A power grid fault intelligent auxiliary disposal method and a disposal system

Abstract

The application discloses a kind of power grid failure intelligent auxiliary handling methods, including extracting and identifying each fault information contained in the preset specific fault description project and the fault handling measure information of the historical fault information;Each fault information and its fault handling measure are expressed and classified in a combined manner according to the information identified;Generate the combination of such fault information and times;All fault information combinations and the combination of times are expressed as matrix;All the fault information and the combination of times obtained are used as the input item of deep convolutional neural network, the optimal 4 fault handling measures corresponding to each fault information combination are used as the output item to train deep convolutional neural network;Generate target fault information combination, pass target fault information combination into deep convolutional neural network, and deep convolutional neural network outputs 4 optimal handling measure information and recommends to user through electronic handling plan.The application can realize the rapid handling of power grid failure.

Description

Technical Field

[0001] This invention relates to the field of power grid fault recovery technology, and specifically to an intelligent auxiliary handling method and system for power grid faults. Background Technology

[0002] With the continuous expansion of regional power grids, power grid accidents will cause power outages for a large number of users, resulting in huge economic losses and more severe impacts. Therefore, the issue of power grid fault recovery is receiving increasing attention from power companies. If power supply is restored to the out-of-load in the shortest possible time after a fault occurs, and the power grid is adjusted to operate in a safe and stable state, the losses from power outages and the impact of accidents can be minimized.

[0003] The regional power grid has a weak structure, numerous electrical devices, and variable operating modes, making it prone to frequent grid failures. If correct and effective restoration measures are taken promptly after a grid failure, the outage time can be significantly shortened, the affected area reduced, and the resulting losses minimized.

[0004] When complex faults occur in the power grid, relying too heavily on the operational experience of dispatching personnel can lead to excessively long fault assessment and handling times, hindering emergency repairs and rapid power restoration. To alleviate the pressure on dispatching personnel during fault handling and avoid human error, implementing auxiliary decision-making mechanisms for power grid fault handling to achieve rapid fault resolution has become an important and effective measure. Summary of the Invention

[0005] 1. The technical problem to be solved:

[0006] To address the aforementioned technical problems, this invention provides an intelligent auxiliary handling method and system for power grid faults. Under different power grid fault information conditions, the system provides targeted auxiliary recommendations for fault measures and electronic contingency plans based on the specific circumstances of the power grid fault, thereby enabling rapid handling of power grid faults.

[0007] 2. Technical Solution:

[0008] A method for intelligent auxiliary handling of power grid faults, characterized by the following steps:

[0009] Step 1: Collect information on all power grid faults that occurred within a preset time period, as well as information on the corresponding handling measures for each fault.

[0010] Step 2: Process the fault information, extract and identify the preset specific fault description items and fault handling measures contained in each fault information; the preset specific fault description items include fault equipment information, fault characteristic information, fault type information, and fault cause information; express each fault information and its fault handling measures in a combination according to the identified information, and classify the combinations with the same corresponding elements into the same fault information combination; count the number of times each identical fault information combination occurs, and generate such fault information and combination of frequency;

[0011] Step 3: Obtain all combinations of all fault information and frequency combinations generated in Step 2; represent all combinations of fault information and frequency combinations as a matrix; use all obtained combinations of fault information and frequency combinations as input terms of a deep convolutional neural network, and the four optimal fault handling measures corresponding to each fault information combination as output terms to train the deep convolutional neural network; the deep convolutional neural network consists of convolutional layers and fully connected layers, with four fully connected layers, each corresponding to an optimal fault handling measure;

[0012] Step 4: Obtain target fault information and identify the preset fault description items contained in the fault information; generate a target fault information combination and input the target fault information combination into the deep convolutional neural network trained in Step 3. The deep convolutional neural network outputs information on 4 optimal handling measures.

[0013] Step 4: Recommend the optimal handling measures to users through the corresponding electronic handling plan.

