Dialogue reasoning and question answering method and system for power system

Abstract

The invention discloses a dialogue reasoning and question-answering method and system for a power system. The method comprises the following steps: (1) obtaining multiple items of configuration data of a target dialogue channel; (2) obtaining each dialogue content corresponding to each previous dialogue of the target dialogue channel; (3) executing each training action on the convolutional neural network to obtain the convolutional neural network completing each training action, and outputting the convolutional neural network as a task prediction model; (4) using the task prediction model to intelligently predict a task type number corresponding to a current dialogue to be started by the target dialogue channel according to the total number of existing device types, the multiple items of configuration data of the target dialogue channel and each dialogue content corresponding to each previous dialogue of the target dialogue channel; (5) converting the task type number obtained by intelligent prediction into a task type corresponding to the current dialogue, and outputting the task type as dialogue reasoning and question and answer information; according to the invention, the reliability and controllability of the power system are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of power systems, and in particular to a dialog reasoning and question-answering method and system for power systems. Background Art

[0002] A dialogue channel can be established between every two power devices of the same or different types in the power system to perform multiple dialogues to complete multiple different tasks, that is, each dialogue performs a specific type of task. Due to the particularity of the power system environment where the power equipment is located and the information relevance and task relevance between different dialogues in time sequence, power system managers hope to study a time sequence analysis algorithm for the particularity of dialogue task data to strengthen the semantic connection and understanding between the dialogue texts to obtain the corresponding dialogue reasoning and question-and-answer information for future dialogues, for example, to obtain the task type corresponding to the future dialogue, so as to configure services and task preparation for future dialogue execution in advance, and enhance the automation and intelligence level of power system management.

[0003] Although the technical solutions involved in the prior art use artificial intelligence models to complete the intelligent analysis of power system task data, they do not involve the dialogue channel between two power equipment, let alone the dialogue reasoning and intelligent analysis of question and answer information involving future dialogues in the dialogue channel. For example, the timing analysis algorithm based on the particularity of the dialogue task data is used to obtain the task type corresponding to the future dialogue, which makes it impossible for the power system manager to configure services and prepare tasks for future dialogues in advance, and thus it is impossible to improve the automation and intelligence level of power system management. Summary of the invention

[0004] Purpose of the invention: The purpose of the present invention is to provide a dialogue reasoning and question-answering method and system for an electric power system, which can make full use of the particularity of the electric power system environment where the electric power equipment is located and the information relevance and task relevance between different dialogues in time sequence, and adopt an artificial intelligence model to intelligently analyze the dialogue reasoning and question-answering information of future dialogues in the dialogue channel between each two electric power equipment based on multiple targeted screened basic data, so as to solve the problems existing in the background technology.

[0005] Technical solution: The dialog reasoning and question-answering method for a power system described in the present invention comprises the following steps:

[0006] (1) acquiring multiple configuration data of a target dialogue channel, where the target dialogue channel is a dialogue channel between a first power device and a second power device, and the multiple configuration data of the target dialogue channel are a device type number of the first power device, a device type number of the second power device, a geographical distance between the first power device and the second power device, and a communication link type number of the target dialogue channel;

[0007] (2) obtaining the conversation contents corresponding to each past conversation of the target conversation channel, wherein each past conversation is before the current conversation to be started on the target conversation channel, and the single conversation content corresponding to each past conversation is an ASCII code data stream of a set number of digits of the conversation text corresponding to the past conversation and a task type number corresponding to the past conversation;

[0008] (3) executing each training action on the convolutional neural network to obtain a convolutional neural network that completes each training action, and outputting the convolutional neural network as a task prediction model, wherein the number of training actions of each training action is inversely correlated with the value of the set number of bits;

[0009] (4) using the task prediction model to intelligently predict the task type number corresponding to the current conversation to be initiated by the target conversation channel based on the total number of existing device types, multiple configuration data of the target conversation channel, and the conversation contents corresponding to each past conversation of the target conversation channel;

[0010] (5) Convert the task type number obtained by intelligent prediction into the task type corresponding to the current dialogue and output it as dialogue reasoning and question-answering information;

[0011] Among them, different task types correspond to different task type numbers, and the task type is one of operation mode change, transmission line power outage, transmission line power supply, relay protection input, relay protection deactivation, parameter change and equipment maintenance;

[0012] Among them, obtaining each conversation content corresponding to each past conversation of the target conversation channel, wherein each past conversation is before the current conversation that is about to be started on the target conversation channel, and the single conversation content corresponding to each past conversation is an ASCII data stream of a set number of digits of the conversation text corresponding to the past conversation and a task type number corresponding to the past conversation, including: after performing character-by-character ASCII conversion on the conversation text corresponding to the past conversation, the number of digits of the ASCII data stream obtained is greater than the set number of digits, and a part of the data stream of the set number of digits that is at the front position in the obtained ASCII data stream is used as the ASCII data stream of the set number of digits of the conversation text corresponding to the past conversation.

[0013] Further, in step (3), each training action is performed on the convolutional neural network to obtain a convolutional neural network that completes each training action, and is output as a task prediction model, wherein the number of training actions of each training action is inversely associated with the value of the set number of bits, including: in each training action performed on the convolutional neural network, the task type number corresponding to a past conversation of the target conversation channel is used as the output content of the convolutional neural network, and the total number of existing device types, multiple configuration data of the target conversation channel, and each conversation content corresponding to each past conversation of the target conversation channel before the past conversation is used as the input content of the convolutional neural network to complete this training action;

[0014] Among them, each conversation content corresponding to each past conversation of the target conversation channel is obtained, and each past conversation is before the current conversation that is about to be started on the target conversation channel. The single conversation content corresponding to each past conversation is an ASCII code data stream of a set number of bits of the conversation text corresponding to the past conversation and the task type number corresponding to the past conversation. It also includes: the number of selected past conversations is positively correlated with the geographical distance between the first power device and the second power device.

[0015] 3. A dialog reasoning and question-answering method for a power system according to claim 2, characterized in that before obtaining multiple configuration data of a target dialog channel, the target dialog channel is a dialog channel between a first power device and a second power device, and the multiple configuration data of the target dialog channel are a device type number of the first power device, a device type number of the second power device, a geographical distance between the first power device and the second power device, and a communication link type number of the target dialog channel, the method further comprises:

[0016] The current positioning information of the first power device and the current positioning information of the second power device are obtained, and the geographical distance between the first power device and the second power device is calculated based on the current positioning information of the first power device and the current positioning information of the second power device.

