A networked order-issuing interactive system for distribution network

Abstract

This application discloses a networked command and interaction system for distribution networks, relating to the field of distribution network control technology. It includes a plant terminal, an operation and maintenance terminal, a dispatch terminal, and a network command issuing terminal communicatively connected to each of the plant terminal, operation and maintenance terminal, and dispatch terminal. The dispatch terminal receives command information input by the dispatcher and generates pre-commands and positive commands based on the command information, then uploads the pre-commands and positive commands to the network command issuing terminal. The network command issuing terminal receives the pre-commands and positive commands generated and uploaded by the dispatch terminal, and simultaneously issues the pre-commands and positive commands to both the plant terminal and the operation and maintenance terminal. The plant terminal executes the pre-commands and positive commands issued by the network command issuing terminal. The dispatch terminal receives the pre-commands and positive commands issued by the network command issuing terminal. This application effectively realizes the construction of a networked dispatch command information interaction management mode, improves the real-time performance and quality of on-site operation and maintenance dispatch control operations, promotes close collaboration between dispatch and operation and maintenance, and enhances the efficiency of distribution network control work.

Description

Technical Field

[0001] This application relates to the field of power distribution network control technology, and in particular to a networked command and interaction system for power distribution networks. Background Technology

[0002] A power distribution network consists of overhead lines, cables, poles, distribution transformers, disconnect switches, reactive power compensators, and some ancillary facilities. It plays a crucial role in distributing electrical energy within the power grid. Power dispatching is an effective management method used to ensure the safe and stable operation of the power grid, reliable power supply to external systems, and the orderly conduct of various power production activities.

[0003] Based on technologies such as computer networks, intelligent telephone voice, and mobile WebSocket + Node, the interaction of dispatching commands between "dispatch -> power plants / distribution networks" can be directly conducted through electronic digitization and message reminders. Power plant users can use the three-zone network WEB application and mobile terminals to receive pre-orders, repeat official orders, apply for commencement of work, report completion, report equipment anomalies, and report grid fault information. Dispatch users can issue orders, grant permissions, and confirm information asynchronously via text, achieving intelligent reminders such as task ringing and real-time push notifications, maximizing the quality and efficiency of interaction between power plant personnel and dispatchers.

[0004] However, due to the poor real-time performance and quality of on-site operation and maintenance dispatching and control command information exchange, it is difficult for dispatching and operation and maintenance to work closely together, which in turn affects the efficiency of distribution network control work and needs to be improved. Summary of the Invention

[0005] In view of this, the purpose of this application is to provide a networked command and interaction system for distribution networks. This system improves the real-time performance and quality of on-site operation and maintenance scheduling and control operations, promoting close collaboration between scheduling and operation and maintenance, thereby enhancing the efficiency of distribution network control. The specific solution is as follows:

[0006] A networked command and interaction system for power distribution network includes a plant end, an operation and maintenance end, a dispatch end, and a network command issuing end that is communicatively connected to the plant end, the operation and maintenance end, and the dispatch end, respectively.

[0007] The scheduling terminal is used to receive instruction information input by the scheduler, generate pre-instructions and positive instructions based on the instruction information, and upload the pre-instructions and positive instructions to the network command issuing terminal;

[0008] The network command terminal is used to receive the pre-instructions and positive instructions generated and uploaded by the scheduling terminal, and simultaneously send the pre-instructions and positive instructions to the plant terminal and the operation and maintenance terminal.

[0009] The factory terminal is used to execute the pre-instructions and positive instructions issued by the network command terminal;

[0010] The scheduling terminal is used to receive pre-instructions and positive instructions issued by the network command terminal.

[0011] Preferably, the scheduling terminal includes a pre-instruction processing module, a positive instruction processing module, and an instruction receipt module;

[0012] The pre-instruction processing module consists of a pre-instruction generation unit and a pre-instruction issuing unit. The pre-instruction generation unit is used to process instruction information to generate pre-instruction information, and the pre-instruction issuing unit is used to send the pre-instruction information to the instruction receipt module.

