Intelligent generation method and device of two votes, computer readable storage medium

Abstract

The application discloses a kind of two ticket intelligent generation method and device, computer readable storage medium.Therein, the method includes: in response to the two ticket generation operation of front-end page, to generate screen recording instruction and information acquisition instruction;According to screen recording instruction, control screen recording component to execute screen recording operation, obtain screen video data;Text extraction is carried out to screen video data, and screen text information in screen video data is obtained;Current operating state and work task information of power system are obtained based on information acquisition instruction;Screen text information, work task information and current operating state are processed using two ticket generation model, to generate work ticket and operation ticket.The present application solves the technical problem that the operation ticket and work ticket are usually generated by artificial handwriting or single and less flexible intelligent algorithm in related art, leading to the technical problem that the filling and approval process of two tickets is complicated and the handling efficiency is low.

Description

Technical Field

[0001] This invention relates to the field of power system two-ticket generation technology, and more specifically, to an intelligent two-ticket generation method and apparatus, and a computer-readable storage medium. Background Technology

[0002] The "two-ticket" system for power grid companies, including operation tickets and work tickets, is a crucial guarantee for safe power production. The accuracy and completeness of the information provided is paramount, and the timing of ticket issuance directly determines work efficiency.

[0003] However, there are two major problems in the current handling of "two tickets" by power grid companies: (1) The handling of "two tickets" is done in the form of handwritten or electronic documents, which is time-consuming, complicated in the approval process, and has low efficiency. This handling mode of "two tickets" directly restricts the smooth operation of various tasks of power grid companies; (2) The level of grassroots employees of power grids is uneven, and there are many subjective factors when handling "two tickets". There are errors in the information filling of "two tickets", the safety measures in the work ticket are not filled in in a standardized manner, and the steps in the operation ticket are reversed, skipped, or omitted.

[0004] Currently, while some existing intelligent algorithm-based ticketing methods have improved ticketing efficiency to some extent, they still have some limitations: 1) Insufficient intelligence: Current ticket generation technologies all use natural language processing to complete text generation tasks. Whether it is template-based, statistical, or deep learning-based text generation technology, it still has the disadvantages of difficulty in training models, easy semantic ambiguity, and difficulty in processing texts of different languages. It cannot simulate the human brain's learning and decision-making process, and cannot identify and encode patterns and relationships in a large amount of data, and then use this information to understand the user's natural language requests or questions and respond with relevant new content; 2) Relatively simple functions, unable to achieve full-process automation of "two tickets": At present, most systems are tailored to specific power grids, and there are problems such as difficulty in establishing knowledge bases, rigid operating rules, insufficient flexibility, and poor universality. Moreover, the operating rule base is huge, and the probability of error is very high, affecting the accuracy of ticket writing. Currently, the system can only handle the issuance of work tickets or operation tickets. However, the entire process of handling these two tickets involves more than just issuing work tickets or operation tickets; it also includes grounding wire drawing, safety technical briefing, survey report, and work measure card. 3) Poor user experience: The current system's user interface and human-computer interaction design may be inadequate, causing inconvenience for users. For example, manual input of work content, time, and site conditions into the power work ticket is still required for automatic generation of safety measures. Maintenance personnel still need to receive dispatch orders and determine whether automated ticketing mode can be enabled based on the availability of the on-duty computer. 4) Insufficient scalability and flexibility: As business needs change, the "two tickets" processing system needs good scalability and flexibility. However, the current system does not provide sufficient data interfaces and functional modules, making it difficult to adapt to changes in business needs. This results in high difficulty and significant time and cost associated with system upgrades and modifications.

[0005] Currently, there is no effective solution to the problem that the above-mentioned technologies typically use manual handwriting or simple, inflexible intelligent algorithms to generate operation tickets and work tickets, resulting in cumbersome filling and approval processes and low efficiency. Summary of the Invention

[0006] This invention provides an intelligent method and apparatus for generating two types of tickets, as well as a computer-readable storage medium, to at least solve the technical problem in related technologies where operation tickets and work tickets are usually generated manually or using simple, inflexible intelligent algorithms, resulting in cumbersome filling and approval processes and low efficiency.

