A large model-based ticketing control method and device

Abstract

The application discloses a kind of based on big model's ticket generation control method and device, obtain the operation ticket demand of user input;Operation ticket format prompt and business content in operation ticket demand are extracted;The operation ticket format prompt and the business content are input into ticket generation model, based on ticket generation model output result, obtain the operation ticket corresponding to operation ticket demand;Wherein, with the historical business content marked with known operation ticket format prompt as input, with ticket generation model as output, train big model.The application analyzes the operation ticket demand input by user, based on ticket generation model trained by big model, can mine the deep meaning of operation ticket demand, so as to obtain more operation ticket demand characteristic information, and then generate ticket generation content based on multidimensional operation ticket demand characteristic information, improve ticket generation control efficiency.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligence, and in particular to a large model-based ticket generation control method and device. BACKGROUND

[0002] In current power grid dispatching services, rule-based ticket generation technology is widely used in dispatching systems. Ticket generation technology is mainly used to generate operation tickets, which contain specific operation steps and instructions to ensure that dispatchers can perform tasks according to the predetermined process, thereby improving the orderliness and safety of device operation in dispatching services.

[0003] Rule-based ticket generation technology usually relies on pre-defined rules and logic, and the automatic execution of these rules is implemented through code writing. Specifically, these rules and logic are embedded in the system, and when an operation ticket needs to be generated, the system will automatically generate the corresponding operation ticket according to the pre-set rules. Rule-based ticket generation technology requires a lot of code writing and maintenance work. Whenever the operation rules change or new rules are added, the system needs to be modified and updated accordingly, and the ticket generation efficiency is low.

[0004] Therefore, there is an urgent need for a ticket generation control strategy to solve the problem of low control efficiency of the preset rule-based ticket generation method. SUMMARY

[0005] The embodiments of the present application provide a large model-based ticket generation control method and device to solve the problem of low control efficiency of the preset rule-based ticket generation method.

[0006] To solve the above problems, an embodiment of the present application provides a large model-based ticket generation control method and device method, comprising:

[0007] Obtain the operation ticket demand input by the user;

[0008] Extract the operation ticket format prompt and business content in the operation ticket demand;

[0009] Input the operation ticket format prompt and the business content into the ticket generation model, and obtain the operation ticket corresponding to the operation ticket demand based on the output result of the ticket generation model; wherein the historical business content marked with the known operation ticket format prompt is input, the ticket generation model is output, and the large model is trained.

[0010] As an improvement of the above scheme, the training of the ticket generation model comprises:

[0011] Obtain a plurality of historical ticket generation data;

[0012] The historical ticket data of the same ticket logic is classified based on a preset ticket logic, and the historical business content marked with an operation ticket format prompt is obtained by marking the historical business content of the same category with the operation ticket format prompt, wherein one ticket logic corresponds to one operation ticket format prompt;

[0013] The historical business content marked with the operation ticket format prompt is input into the large model, and the output content of the large model is evaluated for effectiveness. After the effectiveness evaluation is passed, the ticketing model is obtained.

[0014] As an improvement of the above scheme, the effectiveness evaluation of the output content of the large model comprises:

[0015] After inputting all the historical business content marked with the operation ticket format prompt into the large model, a ticketing content sample corresponding to each historical business content marked with the operation ticket format prompt is obtained;

[0016] Based on the ticketing content sample, the number of successful ticketing, the average ticketing time and the average number of interactions for ticketing are extracted;

[0017] The number of successful ticketing, the average ticketing time and the average number of interactions for ticketing are judged;

[0018] If the number of successful ticketing is greater than or equal to the success threshold, the average ticketing time is less than or equal to the average time threshold, and the average number of interactions for ticketing is less than or equal to the interaction threshold, the effectiveness evaluation is passed;

[0019] Otherwise, the effectiveness evaluation is not passed.

[0020] As an improvement of the above scheme, the operation ticket format prompt comprises: a mode adjustment content extraction prompt, an operation type identification prompt, a plant station line name extraction prompt, an operation ticket order generation prompt and a ticket order sorting prompt.