[0017] Furthermore, after converting the task type number obtained by intelligent prediction into the task type corresponding to the current dialogue and outputting it as dialogue reasoning and question-answering information, the method further includes:

[0018] The received dialogue reasoning and question-answer information are transmitted to a remote power system management server via a wireless communication link;

[0019] Among them, transmitting the received dialogue reasoning and question-and-answer information to a remote power system management server via a wireless communication link includes: the remote power system management server is a cloud computing service node, a big data service node or a blockchain service node.

[0020] Further, obtaining each conversation content corresponding to each past conversation of the target conversation channel, wherein each past conversation is before the current conversation to be started on the target conversation channel, and each single conversation content corresponding to the past conversation is an ASCII data stream of a set number of digits of the conversation text corresponding to the past conversation and a task type number corresponding to the past conversation, further comprising: after performing character-by-character ASCII conversion on the conversation text corresponding to the past conversation, the number of digits of the ASCII data stream obtained is less than the set number of digits, performing zero padding operation on the tail of the obtained ASCII data stream to fill the set number of digits, and obtaining the ASCII data stream of the set number of digits of the conversation text corresponding to the past conversation;

[0021] Wherein, obtaining each conversation content corresponding to each past conversation of the target conversation channel, wherein each past conversation is before the current conversation to be started on the target conversation channel, and each single conversation content corresponding to the past conversation is an ASCII data stream of a set number of digits of the conversation text corresponding to the past conversation and a task type number corresponding to the past conversation, further comprises: after performing character-by-character ASCII conversion on the conversation text corresponding to the past conversation, the number of digits of the ASCII data stream obtained is equal to the set number of digits, and directly using the obtained ASCII data stream as the ASCII data stream of the set number of digits of the conversation text corresponding to the past conversation;

[0022] Wherein, obtaining multiple configuration data of a target dialogue channel, the target dialogue channel is a dialogue channel between a first power device and a second power device, the multiple configuration data of the target dialogue channel are a device type number of the first power device, a device type number of the second power device, a geographical distance between the first power device and the second power device, and a communication link type number of the target dialogue channel, including: different device types of power devices have different corresponding device type numbers;

[0023] Among them, multiple configuration data of the target dialogue channel are obtained, and the target dialogue channel is a dialogue channel between a first power device and a second power device. The multiple configuration data of the target dialogue channel are the equipment type number of the first power device, the equipment type number of the second power device, the geographical distance between the first power device and the second power device, and the communication link type number of the target dialogue channel. The multiple configuration data also include: different communication link types of the target dialogue channel correspond to different communication link type numbers of the target dialogue channel, and the communication link type is a time division duplex communication link, a frequency division duplex communication link, or a power line communication link.

[0024] The dialog reasoning and question-answering system for a power system according to the present invention comprises:

[0025] a first capture module, configured to acquire multiple configuration data of a target dialogue channel, wherein the target dialogue channel is a dialogue channel between a first power device and a second power device, and the multiple configuration data of the target dialogue channel are a device type number of the first power device, a device type number of the second power device, a geographical distance between the first power device and the second power device, and a communication link type number of the target dialogue channel;

[0026] The second capture module is used to obtain the conversation contents corresponding to each past conversation of the target conversation channel, wherein each past conversation is before the current conversation to be started on the target conversation channel, and the single conversation content corresponding to each past conversation is an ASCII code data stream of a set number of digits of the conversation text corresponding to the past conversation and a task type number corresponding to the past conversation;

[0027] A training execution module, used for executing each training action on the convolutional neural network to obtain a convolutional neural network that completes each training action, and outputting it as a task prediction model, wherein the number of training actions of each training action is inversely correlated with the value of the set number of bits;

[0028] a task parsing module, connected to the first capture device, the second capture device and the training execution device respectively, for using the task prediction model to intelligently predict the task type number corresponding to the current conversation to be started by the target conversation channel according to the total number of existing device types, multiple configuration data of the target conversation channel and the conversation contents corresponding to each past conversation of the target conversation channel;

[0029] A data conversion module, connected to the task analysis device, for converting the task type number obtained by intelligent prediction into the task type corresponding to the current dialogue and outputting it as dialogue reasoning and question-answering information;

[0030] Among them, different task types correspond to different task type numbers, and the task type is one of operation mode change, transmission line power outage, transmission line power supply, relay protection input, relay protection deactivation, parameter change and equipment maintenance;

[0031] Among them, obtaining each conversation content corresponding to each past conversation of the target conversation channel, wherein each past conversation is before the current conversation that is about to be started on the target conversation channel, and the single conversation content corresponding to each past conversation is an ASCII data stream of a set number of digits of the conversation text corresponding to the past conversation and a task type number corresponding to the past conversation, including: after performing character-by-character ASCII conversion on the conversation text corresponding to the past conversation, the number of digits of the ASCII data stream obtained is greater than the set number of digits, and a part of the data stream of the set number of digits that is at the front position in the obtained ASCII data stream is used as the ASCII data stream of the set number of digits of the conversation text corresponding to the past conversation.

[0032] Further, in the training execution module, each training action is performed on the convolutional neural network to obtain a convolutional neural network that completes each training action, and is output as a task prediction model, and the number of training actions of each training action is inversely associated with the value of the set number of bits, including: in each training action performed on the convolutional neural network, the task type number corresponding to a past conversation of the target conversation channel is used as the output content of the convolutional neural network, and the total number of existing device types, multiple configuration data of the target conversation channel, and each conversation content corresponding to each past conversation of the target conversation channel before the past conversation is used as the input content of the convolutional neural network to complete this training action;

[0033] Among them, each conversation content corresponding to each past conversation of the target conversation channel is obtained, and each past conversation is before the current conversation that is about to be started on the target conversation channel. The single conversation content corresponding to each past conversation is an ASCII code data stream of a set number of bits of the conversation text corresponding to the past conversation and the task type number corresponding to the past conversation. It also includes: the number of selected past conversations is positively correlated with the geographical distance between the first power device and the second power device.

[0034] Further, the system also includes: a positioning acquisition module, connected to the first capture device, for acquiring current positioning information of the first power device and current positioning information of the second power device, and calculating the geographical distance between the first power device and the second power device based on the current positioning information of the first power device and the current positioning information of the second power device;

[0035] The positioning and collecting device is further used to send the geographical distance between the first power device and the second power device to the first capturing device.

[0036] Furthermore, the system also includes:

[0037] An information transmission device, connected to the data conversion device, for transmitting the received dialogue reasoning and question-answer information to a remote power system management server via a wireless communication link;

[0038] Among them, transmitting the received dialogue reasoning and question-and-answer information to a remote power system management server via a wireless communication link includes: the remote power system management server is a cloud computing service node, a big data service node or a blockchain service node.

