Centralized control station network transfer method and system based on incremental knowledge graph

Abstract

The invention discloses a centralized control station network transfer method and system based on an incremental knowledge graph. The method comprises the following steps: acquiring a scheduling instruction sent by a scheduling instruction source; decomposing the scheduling instruction based on a pre-constructed incremental knowledge graph so as to obtain different operation tasks, and carrying out network transfer according to task types of the operation tasks; according to the operation task of network order transfer, selecting an operation order generation mode to create an operation order, and editing and checking the created operation order to output a formal operation order; executing each step of operation according to the operation sequence of the formal operation order; archiving execution results of all operations, and storing all information of the operation order in each process node state; according to the method, the knowledge graph is dynamically updated through the incremental learning algorithm, so that the running state change of the power grid equipment is captured in real time, the new entity node is added, accurate decomposition and type identification of the operation task in the dispatching instruction are realized, and the reliability of network transfer is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power systems, in particular to a network order conversion method and system for a centralized control station based on an incremental knowledge graph. BACKGROUND

[0002] As a key node of the modern power dispatching system, the centralized control station of the power grid undertakes the core functions of receiving dispatching instructions, converting operation tasks and executing field control. The traditional operation instruction order conversion process highly depends on manual telephone communication, and has problems such as low instruction analysis efficiency, high error rate and insufficient standardization of operation ticket generation. Especially under the background of smart grid construction, the sources of dispatching instructions are increasingly diversified, and the content of the instructions involves multi-dimensional complex information such as device state, operation mode and safety rules, so the manual processing mode has been difficult to meet the business needs of high real-time and high accuracy.

[0003] In recent years, the knowledge graph has shown great potential in the field of electric power. The knowledge graph can store and correlate the triple knowledge in a graphical structure, express the concepts and their complex relationships in the real world, and form a semantic network with strong reasoning ability. However, existing knowledge graph technologies mainly focus on static knowledge graph modeling, which cannot adapt to the dynamic changes of power grid equipment, resulting in an increased error rate in operation task identification. At the same time, the generation of existing operation tickets mainly depends on fixed template libraries, and lacks quantitative evaluation models for device risk, environmental risk and personnel risk. In the multi-task parallel order conversion scene, there is a problem of insufficient detection of time sequence conflicts of operation steps, which may easily trigger mis-triggered chain failures. SUMMARY

[0004] In view of the above shortcomings of the prior art, the present application provides a network order conversion method and system for a centralized control station based on an incremental knowledge graph, which realizes accurate decomposition and type identification of operation tasks in dispatching instructions by using an incremental knowledge graph, improves the order conversion accuracy and work efficiency, and ensures the safety and stability of the power system operation.

[0005] In order to achieve the above purpose, the present application adopts the following technical solutions:

[0006] In the first aspect, the present application provides a network order conversion method for a centralized control station based on an incremental knowledge graph, comprising the following steps:

[0007] S1, obtaining a dispatching instruction sent by a dispatching instruction source;

[0008] S2, decomposing the dispatching instruction based on a pre-constructed incremental knowledge graph to obtain different operation tasks, and performing network order conversion according to the task type of the operation tasks;

[0009] S3, according to the operation task of the network relay, selecting the operation ticket generation mode to create the operation ticket, and editing and auditing the created operation ticket to output the formal operation ticket;

[0010] S4, executing each step of operation according to the operation sequence of the formal operation ticket;

[0011] S5, archiving the execution results of all operations and saving all information of the operation ticket in each process node state.

[0012] In an optional embodiment, the step of constructing the incremental knowledge graph in S2 comprises:

[0013] Obtaining text data and performing data preprocessing on the text data;

[0014] Constructing a BiLSTM-CRF model and extracting the text data after data preprocessing to obtain entities and relationships between entities in the text, representing the entities and their relationships in the form of triples to generate preliminary knowledge triples;

[0015] Performing semantic disambiguation and standardization processing on the preliminary knowledge triples, and performing knowledge completion and fusion on the triples after standardization processing to obtain final knowledge triples;

[0016] Importing the final knowledge triples into the database Neo4j for storage and management to establish a preliminary knowledge graph;

[0017] Based on the incremental learning algorithm, the knowledge graph is dynamically updated to construct an incremental knowledge graph.

[0018] In an optional embodiment, the step of dynamically updating the knowledge graph based on the incremental learning algorithm comprises the following steps:

[0019] Real-time acquisition of the running state data of the power grid equipment, and based on the incremental learning algorithm, new entities are extracted from the changed part of the running state data after the running state of the power grid equipment changes;

[0020] Through the incremental learning algorithm, the similarity between the existing entities in the knowledge graph and the new entities is calculated, and based on the similarity, it is judged whether the new entity is an existing entity in the knowledge graph;

[0021] If the similarity is lower than the similarity threshold, it means that the new entity is a brand new entity, and a new node is added to the knowledge graph to update the knowledge graph;

[0022] If the similarity is higher than the similarity threshold, it means that the new entity is an existing entity, and the knowledge graph does not need to be updated.