[0013] The positive instruction processing module consists of a positive instruction generation unit and a positive instruction issuing unit. The positive instruction generation unit is used to process instruction information to generate positive instruction information, and the positive instruction issuing unit is used to send the positive instruction information to the instruction receipt module.

[0014] The instruction receipt module is used to receive pre-instruction information and positive instruction information, generate pre-instructions and positive instructions according to the pre-instruction information and positive instruction information respectively, and upload the pre-instructions and positive instructions to the network command issuing terminal.

[0015] Preferably, the instruction receipt module consists of an instruction repetition unit, an instruction confirmation unit, a pre-instruction receipt execution unit, and a formal instruction receipt execution unit;

[0016] The instruction repeating unit is used to repeat pre-instruction information and positive instruction information;

[0017] The instruction confirmation unit is used to confirm the pre-instruction information and the positive instruction information repeated by the instruction recitation unit, and to send the pre-instruction information to the pre-instruction receipt and execution unit and the positive instruction information to the positive instruction receipt and execution unit.

[0018] The pre-instruction receipt and execution unit is used to generate pre-instructions for the plant equipment based on the pre-instruction information, and upload the pre-instructions to the network command issuing terminal;

[0019] The positive instruction receipt and execution unit is used to generate positive instructions for the plant equipment based on the positive instruction information and upload the positive instructions to the network command issuing terminal.

[0020] Preferably, the method for generating positive and pre-instructions at the scheduling end includes the following steps:

[0021] S1. Obtain control voice information;

[0022] S2. Input the control voice information into the pre-trained control instruction generation model to obtain the control instruction to be activated corresponding to the control voice information;

[0023] S3. Determine the target user identifier corresponding to the controlled voice information;

[0024] S4. Generate target control instructions based on the target user identifier and the control instructions to be sent.

[0025] Preferably, in S1, the control voice information is issued by the control center staff and includes information on power system control instructions, which are instructions related to the operation and maintenance of the power system.

[0026] Preferably, in S2, the construction of the control command generation model includes the following steps:

[0027] Step 1: Collect dispatch speech data, perform noise reduction and feature sequence extraction on the collected dispatch speech data, and use the extracted feature sequences to construct a dispatch speech corpus, which includes a general corpus, a power industry corpus, and standard speech signals for matching degree testing.

[0028] Step 2: Use the scheduling speech corpus as input data and input it into the trained acoustic model and language model; the trained acoustic model and language model jointly output the scheduling telephone text, and the trained acoustic model and language model use the scheduling speech corpus as the training set and are trained using deep learning algorithms.

[0029] Step 3: Collect regulatory command corpus and vectorize the words in the regulatory command corpus to construct a regulatory command corpus based on the obtained regulatory command word vectors;

[0030] Step 4: Using the word vectors of the control instructions as input data, input them into the domain recognition model and the intent recognition model that have been trained respectively; the domain recognition model outputs the domain matching result, and the intent recognition model outputs the intent recognition result based on the domain matching result. The domain recognition model and the intent recognition model that have been trained use the control instruction corpus as the training set and are trained using deep learning algorithms.

[0031] Step 5: Use the intent recognition result as the input data of the slot recognition model, and the slot recognition model outputs the control instruction matching result. The slot recognition model extracts semantic tags from the scheduling instructions using a rule parsing algorithm.

[0032] Step 6: Based on the matching results of the control instructions, clarify the dispatch instructions to prepare for power dispatch; determine the grid operating status by combining grid power flow calculation, short-circuit current calculation and circuit breaker operation; at the same time, execute power information query, grid automatic calculation and power automatic dispatch response operations based on the matching results of the control instructions.

[0033] Step 7: Input the dispatch call text and the matching results of the control instructions into the speech synthesis engine, convert them into dispatch call voice, and use the dispatch call voice to automatically respond to the dispatch call.