[0021] As an improvement of the above scheme, the operation ticket format prompt comprises: a mode adjustment content extraction prompt, an operation type identification prompt, a plant station line name extraction prompt, an operation ticket order generation prompt and a ticket order sorting prompt.

[0022] The mode adjustment content extraction prompt and the business content are input into the ticketing model to obtain mode adjustment content.

[0023] The operation type identification prompt and the business content are input into the ticketing model to obtain the operation type.

[0024] The plant station line name extraction prompt and the business content are input into the ticketing model to obtain the plant station line name.

[0025] Based on the preset retrieval tool, the operation type and the content are retrieved, and an operation ticket generation case template is obtained; and based on the preset retrieval tool, the station line name is retrieved, and power grid topology information is obtained;

[0026] The operation ticket generation prompt, the operation ticket generation case template and the power grid topology information are input into a ticket generation model, and an initial operation ticket is obtained.

[0027] The ticket order prompt and the initial operation ticket are input into the ticket generation model, and a final operation ticket is obtained, and the final operation ticket is taken as an operation ticket corresponding to the operation ticket demand.

[0028] Correspondingly, an embodiment of the present application also provides a ticket generation control device based on a large model, comprising a data acquisition module, a data extraction module and a result generation module.

[0029] The data acquisition module is used for acquiring an operation ticket demand input by a user.

[0030] The data extraction module is used for extracting an operation ticket format prompt and business content in the operation ticket demand.

[0031] The result generation module is used for inputting the operation ticket format prompt and the business content into a ticket generation model, and outputting a result based on the ticket generation model, so as to obtain an operation ticket corresponding to the operation ticket demand; wherein historical business content marked with a known operation ticket format prompt is taken as input, the ticket generation model is taken as output, and a large model is trained.

[0032] As an improvement of the above-mentioned scheme, the training of the ticket generation model comprises:

[0033] A plurality of historical ticket generation data are acquired.

[0034] Based on a preset ticket generation logic, historical ticket generation data of the same ticket generation logic are classified, and historical business content marked with an operation ticket format prompt is obtained by marking the operation ticket format prompt of the same type of historical ticket generation data; wherein one ticket generation logic corresponds to one operation ticket format prompt.

[0035] The historical business content marked with the operation ticket format prompt is input into the large model, the output content of the large model is evaluated, and the ticket generation model is obtained after the effect evaluation is passed.

[0036] As an improvement of the above-mentioned scheme, the effect evaluation of the output content of the large model comprises:

[0037] After inputting all historical business content marked with operation ticket format prompts into the large model, a sample of the completed ticket content corresponding to each historical business content marked with operation ticket format prompts is obtained.

[0038] Based on the sample of completed ticket content, the number of successful ticket sales, the average time for ticket sales, and the average number of interactions for ticket sales were extracted.

[0039] The system assesses the number of successful ticket transactions, the average time for ticket transactions, and the average number of interactions for ticket transactions.

[0040] If the number of successful ticket sales is greater than or equal to the success number threshold, the average ticket sales time is less than or equal to the average time threshold, and the average number of interactions for ticket sales is less than or equal to the number of interactions threshold, then the performance evaluation is passed.

[0041] Otherwise, the effectiveness evaluation will fail.

[0042] As an improvement to the above scheme, the ticket format prompt includes: a prompt for extracting the content of the method adjustment, a prompt for identifying the operation type, a prompt for extracting the station and line names, a prompt for generating operation tickets, and a prompt for sorting tickets.

[0043] As an improvement to the above solution, the step of inputting the operation ticket format and the business content into the ticket generation model to obtain the operation ticket corresponding to the operation ticket requirements includes:

[0044] The aforementioned method adjustment content extraction prompt and business content are input into the ticketing model to obtain the method adjustment content;

[0045] The operation type identification prompt and business content are input into the ticketing model to obtain the operation type;

[0046] The station and line name extraction prompt and business content are input into the ticketing model to obtain the station and line name;

[0047] Based on the preset search tools, the content of the method adjustment and the operation type are searched to obtain the operation ticket generation case template; and based on the preset search tools, the names of the substations and lines are searched to obtain the power grid topology information.