[0039] Further, obtaining each conversation content corresponding to each past conversation of the target conversation channel, wherein each past conversation is before the current conversation to be started on the target conversation channel, and each single conversation content corresponding to the past conversation is an ASCII data stream of a set number of digits of the conversation text corresponding to the past conversation and a task type number corresponding to the past conversation, further comprising: after performing character-by-character ASCII conversion on the conversation text corresponding to the past conversation, the number of digits of the ASCII data stream obtained is less than the set number of digits, performing zero padding operation on the tail of the obtained ASCII data stream to fill the set number of digits, and obtaining the ASCII data stream of the set number of digits of the conversation text corresponding to the past conversation;

[0040] Wherein, obtaining each conversation content corresponding to each past conversation of the target conversation channel, wherein each past conversation is before the current conversation to be started on the target conversation channel, and each single conversation content corresponding to the past conversation is an ASCII data stream of a set number of digits of the conversation text corresponding to the past conversation and a task type number corresponding to the past conversation, further comprises: after performing character-by-character ASCII conversion on the conversation text corresponding to the past conversation, the number of digits of the ASCII data stream obtained is equal to the set number of digits, and directly using the obtained ASCII data stream as the ASCII data stream of the set number of digits of the conversation text corresponding to the past conversation;

[0041] Wherein, obtaining multiple configuration data of a target dialogue channel, the target dialogue channel is a dialogue channel between a first power device and a second power device, the multiple configuration data of the target dialogue channel are a device type number of the first power device, a device type number of the second power device, a geographical distance between the first power device and the second power device, and a communication link type number of the target dialogue channel, including: different device types of power devices have different corresponding device type numbers;

[0042] Among them, multiple configuration data of the target dialogue channel are obtained, and the target dialogue channel is a dialogue channel between a first power device and a second power device. The multiple configuration data of the target dialogue channel are the equipment type number of the first power device, the equipment type number of the second power device, the geographical distance between the first power device and the second power device, and the communication link type number of the target dialogue channel. The multiple configuration data also include: different communication link types of the target dialogue channel correspond to different communication link type numbers of the target dialogue channel, and the communication link type is a time division duplex communication link, a frequency division duplex communication link, or a power line communication link.

[0043] Beneficial effects: Compared with the prior art, the present invention has the following significant advantages: a dialogue reasoning and question-answering method for an electric power system, relating to the field of electric power systems, the method comprising: using a task prediction model to intelligently predict the task type number corresponding to the current dialogue that is about to be initiated in the target dialogue channel based on the dialogue contents corresponding to each past dialogue in the target dialogue channel; converting the task type number into the task type corresponding to the current dialogue and outputting it as dialogue reasoning and question-answering information; improving the reliability and controllability of the electric power system. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 It is a principle block diagram of the present invention; Figure 2 is a flow chart of the steps of the present invention; Figure 3 is a flow chart of the steps of the present invention; Figure 4 is a flow chart of the steps of the present invention; Figure 5 is a system structure diagram of the present invention; Figure 6 It is the internal structure diagram of the system of the present invention; Figure 7 This is a diagram showing the internal structure of the system information transmission device of the present invention. DETAILED DESCRIPTION

[0050] The technical solution of the present invention is further described below in conjunction with the accompanying drawings.

[0051] like Figure 1 As shown, a technical flow chart of a dialog reasoning and question-answering method and system for a power system according to the present invention is shown.

[0052] The present invention is used to fully utilize the particularity of the power system environment where the power equipment is located and the information relevance and task relevance between different conversations in time sequence, and complete the intelligent analysis of the correlation information of future conversations in the conversation channel between every two power equipment, so as to obtain the conversation reasoning and question-and-answer information of the future conversation, thereby facilitating the power system manager to perform subsequent configuration services and task management.

[0053] like Figure 1 As shown, the present invention is directed to a target dialogue channel between a first power device and a second power device, and the specific technical process of the present invention is as follows:

[0054] Technical process 1: In order to predict the task type corresponding to the future dialogue of the target dialogue channel, multiple basic data are targeted and selected. The task type corresponding to the predicted future dialogue is used as the dialogue reasoning and question-answering information of the future dialogue as the dialogue reasoning result and the backup information for subsequent questions and answers;

[0055] Specifically, the multiple basic data include multiple configuration data of the target dialogue channel and each dialogue content corresponding to each past dialogue of the target dialogue channel;

[0056] More specifically, the multiple configuration data of the target dialogue channel are the device type number of the first power device, the device type number of the second power device, the geographical distance between the first power device and the second power device, and the communication link type number of the target dialogue channel, and the single dialogue content corresponding to each past dialogue is the ASCII code data stream of the set number of digits of the dialogue text corresponding to the past dialogue and the task type number corresponding to the past dialogue;

[0057] In this way, the particularity of the power system environment where the power equipment is located and the information relevance and task relevance between different conversations in time sequence can be fully utilized, thus ensuring the stability and reliability of the intelligent prediction results;

[0058] Technical process 2: Customize a task prediction model with a corresponding structure to predict the task type corresponding to the future dialogue of the target dialogue channel;

[0059] For example, the structural customization of the task prediction model is mainly reflected in the following two aspects:

[0060] First aspect: the task prediction model is a convolutional neural network that completes each training action, and the number of training actions of each training action is inversely correlated with the value of the set digit, thereby providing more training times for shorter dialogue texts to further ensure the stability and reliability of intelligent prediction results;

[0061] Second aspect: in each training action performed on the convolutional neural network, the task type number corresponding to a past conversation of the target conversation channel is used as the output content of the convolutional neural network, and the total number of existing device types, multiple configuration data of the target conversation channel, and each conversation content corresponding to each past conversation of the target conversation channel before the past conversation is used as the input content of the convolutional neural network to complete this training action, thereby ensuring the training effect of each training action performed on the convolutional neural network;

[0062] Technical process three: A task prediction model with customized structural design is used to intelligently predict the task type number corresponding to the current dialogue that is about to be launched in the target dialogue channel based on multiple targeted basic data, and the task type number obtained by intelligent prediction is converted into the task type corresponding to the current dialogue and output as dialogue reasoning and question-answering information, thereby completing the effective prediction of the task type corresponding to the future dialogue of the target dialogue channel.

[0063] The key points of the present invention are: targeted screening of multiple basic data to make full use of the particularity of the power system environment where the power equipment is located and the information relevance and task relevance between different conversations in time sequence, customized structural design of the task prediction model, targeted design of each training action, and specific intelligent analysis of dialogue reasoning and question-and-answer information.

[0064] The following will describe in detail the dialog reasoning and question-answering method and system for a power system according to the present invention in the form of an embodiment.

[0065] Example 1

[0066] Figure 2 The figure is a flowchart of the steps of the dialog reasoning and question-answering method for the power system according to the first embodiment of the present invention.