[0023] In an optional implementation, the BiLSTM-CRF model comprises a BiLSTM module and a CRF module.

[0024] The function expression of the BiLSTM module is:

[0025]

[0026]

[0027]

[0028] wherein, input gate; output gate; is the input at time t; is the hidden state at time t-1; is an activation function; , , are input weights of the forget gate, the input gate and the output gate respectively; , , are bias components of the forget gate, the input gate and the output gate respectively; The function expression of the CRF module is:

[0029]

[0030]

[0031] wherein, S(X, y) is a score when the output label sequence is y for a given input sequence X; is a score of transition from label to label ; is a score of the feature vector output by the BiLSTM module at position i corresponding to label ; n is the length of the output sequence.

[0032] In an optional implementation, the S3 specifically comprises the following steps:

[0033] S301, receiving an operation task of a network transfer order;

[0034] S302, selecting an operation ticket generation mode according to the scene type of the operation task to generate a new operation ticket;

[0035] The operation ticket generation mode comprises an intelligent ticket opening mode and a manual ticket opening mode;

[0036] ​​If the intelligent billing mode is selected, a standard operation template pre-stored in the knowledge graph is called to generate and optimize the newly created operation ticket;

[0037] If the manual billing mode is selected, the newly created operation ticket is manually filled according to the network transfer order content to generate the newly created operation ticket;

[0038] S303, the newly created operation ticket is audited;

[0039] If the newly created operation ticket passes the audit, the newly created operation ticket becomes a formal operation ticket, and the formal operation ticket is output;

[0040] If the newly created operation ticket does not pass the audit, the operation steps and sequence in the newly created operation ticket are modified by manual operation, and the modified newly created operation ticket is re-audited.

[0041] In an optional embodiment, the calling of the standard operation template pre-stored in the knowledge graph to generate and optimize the operation ticket comprises the following steps:

[0042] According to the network transfer order content, the standard operation template pre-stored in the knowledge graph is called to generate the newly created operation ticket;

[0043] Based on the analytic hierarchy process, the operation risk of the newly created operation ticket is evaluated to obtain an operation risk evaluation result;

[0044] The power grid operation state data and the device historical operation data are obtained;

[0045] Based on the power grid operation state, the device historical operation data and the operation risk score, the operation steps and sequence in the newly created operation ticket are optimized by a reinforcement learning algorithm to generate an optimized newly created operation ticket.

[0046] In an optional embodiment, the operation risk of the newly created operation ticket is evaluated based on the analytic hierarchy process to obtain an operation risk evaluation result, which specifically comprises the following steps:

[0047] The operation risk factors in the newly created operation ticket are analyzed, and data of the operation risk factors are obtained;

[0048] The operation risk factors include device risk, operation environment risk and personnel risk;

[0049] A hierarchical structure model of the operation risk factors is constructed, and the operation risk and the operation risk factors are respectively set as the highest layer and the intermediate layer;

[0050] The weights of the operation risk factors are determined based on the analytic hierarchy process, and the operation risk factors are quantified;

[0051] Based on the weights and quantified values of the operation risk factors, the operation risk value of the newly created operation ticket is calculated by a weighted comprehensive evaluation method.

[0052] According to the operation risk value, an operation risk level is divided to obtain an operation risk assessment result.

[0053] The operation risk value is calculated according to the following formula:

[0054]

[0055] In the formula, Rnew is a risk value of a new operation ticket; Rdev, Renv and Rperson are weights of device risk, operation environment risk and personnel risk respectively; and Rdevq, Renvq and Rpersonq are quantified values of the device risk, the operation environment risk and the personnel risk respectively. Rdev, Renv and Rperson are weights of device risk, operation environment risk and personnel risk respectively; and Rdevq, Renvq and Rpersonq are quantified values of the device risk, the operation environment risk and the personnel risk respectively. Rdev, Renv and Rperson are weights of device risk, operation environment risk and personnel risk respectively; and Rdevq, Renvq and Rpersonq are quantified values of the device risk, the operation environment risk and the personnel risk respectively.

[0056] In an optional embodiment, the step S4 specifically comprises the following steps:

[0057] Before the formal operation ticket is executed, a responsibility confirmation information is filled by a relevant personnel;

[0058] The responsibility confirmation information includes an operator, a guardian, a duty person in charge, an order issuer, an order receiver and an order issuing time;

[0059] Based on the filled formal operation ticket, each operation step is executed according to an operation sequence of the operation ticket;

[0060] After each operation step is executed, a start time and an end time of the corresponding operation step are recorded.