[0048] The operation ticket generation prompt, the operation ticket generation case template, and the power grid topology information are input into the ticket generation model to obtain the initial operation ticket;

[0049] The ticket order sorting prompt and the initial operation ticket are input into the ticket generation model to obtain the final operation ticket, and the final operation ticket is used as the operation ticket corresponding to the operation ticket requirement.

[0050] As can be seen from the above, the present invention has the following beneficial effects:

[0051] This invention provides a ticketing control method based on a large model, which involves acquiring user-inputted operation ticket requirements; extracting the operation ticket format prompt and business content from the operation ticket requirements; inputting the operation ticket format prompt and the business content into a ticketing model; and obtaining the operation ticket corresponding to the operation ticket requirements based on the output of the ticketing model. Specifically, historical business content marked with known operation ticket format prompts is used as input, and the ticketing model is used as output to train the large model. This invention analyzes user-inputted operation ticket requirements, and based on the ticketing model trained by the large model, outputs operation tickets based on the analyzed operation ticket format prompt and business content, thus achieving automated ticketing. Compared to extracting content based on preset rules, this application can mine the deeper meaning of operation ticket requirements through the large model, thereby obtaining more characteristic information of operation ticket requirements. Furthermore, based on this multi-dimensional characteristic information of operation ticket requirements, the generated ticket content is improved, thus enhancing the efficiency of ticketing control. Attached Figure Description

[0052] Figure 1 This is a flowchart illustrating a ticketing control method based on a large model provided in an embodiment of the present invention;

[0053] Figure 2 This is a schematic diagram of the structure of a ticket-generating control device based on a large model provided in an embodiment of the present invention;

[0054] Figure 3 This is a schematic diagram of a terminal device structure provided in an embodiment of the present invention;

[0055] Figure 4 This is a schematic diagram of a directed acyclic graph provided in an embodiment of the present invention;

[0056] Figure 5 This is a flowchart illustrating a ticketing control method based on a large model, provided in another embodiment of the present invention.

[0057] Figure 6 This is a schematic diagram of the result of generating an operation ticket according to an embodiment of the present invention. Detailed Implementation

[0058] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0059] Example 1

[0060] See Figure 1 , Figure 1 This is a flowchart illustrating a ticketing control method based on a large model, as provided in an embodiment of the present invention. Figure 1 As shown, this embodiment includes steps 101 to 103, and the specific steps are as follows:

[0061] Step 101: Obtain the user's input operation ticket requirements.

[0062] Step 102: Extract the operation ticket format prompt and business content from the operation ticket requirements.

[0063] In this embodiment, the ticket format prompt includes: a prompt for extracting the method adjustment content, a prompt for identifying the operation type, a prompt for extracting the substation and line names, a prompt for generating operation tickets, and a prompt for sorting tickets.

[0064] In a specific embodiment, the following example is provided for illustration: Maintenance order prompt template:

[0065] Given the following maintenance order information: "The 220kV xxx line at xx station has been switched from operation to maintenance," please write the corresponding power outage operation ticket. The content within the quotation marks is dynamically filled.

[0066] Step 103: Input the operation ticket format prompt and the business content into the ticket generation model, and obtain the operation ticket corresponding to the operation ticket requirement based on the output result of the ticket generation model; wherein, the large model is trained by taking the historical business content marked with the known operation ticket format prompt as input and the ticket generation model as output.

[0067] In this embodiment, the training of the ticket-generating model includes:

[0068] Obtain some historical ticket data;

[0069] Based on the preset ticketing logic, historical ticketing data with the same ticketing logic are classified, and the historical ticketing data of the same category are marked with operation ticket format prompts to obtain several historical business contents marked with operation ticket format prompts; wherein, one ticketing logic corresponds to one operation ticket format prompt.

[0070] The historical business content marked with the operation ticket format prompt is input into the large model, and the output content of the large model is evaluated. After the evaluation is passed, the ticket model is obtained.

[0071] In this embodiment, the training of the ticket-generating model includes:

[0072] Obtain some historical ticket data;

[0073] Based on the preset ticketing logic, historical ticketing data with the same ticketing logic are classified, and the historical ticketing data of the same category are marked with operation ticket format prompts to obtain several historical business contents marked with operation ticket format prompts; wherein, one ticketing logic corresponds to one operation ticket format prompt.