[0014] Furthermore, in step two, the faulty equipment information, fault characteristic information, fault type information, fault cause information, and handling measures information are represented by sets A, B, C, D, and S, respectively. Then we have...

[0015] A = {A1, A2, A3, ..., A...} nA}

[0016] B = {B1, B2, B3, ..., B} nB}

[0017] C = {C1, C2, C3, ..., C} nC}

[0018] D = {D1, D2, D3, ..., D} nD}

[0019] S = {S1, S2, S3, ..., S} nS}

[0020] In the above formula, A represents the set of all nA faulty devices; B represents the set of all nB fault characteristics; C represents the set of all nC fault types; D represents the set of all nD fault causes; and S represents the set of all nS fault handling measures. Each set contains A... n B n C n D n S n Each represents a specific element in its corresponding set;

[0021] The fault information combinations are classified into "2+1", "3+1" and "4+1" combinations based on the number of fault description items identified.

[0022] The “2+1” combination is a combination in which one element is taken from any two sets A, B, C, and D as the first two elements of the combination, and the corresponding treatment measure in the treatment measure set S as the last element.

[0023] The “3+1” combination is a combination in which one element is taken from any three sets A, B, C, and D as the first three elements of the combination, and the corresponding treatment measure in the treatment measure set S as the last element.

[0024] The “4+1” combination is a combination in which one element is taken from each of the sets A, B, C, and D as the first four elements of the combination, and the corresponding treatment measure in the treatment measure set S as the last element.

[0025] Furthermore, in step three, all combinations of the fault information and occurrence frequency generated in step two are obtained. Specifically, if the identified target fault includes two specific fault description items, then all combinations of the fault information and occurrence frequency include all combinations of the "2+1" combination, the "3+1" combination, and the "4+1" combination that include any two of the two specific fault description items.

[0026] If the identified target fault includes 3 specific fault description items, then all combinations that include the type of fault information and the number of occurrences include all combinations in the "2+1" combination that include any two of the 3 specific fault description items, and all combinations in the "3+1" and "4+1" combinations that include any three of the 3 specific fault description items.

[0027] If the identified target fault includes four specific fault description items, then all combinations that include the fault information and the number of occurrences of this type include all combinations in the "2+1" combination that include any two of the four specific fault description items, all combinations in the "3+1" combination that include any three of the four specific fault description items, and all combinations in the "4+1" combination that include any four of the four specific fault description items.

[0028] Furthermore, the matrix is ​​an m×3 dimensional matrix, where m≤nS; each row of the matrix represents the number of times the “2+1” combination, “3+1” combination, and “4+1” combination corresponding to a fault handling measure occur.

[0029] Furthermore, the electronic emergency response plan includes a plan number, fault name, key points of response, response area, response sequence, and stability requirements, providing assistance for user decision-making. The key points of response mainly describe the measures for restoring power grid faults and the adjustment of emergency status. The response area describes the area where power grid fault response is required. The response sequence describes the steps of power grid fault response measures. The stability requirements mainly include the stability regulations that need to be followed in the fault response work.

[0030] A system for intelligent auxiliary handling of power grid faults includes a data storage module for storing combined data of relevant fault information and handling measures information, as well as the frequency of such data storage.

[0031] The emergency response plan recommendation module recommends emergency response plans for power grid faults.

[0032] The data update module enables real-time updates of relevant information in the system.

[0033] Furthermore, the data update module updates the stored combinations and their frequencies in the system by reading the latest relevant fault information and handling measures information input by the user.

[0034] 3. Beneficial effects:

[0035] (1) The present invention can provide targeted recommendations for fault measures and electronic contingency plans based on the fault conditions under different power grid fault information conditions, so as to realize the rapid handling of power grid faults.

[0036] (2) In the case of insufficient fault information, this invention can also provide auxiliary recommendations for fault measures and electronic contingency plans, and is more adaptable to various information conditions. Attached Figure Description

[0037] Figure 1 This is a flowchart illustrating the process of deriving the optimal treatment measures in a specific embodiment.