[0061] In an optional embodiment, the dispatching instruction source includes a dispatching network order issuing system and a power grid resource service middle station; and the operation task includes a remote sequential control operation task and a field operation task.

[0062] The states of the flow nodes include a creating state, an auditing pending state, an auditing completed state, an executing pending state, an executing completed state and an archiving completed state.

[0063] In a second aspect, the application provides a network order issuing system of a centralized control station based on an incremental knowledge graph, which comprises:

[0064] An instruction receiving module is configured to acquire a dispatching instruction sent by a dispatching instruction source;

[0065] An instruction decomposing module is configured to decompose the dispatching instruction based on a pre-constructed incremental knowledge graph to acquire different operation tasks, and to perform network order issuing according to a task type of the operation tasks.

[0066] ​​​​​The operation ticket management module is used to select an operation ticket generation mode according to an operation task of a network transfer order, create an operation ticket, edit and audit the created operation ticket, and output a formal operation ticket after the audit.

[0067] The operation ticket execution module is used to execute each operation step according to an operation sequence of the formal operation ticket.

[0068] The data storage module is used to archive execution results of all operations and save all information of the operation ticket in each process node state.

[0069] The embodiments provided by the present application have the following beneficial effects:

[0070] The present application dynamically updates the knowledge graph through the incremental learning algorithm to capture the changes in the operation state of the power grid equipment in real time and add new entity nodes, ensures that the knowledge graph and the real-time state of the power grid are updated synchronously, realizes accurate decomposition and type identification of operation tasks in the dispatching instruction, effectively overcomes the instruction analysis deviation problem caused by the static knowledge graph model, and improves the reliability of the network transfer order.

[0071] The present application generates an operation ticket by calling the pre-stored standard operation template of the incremental knowledge graph, uses the analytic hierarchy process to evaluate the operation risk of the operation ticket, and optimizes the operation steps and execution timing of the operation ticket based on the reinforcement learning algorithm according to the operation risk evaluation result of the operation ticket, which significantly improves the operation safety and execution efficiency. BRIEF DESCRIPTION OF DRAWINGS

[0072] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or prior art description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other embodiments can also be obtained by those skilled in the art according to these drawings.

[0073] Figure 1 A flow chart of the centralized control station network transfer order method based on the incremental knowledge graph in the embodiments of the present specification is shown;

[0074] Figure 2 A centralized control station network transfer order example in the embodiments of the present specification is shown;

[0075] Figure 3 A centralized control station network transfer order system structure diagram of the knowledge graph in the embodiments of the present specification is shown. DETAILED DESCRIPTION

[0076] The features and exemplary embodiments of the various aspects of the present application will be described in detail below. In the following detailed description, numerous specific details are set forth in order to provide a thorough understanding of the present application. However, it will be apparent to one of ordinary skill in the art that the present application can be practiced without some or all of these specific details. The description of the embodiments is merely illustrative of the present application and is not intended to limit the present application.

[0077] Manual telephone order has been difficult to meet the requirements of safe operation of smart grid, and power system urgently needs to realize intelligent order and intelligent receiving order by means of advanced technologies such as artificial intelligence, so as to reduce personnel burden, standardize operation process, prevent operation accidents caused by manual telephone order errors, and ensure the safety and stability of power system operation. At present, the application and development of artificial intelligence technology are profoundly changing the production and life of human beings. Among them, the knowledge graph stores and correlates the triple knowledge in the form of graph structure, expresses the concepts and complex relationships in the real world, forms a semantic network with strong reasoning ability, and is a technical direction with machine cognitive intelligence. It provides a new idea for solving the problems faced by intelligent order and intelligent receiving order. Based on this, the present application provides a network order scheme for a centralized control station based on an incremental knowledge graph, so as to improve the order accuracy and work efficiency, prevent operation accidents caused by manual telephone order errors, ensure the safety and stability of power system operation, and improve the practical level of one-key sequence control of the centralized control station.

[0078] The technical solutions of the present application will be described in detail below in combination with specific embodiments.

[0079] Embodiment 1

[0080] As shown in the accompanying drawings, Figure 1 the present application provides a network order method for a centralized control station based on an incremental knowledge graph, which comprises the following steps:

[0081] S1, obtaining a dispatching instruction sent by a dispatching instruction source;

[0082] The dispatching instruction source includes but is not limited to a dispatching network order system and a power grid resource business platform.

