to an embodiment of the present invention;

[0045] Figure 2This is a diagram illustrating the architecture of a post-disaster transformer availability assessment system, as provided in an embodiment of the present invention. Detailed Implementation

[0046] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0047] Please see Figure 1 This embodiment provides a method for assessing the availability of transformers after a disaster, including the following steps:

[0048] S1: Obtain post-disaster transformer condition monitoring data;

[0049] S2: Obtain transformer fault tag values ​​based on condition monitoring data;

[0050] S3: Classify the transformer health status based on the transformer fault tag value, and determine the component availability range under each health status;

[0051] S4: Based on all transformer operation information after the Baoxin system was put into operation, the current component availability of the transformer is corrected;

[0052] S5: Determine transformer availability based on the corrected component availability.

[0053] The post-disaster transformer availability assessment method provided in this embodiment fully considers the transformer protection operation, making it more reliable and facilitating the rapid formulation of post-disaster recovery strategies for the system, thereby effectively improving the reliability and security of the power system.

[0054] Furthermore, the correction of the current component availability rate of the transformer, based on all transformer operation information after the commissioning of the information guarantee system, is specifically as follows:

[0055] Based on all transformer operation information after the Baoxin system is put into operation, if the average number of operations per year for the current transformer component is less than 0.5, it indicates that the component has high reliability, and its highest availability rate under its health condition is taken as its current availability rate; if the average number of operations per year for the current transformer component is greater than or equal to 0.5, it indicates that the component has low reliability, and its availability rate is corrected by reducing the availability rate by a% for each operation under its health condition and at its highest availability rate.

[0056] It should be noted that parameter 'a' can be set according to actual needs.

[0057] Furthermore, the acquisition of transformer fault tag values ​​based on condition monitoring data specifically includes:

[0058] Condition monitoring data is used as input to a pre-defined deep neural network. This network is trained and used for prediction and verification based on the condition monitoring data to obtain transformer fault label values. Specifically, the deep neural network is used to train and fit the relationship between transformer fault label values ​​and transformer conditions. Its network input is the transformer condition monitoring data, and its output is the transformer fault label value.

[0059] y = f(x)

[0060] Where: x is the network input data, y is the fault label value, 0 represents normal, 1 represents medium and low temperature overheating, 2 represents high temperature overheating, 3 represents low energy discharge, 4 represents high energy discharge, and 5 represents partial discharge.

[0061] In this embodiment, for a preset deep neural network, 16 samples can be selected from various sample sets to form a training set, and 8 samples can be selected to form a test set for training and prediction verification of the deep neural network. During training and prediction verification, the sample data can be normalized as follows:

[0062]

[0063] In the formula: x istd Let x be the normalized value of the i-th feature in the sample. i Let x be the value of the i-th feature in the sample. imax and x imin These are the maximum and minimum values ​​corresponding to this feature in the sample, respectively.

[0064] Furthermore, the process of classifying the transformer's health status based on its fault tag value and determining the component availability range for each health status is specifically expressed as follows:

[0065]

[0066] In the formula, R represents the availability range; y represents the transformer fault label value, where 0 represents normal, 1 represents medium and low temperature overheating, 2 represents high temperature overheating, 3 represents low energy discharge, 4 represents high energy discharge, and 5 represents partial discharge.

[0067] Furthermore, the determination of transformer availability based on the corrected component availability rate is specifically expressed as follows:

[0068]

[0069] In the formula: P represents the transformer availability, where 1 indicates the transformer is available and 0 indicates it is unavailable; r represents the corrected component availability rate, expressed as:

[0070] r = R max -a%*n

[0071] In the formula: R max This represents the highest availability rate at the current health level, and n represents the average number of transformer protection operations per year (rounded up).

[0072] This embodiment provides a method for assessing the availability of transformers after a disaster. After acquiring the condition monitoring data of the transformers after a disaster, a deep neural network (DNN) is used for classification training and prediction verification based on the condition monitoring data. The network input is the transformer condition monitoring data, and the output is the transformer fault label value. The transformers are then classified into health levels based on the fault label values, and the availability range of components under each health level is determined. Finally, the current component availability of the transformers is corrected with reference to all transformer operation information after the system's commissioning. The transformer availability is determined according to the principle that a transformer availability rate higher than 0.85 is considered available, and a rate lower than 0.85 is considered unavailable. This availability assessment method fully considers the operation of transformer protection during the assessment process, making the assessment results more reliable and contributing to the formulation of post-disaster recovery strategies and improving the system's safety and reliability.