[0067] like Figure 2 As shown, the dialog reasoning and question-answering method for the power system includes the following steps:

[0068] STEP 1: Acquire multiple configuration data of a target dialogue channel, where the target dialogue channel is a dialogue channel between a first power device and a second power device, and the multiple configuration data of the target dialogue channel are a device type number of the first power device, a device type number of the second power device, a geographical distance between the first power device and the second power device, and a communication link type number of the target dialogue channel;

[0069] For example, different types of data acquisition components may be used to respectively acquire multiple pieces of configuration data of the target dialogue channel, including the device type number of the first power device, the device type number of the second power device, the geographical distance between the first power device and the second power device, and the communication link type number of the target dialogue channel;

[0070] STEP 2: Obtain the conversation contents corresponding to each past conversation of the target conversation channel. The past conversations are before the current conversation to be started on the target conversation channel. The single conversation contents corresponding to each past conversation are the ASCII code data stream of the set number of digits of the conversation text corresponding to the past conversation and the task type number corresponding to the past conversation.

[0071] Specifically, obtaining each conversation content corresponding to each past conversation of the target conversation channel, wherein each past conversation is before the current conversation to be started by the target conversation channel, and each single conversation content corresponding to each past conversation is an ASCII code data stream of a set number of digits of the conversation text corresponding to the past conversation and a task type number corresponding to the past conversation, including: different conversations of the target conversation channel correspond to different conversation IDs;

[0072] STEP 3: Execute each training action on the convolutional neural network to obtain a convolutional neural network that completes each training action, and output it as a task prediction model, wherein the number of training actions of each training action is inversely correlated with the value of the set number of bits;

[0073] For example, each training action is performed on the convolutional neural network to obtain a convolutional neural network that completes each training action, and is output as a task prediction model, and the number of training actions of each training action is inversely associated with the value of the set number of bits, including: the value of the set number of bits is the number of bits corresponding to 1M data volume, the number of training actions of each training action is 100 times, the value of the set number of bits is the number of bits corresponding to 5M data volume, the number of training actions of each training action is 150 times, and the value of the set number of bits is the number of bits corresponding to 10M data volume, the number of training actions of each training action is 200 times, and so on;

[0074] STEP 4: Using the task prediction model, intelligently predict the task type number corresponding to the current conversation to be started by the target conversation channel according to the total number of existing device types, multiple configuration data of the target conversation channel, and the conversation contents corresponding to each past conversation of the target conversation channel;

[0075] For example, a numerical simulation mode may be selected to perform simulation and testing of a data processing process of using the task prediction model to intelligently predict a task type number corresponding to a current conversation to be started on the target conversation channel based on the total number of existing device types, multiple configuration data of the target conversation channel, and conversation contents corresponding to each past conversation on the target conversation channel;

[0076] STEP 5: Convert the task type number obtained by intelligent prediction into the task type corresponding to the current dialogue and output it as dialogue reasoning and question-answering information;

[0077] In this way, the conversation reasoning and question-answering information obtained for the current conversation is used as the type reasoning result of the current conversation that is about to be started and the reserve information for subsequent questions and answers, thereby improving the responsiveness and intelligence level of the entire conversation channel;

[0078] Among them, different task types correspond to different task type numbers, and the task type is one of operation mode change, transmission line power outage, transmission line power supply, relay protection input, relay protection deactivation, parameter change and equipment maintenance;

[0079] Wherein, obtaining each conversation content corresponding to each past conversation of the target conversation channel, wherein each past conversation is before the current conversation to be started on the target conversation channel, and each single conversation content corresponding to the past conversation is an ASCII code data stream of a set number of digits of the conversation text corresponding to the past conversation and a task type number corresponding to the past conversation, including: after executing ASCII code conversion character by character on the conversation text corresponding to the past conversation, the number of digits of the ASCII code data stream obtained is greater than the set number of digits, and a part of the data stream of the set number of digits in the front position of the obtained ASCII code data stream is used as the ASCII code data stream of the set number of digits of the conversation text corresponding to the past conversation;

[0080] Among them, executing each training action on the convolutional neural network to obtain the convolutional neural network that completes each training action, and outputting it as a task prediction model, the number of training actions of each training action is inversely associated with the value of the set number of bits, including: in each training action executed on the convolutional neural network, taking the task type number corresponding to a past conversation of the target conversation channel as the output content of the convolutional neural network, taking the total number of existing device types, multiple configuration data of the target conversation channel, and each conversation content corresponding to each past conversation of the target conversation channel before the past conversation as the input content of the convolutional neural network, to complete this training action;

[0081] And wherein, each conversation content corresponding to each past conversation of the target conversation channel is obtained, wherein each past conversation is before the current conversation that is about to be started on the target conversation channel, and the single conversation content corresponding to each past conversation is an ASCII code data stream of a set number of digits of the conversation text corresponding to the past conversation and the task type number corresponding to the past conversation, and also includes: the number of selected past conversations is positively correlated with the geographical distance between the first power device and the second power device.

[0082] Example 2

[0083] Figure 3 The figure is a flowchart of the steps of the dialog reasoning and question-answering method for the power system according to the second embodiment of the present invention.

[0084] like Figure 3 As shown, before obtaining multiple configuration data of a target dialogue channel, the target dialogue channel is a dialogue channel between a first power device and a second power device, and the multiple configuration data of the target dialogue channel are a device type number of the first power device, a device type number of the second power device, a geographical distance between the first power device and the second power device, and a communication link type number of the target dialogue channel, that is, before STEP 1, the dialogue reasoning and question-answering method for a power system further includes:

[0085] STEP 6: Obtain current positioning information of the first power device and current positioning information of the second power device, and calculate the geographical distance between the first power device and the second power device based on the current positioning information of the first power device and the current positioning information of the second power device;

[0086] Specifically, obtaining the current positioning information of the first power device and the current positioning information of the second power device, and calculating the geographical distance between the first power device and the second power device based on the current positioning information of the first power device and the current positioning information of the second power device include: the current positioning information of the first power device and the current positioning information of the second power device are both Beidou positioning data or both are Galileo positioning data.

[0087] Example 3

[0088] Figure 4 The figure is a flowchart of the steps of the dialog reasoning and question-answering method for the power system according to the third embodiment of the present invention.

[0089] like Figure 4As shown, after converting the task type number obtained by intelligent prediction into the task type corresponding to the current dialogue and outputting it as dialogue reasoning and question-answering information, that is, after STEP 5, the dialogue reasoning and question-answering method for the power system further includes:

[0090] STEP 7: Transmit the received dialogue reasoning and question-answer information to a remote power system management server via a wireless communication link;

[0091] Specifically, transmitting the received dialogue reasoning and question-answer information to a remote power system management server via a wireless communication link includes: the wireless communication link is a 5G communication link;

[0092] Among them, transmitting the received dialogue reasoning and question-and-answer information to a remote power system management server via a wireless communication link includes: the remote power system management server is a cloud computing service node, a big data service node or a blockchain service node.