[0034] Preferably, step 2 includes:

[0035] Step 2.1: Use the scheduled speech corpus as input data and input it into the trained acoustic model and language model;

[0036] Step 2.2: The trained acoustic model performs noise processing and acoustic feature extraction on the scheduled speech corpus to output an acoustic feature sequence; the matching degree of the acoustic feature sequence is obtained using a standard speech signal.

[0037] Step 2.3: The trained language model performs noise processing and semantic feature extraction on the scheduled speech corpus to output a semantic feature sequence; using standard speech signals, the matching degree of the semantic feature sequence is obtained, i.e., the language model score.

[0038] Step 2.4: The feature sequence with the highest combined score of the acoustic model and language model is used as the dispatch telephone feature sequence;

[0039] Step 2.5: Based on the power industry corpus, perform language decoding on the dispatch telephone feature sequence to obtain the dispatch telephone text.

[0040] Preferably: In step 2.2, a standard speech signal is used as the matching degree test input data and input into the trained acoustic model, and the acoustic model outputs a standard acoustic feature sequence; the matching degree between the acoustic feature sequence and the standard acoustic feature sequence is calculated based on the distance algorithm; In step 2.3, a standard speech signal is used as the matching degree test input data and input into the trained language model, and the language model outputs a standard semantic feature sequence; the matching degree between the semantic feature sequence and the standard semantic feature sequence is calculated based on the distance algorithm.

[0041] Preferably, step 6 includes integrating and summarizing power flow, grid events, operating modes, risk warnings, equipment maintenance, accident plans, and meteorological information, allowing users to query and view various types of power information on a single interface; enabling self-customization of reports, self-generation of screens, and self-answering of telephone calls through interaction with the control knowledge base; and realizing intelligent ticket generation, automatic execution of operation tickets, and confirmation of disconnector positions based on the switching operation knowledge graph and disconnector position image recognition results.

[0042] Preferably, in S3, the target user identifier is the identifier corresponding to the user who issued the control voice, and the corresponding target user's associated information is determined through the corresponding target user identifier.

[0043] As can be seen from the above solutions, this application provides a networked command and interaction system for power distribution networks, which has the following beneficial effects:

[0044] 1. After acquiring the control voice information, the process of further identifying the target user identifier corresponding to the information is a process in which the control center automatically sorts out the relationship between the control instructions and the dispatcher, ensuring the traceability of each control instruction, avoiding the tedious process of manual verification, and improving the level of intelligence.

[0045] 2. After acquiring the control voice information, the corresponding control command to be sent is obtained based on the pre-trained control command generation model. The target user identifier corresponding to the control voice information is determined. Based on the target user identifier and the control command to be sent, the target control command is generated, which significantly improves the convenience of dispatchers issuing control commands, enhances the intelligence of human-computer interaction in the control system, reduces the consumption of human resources and time in the process of command transmission, and thus improves the response speed of scheduling tasks.

[0046] 3. The dispatcher inputs instruction information into the dispatch terminal, which then generates pre-instructions and positive instructions based on the instruction information and uploads them to the network command terminal. The plant and maintenance terminals interact with the network command terminal via wireless communication. The plant executes the pre-instructions and positive instructions issued by the network command terminal, while the maintenance terminal receives the pre-instructions and positive instructions pushed by the network command terminal. This enables the construction of a networked dispatch instruction information interaction management mode, allowing the dispatch terminal, plant terminal, and maintenance terminal to display information synchronously and improving the interaction capabilities between the terminals. Attached Figure Description

[0047] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of this application. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0048] Figure 1 This is a system block diagram of the networked command and interaction system for power distribution networks disclosed in this application;

[0049] Figure 2 This is a schematic block diagram of the scheduling terminal disclosed in this application;

[0050] Figure 3 This is a flowchart of the scheduling terminal disclosed in this application. Detailed Implementation

[0051] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0052] It should be mentioned that the pre-command generation unit of this application is the same as the pre-instruction generation unit, and the pre-command issuance unit is the same as the pre-instruction issuance unit; the regular command generation unit itself is the same as the regular instruction generation unit, and the regular command issuance unit is the same as the regular instruction issuance unit.