[0007] According to one aspect of the present invention, an intelligent method for generating two tickets is provided, comprising: responding to a two-ticket generation operation acting on a front-end page to generate a screen recording instruction and an information acquisition instruction, wherein the two tickets include: a work ticket and an operation ticket, the work ticket being used to record a maintenance task for maintaining power equipment, and the operation ticket being used to record an operation task for performing an operation on the power equipment; controlling a screen recording component to perform a screen recording operation according to the screen recording instruction to obtain screen video data, wherein the screen recording operation refers to controlling the screen recording component to record the operation sensed by the front-end page; extracting text from the screen video data to obtain screen text information in the screen video data; based on The information acquisition instruction acquires the current operating status and work task information of the power system, wherein the work task information refers to the information of the work task that needs to be recorded in the work ticket and the operation ticket; the two-ticket generation model is used to process the screen text information, the work task information and the current operating status to generate the work ticket and the operation ticket, wherein the two-ticket generation model is trained using multiple sets of training data through deep learning, each set of multiple sets of training data includes: sample input data, sample work ticket and sample operation ticket corresponding to the sample input data, and the sample input data includes: sample screen text information, sample work task information and sample current operating status.

[0008] Optionally, after responding to the two-ticket generation operation applied to the front-end page to generate a screen recording instruction and an information retrieval instruction, the intelligent generation method for the two tickets further includes: obtaining the sensing habits of the front-end page based on the sensing identifier of the two-ticket generation operation sensed by the front-end page, wherein the sensing habits refer to the habits when the two-ticket generation operation is applied to the front-end page; optimizing the screen recording instruction based on the sensing habits to obtain an optimized screen recording instruction; obtaining the sensing time of the front-end page sensing the two-ticket generation operation; determining the operating habits of the power system based on the time period in which the sensing time is located, wherein the operating habits refer to the operating state of the power system in different time periods; and optimizing the information retrieval instruction based on the operating habits to obtain an optimized information retrieval instruction.

[0009] Optionally, text extraction is performed on the screen video data to obtain screen text information in the screen video data, including: decomposing the screen video data to obtain multiple video frame images; identifying the text direction of the text in each video frame image to obtain an identification result, wherein the text direction refers to the direction of the center line of the text; if the identification result indicates that there is no offset between the text direction and a specific direction, the video frame image is determined to be a target video frame image; if the identification result indicates that there is an offset between the text direction and the specific direction, the video frame image is corrected until the text direction of the text in the video frame image is the same as the specific direction, and the video frame image is determined to be the target video frame image, wherein the specific direction refers to the vertical direction; and text extraction is performed on each target video frame image to obtain the screen text information.

[0010] Optionally, before processing the screen text information, the work task information, and the current running status using the two-ticket generation model to generate the work ticket and the operation ticket, the intelligent generation method for the two tickets further includes: acquiring multiple sets of sample data, wherein each set of the multiple sets of sample data includes: sample input data, a sample work ticket and a sample operation ticket corresponding to the sample input data, wherein the sample input data includes: sample screen text information, sample work task information, and sample current running status; dividing the multiple sets of sample data to obtain sample training data and sample test data, wherein the sample training data is used to train the two-ticket generation model, and the sample test data is used to test the two-ticket generation model; and using the multiple sets of sample training data... The AI-generated content model is iteratively trained until an iteration termination condition is met. The resulting AI-generated content model is then identified as the initial two-vote generation model. This AI-generated content model is a deep learning model that automatically generates various types of content based on input information. These various types of content include at least one of the following: text and images. The iteration termination condition includes at least one of the following: the number of iterations reaches a preset number of iterations, and the model output error is below an error threshold. The initial two-vote generation model is then validated using the sample test data to obtain validation results. If the validation results indicate that the performance of the initial two-vote generation model meets a preset performance standard, the initial two-vote generation model is identified as the two-vote generation model.