[0083] S2, decomposing the dispatching instruction based on a pre-constructed incremental knowledge graph to obtain different operation tasks, and performing network order according to the task type of the operation tasks; wherein the operation tasks include remote sequence control operation tasks and on-site operation tasks;

[0084] Exemplarily, the construction steps of the incremental knowledge graph in S2 include:

[0085] obtaining text data and performing data preprocessing on the text data;

[0086] wherein the text data comprises power grid model information, power grid operation mode information and one-key sequence control operation list;

[0087] A BiLSTM-CRF model is constructed, and the text data after data preprocessing is extracted to obtain entities in the text and the relationships between the entities, and the entities and the relationships are represented in the form of triplets to generate preliminary knowledge triplets; wherein the BiLSTM-CRF model comprises a BiLSTM module and a CRF module.

[0088] It should be noted that BiLSTM (Bi-directional Long Short Term Memory) contains two independent LSTM layers to extract features and encode context information for the input sequence. CRF (Conditional Random Field) is a graph model of joint probability distribution represented by an undirected graph, which is used to globally optimize the sequence labeling result to obtain the optimal label sequence.

[0089] In this embodiment, the text data after data preprocessing is converted into a word vector sequence H=[h1, h2, …, h n ] through a pre-trained language model, and the BiLSTM module in the BiLSTM-CRF model is used to extract features from the word vector sequence to output a feature vector sequence X=[x1, x2, …, x n ]; the feature vector sequence X is used as the given input sequence X of the CRF module to perform sequence labeling on the feature vector sequence X, and the local features are normalized to global features by calculating the probability distribution of the overall sequence to solve the local labeling bias problem, thereby obtaining the globally optimal label sequence to output the entity recognition result in the text data; after obtaining the entity recognition result, the dependency paths between the entities are extracted through the dependency analysis for the sentences containing the target entities in the text data to obtain the relationships between the entities.

[0090] Specifically, the function expression of the BiLSTM module is:

[0091]

[0092]

[0093]

[0094] wherein, input gate; output gate; input at time t; hidden state at time t-1; is an activation function; , , are input weights of the forget gate, the input gate and the output gate respectively; , , are bias components of the forget gate, the input gate and the output gate respectively;

[0095] Specifically, the function expression of the CRF module is as follows:

[0096]

[0097] wherein S(X, y) is a score when the output label sequence is y for a given input sequence X, and the input sequence X is a feature vector sequence; is a score representing transition from label to label in the CRF state transition matrix; is a score corresponding to label output by the BiLSTM module at position i; and n is the length of the output sequence.

[0098] Specifically, the function expression of the optimal label sequence is as follows:

[0099]

[0100] wherein is the optimal label sequence, and Y is a set of valid label sequences;

[0101] The preliminary knowledge triplets are subjected to semantic disambiguation and standardization, and the triplets after the standardization are subjected to knowledge completion and fusion to obtain final knowledge triplets;

[0102] The final knowledge triplets are imported into a database Neo4j for storage and management to establish a preliminary knowledge graph;

[0103] The knowledge graph is dynamically updated based on an incremental learning algorithm to construct an incremental knowledge graph.

[0104] Specifically, the knowledge graph is dynamically updated based on the incremental learning algorithm, and the specific steps include the following:

[0105] The running state data of the power grid equipment are acquired in real time, and after the running state of the power grid equipment changes, new entities are extracted from the changed part of the running state data based on the incremental learning algorithm;

[0106] ​​The similarity between the existing entity in the knowledge graph and the new entity is calculated through an incremental learning algorithm, and whether the new entity is an existing entity in the knowledge graph is determined based on the similarity;

[0107] If the similarity is lower than the similarity threshold, it means that the new entity is a brand new entity, and a new node is added to the knowledge graph to update the knowledge graph;

[0108] If the similarity is higher than the similarity threshold, it means that the new entity is an existing entity, and the knowledge graph does not need to be updated.

[0109] It should be noted that the scheduling instruction content is based on the instruction decomposition of the incremental knowledge graph to identify whether the task type of the operation task is a remote operation task or an on-site operation task, and then perform network transfer according to the task type; as Figure 2 shown, taking the scheduling instruction 220kV xx line 606 from II mother operation to maintenance as an example, the process of scheduling instruction based on knowledge graph analysis and reasoning to obtain operation task and network transfer is explained:

[0110] The first step is to analyze the structure of the scheduling instruction:

[0111] The scheduling instruction 220kV xx line 606 from II mother operation to maintenance is analyzed as follows: the operation object is 220kV xx line 606; the current state of the operation object is II mother operation; the target state of the operation object is maintenance;

[0112] The second step is to decompose the task of the scheduling instruction:

[0113] According to the operation object 220kV xx line 606, the pre-defined task template in the incremental knowledge graph is matched, that is, the state transition path of running→cold standby→maintenance in the knowledge graph is matched, and the scheduling instruction is decomposed into the following first operation task and second operation task:

[0114] 220kV xx line 606 from II mother operation to cold standby (first operation task);

[0115] 220kV xx line 606 from cold standby to maintenance (second operation task);

[0116] The third step is to dynamically reason and decide the task sequence:

[0117] The sub-task sequence is determined through the knowledge graph relationship (sequential strong constraint), that is, the task sequence must first reach the cold standby state and then reach the maintenance state;

[0118] The fourth step is to determine the task type and network transfer:

[0119] The first operation task and / or the second operation task are determined whether to be remote operation based on the incremental knowledge graph, if the operation task can be remote operation, it is determined as a remote operation task, otherwise it is determined as an on-site operation task;

[0120] The network transfer order is performed according to the task type of the first operation task and the second operation task.