[0074] The historical business content marked with the operation ticket format prompt is input into the large model, and the output content of the large model is evaluated. After the evaluation is passed, the ticket model is obtained.

[0075] In a specific embodiment, a multi-round interaction mechanism is used to adjust and correct the generated operation ticket, enabling it to flexibly support the needs of real-world scenarios. The specific steps are as follows:

[0076] (1) Preliminary generation: The model generates preliminary operation tickets based on the input prompt;

[0077] (2) User feedback: Users review the initially generated operation tickets and provide feedback through simple instructions (such as modification, supplementation, deletion, etc.);

[0078] (3) Model adjustment: The model is adjusted based on user feedback to generate a new version of the operation ticket;

[0079] (4) Multi-round interaction: Repeat the above process until the generated operation ticket meets the user's needs.

[0080] It should be noted that the ability to adjust ticket orders via simple commands is provided, improving the efficiency of generating accurate ticket orders. The specific steps are as follows:

[0081] (1) Interactive interface design: Design a user-friendly interactive interface so that users can easily input commands and view the generated results;

[0082] (2) Instruction parsing: The model can parse simple instructions input by the user and adjust the generated operation ticket according to the instructions;

[0083] (3) Feedback mechanism: The user's historical sessions can be remembered by the model and used in subsequent generation processes to improve generation efficiency and accuracy.

[0084] In this embodiment, the evaluation of the output of the large model includes:

[0085] After inputting all historical business content marked with operation ticket format prompts into the large model, a sample of the completed ticket content corresponding to each historical business content marked with operation ticket format prompts is obtained.

[0086] Based on the sample of completed ticket content, the number of successful ticket sales, the average time for ticket sales, and the average number of interactions for ticket sales were extracted.

[0087] The system assesses the number of successful ticket transactions, the average time for ticket transactions, and the average number of interactions for ticket transactions.

[0088] If the number of successful ticket sales is greater than or equal to the success number threshold, the average ticket sales time is less than or equal to the average time threshold, and the average number of interactions for ticket sales is less than or equal to the number of interactions threshold, then the performance evaluation is passed.

[0089] Otherwise, the effectiveness evaluation will fail.

[0090] In one specific embodiment, the effect evaluation includes three metrics:

[0091] 1. Accuracy Metric: Precision is used as the metric. Assuming the number of samples where a ticket perfectly meets the requirements in a single transaction is N_correct, and the total number of samples is N, then the precision P can be expressed as:

[0092] P = N_correct / N * 100%

[0093] It should be noted that the higher the P-value, the higher the accuracy of ticket issuance, indicating a better model performance.

[0094] 2. Efficiency Indicators: The efficiency of the ticketing system will be measured using the following two methods:

[0095] (1) Average Generation Time: Average generation time measures the average time required for the system to generate one operation ticket. Assuming the total time for the system to generate N operation tickets is T, the average generation time T_avg can be expressed as:

[0096] T_avg = T / N

[0097] It should be noted that the smaller the T_avg value, the shorter the average generation time and the higher the generation efficiency.

[0098] (2) Average Interaction Rounds: Average interaction rounds measure the average number of interactions required for the system to generate a satisfactory operation ticket. Assuming the total number of interactions required to generate N operation tickets is I, then the average number of interactions I_avg can be expressed as:

[0099] I_avg = I / N

[0100] It should be noted that the smaller I_avg is, the fewer interactions are required to generate an accurate operation ticket, and the higher the generation efficiency. For example, if three rounds of dialogue are required to generate an accurate operation ticket, then the number of interactions for ticket generation is 3. The fewer the number of interactions, the higher the efficiency.

[0101] In one specific embodiment, the reuse of factual and contextual information is enhanced, making complex tasks decomposable, while the reliability of the generated content is improved by leveraging factual information data and logical calculation tools. The specific steps are as follows:

[0102] (1) Memory mechanism design: Design a memory mechanism so that the model can remember contextual information and factual data.

[0103] (2) Information reuse: When generating operation tickets, the model can reuse previous context information and factual data to improve the consistency and accuracy of the generated content.