[0011] Optionally, the intelligent generation method for the two votes further includes: before iteratively training the AI-generated content model using multiple sets of sample training data, randomly initializing the model parameters of the AI-generated content model using an improved chaotic mapping, wherein the first expression of the improved chaotic mapping is: x i Let x represent the state value of the i-th generation population. i+1 The state value represents the next generation population of the i-th generation population; during the iterative training of the AI-generated content model using multiple sets of sample training data, the learning rate of the AI-generated content model is optimized using a first formula, wherein the first formula is: GP represents the learning rate, and n represents the current iteration number. max The maximum number of iterations is indicated, and the learning rate is used to control the step size for updating the model parameters.

[0012] Optionally, the intelligent generation method for the two tickets further includes: during the process of generating the work ticket and the operation ticket using the two-ticket generation model, dynamically adjusting the regularization parameter of the two-ticket generation model using a second formula to adaptively optimize the two-ticket generation model, wherein the regularization parameter, according to the second formula, is: 'a' represents the regularization parameter, which is used to determine the degree of penalty for the two-vote generation model.

[0013] Optionally, after processing the screen text information, the work task information, and the current running status using the two-ticket generation model to generate the work ticket and the operation ticket, the intelligent generation method for the two tickets further includes: sending the work ticket and the operation ticket to the front-end page for visual presentation to review the work ticket and the operation ticket; upon receiving a two-ticket adjustment request, parsing the two-ticket adjustment request to obtain an adjustment strategy for adjusting the work ticket and / or the operation ticket, and receiving update information, wherein the two-ticket adjustment request is an adjustment request generated by the adjustment component when the front-end page senses that the work ticket and the operation ticket have failed the review, and the update information is information that needs to be presented on the work ticket and / or the operation ticket; adjusting the work ticket and / or the operation ticket according to the adjustment strategy based on the update information to obtain the target work ticket and / or the target operation ticket.

[0014] According to another aspect of the present invention, an intelligent two-ticket generation device is also provided, comprising: a first generation unit, configured to respond to a two-ticket generation operation acting on a front-end page to generate a screen recording instruction and an information acquisition instruction, wherein the two tickets include: a work ticket and an operation ticket, the work ticket being used to record a maintenance task for maintaining power equipment, and the operation ticket being used to record an operation task for performing an operation on the power equipment; a first acquisition unit, configured to control a screen recording component to perform a screen recording operation according to the screen recording instruction to obtain screen video data, wherein the screen recording operation refers to controlling the screen recording component to record the operation sensed by the front-end page; and a second acquisition unit, configured to extract text from the screen video data to obtain screen text in the screen video data. The system comprises: a third acquisition unit, configured to acquire the current operating status and work task information of the power system based on the information acquisition instruction, wherein the work task information refers to the information of the work task that needs to be recorded in the work ticket and the operation ticket; and a second generation unit, configured to process the screen text information, the work task information, and the current operating status using a two-ticket generation model to generate the work ticket and the operation ticket, wherein the two-ticket generation model is trained using multiple sets of training data through deep learning, and each set of training data includes: sample input data, sample work tickets and sample operation tickets corresponding to the sample input data, wherein the sample input data includes: sample screen text information, sample work task information, and sample current operating status.

[0015] Optionally, the intelligent generation device for the two tickets further includes: a fourth acquisition unit, used to acquire the sensing habits of the front-end page based on the sensing identifier of the front-end page sensing the two ticket generation operation after responding to the two ticket generation operation applied to the front-end page to generate a screen recording instruction and an information acquisition instruction, wherein the sensing habits refer to the habits when the two ticket generation operation is applied to the front-end page; a fifth acquisition unit, used to optimize the screen recording instruction based on the sensing habits to obtain an optimized screen recording instruction; a sixth acquisition unit, used to acquire the sensing time of the front-end page sensing the two ticket generation operation; a first determination unit, used to determine the operating habits of the power system based on the time period in which the sensing time is located, wherein the operating habits refer to the operating state of the power system in different time periods; and a seventh acquisition unit, used to optimize the information acquisition instruction based on the operating habits to obtain an optimized information acquisition instruction.