[0073] To more clearly illustrate the technical features of this invention and highlight its technical advantages, this embodiment provides a specific application of a post-disaster transformer availability assessment method, specifically as follows:

[0074] Step 1: Obtain transformer condition monitoring data, extract training and test sets from the dataset, and normalize the label values ​​of each sample. In this example, the training set size is 96 and the test set size is 48.

[0075] Step 2: Use a DNN to train the training set data, and use the test set to predict the fault labels of the samples in the prediction set.

[0076] Step 3: Consider the protection action information to correct the component availability. In this embodiment, parameter a is set to 5, and the availability of the transformer is determined. This embodiment selects 10 samples from the test set as examples, and their current status classification and protection historical action information are shown in Table 1:

[0077] Table 1 Current Status Classification and Protection Historical Action Information

[0078]

[0079]

[0080] It can be seen that after considering the number of protection actions, the availability of samples 7, 8, and 10 will change from available before the correction to unavailable after the correction. This is due to the unreliability of the historical state of the equipment, thus ensuring the safety and reliability of the recovery process.

[0081] This invention also provides a post-disaster transformer availability assessment system for implementing a post-disaster transformer availability assessment method, comprising:

[0082] The data acquisition module is used to acquire post-disaster transformer status monitoring data.

[0083] The fault tag value acquisition module is used to acquire transformer fault tag values ​​based on condition monitoring data;

[0084] The availability determination module is used to classify the health status of transformers based on transformer fault tag values ​​and determine the availability range of components under each health status.

[0085] The availability correction module is used to correct the current availability of transformer components by referencing all transformer operation information after the Baoxin system was put into operation.

[0086] The availability determination module is used to determine the transformer availability based on the corrected component availability.

[0087] The system provided by the above solution can quickly and accurately assess the availability of transformers after a disaster. The system has a simple architecture and is easy to implement. Its assessment results fully consider the operation of transformer protection and are more reliable. It helps to quickly formulate post-disaster recovery strategies and effectively improves the reliability and security of the power system.

[0088] Furthermore, the availability correction module is used to correct the current component availability of the transformer by referring to all transformer operation information after the commissioning of the information guarantee system, specifically as follows:

[0089] Based on all transformer operation information after the Baoxin system is put into operation, if the average number of operations per year for the current transformer component is less than 0.5, it indicates that the component has high reliability, and its highest availability rate under its health condition is taken as its current availability rate; if the average number of operations per year for the current transformer component is greater than or equal to 0.5, it indicates that the component has low reliability, and its availability rate is corrected by reducing the availability rate by a% for each operation under its health condition and at its highest availability rate.

[0090] Furthermore, the fault tag value acquisition module is used to acquire transformer fault tag values ​​based on condition monitoring data, specifically as follows:

[0091] The condition monitoring data is used as the input to a pre-set deep neural network, which is then trained and used for prediction and verification based on the condition monitoring data to obtain transformer fault label values.

[0092] Furthermore, the availability determination module is used to classify the transformer's health status based on the transformer fault tag value, and determine the component availability range under each health status, specifically expressed as follows:

[0093]

[0094] In the formula, R represents the availability range; y represents the transformer fault label value, where 0 represents normal, 1 represents medium and low temperature overheating, 2 represents high temperature overheating, 3 represents low energy discharge, 4 represents high energy discharge, and 5 represents partial discharge.

[0095] Furthermore, the availability determination module is used to determine the transformer availability based on the corrected component availability rate, specifically as follows:

[0096]

[0097] In the formula: P represents the transformer availability, where 1 indicates the transformer is available and 0 indicates it is unavailable; r represents the corrected component availability rate, expressed as:

[0098] r = R max -a%*n

[0099] In the formula: R max This represents the highest availability rate at the current health level, and n represents the average number of transformer protection operations per year (rounded up).