[0093] Next, the dialog reasoning and question-answering method for a power system according to various embodiments of the present invention is further described.

[0094] In any of the above embodiments 1-3, optionally, in the dialog reasoning and question-answering method for a power system:

[0095] Acquiring each conversation content corresponding to each past conversation of the target conversation channel, wherein each past conversation is before the current conversation to be started by the target conversation channel, and the single conversation content corresponding to each past conversation is an ASCII data stream of a set number of bits of the conversation text corresponding to the past conversation and a task type number corresponding to the past conversation, further comprising: when the number of bits of the ASCII data stream obtained after performing character-by-character ASCII conversion on the conversation text corresponding to the past conversation is less than the set number of bits, performing a zero padding operation on the tail of the obtained ASCII data stream to pad the set number of bits, and obtaining the ASCII data stream of the set number of bits of the conversation text corresponding to the past conversation;

[0096] Wherein, obtaining each conversation content corresponding to each past conversation of the target conversation channel, wherein each past conversation is before the current conversation to be started on the target conversation channel, and each single conversation content corresponding to the past conversation is an ASCII data stream of a set number of digits of the conversation text corresponding to the past conversation and a task type number corresponding to the past conversation, further comprises: after performing character-by-character ASCII conversion on the conversation text corresponding to the past conversation, the number of digits of the ASCII data stream obtained is equal to the set number of digits, and directly using the obtained ASCII data stream as the ASCII data stream of the set number of digits of the conversation text corresponding to the past conversation;

[0097] In this way, by performing bit normalization processing on the ASCII code data stream, reliable and effective basic data in normalized processing form is provided for subsequent intelligent analysis;

[0098] Wherein, obtaining multiple configuration data of a target dialogue channel, the target dialogue channel is a dialogue channel between a first power device and a second power device, the multiple configuration data of the target dialogue channel are a device type number of the first power device, a device type number of the second power device, a geographical distance between the first power device and the second power device, and a communication link type number of the target dialogue channel, including: different device types of power devices have different corresponding device type numbers;

[0099] And wherein, multiple configuration data of the target dialogue channel are obtained, the target dialogue channel is a dialogue channel between a first power device and a second power device, the multiple configuration data of the target dialogue channel are the equipment type number of the first power device, the equipment type number of the second power device, the geographical distance between the first power device and the second power device, and the communication link type number of the target dialogue channel also include: the communication link type of the target dialogue channel is different, and the corresponding communication link type number of the target dialogue channel is different, and the communication link type is a time division duplex communication link, a frequency division duplex communication link or a power line communication link.

[0100] Example 4

[0101] Figure 5 1 is an internal structure diagram of a dialog reasoning and question-answering system for a power system according to Embodiment 4 of the present invention.

[0102] like Figure 5 As shown, the dialogue reasoning and question answering system for the power system includes the following components:

[0103] a first capture device, used to obtain multiple configuration data of a target dialogue channel, wherein the target dialogue channel is a dialogue channel between a first power device and a second power device, and the multiple configuration data of the target dialogue channel are a device type number of the first power device, a device type number of the second power device, a geographical distance between the first power device and the second power device, and a communication link type number of the target dialogue channel;

[0104] For example, different types of data acquisition components may be used to respectively acquire multiple pieces of configuration data of the target dialogue channel, including the device type number of the first power device, the device type number of the second power device, the geographical distance between the first power device and the second power device, and the communication link type number of the target dialogue channel;

[0105] The second capture device is used to obtain each conversation content corresponding to each past conversation of the target conversation channel, wherein each past conversation is before the current conversation to be started on the target conversation channel, and each single conversation content corresponding to each past conversation is an ASCII code data stream of a set number of digits of the conversation text corresponding to the past conversation and a task type number corresponding to the past conversation;

[0106] Specifically, obtaining each conversation content corresponding to each past conversation of the target conversation channel, wherein each past conversation is before the current conversation to be started by the target conversation channel, and each single conversation content corresponding to each past conversation is an ASCII code data stream of a set number of digits of the conversation text corresponding to the past conversation and a task type number corresponding to the past conversation, including: different conversations of the target conversation channel correspond to different conversation IDs;

[0107] A training execution device, used for executing each training action on the convolutional neural network to obtain a convolutional neural network that completes each training action, and outputting it as a task prediction model, wherein the number of training actions of each training action is inversely correlated with the value of the set number of bits;

[0108] For example, each training action is performed on the convolutional neural network to obtain a convolutional neural network that completes each training action, and is output as a task prediction model, and the number of training actions of each training action is inversely associated with the value of the set number of bits, including: the value of the set number of bits is the number of bits corresponding to 1M data volume, the number of training actions of each training action is 100 times, the value of the set number of bits is the number of bits corresponding to 5M data volume, the number of training actions of each training action is 150 times, and the value of the set number of bits is the number of bits corresponding to 10M data volume, the number of training actions of each training action is 200 times, and so on;

[0109] a task analysis device, connected to the first capture device, the second capture device and the training execution device respectively, for using the task prediction model to intelligently predict the task type number corresponding to the current conversation to be started by the target conversation channel according to the total number of existing device types, multiple configuration data of the target conversation channel and the conversation contents corresponding to each past conversation of the target conversation channel;

[0110] For example, a numerical simulation mode may be selected to perform simulation and testing of a data processing process of using the task prediction model to intelligently predict a task type number corresponding to a current conversation to be started on the target conversation channel based on the total number of existing device types, multiple configuration data of the target conversation channel, and conversation contents corresponding to each past conversation on the target conversation channel;

[0111] A data conversion device, connected to the task analysis device, for converting the task type number obtained by intelligent prediction into the task type corresponding to the current dialogue and outputting it as dialogue reasoning and question-answering information;

[0112] In this way, the conversation reasoning and question-answering information obtained for the current conversation is used as the type reasoning result of the current conversation that is about to be started and the reserve information for subsequent questions and answers, thereby improving the responsiveness and intelligence level of the entire conversation channel;

[0113] Among them, different task types correspond to different task type numbers, and the task type is one of operation mode change, transmission line power outage, transmission line power supply, relay protection input, relay protection deactivation, parameter change and equipment maintenance;