[0121] S3, according to the operation task of the network transfer order, selecting an operation ticket generation mode to create an operation ticket, and editing and auditing the created operation ticket to output a formal operation ticket;

[0122] Exemplarily, S3 specifically includes the following steps:

[0123] S301, receiving an operation task of network transfer order;

[0124] S302, according to the scene type of the operation task, selecting an operation ticket generation mode to generate a new operation ticket;

[0125] The operation ticket generation mode includes an intelligent ticketing mode and a manual ticketing mode;

[0126] If the intelligent ticketing mode is selected, the standard operation template pre-stored in the knowledge graph is called to generate and optimize the new operation ticket;

[0127] If the manual ticketing mode is selected, manual ticketing is performed according to the network transfer order content to generate a new operation ticket;

[0128] In this embodiment, if the scene type of the operation task is a regular switching operation, the intelligent ticketing mode can be selected, and the operation ticket is intelligently generated by the pre-stored standard template of the incremental knowledge graph, which can improve the ticketing efficiency; if the scene type of the operation task is emergency fault handling or first-time operation of new equipment, the manual ticketing mode is selected, because the intelligent ticketing mode cannot dynamically decide in emergency fault handling, and the first-time operation of new equipment lacks historical data and cannot support intelligent modeling, which needs to rely on manual experience.

[0129] S303, auditing the new operation ticket;

[0130] If the new operation ticket passes the audit, the new operation ticket becomes a formal operation ticket, and the formal operation ticket is output;

[0131] If the new operation ticket does not pass the audit, the operation steps and sequence in the new operation ticket are modified by manual, and the modified new operation ticket is re-audited.

[0132] In the embodiment, if the manual billing mode is selected, the operation steps that have passed the audit will be locked to prevent false modification. If the formal operation ticket that has passed the audit needs to be modified subsequently, the process state must be rolled back to the editable state, and then the authorized personnel modifies it and triggers the complete audit process again.

[0133] It should be noted that the intelligent billing mode and the manual billing mode both need to allocate a continuous and unique temporary ticket number when generating a new ticket; the core functions of the temporary ticket number include: unique identification, ensuring that each operation ticket has a unique identification; security time control, the temporary ticket number has a validity period, and if it is not used within the validity period, it will be automatically invalidated, and a new operation ticket needs to be issued; and repeated operation prevention, the system or operator of the control station operation ticket will refuse to execute the instruction of using the same temporary ticket number repeatedly, effectively avoiding the misexecution of historical operation tickets.

[0134] Wherein, when the new operation ticket is deleted or converted into a formal operation ticket, the temporary ticket number corresponding to the new ticket will be deleted.

[0135] Specifically, the calling of the pre-stored standard operation template of the knowledge graph to generate and optimize the operation ticket includes the following steps:

[0136] According to the network transfer order content, a new operation ticket is generated by calling the pre-stored standard operation template of the knowledge graph;

[0137] Based on the analytic hierarchy process, the operation risk of the new operation ticket is evaluated to obtain an operation risk evaluation result;

[0138] Obtain power grid operation state data and device historical operation data;

[0139] Based on the power grid operation state, the device historical operation data and the operation risk score, the operation steps and sequence in the new operation ticket are optimized through a reinforcement learning algorithm to generate an optimized new operation ticket.

[0140] Specifically, the operation risk of the new operation ticket is evaluated based on the analytic hierarchy process to obtain an operation risk evaluation result, which specifically includes the following steps:

[0141] Analyzing the operation risk factors in the new operation ticket and obtaining data of the operation risk factors;

[0142] Wherein, the operation risk factors include device risk, operation environment risk and personnel risk;

[0143] In some embodiments, the operation object information (operation object information includes device number, operation position and operation personnel information) of the new operation ticket is extracted by analyzing the new operation ticket, and the data of the corresponding operation risk factors is obtained from the integrated system of the control station;

[0144] In the embodiment, the equipment risk data includes but is not limited to real-time monitoring data and historical maintenance data; the operating environment risk data includes but is not limited to real-time gas monitoring data and real-time microclimate monitoring data; the personnel risk data includes but is not limited to the operation certificate validity period of the operator, the safety regulation test score of the operator and the historical number of times of violation of rules by the operator;

[0145] A hierarchical model of the operation risk factors is constructed, and the operation risk and the operation risk factors are respectively set as the highest layer and the intermediate layer;

[0146] The weight of each operation risk factor is determined based on the analytic hierarchy process, and each operation risk factor is quantified;

[0147] In the embodiment, the quantification of the equipment risk adopts an aging index, that is, the ratio of the service life of the equipment to the design life thereof, for example, when the ratio is greater than 80%, the equipment can be evaluated as a high risk level.