[0038] Figure 2 A schematic diagram of a smart auxiliary system for handling power grid faults;

[0039] Figure 3 This is a schematic diagram of the deep convolutional neural network in this invention. Detailed Implementation

[0040] The present invention will now be described in detail with reference to the accompanying drawings.

[0041] A method for intelligent auxiliary handling of power grid faults, characterized by the following steps:

[0042] Step 1: Collect information on all power grid faults that occurred within the preset time period in the past, as well as the corresponding handling measures for each fault.

[0043] Step 2: Process the fault information, extract and identify the preset specific fault description items and fault handling measures contained in each fault information; the preset specific fault description items include fault equipment information, fault characteristic information, fault type information, and fault cause information; express each fault information and its fault handling measures in a combination according to the identified information, and classify the combinations with the same corresponding elements into the same fault information combination; count the number of times each identical fault information combination occurs, and generate such fault information and combinations of occurrence counts;

[0044] Furthermore, if we represent the faulty equipment information, fault characteristic information, fault type information, fault cause information, and handling measure information using sets A, B, C, D, and S respectively, then we have:

[0045] A = {A1, A2, A3, ..., A...} nA}

[0046] B = {B1, B2, B3, ..., B} nB}

[0047] C = {C1, C2, C3, ..., C} nC}

[0048] D = {D1, D2, D3, ..., D} nD}

[0049] S = {S1, S2, S3, ..., S} nS ,}

[0050] In the above formula, A represents the set of all nA faulty devices; B represents the set of all nB fault characteristics; C represents the set of all nC fault types; D represents the set of all nD fault causes; and S represents the set of all nS fault handling measures. Each set contains A...n B n C n D n S n Each represents a specific element in its corresponding set.

[0051] For example: extracting specific combinations of relevant fault information and corresponding handling measures from historical data, including...<A1、B1、S1> ,<A2、B1、S2> ,<B1、D1、S2> ,<A1、B1、C1、S2> ,<A1、B1、C1、D1、S1> wait.

[0052] The fault information combinations are classified into "2+1", "3+1" and "4+1" combinations based on the number of fault description items identified.

[0053] The “2+1” combination is a combination in which one element is taken from any two sets A, B, C, and D as the first two elements of the combination, and the corresponding treatment measure in the treatment measure set S as the last element.

[0054] The “3+1” combination is a combination in which one element is taken from any three sets A, B, C, and D as the first three elements of the combination, and the corresponding treatment measure in the treatment measure set S as the last element.

[0055] The “4+1” combination is a combination in which one element is taken from each of the sets A, B, C, and D as the first four elements of the combination, and the corresponding treatment measure in the treatment measure set S as the last element.

[0056] Therefore, the "2+1" combination includes:<A’、B’、S’> ,<A’、C’、S’> ,<A’、D’、S’> ,<B’、C’、S’> ,<B’、D’、S’> ,<C’、D’、S’> combination;

[0057] The “3+1” combination includes:<A’、B’、C’、S’> ,<A’、B’、D’、S’> ,<B’、C’、D’、S’> ,<A’、C’、D’、S’> combination;

[0058] The “4+1” combination includes:<A’、B’、C’、D’、S’> combination.

[0059] A', B', C', D', and S' above represent an element in the sets A, B, C, D, and S, respectively.

[0060] Step 3: Obtain all combinations of all fault information and frequency combinations generated in Step 2; represent all combinations of fault information and frequency combinations as a matrix; train the deep convolutional neural network using all obtained fault information and frequency combinations as input terms and the four optimal fault handling measures corresponding to each fault information combination as output terms; as shown in the appendix. Figure 3 As shown, this deep convolutional neural network consists of convolutional layers and fully connected layers. There are four fully connected layers, each corresponding to an optimal fault handling measure.

[0061] As attached Figure 1 As shown, if the identified target fault includes 3 specific fault description items, then all combinations including the type of fault information and the number of occurrences include all combinations in the "2+1" combination that include any two of the 3 specific fault description items, and all combinations in the "3+1" combination and the "4+1" combination that include any three of the 3 specific fault description items.