[0053] like Figure 1 As shown, a networked command and control system for power distribution networks includes a plant terminal, an operation and maintenance terminal, a dispatch terminal, and a network command issuing terminal communicatively connected to each of the plant terminal, operation and maintenance terminal, and dispatch terminal. The dispatch terminal receives command information input by the dispatcher and generates pre-commands and active commands based on the command information, then uploads the pre-commands and active commands to the network command issuing terminal. The network command issuing terminal receives the pre-commands and active commands generated and uploaded by the dispatch terminal, and simultaneously issues the pre-commands and active commands to both the plant terminal and the operation and maintenance terminal. The plant terminal executes the pre-commands and active commands issued by the network command issuing terminal. The dispatch terminal receives the pre-commands and active commands issued by the network command issuing terminal.

[0054] like Figure 2 , Figure 3 As shown, the dispatching terminal includes a pre-instruction processing module, a positive instruction processing module, and an instruction receipt module. The pre-instruction processing module consists of a pre-instruction generation unit and a pre-instruction issuance unit. The pre-instruction generation unit processes instruction information to generate pre-instruction information, and the pre-instruction issuance unit sends the pre-instruction information to the instruction receipt module. The positive instruction processing module consists of a positive instruction generation unit and a positive instruction issuance unit. The positive instruction generation unit processes instruction information to generate positive instruction information, and the positive instruction issuance unit sends the positive instruction information to the instruction receipt module. The instruction receipt module receives the pre-instruction information and the positive instruction information, generates corresponding pre-instructions and positive instructions based on the pre-instruction information and the positive instruction information, and uploads the pre-instructions and positive instructions to the network dispatching terminal.

[0055] It should be noted that the instruction receipt module consists of an instruction repetition unit, an instruction confirmation unit, a pre-instruction receipt execution unit, and a formal instruction receipt execution unit. The instruction repetition unit repetitively reads the pre-instruction and formal instruction information. The instruction confirmation unit confirms the pre-instruction and formal instruction information repetitively read by the instruction repetition unit, and sends the pre-instruction information to the pre-instruction receipt execution unit and the formal instruction information to the formal instruction receipt execution unit. The pre-instruction receipt execution unit generates pre-instructions for the plant equipment based on the pre-instruction information and uploads the pre-instructions to the network command transmitter. The formal instruction receipt execution unit generates formal instructions for the plant equipment based on the formal instruction information and uploads the formal instructions to the network command transmitter.

[0056] Therefore, the dispatcher inputs instruction information into the dispatch terminal, which then generates pre-instructions and positive instructions based on the instruction information and uploads them to the network command terminal. The plant and maintenance terminals interact with the network command terminal via wireless communication, with the plant executing the pre-instructions and positive instructions issued by the network command terminal and the maintenance terminal receiving the pre-instructions and positive instructions pushed by the network command terminal. This enables the construction of a networked dispatch instruction information interaction management mode, allowing the dispatch terminal, plant terminal, and maintenance terminal to display information synchronously and improving the interaction capabilities between the terminals.

[0057] It should be noted that the method for generating positive instructions and pre-instructions at the scheduling end includes the following steps:

[0058] S1. Obtain control voice information;

[0059] Specifically, the control voice information in S1 is issued by the control center staff, and in addition to the information of power system control instructions, it can also include information such as time, location and personnel related to the control instructions. Among them, power system control instructions are instructions related to the operation and maintenance of the power system.

[0060] In addition, there are multiple ways to acquire control voice information, one of which is to acquire control voice information based on a voice acquisition device. This voice acquisition device can be any voice acquisition device equipped with a microphone, such as a device connected to a voice transmission device, and includes an audio input interface, a gain amplifier, and multiple audio processing modules; furthermore, the voice acquisition device can be connected to the system via wired or wireless means.

[0061] After acquiring the control voice information, the system can either further process it or store it on a server or in the cloud for later forwarding. For example, after acquiring the control voice information issued by the dispatcher, the system can use algorithms to process the noise in the audio (i.e., noise reduction), or it can directly forward the voice information to the lower-level control center or the work site.