[0148] The quantification of the operating environment risk is performed according to the key environmental parameters to set the risk scores, for example, when the real-time wind speed of the operating environment exceeds 10 m / s, the highest risk score is given (the operation is prohibited when the real-time wind speed of the operating environment exceeds 10 m / s); when the relative humidity of the operating environment exceeds 80%, the corresponding risk score is given (the relative humidity exceeding 80% can increase the risk of pollution flash); when the SF6 gas leakage concentration (volume fraction) in the operating environment exceeds 1000 μL / L, the highest risk score is given. The personnel risk is quantified by using a personnel risk quantification formula.

[0149] The personnel risk quantification formula is G = F·(β-E).

[0150] In the formula, G is the personnel risk quantification value; F is the operation personnel qualification risk, which is calculated from the personnel risk data; (β-E) is the working experience gap; β is the maximum working experience coefficient; E is the working experience coefficient of the operation personnel, which reflects the weakening effect of the experience of the operation personnel on the risk, wherein E = actual working life of the operation personnel / base working life.

[0151] Based on the weight and the quantification value of each operation risk factor, the operation risk value of the newly created operation ticket is calculated by using a weighted comprehensive evaluation method;

[0152] According to the operation risk value, the operation risk level is divided to obtain an operation risk evaluation result;

[0153] The calculation formula of the operation risk value is as follows:

[0154]

[0155] In the formula, R is the risk value of the newly created operation ticket; and ,​ , are weights of the equipment risk, the operating environment risk and the personnel risk, respectively; , , are quantitative values of the equipment risk, the operating environment risk and the personnel risk, respectively;

[0156] In the embodiment, the operation risk level is evaluated by dividing the operation risk level according to the operation risk value based on preset different risk level thresholds; if the risk value R is in the range of 0-30, it is determined as a low risk level; if the risk value R is in the range of 31-60, it is determined as a medium risk level; if the risk value R is in the range of 61-100, it is determined as a high risk level.

[0157] In some embodiments, corresponding risk response measures can be provided to the operator according to different risk levels; for example, when the operation involves important equipment or complex operation, the system automatically prompts the operator to pay attention to matters and provides detailed operation guidance to ensure safe and reliable operation.

[0158] S4, performing each step of operation according to the operation sequence of the formal operation ticket;

[0159] For example, step S4 specifically includes the following steps:

[0160] Before the formal operation ticket is executed, the responsibility confirmation information is filled out by the relevant personnel;

[0161] The responsibility confirmation information includes the operator, the guardian, the on-duty person in charge, the order issuer, the order receiver and the order issuing time;

[0162] Based on the formal operation ticket after the filling is completed, each step of operation is performed according to the operation sequence of the operation ticket;

[0163] After each operation is completed, the start time and the end time of the corresponding operation are recorded.

[0164] S5, archiving the execution results of all operations and saving all information of the operation ticket in each process node state.

[0165] The process node states include a creation state, a to-be-audited state, an audited state, a to-be-executed state, an executed state and an archived state.

[0166] In some embodiments, the operation ticket information in all states can be queried and counted by the user; the graphical user interface can be used to intuitively show the operation ticket information and the state flow process to the user, and the user can conveniently query the detailed information of each type of operation ticket, filter the data in a specific state, or select to view specific content by clicking buttons, dragging sliders or inputting texts.

[0167] In a second aspect, the present application provides a centralized station network order transfer system 100 based on incremental knowledge graph, comprising:

[0168] An instruction receiving module 101 is configured to obtain a dispatching instruction sent by a dispatching instruction source;

[0169] An instruction decomposing module 102 is configured to decompose the dispatching instruction based on a pre-constructed incremental knowledge graph to obtain different operation tasks, and perform network order transfer according to the task type of the operation tasks;

[0170] An operation order management module 103 is configured to select an operation order generation mode to create an operation order according to the operation tasks of the network order transfer, and edit and audit the created operation order to output a formal operation order after the audit;

[0171] An operation order executing module 104 is configured to execute each step of operation according to the operation sequence of the formal operation order;

[0172] A data storage module 105 is configured to archive the execution results of all operations and save all information of the operation order in each process node state.