[0100] This embodiment provides a post-disaster transformer availability assessment system. After acquiring post-disaster transformer condition monitoring data, it uses a deep neural network (DNN) for classification training and prediction verification based on the condition monitoring data. The network input is the transformer condition monitoring data, and the output is the transformer fault label value. The system then classifies the transformer's health level based on the fault label value, determines the component availability range for each health level, and finally corrects the current component availability rate of the transformer using all transformer operation information after the system's commissioning as a reference. The system determines the transformer's availability status according to the principle that a transformer availability rate above 0.85 is considered available, and below 0.85 is considered unavailable. This availability assessment system fully considers the operation of transformer protection during the assessment process, making the assessment results more reliable and contributing to the formulation of post-disaster recovery strategies and improving the system's safety and reliability.

[0101] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A method for assessing the availability of transformers after a disaster, characterized in that, The method comprises the following steps: obtain state monitoring data of the transformer after the disaster; obtain a transformer fault label value based on the state monitoring data; divide the health degree of the transformer according to the transformer fault label value, and determine the component availability rate range under each health degree, which is specifically expressed as: In the formula, R represents the available rate range; y represents the transformer fault label value, wherein 0 represents normal, 1 is medium and low temperature overheating, 2 represents high temperature overheating, 3 represents low energy discharge, 4 represents high energy discharge, and 5 represents partial discharge; correct the current component availability rate of the transformer with reference to all transformer action information after the operation of the protection and communication system, which is specifically: based on all transformer action information after the operation of the protection and communication system, if the average annual action frequency of the current component of the transformer is less than 0.5, it indicates that the component has high reliability, and the highest availability rate under its health degree is taken as the current availability rate; if the average annual action frequency of the current component of the transformer is greater than or equal to 0.5, it indicates that the component has low reliability, and the availability rate of the component is corrected in the manner of reducing a% availability rate per action under the highest availability rate under its health degree; determine the availability of the transformer based on the corrected component availability rate, which is specifically expressed as: In the formula, P represents the transformer availability, which is 1 when the transformer is available and 0 when the transformer is not available; represents the corrected element availability, which is represented by: In the formula: represents the highest available rate of the current health level, represents the integer value of the average number of actions of the transformer protection per year.

2. The method of claim 1, wherein, The transformer fault label value is obtained based on the state monitoring data, which is specifically: the state monitoring data is taken as the input of the preset deep neural network, the deep neural network is trained and prediction verified based on the state monitoring data, and the transformer fault label value is obtained.

3. A post-disaster transformer availability assessment system, characterized by, It comprises: a data acquisition module for obtaining state monitoring data of the transformer after the disaster; a fault label value acquisition module for obtaining a transformer fault label value based on the state monitoring data; an availability rate determination module for dividing the health degree of the transformer according to the transformer fault label value, and determining the component availability rate range under each health degree, which is specifically expressed as: In the formula, R represents the available rate range; y represents the transformer fault label value, wherein 0 represents normal, 1 is medium and low temperature overheating, 2 represents high temperature overheating, 3 represents low energy discharge, 4 represents high energy discharge, and 5 represents partial discharge; an availability rate correction module for correcting the current component availability rate of the transformer with reference to all transformer action information after the operation of the protection and communication system, which is specifically: based on all transformer action information after the operation of the protection and communication system, if the average annual action frequency of the current component of the transformer is less than 0.5, it indicates that the component has high reliability, and the highest availability rate under its health degree is taken as the current availability rate; if the average annual action frequency of the current component of the transformer is greater than or equal to 0.5, it indicates that the component has low reliability, and the availability rate of the component is corrected in the manner of reducing a% availability rate per action under the highest availability rate under its health degree; an availability determination module for determining the availability of the transformer based on the corrected component availability rate, which is specifically expressed as: In the formula, P represents the transformer availability, which is 1 when the transformer is available and 0 when the transformer is not available; represents the corrected element availability, which is represented by: In the formula: represents the highest available rate of the current health level, represents the integer value of the average number of actions of the transformer protection per year.

4. The post-disaster transformer availability assessment system of claim 3, wherein, The fault label value acquisition module is used to obtain a transformer fault label value based on the state monitoring data, which is specifically: the state monitoring data is taken as the input of the preset deep neural network, the deep neural network is trained and prediction verified based on the state monitoring data, and the transformer fault label value is obtained.

Images