[0062] S2. Input the control voice information into the pre-trained control instruction generation model to obtain the control instruction to be activated corresponding to the control voice information;

[0063] Specifically, the control command generation model can include various algorithm models that can be used for semantic recognition to generate control commands to be sent based on the input control speech information. This allows the system to use the acquired control speech information as input and obtain the corresponding control commands to be sent through the trained model.

[0064] In addition, the construction of the control command generation model includes the following steps:

[0065] Step 1: Collect dispatch speech data, perform noise reduction and feature sequence extraction on the collected dispatch speech data, and use the extracted feature sequences to construct a dispatch speech corpus, which includes a general corpus, a power industry corpus, and standard speech signals for matching degree testing.

[0066] Step 2: Use the scheduling speech corpus as input data and input it into the trained acoustic model and language model; the trained acoustic model and language model jointly output the scheduling telephone text, and the trained acoustic model and language model use the scheduling speech corpus as the training set and are trained using deep learning algorithms.

[0067] Step 3: Collect regulatory command corpus and vectorize the words in the regulatory command corpus to construct a regulatory command corpus based on the obtained regulatory command word vectors;

[0068] Step 4: Using the word vectors of the control instructions as input data, input them into the domain recognition model and the intent recognition model that have been trained respectively; the domain recognition model outputs the domain matching result, and the intent recognition model outputs the intent recognition result based on the domain matching result. The domain recognition model and the intent recognition model that have been trained use the control instruction corpus as the training set and are trained using deep learning algorithms.

[0069] Step 5: Use the intent recognition result as the input data of the slot recognition model, and the slot recognition model outputs the control instruction matching result. The slot recognition model extracts semantic tags from the scheduling instructions using a rule parsing algorithm.

[0070] Step 6: Based on the matching results of the control instructions, clarify the dispatch instructions to prepare for power dispatch; determine the grid operating status by combining grid power flow calculation, short-circuit current calculation and circuit breaker operation; at the same time, execute power information query, grid automatic calculation and power automatic dispatch response operations based on the matching results of the control instructions.

[0071] Step 7: Input the dispatch call text and the matching results of the control instructions into the speech synthesis engine, convert them into dispatch call voice, and use the dispatch call voice to automatically respond to the dispatch call.

[0072] S3. Determine the target user identifier corresponding to the controlled voice information;

[0073] Specifically, the target user identifier is the identifier corresponding to the user who issued the control voice, and the relevant target user's associated information is determined through the corresponding target user identifier.

[0074] In addition, by identifying the dispatcher, we can obtain the dispatcher's number in the control center, the area they are responsible for, and the historical records of the control instructions they have issued. Furthermore, the target user identifier can also be the identifier of the staff member receiving the control voice information.

[0075] At the same time, when the target user identifier is the identifier corresponding to the user who issued the control voice, there are at least two ways to determine the target user identifier, specifically including:

[0076] 1) After analyzing and recording the control voice information, determine the control instructions in the information, and determine the corresponding target user (dispatcher) identifier through the content of the control instructions. For example, first, for different lines in each area, assign the corresponding dispatchers in the control center.

[0077] 2) Deploy a voice input device for each user (dispatcher) in the control center. Each device stores the corresponding user information and associates it with the user identifier. At the same time, the control voice information output by each device also carries the identifier associated with the corresponding device. When the system obtains the control voice information, it can determine the corresponding voice input device through the identifier carried in the information, and then determine the target user identifier associated with it through the device.

[0078] After obtaining the control voice information, the process of further determining the target user identifier corresponding to the information is the process by which the control center automatically sorts out the relationship between the control instructions and the dispatcher; and when the system cannot determine the target user identifier through the control voice information, the control instructions to be sent corresponding to the control voice information are forwarded to a specific server for storage, so that the system or staff can conduct secondary verification and confirmation.

[0079] S4. Generate target control instructions based on the target user identifier and the control instructions to be sent.