[0173] Embodiment 3

[0174] The electronic device includes at least one control processor and a memory in communication with the at least one control processor;

[0175] The memory, as a non-transitory computer readable storage medium, can be used to store non-transitory software programs and non-transitory computer executable programs. In addition, the memory can include a high-speed random access memory and can also include a non-transitory memory, such as at least one magnetic disk storage device, a flash memory device, or other non-transitory solid-state memory device. In some embodiments, the memory can optionally include a memory remotely disposed relative to the processor, and these remote memories can be connected to the processor through a network. Examples of the above-mentioned network include but are not limited to the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0176] The non-transitory software programs and instructions required for implementing the above-mentioned embodiments of the centralized station network order transfer method based on incremental knowledge graph are stored in the memory, and when executed by the processor, the above-mentioned embodiments of the centralized station network order transfer method based on incremental knowledge graph are executed, for example, the method steps S1 to S5 in the above-described Figure 1 are executed.

[0177] The system embodiments described above are merely illustrative, wherein the units described as separate components can or can not be physically separated, i.e., can be located in one place, or can be distributed to multiple network units. Part or all of the units can be selected according to actual needs to achieve the purposes of the embodiments.

[0178] Embodiment 4

[0179] The embodiment provides a computer readable storage medium, the computer readable storage medium stores computer executable instructions, and the computer executable instructions are used for causing a computer to execute a centralized station network order changing method based on an incremental knowledge graph.

[0180] It should be noted that the computer readable storage medium of the embodiment can be a computer readable signal medium or a computer readable storage medium, or any combination of the two. The computer readable storage medium may, for example, but is not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or apparatus, or any combination of the above. More specific examples of the computer readable storage medium can include, but are not limited to, an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0181] In the embodiment, the computer readable storage medium can be any tangible medium that contains or stores a program, which can be used by or in conjunction with an instruction execution system, device or apparatus. In the embodiment, the computer readable signal medium can include a data signal carried in a baseband or as a part of a carrier wave, which carries computer readable programs. Such a propagated data signal can take various forms, including but not limited to an electromagnetic signal, an optical signal or any suitable combination of the above. The computer readable signal medium can also be any computer readable medium other than the computer readable storage medium, which can send, propagate or transmit programs for use by or in conjunction with an instruction execution system, device or apparatus. The computer programs contained in the computer readable storage medium can be transmitted by any suitable medium, including but not limited to a wire, an optical cable, an RF (radio frequency) or the like, or any suitable combination of the above.

[0182] The computer readable storage medium described above can be written in any one or combination of one or more programming languages, including an object oriented programming language such as Python, Java, C++ or conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program can execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the latter scenario, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection can be made to an external computer (for example, through the Internet using an Internet Service Provider).

[0183] The above description is only preferred embodiments of the present application, and it is obvious for those skilled in the art that the present application is not limited to the details of the above exemplary embodiments, but can be implemented in other concrete forms without departing from the spirit or essential characteristics of the present application. Therefore, the embodiments should be considered in all aspects as illustrative and not restrictive, and the scope of the present application is defined by the appended claims rather than the above description, and it is intended to embrace all changes falling within the meaning and range of equivalents of the claims.

Claims

1. A method for transferring network commands in a centralized control station based on an incremental knowledge graph, characterized in that: The following steps are involved: S1. Obtain a scheduling instruction sent by a scheduling instruction source; S2. Decompose the scheduling instructions based on the pre-built incremental knowledge graph to obtain different operation tasks, and perform network transfer according to the task type of the operation task; S3. Select an operation ticket generation mode based on the operation task transferred from the network to create an operation ticket, and edit and review the created operation ticket to output a formal operation ticket; S4. Execute each operation step according to the operation sequence of the formal operation ticket; S5. Archive the execution results of all operations and save all information of the operation ticket in each process node status.

2. The method according to claim 1, characterized in that The steps for constructing the incremental knowledge graph in S2 include: Acquire text data and perform data preprocessing on the text data; A BiLSTM-CRF model is constructed and the pre-processed text data is extracted to obtain entities in the text and the relationships between entities. The entities and their relationships are represented by triples to generate preliminary knowledge triples. The preliminary knowledge triples are semantically disambiguated and standardized, and the standardized triples are supplemented and integrated to obtain the final knowledge triples. Import the final knowledge triples into the Neo4j database for storage and management to build a preliminary knowledge graph; The knowledge graph is dynamically updated based on an incremental learning algorithm to construct an incremental knowledge graph.

3. The method according to claim 2, characterized in that The dynamic updating of the knowledge graph based on the incremental learning algorithm specifically includes the following steps: Acquire the operating status data of the power grid equipment in real time, and extract new entities from the changed part of the operating status data based on the incremental learning algorithm after the operating status of the power grid equipment changes; Calculate the similarity between existing entities in the knowledge graph and the new entity through an incremental learning algorithm, and determine whether the new entity is an existing entity in the knowledge graph based on the similarity; If the similarity is lower than the similarity threshold, it means that the new entity is a brand new entity, and a new node is added to the knowledge graph to update the knowledge graph; If the similarity is higher than the similarity threshold, it means that the new entity is an existing entity and the knowledge graph does not need to be updated.