[0062] If the identified target fault includes four specific fault description items, then all combinations that include the fault information and the number of occurrences of this type include all combinations in the "2+1" combination that include any two of the four specific fault description items, all combinations in the "3+1" combination that include any three of the four specific fault description items, and all combinations in the "4+1" combination that include any four of the four specific fault description items.

[0063] For example, if the fault information obtained is<A1、B1> Then it is necessary to read the fault information.<A1、B1> The "2+1" combination data includes: [<A1、B1、S1> [2 times], [<A1、B1、S3> , 5 times], [<A1、B1、S4> , 6 times], [<A1、B1、S7> [2 times] etc.

[0064] At the same time, it is necessary to read the fault information.<A1、B1> The "3+1" combination data includes: [<A1、B1、C1、S1> [2 times], [<A1、B1、D2、S1> , 4 times], [<A1、B1、D4、S2> [6 times] etc.

[0065] At the same time, it is necessary to read the fault information.<A1、B1> The "4+1" combination data includes: [<A1、B1、C2、D2、S1> [1 time], [<A1、B1、C1、D1、S2> , 5 times], [<A1、B1、C1、D4、S3> [2 times] etc.

[0066] Step 4: Recommend the optimal handling measures to users through the corresponding electronic handling plan.

[0067] Furthermore, the electronic emergency response plan includes a plan number, fault name, key points of response, response area, response sequence, and stability requirements, providing assistance for user decision-making. The key points of response mainly describe the measures for restoring power grid faults and the adjustment of emergency status. The response area describes the area where power grid fault response is required. The response sequence describes the steps of power grid fault response measures. The stability requirements mainly include the stability regulations that need to be followed in the fault response work.

[0068] Furthermore, the matrix is ​​an m×3 dimensional matrix, where m≤nS; each row of the matrix represents the number of times the “2+1” combination, “3+1” combination, and “4+1” combination corresponding to a fault handling measure occur.

[0069] The table below shows an m*3 dimensional matrix table. In the table, x represents the statistical frequency of different fault handling measures under each combination, and x_max represents the largest value among these frequencies. The three matrices in the table represent the matrix representation of different combinations of fault information. When x_max is in the first column, it means that the fault information combination is a "2+1" combination; when x_max is in the second column, it means that the fault information combination is a "3+1" combination; and when x_max is in the third column, it means that the fault information combination is a "4+1" combination, corresponding to the three sub-tables a, b, and c in the table, respectively.

[0070] m×3 dimensional matrix

[0071]

[0072] (a) A matrix with a "2+1" combination

[0073]

[0074] (b) Matrix of “3+1” combination

[0075]

[0076] (c) Matrix of the “4+1” combination

[0077] Furthermore, the data storage module is used to store combined data of relevant fault information and handling measures information, as well as the frequency thereof;

[0078] The emergency response plan recommendation module recommends emergency response plans for power grid faults.

[0079] The data update module enables real-time updates of relevant information in the system.

[0080] As attached Figure 2As shown, the data update module updates the stored combination content and its frequency by reading the latest relevant fault information and handling measures information input by the user.

[0081] Although the present invention has been disclosed above with reference to preferred embodiments, these are not intended to limit the invention. Any person skilled in the art can make various changes or modifications without departing from the spirit and scope of the invention. Therefore, the scope of protection of the present invention should be defined by the scope of the claims of this application.