[0080] Specifically, target control instructions are those identified from the control instructions to be sent, whose source can be clearly determined, and which also need to be transmitted to lower-level control centers or on-site personnel.

[0081] It should be mentioned that step 2 includes:

[0082] Step 2.1: Use the scheduled speech corpus as input data and input it into the trained acoustic model and language model;

[0083] Step 2.2: The trained acoustic model performs noise processing and acoustic feature extraction on the scheduled speech corpus to output an acoustic feature sequence; using a standard speech signal, the matching degree of the acoustic feature sequence is obtained; specifically, the standard speech signal is used as the matching degree test input data and input into the trained acoustic model, which outputs a standard acoustic feature sequence; based on the distance algorithm, the matching degree between the acoustic feature sequence and the standard acoustic feature sequence is calculated.

[0084] Step 2.3: The trained language model performs noise processing and semantic feature extraction on the scheduled speech corpus to output a semantic feature sequence; using the standard speech signal, the matching degree of the semantic feature sequence is obtained, i.e., the language model score; specifically, the standard speech signal is used as the matching degree test input data and input into the trained language model, and the language model outputs a standard semantic feature sequence; based on the distance algorithm, the matching degree between the semantic feature sequence and the standard semantic feature sequence is calculated;

[0085] Step 2.4: The feature sequence with the highest combined score of the acoustic model and language model is used as the dispatch telephone feature sequence;

[0086] Step 2.5: Based on the power industry corpus, perform language decoding on the dispatch telephone feature sequence to obtain the dispatch telephone text.

[0087] Meanwhile, step 6 includes integrating and summarizing power flow, grid events, operating modes, risk warnings, equipment maintenance, emergency plans, and meteorological information, allowing users to query and view various types of power information on a single interface; it also enables self-customized reports, self-generated screens, and self-answering telephone calls through interaction with the control knowledge base; and based on the switching operation knowledge graph and disconnector position image recognition results, it achieves intelligent ticket generation, automatic execution of operation tickets, and confirmation of disconnector positions. The self-generated screen integrates and summarizes power flow, grid events, operating modes, risk warnings, equipment maintenance, emergency plans, and meteorological information, allowing users to view various types of information on a single interface. Step 7 includes obtaining the matching results of dispatch telephone text and control instructions; by setting synthetic speech feature parameters, using a speech synthesis engine, both the dispatch telephone text and the matching results of control instructions are converted into speech; the synthesized speech is then broadcast to the control personnel.

[0088] Therefore, in this embodiment, the pre-instruction is the instruction output after the control voice information is processed by the control instruction generation model, and the pre-instruction is expressed in language that conforms to the power system operation standards. Meanwhile, since multiple dispatchers in the dispatch center may issue control instructions at a certain time, even for the same dispatcher, a single notification to on-site workers may include multiple control instructions. Therefore, by inputting the control voice information into the model for processing, the resulting control instructions to be sent can be one or multiple.

[0089] In summary, this application provides a networked command and interaction system for distribution networks. This system, after acquiring control voice information, further determines the target user identifier corresponding to the information. This process automatically establishes the relationship between control commands and dispatchers, ensuring the traceability of each control command, avoiding the tedious manual verification process, and improving the level of intelligence. Furthermore, after acquiring control voice information, it obtains the corresponding control command to be sent based on a pre-trained control command generation model, determines the target user identifier corresponding to the control voice information, and generates the target control command based on the target user identifier and the control command to be sent. This significantly improves the convenience for dispatchers to issue control commands, enhances the intelligence of human-computer interaction in the control system, reduces the consumption of human resources and time during command transmission, and thus improves the response speed of dispatching tasks. Meanwhile, the dispatcher inputs instruction information into the dispatch terminal, which then generates pre-instructions and positive instructions based on the instruction information and uploads them to the network command terminal. The plant and maintenance terminals interact with the network command terminal via wireless communication, with the plant executing the pre-instructions and positive instructions issued by the network command terminal and the maintenance terminal receiving the pre-instructions and positive instructions pushed by the network command terminal. This enables the construction of a networked dispatch instruction information interaction management mode, allowing the dispatch terminal, plant terminal, and maintenance terminal to display information synchronously and improving the interaction capabilities between the terminals.