[0114] Wherein, obtaining each conversation content corresponding to each past conversation of the target conversation channel, wherein each past conversation is before the current conversation to be started on the target conversation channel, and each single conversation content corresponding to the past conversation is an ASCII code data stream of a set number of digits of the conversation text corresponding to the past conversation and a task type number corresponding to the past conversation, including: after executing ASCII code conversion character by character on the conversation text corresponding to the past conversation, the number of digits of the ASCII code data stream obtained is greater than the set number of digits, and a part of the data stream of the set number of digits in the front position of the obtained ASCII code data stream is used as the ASCII code data stream of the set number of digits of the conversation text corresponding to the past conversation;

[0115] Among them, executing each training action on the convolutional neural network to obtain the convolutional neural network that completes each training action, and outputting it as a task prediction model, the number of training actions of each training action is inversely associated with the value of the set number of bits, including: in each training action executed on the convolutional neural network, taking the task type number corresponding to a past conversation of the target conversation channel as the output content of the convolutional neural network, taking the total number of existing device types, multiple configuration data of the target conversation channel, and each conversation content corresponding to each past conversation of the target conversation channel before the past conversation as the input content of the convolutional neural network, to complete this training action;

[0116] And wherein, each conversation content corresponding to each past conversation of the target conversation channel is obtained, wherein each past conversation is before the current conversation that is about to be started on the target conversation channel, and the single conversation content corresponding to each past conversation is an ASCII code data stream of a set number of digits of the conversation text corresponding to the past conversation and the task type number corresponding to the past conversation, and also includes: the number of selected past conversations is positively correlated with the geographical distance between the first power device and the second power device.

[0117] Example 5

[0118] Figure 6 1 is an internal structure diagram of a dialog reasoning and question-answering system for a power system according to Embodiment 5 of the present invention.

[0119] like Figure 6 As shown, the dialogue reasoning and question-answering system for the power system also includes:

[0120] a positioning acquisition device, connected to the first capture device, and used to obtain current positioning information of the first power device and current positioning information of the second power device, and calculate the geographical distance between the first power device and the second power device based on the current positioning information of the first power device and the current positioning information of the second power device;

[0121] Specifically, obtaining the current positioning information of the first power device and the current positioning information of the second power device, and calculating the geographical distance between the first power device and the second power device based on the current positioning information of the first power device and the current positioning information of the second power device includes: the current positioning information of the first power device and the current positioning information of the second power device are both Beidou positioning data or both are Galileo positioning data;

[0122] The positioning and collecting device is further used to send the geographical distance between the first power device and the second power device to the first capturing device.

[0123] Example 6

[0124] Figure 7 1 is an internal structure diagram of a dialog reasoning and question-answering system for a power system according to Embodiment 6 of the present invention.

[0125] like Figure 7 As shown, the dialogue reasoning and question-answering system for the power system also includes:

[0126] An information transmission device, connected to the data conversion device, for transmitting the received dialogue reasoning and question-answer information to a remote power system management server via a wireless communication link;

[0127] Specifically, transmitting the received dialogue reasoning and question-answer information to a remote power system management server via a wireless communication link includes: the wireless communication link is a 5G communication link;

[0128] Among them, transmitting the received dialogue reasoning and question-and-answer information to a remote power system management server via a wireless communication link includes: the remote power system management server is a cloud computing service node, a big data service node or a blockchain service node.

[0129] Next, the dialog reasoning and question-answering system for a power system according to various embodiments of the present invention is further described.

[0130] In any of the above embodiments 4-6, optionally, in the dialog reasoning and question-answering system for a power system:

[0131] Acquiring each conversation content corresponding to each past conversation of the target conversation channel, wherein each past conversation is before the current conversation to be started by the target conversation channel, and the single conversation content corresponding to each past conversation is an ASCII data stream of a set number of bits of the conversation text corresponding to the past conversation and a task type number corresponding to the past conversation, further comprising: when the number of bits of the ASCII data stream obtained after performing character-by-character ASCII conversion on the conversation text corresponding to the past conversation is less than the set number of bits, performing a zero padding operation on the tail of the obtained ASCII data stream to pad the set number of bits, and obtaining the ASCII data stream of the set number of bits of the conversation text corresponding to the past conversation;

[0132] Wherein, obtaining each conversation content corresponding to each past conversation of the target conversation channel, wherein each past conversation is before the current conversation to be started on the target conversation channel, and each single conversation content corresponding to the past conversation is an ASCII data stream of a set number of digits of the conversation text corresponding to the past conversation and a task type number corresponding to the past conversation, further comprises: after performing character-by-character ASCII conversion on the conversation text corresponding to the past conversation, the number of digits of the ASCII data stream obtained is equal to the set number of digits, and directly using the obtained ASCII data stream as the ASCII data stream of the set number of digits of the conversation text corresponding to the past conversation;

[0133] In this way, by performing bit normalization processing on the ASCII code data stream, reliable and effective basic data in normalized processing form is provided for subsequent intelligent analysis;

[0134] Wherein, obtaining multiple configuration data of a target dialogue channel, the target dialogue channel is a dialogue channel between a first power device and a second power device, the multiple configuration data of the target dialogue channel are a device type number of the first power device, a device type number of the second power device, a geographical distance between the first power device and the second power device, and a communication link type number of the target dialogue channel, including: different device types of power devices have different corresponding device type numbers;

[0135] And wherein, multiple configuration data of the target dialogue channel are obtained, the target dialogue channel is a dialogue channel between a first power device and a second power device, the multiple configuration data of the target dialogue channel are the equipment type number of the first power device, the equipment type number of the second power device, the geographical distance between the first power device and the second power device, and the communication link type number of the target dialogue channel also include: the communication link type of the target dialogue channel is different, and the corresponding communication link type number of the target dialogue channel is different, and the communication link type is a time division duplex communication link, a frequency division duplex communication link or a power line communication link.

[0136] In addition, the prominent features and significant progress of the conversational reasoning and question-answering method and system for power systems according to the present invention can be further characterized from the following aspects:

[0137] The number of selected past conversations is positively correlated with the geographical distance between the first power device and the second power device, including: the greater the geographical distance between the first power device and the second power device, the greater the number of selected past conversations;

[0138] For example, a first conversion function may be used to represent a numerical conversion relationship of a positive correlation between the number of selected past conversations and the geographical distance between the first power device and the second power device, in which the geographical distance between the first power device and the second power device is input content, and the number of selected past conversations is output content;

[0139] The convolutional neural network is subjected to each training action to obtain a convolutional neural network that completes each training action, and the convolutional neural network is output as a task prediction model, and the number of training actions of each training action is inversely related to the value of the set number of bits, including: the larger the value of the set number of bits, the smaller the number of training actions of each training action;

[0140] For example, a second conversion function may be used to represent a numerical conversion relationship in which the number of training actions of each training action is inversely associated with the value of the set number of bits, in which the value of the set number of bits is input content, and the number of training actions of each training action is output content;

[0141] Among them, multiple configuration data of the target dialogue channel are obtained, and the target dialogue channel is a dialogue channel between a first power device and a second power device. The multiple configuration data of the target dialogue channel are the equipment type number of the first power device, the equipment type number of the second power device, the geographical distance between the first power device and the second power device, and the communication link type number of the target dialogue channel. It also includes: binary values ​​can be used to perform numerical representations of various type numbers, including but not limited to equipment type numbers and communication link type numbers.