[0016] Optionally, the second acquisition unit includes: a first acquisition module, configured to decompose the screen video data to obtain multiple video frame images; a second acquisition module, configured to identify the text direction of the text in each of the video frame images to obtain an identification result, wherein the text direction refers to the direction of the center line of the text; a first determination module, configured to determine the video frame image as a target video frame image when the identification result indicates that there is no offset between the text direction and the specific direction; a second determination module, configured to correct the video frame image when the identification result indicates that there is an offset between the text direction and the specific direction, until the text direction of the text in the video frame image is the same as the specific direction, and determine the video frame image as the target video frame image, wherein the specific direction refers to the vertical direction; and a third acquisition module, configured to extract text from each of the target video frame images to obtain the screen text information.

[0017] Optionally, the intelligent generation device for the two tickets further includes: an eighth acquisition unit, configured to acquire multiple sets of sample data before processing the screen text information, the work task information, and the current running status using the two-ticket generation model to generate the work ticket and the operation ticket, wherein each set of the multiple sets of sample data includes: sample input data, a sample work ticket and a sample operation ticket corresponding to the sample input data, wherein the sample input data includes: sample screen text information, sample work task information, and sample current running status; a ninth acquisition unit, configured to divide the multiple sets of sample data to obtain sample training data and sample test data, wherein the sample training data is used to train the two-ticket generation model, and the sample test data is used to test the two-ticket generation model; and a second determination unit, configured to use the multiple sets of sample data... The training data is used to iteratively train the AI-generated content model until the iteration termination condition is met. The AI-generated content model is then determined to be the initial two-vote generation model. This AI-generated content model is a deep learning model that automatically generates various types of content based on input information. These various types of content include at least one of the following: text and images. The iteration termination condition includes at least one of the following: the number of iterations reaches a preset number of iterations, and the model output error is lower than an error threshold. The tenth acquisition unit is used to perform performance verification on the initial two-vote generation model using the sample test data and obtain verification results. The third determination unit is used to determine the initial two-vote generation model as the two-vote generation model if the verification result indicates that the performance of the initial two-vote generation model meets a preset performance standard.

[0018] Optionally, the intelligent generation device for the two votes further includes: an initialization unit, used to randomly initialize the model parameters of the AI-generated content model using an improved chaotic mapping before iteratively training the AI-generated content model using multiple sets of sample training data, wherein the first expression of the improved chaotic mapping is: xi represents the state value of the i-th generation population, x i+1 The state value represents the next generation population of the i-th generation population; the first optimization unit is used to optimize the learning rate of the AI-generated content model using a first formula during the iterative training of the AI-generated content model using multiple sets of sample training data, wherein the first formula is: GP represents the learning rate, and n represents the current iteration number. max The maximum number of iterations is indicated, and the learning rate is used to control the step size for updating the model parameters.

[0019] Optionally, the intelligent generation device for the two tickets further includes: a second optimization unit, used to dynamically adjust the regularization parameter of the two-ticket generation model using a second formula during the process of generating the work ticket and the operation ticket using the two-ticket generation model, so as to adaptively optimize the two-ticket generation model, wherein the regularization parameter, the second formula is: 'a' represents the regularization parameter, which is used to determine the degree of penalty for the two-vote generation model.

[0020] Optionally, the intelligent generation device for the two tickets further includes: a sending unit, configured to, after processing the screen text information, the work task information, and the current running status using the two-ticket generation model to generate the work ticket and the operation ticket, send the work ticket and the operation ticket to the front-end page for visual presentation to review the work ticket and the operation ticket; a receiving unit, configured to, upon receiving a two-ticket adjustment request, parse the two-ticket adjustment request to obtain an adjustment strategy for adjusting the work ticket and / or the operation ticket, and receive update information, wherein the two-ticket adjustment request is an adjustment request generated by the adjustment component when the front-end page senses that the work ticket and the operation ticket have failed review, and the update information is information that needs to be presented on the work ticket and / or the operation ticket; and an eleventh obtaining unit, configured to adjust the work ticket and / or the operation ticket according to the adjustment strategy based on the update information to obtain a target work ticket and / or a target operation ticket.

[0021] According to another aspect of the present invention, an intelligent two-vote generation system is also provided, wherein the intelligent two-vote generation system uses any of the above-described intelligent two-vote generation methods.