[0104] In this embodiment, inputting the operation ticket format and the business content into the ticket generation model to obtain the operation ticket corresponding to the operation ticket requirements includes:

[0105] The aforementioned method adjustment content extraction prompt and business content are input into the ticketing model to obtain the method adjustment content;

[0106] The operation type identification prompt and business content are input into the ticketing model to obtain the operation type;

[0107] The station and line name extraction prompt and business content are input into the ticketing model to obtain the station and line name;

[0108] Based on the preset search tools, the content of the method adjustment and the operation type are searched to obtain the operation ticket generation case template; and based on the preset search tools, the names of the substations and lines are searched to obtain the power grid topology information.

[0109] The operation ticket generation prompt, the operation ticket generation case template, and the power grid topology information are input into the ticket generation model to obtain the initial operation ticket;

[0110] The ticket order sorting prompt and the initial operation ticket are input into the ticket generation model to obtain the final operation ticket, and the final operation ticket is used as the operation ticket corresponding to the operation ticket requirement.

[0111] In one specific embodiment, the business flow is transformed into a call flow for LLM (Large Language Model), RAG (Retrieval Augmentation Generation), tools, etc. The specific steps are as follows:

[0112] (1) Business process analysis: Analyze the ticketing business process and determine the input and output relationships of each link.

[0113] (2) Agent construction: Based on the business process, construct an intelligent agent (i.e. the subject that executes this invention) so that it can call LLM, RAG and other tools to complete the ticketing task.

[0114] (3) Call Flow Design: The call flow is designed to enable efficient collaboration among various stages, ensuring a smooth and accurate ticketing process. When designing the call flow, we use a directed graph from graph theory to represent the business process. Assuming the business process can be represented as a directed acyclic graph (G = (V, E)), where (V) is the set of nodes and (E) is the set of edges, then the path from node (i) to node (j) can be represented as:

[0115] P_{ij}={v_1,v_2,...,v_n}

[0116] Where (v_1=i), (v_n=j), and for any (k\in{1,2,...,n-1}), ((v_k,v_{k+1})\in E).

[0117] For a better explanation, see [link to relevant documentation]. Figure 4 The following examples illustrate directed acyclic graphs:

[0118] Figure 4 The content within the rectangle represents the nodes (V) of the directed acyclic graph. The nodes are connected by directed edges. The calling process is as follows:

[0119] 1. Based on the user input query (i.e., the operation ticket requirement described in this invention), the content extraction prompt is adjusted according to the construction method:

[0120]

[0121] It should be noted that the content extraction prompt is entered into LLM with the method adjusted, and the content extraction method is adjusted.

[0122] 2. Construct an operation type recognition prompt:

[0123]

[0124] It should be noted that the operation type identification prompt is input into LLM to identify the operation type.

[0125] 3. Based on the previously obtained information such as the method of adjusting content and operation type, call the case template retrieval tool to obtain the corresponding power grid topology information and the case template for generating operation tickets.

[0126] 4. Construct a prompt for extracting plant and substation line names:

[0127]

[0128] It should be noted that the prompt input for extracting plant and line names is LLM, which extracts the plant and line name information. Based on the plant and line names, the corresponding power grid topology information is obtained by calling the interface.

[0129] 5. Based on the obtained power grid topology information and operation ticket generation case template, construct an operation ticket generation prompt:

[0130]

[0131] It should be noted that the operation ticket prompt is input into the LLM to generate the operation ticket command.

[0132] 6. Based on the generated operation tickets, construct a ticket ordering prompt:

[0133]

[0134] It should be noted that the sorted ticket prompt is input into LLM to generate the sorted ticket prompt.

[0135] In one specific embodiment, see Figure 5 ,like Figure 5 As shown, the ticketing process based on the large model agent is as follows:

[0136] Step 1: Based on the input query, identify the intent. If it is a request to generate an operation ticket, extract the parameters; otherwise, call the LLM output.

[0137] Step 2: Use the tool to retrieve OCS data, and simultaneously search for cases and rules to obtain the operation ticket template. If the tool fails to retrieve the required data, it will return a prompt message; otherwise, proceed to the next step.