Claims

1. A conversational reasoning and question-answering method for power systems, characterized in that: The following steps are involved: (1) acquiring multiple configuration data of a target dialogue channel, where the target dialogue channel is a dialogue channel between a first power device and a second power device, and the multiple configuration data of the target dialogue channel are a device type number of the first power device, a device type number of the second power device, a geographical distance between the first power device and the second power device, and a communication link type number of the target dialogue channel; (2) obtaining the conversation contents corresponding to each past conversation of the target conversation channel, wherein each past conversation is before the current conversation to be started on the target conversation channel, and the single conversation content corresponding to each past conversation is an ASCII code data stream of a set number of digits of the conversation text corresponding to the past conversation and a task type number corresponding to the past conversation; (3) executing each training action on the convolutional neural network to obtain a convolutional neural network that completes each training action, and outputting the convolutional neural network as a task prediction model, wherein the number of training actions of each training action is inversely correlated with the value of the set number of bits; (4) using the task prediction model to intelligently predict the task type number corresponding to the current conversation to be initiated by the target conversation channel based on the total number of existing device types, multiple configuration data of the target conversation channel, and the conversation contents corresponding to each past conversation of the target conversation channel; (5) Convert the task type number obtained by intelligent prediction into the task type corresponding to the current dialogue and output it as dialogue reasoning and question-answering information; Among them, different task types correspond to different task type numbers, and the task type is one of operation mode change, transmission line power outage, transmission line power supply, relay protection input, relay protection deactivation, parameter change and equipment maintenance; Among them, obtaining each conversation content corresponding to each past conversation of the target conversation channel, wherein each past conversation is before the current conversation that is about to be started on the target conversation channel, and the single conversation content corresponding to each past conversation is an ASCII data stream of a set number of digits of the conversation text corresponding to the past conversation and a task type number corresponding to the past conversation, including: after performing character-by-character ASCII conversion on the conversation text corresponding to the past conversation, the number of digits of the ASCII data stream obtained is greater than the set number of digits, and a part of the data stream of the set number of digits that is at the front position in the obtained ASCII data stream is used as the ASCII data stream of the set number of digits of the conversation text corresponding to the past conversation.

2. A dialog reasoning and question-answering method for a power system according to claim 1, characterized in that: In step (3), each training action is performed on the convolutional neural network to obtain a convolutional neural network that completes each training action, and outputs the convolutional neural network as a task prediction model, wherein the number of training actions of each training action is inversely associated with the value of the set number of bits, including: in each training action performed on the convolutional neural network, the task type number corresponding to a past conversation of the target conversation channel is used as the output content of the convolutional neural network, and the total number of existing device types, multiple configuration data of the target conversation channel, and each conversation content corresponding to each past conversation of the target conversation channel before the past conversation is used as the input content of the convolutional neural network to complete this training action; Among them, each conversation content corresponding to each past conversation of the target conversation channel is obtained, and each past conversation is before the current conversation that is about to be started on the target conversation channel. The single conversation content corresponding to each past conversation is an ASCII code data stream of a set number of bits of the conversation text corresponding to the past conversation and the task type number corresponding to the past conversation. It also includes: the number of selected past conversations is positively correlated with the geographical distance between the first power device and the second power device.

3. A dialog reasoning and question-answering method for a power system according to claim 2, characterized in that: Before acquiring multiple configuration data of a target dialogue channel, the target dialogue channel being a dialogue channel between a first power device and a second power device, the multiple configuration data of the target dialogue channel being a device type number of the first power device, a device type number of the second power device, a geographical distance between the first power device and the second power device, and a communication link type number of the target dialogue channel, the method further includes: The current positioning information of the first power device and the current positioning information of the second power device are obtained, and the geographical distance between the first power device and the second power device is calculated based on the current positioning information of the first power device and the current positioning information of the second power device.

4. A dialog reasoning and question-answering method for a power system according to claim 2, characterized in that: After converting the task type number obtained by intelligent prediction into the task type corresponding to the current dialogue and outputting it as dialogue reasoning and question-answering information, the method further includes: The received dialogue reasoning and question-answer information are transmitted to a remote power system management server via a wireless communication link; Among them, transmitting the received dialogue reasoning and question-and-answer information to a remote power system management server via a wireless communication link includes: the remote power system management server is a cloud computing service node, a big data service node or a blockchain service node.

5. A dialog reasoning and question-answering method for a power system according to claim 4, characterized in that: Acquiring each conversation content corresponding to each past conversation of the target conversation channel, wherein each past conversation is before the current conversation to be started by the target conversation channel, and the single conversation content corresponding to each past conversation is an ASCII data stream of a set number of bits of the conversation text corresponding to the past conversation and a task type number corresponding to the past conversation, further comprising: when the number of bits of the ASCII data stream obtained after performing character-by-character ASCII conversion on the conversation text corresponding to the past conversation is less than the set number of bits, performing a zero padding operation on the tail of the obtained ASCII data stream to pad the set number of bits, and obtaining the ASCII data stream of the set number of bits of the conversation text corresponding to the past conversation; Wherein, obtaining each conversation content corresponding to each past conversation of the target conversation channel, wherein each past conversation is before the current conversation to be started on the target conversation channel, and each single conversation content corresponding to the past conversation is an ASCII data stream of a set number of digits of the conversation text corresponding to the past conversation and a task type number corresponding to the past conversation, further comprises: after performing character-by-character ASCII conversion on the conversation text corresponding to the past conversation, the number of digits of the ASCII data stream obtained is equal to the set number of digits, and directly using the obtained ASCII data stream as the ASCII data stream of the set number of digits of the conversation text corresponding to the past conversation; Wherein, obtaining multiple configuration data of a target dialogue channel, the target dialogue channel is a dialogue channel between a first power device and a second power device, the multiple configuration data of the target dialogue channel are a device type number of the first power device, a device type number of the second power device, a geographical distance between the first power device and the second power device, and a communication link type number of the target dialogue channel, including: different device types of power devices have different corresponding device type numbers; Among them, multiple configuration data of the target dialogue channel are obtained, and the target dialogue channel is a dialogue channel between a first power device and a second power device. The multiple configuration data of the target dialogue channel are the equipment type number of the first power device, the equipment type number of the second power device, the geographical distance between the first power device and the second power device, and the communication link type number of the target dialogue channel. The multiple configuration data also include: different communication link types of the target dialogue channel correspond to different communication link type numbers of the target dialogue channel, and the communication link type is a time division duplex communication link, a frequency division duplex communication link, or a power line communication link.