[0022] According to another aspect of the present invention, a computer-readable storage medium is also provided, the computer-readable storage medium including a stored program, wherein the program executes any of the above-described intelligent two-vote generation methods.

[0023] According to another aspect of the present invention, a processor is also provided, the processor being used to run a program, wherein the program, when running, executes any of the above-described intelligent two-vote generation methods.

[0024] According to another aspect of the present invention, a computer program product is also provided, including computer instructions, which, when executed by a processor, perform any of the above-described intelligent two-vote generation methods.

[0025] In this embodiment of the invention, in response to the two-ticket generation operation applied to the front-end page, a screen recording instruction and an information acquisition instruction are generated. The two tickets include a work ticket and an operation ticket. The work ticket records maintenance tasks for power equipment, and the operation ticket records operational tasks for power equipment. The screen recording component is controlled to perform a screen recording operation according to the screen recording instruction to obtain screen video data. The screen recording operation refers to controlling the screen recording component to record operations sensed by the front-end page. Text extraction is performed on the screen video data to obtain screen text information. The current operating status and work task information of the power system are obtained based on the information acquisition instruction. The work task information refers to the information of the work tasks that need to be recorded in the work ticket and operation ticket. The two-ticket generation model is used to process the screen text information, work task information, and current operating status to generate work tickets and operation tickets. The two-ticket generation model is trained using multiple sets of training data through deep learning. Each set of training data includes: sample input data, sample work tickets, and sample operation tickets corresponding to the sample input data. The sample input data includes: sample screen text information, sample work task information, and sample current operating status. The above technical solution achieves the goal of recording the operations performed by maintenance personnel at the front end, extracting the text, obtaining the work task information submitted by the maintenance personnel and the current operating status of the power system, and then using a trained two-ticket generation model to process the obtained information to generate work tickets and operation tickets. This achieves a more comprehensive acquisition of the content required for generating two tickets from multiple dimensions, and improves the efficiency and accuracy of generating two tickets by using the two-ticket generation model. It also simplifies the two-ticket processing procedure, thereby solving the technical problem that the current technology often uses manual handwriting or simple and inflexible intelligent algorithms to generate operation tickets and work tickets, resulting in cumbersome filling and approval processes and low processing efficiency. Attached Figure Description

[0026] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings:

[0027] Figure 1 This is a hardware structure block diagram of a mobile terminal for an intelligent two-vote generation method according to an embodiment of the present invention.

[0028] Figure 2 This is a flowchart of an intelligent two-vote generation method according to an embodiment of the present invention;

[0029] Figure 3 This is a schematic diagram of the code for optimizing text direction according to an embodiment of the present invention;

[0030] Figure 4(a) is a scatter plot of the Circle chaos map according to an embodiment of the present invention;

[0031] Figure 4(b) is a scatter plot of the improved Circle chaotic map according to an embodiment of the present invention;

[0032] Figure 5(a) is a distribution histogram of the Circle chaotic map according to an embodiment of the present invention;

[0033] Figure 5(b) is a distribution histogram of the improved Circle chaotic map according to an embodiment of the present invention;

[0034] Figure 6 This is a schematic diagram illustrating the change of the learning rate parameter according to an embodiment of the present invention;

[0035] Figure 7 This is a schematic diagram illustrating the variation of regularization parameters according to an embodiment of the present invention;

[0036] Figure 8 This is a schematic diagram of an intelligent two-vote generation device according to an embodiment of the present invention. Detailed Implementation

[0037] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. 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 should fall within the scope of protection of the present invention.

[0038] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0039] As described in the background section, related technologies typically employ manual handwriting or simplistic, inflexible intelligent algorithms to generate operation tickets and work tickets, resulting in cumbersome completion and approval processes and low efficiency. To address these shortcomings, embodiments of the present invention provide an intelligent generation method and apparatus for these two tickets, as well as a computer-readable storage medium.