[0090] The terms “first,” “second,” “third,” “fourth,” etc., used in this application (if applicable) are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms “comprising” and “having,” and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, or apparatus that includes a series of steps or units is not necessarily limited to those explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, or apparatus.

[0091] It should be noted that the use of terms such as "first" and "second" in this application is for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include at least one of those features. Furthermore, the technical solutions of the various embodiments can be combined with each other, but this must be based on the ability of those skilled in the art to implement them. If the combination of technical solutions is contradictory or impossible to implement, such a combination of technical solutions should be considered non-existent and not within the scope of protection claimed in this application.

[0092] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A networked command and interaction system for power distribution networks, characterized in that: It includes a plant end, an operation and maintenance end, a dispatch end, and a network command issuing end that is communicatively connected to the plant end, the operation and maintenance end, and the dispatch end, respectively; The scheduling terminal is used to receive instruction information input by the scheduler, generate pre-instructions and positive instructions based on the instruction information, and upload the pre-instructions and positive instructions to the network command issuing terminal; The network command terminal is used to receive the pre-instructions and positive instructions generated and uploaded by the scheduling terminal, and simultaneously send the pre-instructions and positive instructions to the plant terminal and the operation and maintenance terminal. The factory terminal is used to execute the pre-instructions and positive instructions issued by the network command terminal; The operation and maintenance terminal is used to receive pre-instructions and positive instructions issued by the network command terminal; The method for generating positive and pre-instructions at the scheduling end includes the following steps: S1. Obtain control voice information; S2. Input the control voice information into the pre-trained control instruction generation model to obtain the control instruction to be activated corresponding to the control voice information; S3. Determine the target user identifier corresponding to the controlled voice information; S4. Generate the target control instruction based on the target user identifier and the control instruction to be sent; In S2, the construction of the control command generation model includes the following steps: Step 1: Collect dispatch speech data, perform noise reduction and feature sequence extraction on the collected dispatch speech data, and use the extracted feature sequences to construct a dispatch speech corpus, which includes a general corpus, a power industry corpus, and standard speech signals for matching degree testing. Step 2: Use the scheduled speech corpus as input data and input it into the trained acoustic model and language model; The dispatch telephone text is jointly output by the trained acoustic model and language model, and the trained acoustic model and language model are trained using a dispatch speech corpus as the training set and deep learning algorithms. Step 3: Collect regulatory command corpus and vectorize the words in the regulatory command corpus to construct a regulatory command corpus based on the obtained regulatory command word vectors; Step 4: Use the control instruction word vectors as input data and input them into the trained domain recognition model and intent recognition model respectively; The domain recognition model outputs the domain matching result, and the intent recognition model outputs the intent recognition result based on the domain matching result. The trained domain recognition model and intent recognition model are trained using a control command corpus as the training set and deep learning algorithms. Step 5: Use the intent recognition result as the input data of the slot recognition model, and the slot recognition model outputs the control instruction matching result. The slot recognition model extracts semantic tags from the scheduling instructions using a rule parsing algorithm. Step 6: Based on the matching results of the control instructions, clarify the dispatch instructions in preparation for carrying out power dispatch; By combining power flow calculation, short-circuit current calculation, and circuit breaker operation, the operating status of the power grid is determined; at the same time, based on the matching results of control commands, power information query, automatic power grid calculation, and automatic power dispatch response operations are performed. Step 7: Input the dispatch call text and the matching results of the control instructions into the speech synthesis engine, convert them into dispatch call voice, and use the dispatch call voice to automatically respond to the dispatch call.