6. A conversational reasoning and question-answering system for power systems, characterized in that: include: a first capture module, configured to acquire multiple configuration data of a target dialogue channel, wherein the target dialogue channel is a dialogue channel between a first power device and a second power device, and the multiple configuration data of the target dialogue channel are a device type number of the first power device, a device type number of the second power device, a geographical distance between the first power device and the second power device, and a communication link type number of the target dialogue channel; The second capture module is used to obtain the conversation contents corresponding to each past conversation of the target conversation channel, wherein each past conversation is before the current conversation to be started on the target conversation channel, and the single conversation content corresponding to each past conversation is an ASCII code data stream of a set number of digits of the conversation text corresponding to the past conversation and a task type number corresponding to the past conversation; A training execution module, used for executing each training action on the convolutional neural network to obtain a convolutional neural network that completes each training action, and outputting it as a task prediction model, wherein the number of training actions of each training action is inversely correlated with the value of the set number of bits; a task parsing module, connected to the first capture device, the second capture device and the training execution device respectively, for using the task prediction model to intelligently predict the task type number corresponding to the current conversation to be started by the target conversation channel according to the total number of existing device types, multiple configuration data of the target conversation channel and the conversation contents corresponding to each past conversation of the target conversation channel; A data conversion module, connected to the task analysis device, for converting the task type number obtained by intelligent prediction into the task type corresponding to the current dialogue and outputting it as dialogue reasoning and question-answering information; Among them, different task types correspond to different task type numbers, and the task type is one of operation mode change, transmission line power outage, transmission line power supply, relay protection input, relay protection deactivation, parameter change and equipment maintenance; Among them, obtaining each conversation content corresponding to each past conversation of the target conversation channel, wherein each past conversation is before the current conversation that is about to be started on the target conversation channel, and the single conversation content corresponding to each past conversation is an ASCII data stream of a set number of digits of the conversation text corresponding to the past conversation and a task type number corresponding to the past conversation, including: after performing character-by-character ASCII conversion on the conversation text corresponding to the past conversation, the number of digits of the ASCII data stream obtained is greater than the set number of digits, and a part of the data stream of the set number of digits that is at the front position in the obtained ASCII data stream is used as the ASCII data stream of the set number of digits of the conversation text corresponding to the past conversation.

7. A dialog reasoning and question-answering system for power system according to claim 6, characterized in that: In the training execution module, each training action is performed on the convolutional neural network to obtain a convolutional neural network that completes each training action, and is output as a task prediction model, wherein the number of training actions of each training action is inversely associated with the value of the set number of bits, including: in each training action performed on the convolutional neural network, the task type number corresponding to a past conversation of the target conversation channel is used as the output content of the convolutional neural network, and the total number of existing device types, multiple configuration data of the target conversation channel, and each conversation content corresponding to each past conversation of the target conversation channel before the past conversation is used as the input content of the convolutional neural network to complete this training action; Among them, each conversation content corresponding to each past conversation of the target conversation channel is obtained, and each past conversation is before the current conversation that is about to be started on the target conversation channel. The single conversation content corresponding to each past conversation is an ASCII code data stream of a set number of bits of the conversation text corresponding to the past conversation and the task type number corresponding to the past conversation. It also includes: the number of selected past conversations is positively correlated with the geographical distance between the first power device and the second power device.

8. The dialog reasoning and question-answering system for power system according to claim 6, characterized in that: The system further includes: a positioning acquisition module, connected to the first capture device, for acquiring current positioning information of the first power device and current positioning information of the second power device, and calculating the geographical distance between the first power device and the second power device based on the current positioning information of the first power device and the current positioning information of the second power device; The positioning and collecting device is further used to send the geographical distance between the first power device and the second power device to the first capturing device.

9. The dialog reasoning and question-answering system for power system according to claim 6, characterized in that: The system further comprises: An information transmission device, connected to the data conversion device, for transmitting the received dialogue reasoning and question-answer information to a remote power system management server via a wireless communication link; Among them, transmitting the received dialogue reasoning and question-and-answer information to a remote power system management server via a wireless communication link includes: the remote power system management server is a cloud computing service node, a big data service node or a blockchain service node.

10. The dialog reasoning and question-answering system for power system according to claim 6, characterized in that: Acquiring each conversation content corresponding to each past conversation of the target conversation channel, wherein each past conversation is before the current conversation to be started by the target conversation channel, and the single conversation content corresponding to each past conversation is an ASCII data stream of a set number of bits of the conversation text corresponding to the past conversation and a task type number corresponding to the past conversation, further comprising: when the number of bits of the ASCII data stream obtained after performing character-by-character ASCII conversion on the conversation text corresponding to the past conversation is less than the set number of bits, performing a zero padding operation on the tail of the obtained ASCII data stream to pad the set number of bits, and obtaining the ASCII data stream of the set number of bits of the conversation text corresponding to the past conversation; Wherein, obtaining each conversation content corresponding to each past conversation of the target conversation channel, wherein each past conversation is before the current conversation to be started on the target conversation channel, and each single conversation content corresponding to the past conversation is an ASCII data stream of a set number of digits of the conversation text corresponding to the past conversation and a task type number corresponding to the past conversation, further comprises: after performing character-by-character ASCII conversion on the conversation text corresponding to the past conversation, the number of digits of the ASCII data stream obtained is equal to the set number of digits, and directly using the obtained ASCII data stream as the ASCII data stream of the set number of digits of the conversation text corresponding to the past conversation; Wherein, obtaining multiple configuration data of a target dialogue channel, the target dialogue channel is a dialogue channel between a first power device and a second power device, the multiple configuration data of the target dialogue channel are a device type number of the first power device, a device type number of the second power device, a geographical distance between the first power device and the second power device, and a communication link type number of the target dialogue channel, including: different device types of power devices have different corresponding device type numbers; Among them, multiple configuration data of the target dialogue channel are obtained, and the target dialogue channel is a dialogue channel between a first power device and a second power device. The multiple configuration data of the target dialogue channel are the equipment type number of the first power device, the equipment type number of the second power device, the geographical distance between the first power device and the second power device, and the communication link type number of the target dialogue channel. The multiple configuration data also include: different communication link types of the target dialogue channel correspond to different communication link type numbers of the target dialogue channel, and the communication link type is a time division duplex communication link, a frequency division duplex communication link, or a power line communication link.

Images