[0040] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0041] The methods and embodiments provided in this invention can be executed on a mobile terminal, a computer terminal, or a similar computing device. Taking running on a mobile terminal as an example, Figure 1 This is a hardware structure block diagram of a mobile terminal for an intelligent two-vote generation method according to an embodiment of the present invention. Figure 1 As shown, a mobile terminal may include one or more ( Figure 1 Only one is shown in the diagram. A processor 102 (which may include, but is not limited to, a microprocessor MCU or a programmable logic device FPGA, etc.) and a memory 104 for storing data are also shown. The mobile terminal may further include a transmission device 106 for communication functions and an input / output device 108. Those skilled in the art will understand that... Figure 1 The structure shown is for illustrative purposes only and does not limit the structure of the mobile terminal described above. For example, the mobile terminal may also include components that are more... Figure 1 The more or fewer components shown, or having the same Figure 1 The different configurations shown.

[0042] The memory 104 can be used to store computer programs, such as application software programs and modules, like the computer program corresponding to the intelligent two-vote generation method in this embodiment of the invention. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, thereby implementing the above-described method. The memory 104 may include high-speed random access memory and non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 104 may further include memory remotely located relative to the processor 102, and these remote memories can be connected to the mobile terminal via a network. Examples of the aforementioned networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof. The transmission device 106 is used to receive or send data via a network. Specific examples of the aforementioned networks may include wireless networks provided by the mobile terminal's communication provider. In one example, the transmission device 106 includes a network interface controller (NIC), which can be connected to other network devices via a base station to communicate with the Internet. In one example, the transmission device 106 may be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0043] According to an embodiment of the present invention, a method embodiment of an intelligent two-vote generation method is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0044] Figure 2 This is a flowchart of an intelligent two-vote generation method according to an embodiment of the present invention, such as... Figure 2 As shown, the method includes the following steps:

[0045] Step S202: Respond to the two-ticket generation operation applied to the front-end page to generate a screen recording instruction and an information acquisition instruction. The two tickets include a work ticket and an operation ticket. The work ticket is used to record maintenance tasks for power equipment, and the operation ticket is used to record operation tasks for power equipment.

[0046] In this embodiment, the system backend can generate screen recording instructions and information retrieval instructions when maintenance personnel trigger the operation of processing two tickets (including work tickets and operation tickets) through the previous page, so as to obtain the relevant data required for the subsequent generation of the two tickets based on the instructions.

[0047] Step S204: Control the screen recording component to perform screen recording operation according to the screen recording instruction to obtain screen video data. The screen recording operation refers to controlling the screen recording component to record the operations sensed by the front-end page.

[0048] In this embodiment, a real-time screen information acquisition system can be built. This system can use the ScreenCaptureKit provided by the operating system to record the applications opened by maintenance personnel during the "two tickets" processing and record the application window information. Here, ScreenCaptureKit refers to a toolkit or framework that can implement screen recording. In different operating systems or programming environments, the tools or libraries that implement screen recording may have different names.

[0049] Step S206: Extract text from the screen video data to obtain screen text information from the screen video data.

[0050] In this embodiment, an Optical Character Recognition (OCR) model is used to extract text information from the video data using an image recognition algorithm, and the extracted information is stored to achieve real-time information sharing and unified management.

[0051] Step S208: Obtain the current operating status and work task information of the power system based on the information acquisition instruction. The work task information refers to the information of the work tasks that need to be recorded in the work order and operation order.

[0052] In this embodiment, the system can also obtain work task information related to the two tickets submitted or uploaded by the operation and maintenance personnel based on the information acquisition instruction, and obtain the current operating status of the power system in real time, so as to provide a data basis for the subsequent generation of the two tickets.

[0053] Step S210: The two-ticket generation model is used to process the screen text information, work task information and current running status to generate work tickets and operation tickets. The two-ticket generation model is trained using multiple sets of training data through deep learning. Each set of training data includes: sample input data, sample work tickets and sample operation tickets corresponding to the sample input data. The sample input data includes: sample screen text information, sample work task information and sample current running status.

[0054] In this embodiment, a trained two-ticket generation model can be used to process the obtained screen text information, work task information, and the current operating status of the power system to generate work tickets and operation tickets.