[0138] Step 3: Based on the data obtained in the previous steps, construct an operation ticket, generate a prompt, and input it into the LLM to obtain the operation ticket command;

[0139] Step 4: Users provide modification suggestions for the generated ticket order based on their actual needs, and the large model modifies the ticket order according to the suggestions.

[0140] In one specific embodiment, a method is provided Figure 6 The generated operation ticket is illustrated with an example.

[0141] See Figure 2 , Figure 2 This is a schematic diagram of a ticket-generating control device based on a large model according to an embodiment of the present invention, including: a data acquisition module 201, a data extraction module 202, and a result generation module 203;

[0142] The data acquisition module is used to acquire the operation ticket requirements input by the user;

[0143] The data extraction module is used to extract the operation ticket format prompt and business content from the operation ticket requirements;

[0144] The result generation module is used to input the operation ticket format prompt and the business content into the ticket generation model, and output the result based on the ticket generation model to obtain the operation ticket corresponding to the operation ticket requirement; wherein, the large model is trained by taking historical business content marked with a known operation ticket format prompt as input and the ticket generation model as output.

[0145] As an improvement to the above scheme, the training of the ticket-generating model includes:

[0146] Obtain some historical ticket data;

[0147] Based on the preset ticketing logic, historical ticketing data with the same ticketing logic are classified, and the historical ticketing data of the same category are marked with operation ticket format prompts to obtain several historical business contents marked with operation ticket format prompts; wherein, one ticketing logic corresponds to one operation ticket format prompt.

[0148] The historical business content marked with the operation ticket format prompt is input into the large model, and the output content of the large model is evaluated. After the evaluation is passed, the ticket model is obtained.

[0149] As an improvement to the above solution, the evaluation of the output of the large model includes:

[0150] After inputting all historical business content marked with operation ticket format prompts into the large model, a sample of the completed ticket content corresponding to each historical business content marked with operation ticket format prompts is obtained.

[0151] Based on the sample of completed ticket content, the number of successful ticket sales, the average time for ticket sales, and the average number of interactions for ticket sales were extracted.

[0152] The system assesses the number of successful ticket transactions, the average time for ticket transactions, and the average number of interactions for ticket transactions.

[0153] If the number of successful ticket sales is greater than or equal to the success number threshold, the average ticket sales time is less than or equal to the average time threshold, and the average number of interactions for ticket sales is less than or equal to the number of interactions threshold, then the performance evaluation is passed.

[0154] Otherwise, the effectiveness evaluation will fail.

[0155] As an improvement to the above scheme, the ticket format prompt includes: a prompt for extracting the content of the method adjustment, a prompt for identifying the operation type, a prompt for extracting the station and line names, a prompt for generating operation tickets, and a prompt for sorting tickets.

[0156] As an improvement to the above solution, the step of inputting the operation ticket format and the business content into the ticket generation model to obtain the operation ticket corresponding to the operation ticket requirements includes:

[0157] The aforementioned method adjustment content extraction prompt and business content are input into the ticketing model to obtain the method adjustment content;

[0158] The operation type identification prompt and business content are input into the ticketing model to obtain the operation type;

[0159] The station and line name extraction prompt and business content are input into the ticketing model to obtain the station and line name;

[0160] Based on the preset search tools, the content of the method adjustment and the operation type are searched to obtain the operation ticket generation case template; and based on the preset search tools, the names of the substations and lines are searched to obtain the power grid topology information.

[0161] The operation ticket generation prompt, the operation ticket generation case template, and the power grid topology information are input into the ticket generation model to obtain the initial operation ticket;

[0162] The ticket order sorting prompt and the initial operation ticket are input into the ticket generation model to obtain the final operation ticket, and the final operation ticket is used as the operation ticket corresponding to the operation ticket requirement.

[0163] This embodiment obtains user-inputted operation ticket requirements; extracts the operation ticket format prompt and business content from the operation ticket requirements; inputs the operation ticket format prompt and business content into a ticket generation model; and obtains the operation ticket corresponding to the operation ticket requirements based on the output results of the ticket generation model. Specifically, historical business content marked with known operation ticket format prompts is used as input, and the ticket generation model is used as output to train a large-scale model. This invention analyzes user-inputted operation ticket requirements, and based on a ticket generation model trained on a large-scale model, outputs operation tickets from the analyzed operation ticket format prompt and business content, thus achieving automated ticket generation. Compared to extracting content based on preset rules, this application, through a large-scale model, can mine the deeper meaning of operation ticket requirements, thereby obtaining more characteristic information of operation ticket requirements. Furthermore, based on this multi-dimensional characteristic information of operation ticket requirements, the generated ticket content is improved, enhancing the efficiency of ticket generation control.