Claims

1. A method for intelligent auxiliary handling of power grid faults, characterized in that: Includes the following steps: Step 1: Collect information on all power grid faults that occurred within a preset time period, as well as information on the corresponding handling measures for each fault. Step 2: Process the fault information, extract and identify the preset specific fault description items and fault handling measures contained in each fault information; the preset specific fault description items include fault equipment information, fault characteristic information, fault type information, and fault cause information; express each fault information and its fault handling measures in a combination according to the identified information, and classify the combinations with the same corresponding elements into the same fault information combination; count the number of times each identical fault information combination occurs, and generate such fault information and combination of frequency; Step 3: Obtain all combinations of all fault information and frequency combinations generated in Step 2; represent all combinations of fault information and frequency combinations as a matrix; use all obtained combinations of fault information and frequency combinations as input terms of a deep convolutional neural network, and the four optimal fault handling measures corresponding to each fault information combination as output terms to train the deep convolutional neural network; the deep convolutional neural network consists of convolutional layers and fully connected layers, with four fully connected layers, each corresponding to an optimal fault handling measure; Step 4: Obtain target fault information and identify the preset fault description items contained in the fault information; generate a target fault information combination and input the target fault information combination into the deep convolutional neural network trained in Step 3. The deep convolutional neural network outputs information on 4 optimal handling measures. Step 5: Recommend the optimal handling measures to users through the corresponding electronic handling plan; In step two, the faulty equipment information, fault characteristic information, fault type information, fault cause information, and handling measures information are represented by sets A, B, C, D, and S, respectively. Then: ； In the above formula, A represents the set of all nA faulty devices; B represents the set of all nB fault characteristics; C represents the set of all nC fault types; D represents the set of all nD fault causes; and S represents the set of all nS fault handling measures. Each set contains... Each represents a specific element in its corresponding set; The fault information combinations are classified into "2+1" combinations, "3+1" combinations, and "4+1" combinations based on the number of fault description items identified. The "2+1" combination is a combination in which one element is taken from any two sets A, B, C, and D as the first two elements of the combination, and the corresponding treatment measure in the treatment measure set S as the last element. The "3+1" combination is a combination in which one element is taken from any three sets A, B, C, and D as the first three elements of the combination, and the corresponding treatment measure in the treatment measure set S as the last element. The "4+1" combination is a combination in which one element is taken from each of the sets A, B, C, and D as the first four elements of the combination, and the corresponding treatment measure in the treatment measure set S as the last element. The matrix is ​​an m×3 matrix, where m≤nS; each row of the matrix represents the number of times the "2+1", "3+1" and "4+1" combinations corresponding to a fault handling measure occur.

2. The intelligent auxiliary handling method for power grid faults according to claim 1, characterized in that: In step three, all combinations of the fault information and occurrence count generated in step two are obtained. Specifically, if the identified target fault includes two specific fault description items, then all combinations of the fault information and occurrence count include all combinations of the "2+1" combination, "3+1" combination, and "4+1" combination that include any two of the two specific fault description items. If the identified target fault includes 3 specific fault description items, then all combinations that include the fault information and the number of occurrences of this type include all combinations in the "2+1" combination that include any two of the 3 specific fault description items, and all combinations in the "3+1" and "4+1" combinations that include any three of the 3 specific fault description items. If the identified target fault includes four specific fault description items, then all combinations that include the fault information and the number of occurrences of this type include all combinations in the "2+1" combination that include any two of the four specific fault description items, all combinations in the "3+1" combination that include any three of the four specific fault description items, and all combinations in the "4+1" combination that include any four of the four specific fault description items.

3. The intelligent auxiliary handling method for power grid faults according to claim 2, characterized in that: The electronic emergency response plan includes a plan number, fault name, key points of response, response area, response sequence, and stability requirements, providing assistance for user decision-making. The key points of response mainly describe the measures for restoring the power grid fault and the adjustment of the emergency state. The response area describes the area where the power grid fault needs to be handled. The response sequence describes the steps of the power grid fault handling measures. The stability requirements mainly include the stability regulations that need to be followed in the fault handling work.

4. A fault handling system applying the intelligent auxiliary fault handling method as described in any one of claims 1-3, characterized in that: The data storage module is used to store combined data of relevant fault information and handling measures information, as well as the frequency of such data storage. The emergency response plan recommendation module recommends emergency response plans for power grid faults. The data update module enables real-time updates of relevant information in the system.

5. The treatment system according to claim 4, characterized in that: The data update module updates the stored combination content and its frequency by reading the latest relevant fault information and handling measures information input by the user.

Images