4. The method according to claim 2, characterized in that The BiLSTM-CRF model includes a BiLSTM module and a CRF module; The function expression of the BiLSTM module is: ； ； ； Where, Input gate; Output gate; is the input at time t; is the implicit state at time t-1; is the activation function; 、 、 They are forget gate, input gate and output gate respectively The input weight of 、 、 They are forget gate, input gate and output gate respectively The bias component of The function expression of the CRF module is: ； Where S(X,y) is the score when the output label sequence is y for a given input sequence X; For label Transfer to label score; The feature vector output by the BiLSTM module at position i corresponds to the label The fraction of ; n is the length of the output sequence.

5. The method according to claim 1, wherein The S3 specifically includes the following steps: S301, receiving an operation task of a network transfer command; S302: Select an operation ticket generation mode according to the scenario type of the operation task to generate a new operation ticket; The operation ticket generation mode includes intelligent invoicing mode and manual invoicing mode; If the intelligent invoicing mode is selected, the standard operation template pre-stored in the knowledge graph is called to generate and optimize the new operation ticket; If manual invoicing mode is selected, manual invoice filling will be performed based on the network transfer order content to generate a new operation ticket; S303: Review the newly created operation ticket; If the newly created operation ticket passes the review, it will be converted into a formal operation ticket and output as a formal operation ticket; If the newly created operation ticket fails to pass the review, the operation steps and sequence in the newly created operation ticket will be manually modified, and the modified new operation ticket will be reviewed again.

6. The method according to claim 5, characterized in that The calling of the pre-stored standard operation template in the knowledge graph to generate and optimize the operation ticket includes the following steps: According to the content of the network transfer order, the standard operation template pre-stored in the knowledge graph is called to generate a new operation ticket; Conduct operational risk assessment on newly created operation tickets based on the analytic hierarchy process to obtain operational risk assessment results; Obtain grid operation status data and equipment historical operation data; Based on the grid operation status, historical equipment operation data and operation risk assessment results, the operation steps and sequence in the new operation ticket are optimized through the reinforcement learning algorithm to generate an optimized new operation ticket.

7. The method according to claim 6, characterized in that The operation risk assessment of the newly created operation ticket based on the hierarchical analysis method to obtain the operation risk assessment result specifically includes the following steps: Analyze the operational risk factors in the newly created operation ticket and obtain the data of the operational risk factors; Said operational risk factors include equipment risk, operating environment risk and personnel risk; Construct a hierarchical model of operational risk factors to set operational risk and operational risk factors as the highest level and middle level respectively; Determine the weight of each operational risk factor based on the analytic hierarchy process and quantify each operational risk factor; Based on the weights and quantitative values ​​of each operational risk factor, the operational risk value of the newly created operation ticket is calculated using a weighted comprehensive evaluation method; Classify the operational risk level according to the operational risk value to obtain an operational risk assessment result for a newly created operation ticket; The calculation formula of the operational risk value is: ； Where, The risk value of the newly created operation ticket; 、 、 are the weights of equipment risk, operating environment risk and personnel risk respectively; 、 、 They are the quantitative values ​​of equipment risk, operating environment risk and personnel risk respectively.

8. The method according to claim 1, characterized in that Step S4 specifically includes the following steps: Before the formal operation ticket is executed, the relevant personnel shall fill in the responsibility confirmation information; The responsibility confirmation information includes the operator, guardian, person in charge on duty, person who issues the order, person who receives the order and the time of issuance; Based on the completed formal operation ticket, perform each step according to the operation sequence of the operation ticket; After each operation is completed, the start time and end time of the corresponding operation are recorded.

9. The method according to claim 1, characterized in that The dispatch instruction source includes a dispatch network command system and a power grid resource business middle station; The operation tasks include remote sequential control operation tasks and on-site operation tasks; The status of each process node includes creation status, pending review status, reviewed status, pending execution status, executed status, and archived status.

10. A centralized control station network transfer system based on incremental knowledge graph, characterized in that: include: An instruction receiving module is used to obtain the scheduling instruction sent by the scheduling instruction source; An instruction decomposition module is used to decompose scheduling instructions based on a pre-built incremental knowledge graph to obtain different operation tasks, and perform network transfer according to the task type of the operation task; The operation ticket management module is used to select the operation ticket generation mode to create the operation ticket according to the operation task transferred by the network, and edit and review the created operation ticket to output the formal operation ticket after review; The operation ticket execution module is used to execute each operation step according to the operation sequence of the formal operation ticket; The data storage module is used to archive the execution results of all operations and save all information of the operation ticket in each process node status.

Images