[0164] Example 2

[0165] See Figure 3 , Figure 3 This is a schematic diagram of the terminal device structure provided in an embodiment of the present invention.

[0166] One terminal device in this embodiment includes: a processor 301, a memory 302, and a computer program stored in the memory 302 and executable on the processor 301. When the processor 301 executes the computer program, it implements the steps of the various large-model-based ticketing control methods described above in this embodiment, for example... Figure 1 The illustrated method for ticket issuance control based on a large model includes all steps. Alternatively, when the processor executes the computer program, it implements the functions of each module in the above-described device embodiments, for example: Figure 2 The diagram shows all modules of the ticket control device based on a large model.

[0167] In addition, embodiments of the present invention also provide a computer-readable storage medium, the computer-readable storage medium including a stored computer program, wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to perform the ticketing control method based on a large model as described in any of the above embodiments.

[0168] Those skilled in the art will understand that the schematic diagram is merely an example of a terminal device and does not constitute a limitation on the terminal device. It may include more or fewer components than shown in the diagram, or combine certain components, or different components. For example, the terminal device may also include input / output devices, network access devices, buses, etc.

[0169] The processor 301 may be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor. The processor 301 is the control center of the terminal device, connecting various parts of the terminal device through various interfaces and lines.

[0170] The memory 302 can be used to store the computer programs and / or modules. The processor 301 implements various functions of the terminal device by running or executing the computer programs and / or modules stored in the memory and calling the data stored in the memory 302. The memory 302 may mainly include a program storage area and a data storage area. The program storage area may store the operating system, at least one application program required for a function (such as sound playback function, image playback function, etc.), etc.; the data storage area may store data created according to the use of the mobile phone (such as audio data, phonebook, etc.). In addition, the memory may include high-speed random access memory, and may also include non-volatile memory, such as hard disk, memory, plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, at least one disk storage device, flash memory device, or other volatile solid-state storage device.

[0171] Wherein, if the modules / units integrated in the terminal device are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments of the present invention can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when the computer program is executed by a processor, it can implement the steps of the various method embodiments described above. Wherein, the computer program includes computer program code, which can be in the form of source code, object code, executable file, or some intermediate form, etc. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, portable hard drive, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal, and software distribution medium, etc.

[0172] It should be noted that the device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Furthermore, in the accompanying drawings of the device embodiments provided by this invention, the connection relationships between modules indicate that they have communication connections, which can be specifically implemented as one or more communication buses or signal lines. Those skilled in the art can understand and implement this without any creative effort.

[0173] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A ticket issuance control method based on a large model, characterized in that, include: The requirement is to obtain the user's input operation ticket. Extract the operation ticket format prompt and business content from the operation ticket requirements; wherein, the operation ticket format prompt includes: mode adjustment content extraction prompt, operation type identification prompt, substation line name extraction prompt, operation ticket order generation prompt, and ticket order sorting prompt; The aforementioned method adjustment content extraction prompt and business content are input into the ticketing model to obtain the method adjustment content; The operation type identification prompt and business content are input into the ticketing model to obtain the operation type; The station and line name extraction prompt and business content are input into the ticketing model to obtain the station and line name; Based on the preset search tools, the content of the method adjustment and the operation type are searched to obtain the operation ticket generation case template; and based on the preset search tools, the names of the substations and lines are searched to obtain the power grid topology information. The operation ticket generation prompt, the operation ticket generation case template, and the power grid topology information are input into the ticket generation model to obtain the initial operation ticket; The ticket order sorting prompt and the initial operation ticket are input into the ticket generation model to obtain the final operation ticket, and the final operation ticket is used as the operation ticket corresponding to the operation ticket requirement; wherein, historical business content marked with a known operation ticket format prompt is used as input, and the ticket generation model is used as output to train the large model.