2. The networked command and interaction system for power distribution networks according to claim 1, characterized in that: The scheduling terminal includes a pre-instruction processing module, a positive instruction processing module, and an instruction receipt module; The pre-instruction processing module consists of a pre-instruction generation unit and a pre-instruction issuing unit. The pre-instruction generation unit is used to process instruction information to generate pre-instruction information, and the pre-instruction issuing unit is used to send the pre-instruction information to the instruction receipt module. The positive instruction processing module consists of a positive instruction generation unit and a positive instruction issuing unit. The positive instruction generation unit is used to process instruction information to generate positive instruction information, and the positive instruction issuing unit is used to send the positive instruction information to the instruction receipt module. The instruction receipt module is used to receive pre-instruction information and positive instruction information, generate pre-instructions and positive instructions according to the pre-instruction information and positive instruction information respectively, and upload the pre-instructions and positive instructions to the network command issuing terminal.

3. The networked command and interaction system for power distribution networks according to claim 2, characterized in that: The instruction receipt module consists of an instruction repetition unit, an instruction confirmation unit, a pre-instruction receipt execution unit, and a positive instruction receipt execution unit. The instruction repeating unit is used to repeat pre-instruction information and positive instruction information; The instruction confirmation unit is used to confirm the pre-instruction information and the positive instruction information repeated by the instruction recitation unit, and to send the pre-instruction information to the pre-instruction receipt and execution unit and the positive instruction information to the positive instruction receipt and execution unit. The pre-instruction receipt and execution unit is used to generate pre-instructions for the plant equipment based on the pre-instruction information, and upload the pre-instructions to the network command issuing terminal; The positive instruction receipt and execution unit is used to generate positive instructions for the plant equipment based on the positive instruction information and upload the positive instructions to the network command issuing terminal.

4. The networked command and interaction system for power distribution networks according to claim 1, characterized in that: In S1, the control voice information is issued by the control center staff and includes information on power system control instructions, which are instructions related to the operation and maintenance of the power system.

5. A networked command and interaction system for power distribution networks according to claim 1, characterized in that, Step 2 includes: Step 2.1: Use the scheduled speech corpus as input data and input it into the trained acoustic model and language model; Step 2.2: The trained acoustic model performs noise processing and acoustic feature extraction on the scheduled speech corpus to output an acoustic feature sequence; the matching degree of the acoustic feature sequence is obtained using a standard speech signal. Step 2.3: The trained language model performs noise processing and semantic feature extraction on the scheduled speech corpus to output a semantic feature sequence; using standard speech signals, the matching degree of the semantic feature sequence is obtained, i.e., the language model score. Step 2.4: The feature sequence with the highest combined score of the acoustic model and language model is used as the dispatch telephone feature sequence; Step 2.5: Based on the power industry corpus, perform language decoding on the dispatch telephone feature sequence to obtain the dispatch telephone text.

6. A networked command and interaction system for power distribution networks according to claim 5, characterized in that: In step 2.2, a standard speech signal is used as the input data for the matching degree test and is input into the trained acoustic model. The acoustic model outputs a standard acoustic feature sequence. Based on the distance algorithm, the matching degree between the acoustic feature sequence and the standard acoustic feature sequence is calculated. In step 2.3, the standard speech signal is used as the matching degree test input data and input into the trained language model. The language model outputs the standard semantic feature sequence. Based on the distance algorithm, the matching degree between the semantic feature sequence and the standard semantic feature sequence is calculated.

7. A networked command and interaction system for power distribution networks according to claim 1, characterized in that: Step 6 includes integrating and summarizing power flow, power grid events, operating modes, risk warnings, equipment maintenance, accident plans, and meteorological information, and querying and viewing various types of power information on the same interface; It also enables self-customized reports, self-generated screens, and self-answering telephone calls through interaction with the control knowledge base; and realizes intelligent ticket generation, automatic execution of operation tickets, and confirmation of disconnector position based on the switching operation knowledge graph and disconnector position image recognition results.

8. A networked command and interaction system for power distribution networks according to claim 1, characterized in that: In S3, the target user identifier is the identifier corresponding to the user who issued the control voice, and the corresponding target user's associated information is determined through the corresponding target user identifier.

Images