2. The ticket issuance control method based on a large model according to claim 1, characterized in that, The training of the ticket-generating model includes: Obtain some historical ticket data; Based on the preset ticketing logic, historical ticketing data with the same ticketing logic are classified, and the historical ticketing data of the same category are marked with operation ticket format prompts to obtain several historical business contents marked with operation ticket format prompts; wherein, one ticketing logic corresponds to one operation ticket format prompt. The historical business content marked with the operation ticket format prompt is input into the large model, and the output content of the large model is evaluated. After the evaluation is passed, the ticket model is obtained.

3. The ticket issuance control method based on a large model according to claim 2, characterized in that, The evaluation of the output of the large model includes: After inputting all historical business content marked with operation ticket format prompts into the large model, a sample of the completed ticket content corresponding to each historical business content marked with operation ticket format prompts is obtained. Based on the sample of completed ticket content, the number of successful ticket sales, the average time for ticket sales, and the average number of interactions for ticket sales were extracted. The system assesses the number of successful ticket transactions, the average time for ticket transactions, and the average number of interactions for ticket transactions. If the number of successful ticket sales is greater than or equal to the success number threshold, the average ticket sales time is less than or equal to the average time threshold, and the average number of interactions for ticket sales is less than or equal to the number of interactions threshold, then the performance evaluation is passed. Otherwise, the effectiveness evaluation will fail.

4. A ticket-generating control device based on a large model, characterized in that, include: Data acquisition module, data extraction module, and result generation module; The data acquisition module is used to acquire the operation ticket requirements input by the user; The data extraction module is used to extract the operation ticket format prompt and business content from the operation ticket requirements; wherein, the operation ticket format prompt includes: mode adjustment content extraction prompt, operation type identification prompt, substation line name extraction prompt, operation ticket order generation prompt, and ticket order sorting prompt; The result generation module is used to input the extracted prompt of the method adjustment content and the business content into the ticketing model to obtain the method adjustment content; The operation type identification prompt and business content are input into the ticketing model to obtain the operation type; The station and line name extraction prompt and business content are input into the ticketing model to obtain the station and line name; Based on the preset search tools, the content of the method adjustment and the operation type are searched to obtain the operation ticket generation case template; and based on the preset search tools, the names of the substations and lines are searched to obtain the power grid topology information. The operation ticket generation prompt, the operation ticket generation case template, and the power grid topology information are input into the ticket generation model to obtain the initial operation ticket; The ticket order sorting prompt and the initial operation ticket are input into the ticket generation model to obtain the final operation ticket, and the final operation ticket is used as the operation ticket corresponding to the operation ticket requirement; wherein, historical business content marked with a known operation ticket format prompt is used as input, and the ticket generation model is used as output to train the large model.

5. The ticket-generating control device based on a large model according to claim 4, characterized in that, The training of the ticket-generating model includes: Obtain some historical ticket data; Based on the preset ticketing logic, historical ticketing data with the same ticketing logic are classified, and the historical ticketing data of the same category are marked with operation ticket format prompts to obtain several historical business contents marked with operation ticket format prompts; wherein, one ticketing logic corresponds to one operation ticket format prompt. The historical business content marked with the operation ticket format prompt is input into the large model, and the output content of the large model is evaluated. After the evaluation is passed, the ticket model is obtained.

6. The ticket-generating control device based on a large model according to claim 5, characterized in that, The evaluation of the output of the large model includes: After inputting all historical business content marked with operation ticket format prompts into the large model, a sample of the completed ticket content corresponding to each historical business content marked with operation ticket format prompts is obtained. Based on the sample of completed ticket content, the number of successful ticket sales, the average time for ticket sales, and the average number of interactions for ticket sales were extracted. The system assesses the number of successful ticket transactions, the average time for ticket transactions, and the average number of interactions for ticket transactions. If the number of successful ticket sales is greater than or equal to the success number threshold, the average ticket sales time is less than or equal to the average time threshold, and the average number of interactions for ticket sales is less than or equal to the number of interactions threshold, then the performance evaluation is passed. Otherwise, the effectiveness evaluation will fail.

Images