[0055] As described above, the technical solution provided by the above embodiments of the present invention can respond to the two-ticket generation operation acting on the front-end page to generate screen recording instructions and information acquisition instructions. The two tickets include a work ticket and an operation ticket. The work ticket records maintenance tasks for power equipment, and the operation ticket records operation tasks for power equipment. According to the screen recording instruction, the screen recording component is controlled to perform a screen recording operation to obtain screen video data. The screen recording operation refers to controlling the screen recording component to record operations sensed by the front-end page. Text extraction is performed on the screen video data to obtain screen text information. Based on the information acquisition instruction, the current operating status and work task information of the power system are obtained. The work task information refers to the information of the work tasks that need to be recorded in the work ticket and operation ticket. The two-ticket generation model is used to process the screen text information, work task information, and current status of the power system. The system processes the operational status to generate work tickets and operation tickets. The ticket generation model is trained using deep learning with multiple sets of training data. Each set includes sample input data, a corresponding sample work ticket, and a sample operation ticket. The sample input data includes sample screen text information, sample work task information, and the sample's current operational status. This achieves the goal of recording operations performed by maintenance personnel at the front end, extracting text, obtaining the submitted work task information and the current operational status of the power system, and then using the trained ticket generation model to process the information to generate work tickets and operation tickets. This achieves a more comprehensive acquisition of the content needed to generate the two tickets from multiple dimensions, improving the efficiency and accuracy of ticket generation and simplifying the ticket processing procedure.

[0056] Therefore, the technical solution provided by the above embodiments of the present invention solves the technical problem in the related art that the operation tickets and work tickets are usually generated by manual handwriting or by a single, inflexible intelligent algorithm, which leads to a cumbersome filling and approval process and low processing efficiency.

[0057] According to the above embodiments of the present invention, after generating a screen recording instruction and an information retrieval instruction in response to a two-ticket generation operation applied to the front-end page, the intelligent generation method of the two tickets further includes: obtaining the sensing habits of the front-end page based on the sensing identifier of the two-ticket generation operation, wherein the sensing habits refer to the habits when the two-ticket generation operation is applied to the front-end page; optimizing the screen recording instruction based on the sensing habits to obtain an optimized screen recording instruction; obtaining the sensing time of the two-ticket generation operation on the front-end page; determining the operating habits of the power system based on the time period in which the sensing time is located, wherein the operating habits refer to the operating state of the power system in different time periods; and optimizing the information retrieval instruction based on the operating habits to obtain an optimized information retrieval instruction.

[0058] Specifically, after generating screen recording and information retrieval commands, the screen recording commands can be optimized based on the sensing habits of the maintenance personnel recorded by the system. For example, each maintenance personnel may have different sensing habits (here, the sensing habits of maintenance personnel are equivalent to the sensing habits of the front-end page when it senses the two ticket generation operations). Some people may be more accustomed to using keyboard shortcuts, while others may rely more on mouse clicks. The system identifies and stores these habits to more accurately generate screen recording instructions based on the recorded sensory habits when maintenance personnel operate the screen. For example, recording is triggered only when critical information is accessed, avoiding the recording of a large amount of irrelevant content, thus saving storage space and processing time. It can also automatically adjust the priority and strategy for acquiring power system status information based on the current time period. For instance, the power system may have different operating states and operational needs at different times, and operating procedures during nighttime or peak hours may differ from those during the day. By identifying these specific time-specific operating procedures, the system can automatically adjust the priority and strategy for information acquisition. For example, it can acquire all necessary equipment status and safety specification information before critical operations, ensuring that the generated two-ticket content conforms to the current operating environment. By matching the power system's specific time-specific operating procedures, the system can more quickly obtain all the information needed to generate two-tickets, avoiding repeated queries and confirmations, thereby significantly improving the efficiency of two-ticket generation.

[0059] According to the above embodiments of the present invention, text extraction from screen video data to obtain screen text information in the screen video data includes: decomposing the screen video data to obtain multiple video frame images; identifying the text direction of the text in each video frame image to obtain an identification result, wherein the text direction refers to the direction of the center line of the text; determining the video frame image as a target video frame image when there is no offset between the identification result indicating the text direction and the specific direction; correcting the video frame image until the text direction of the text in the video frame image is the same as the specific direction, wherein the specific direction refers to the vertical direction; and extracting text from each target video frame image to obtain screen text information.

[0060] The following is combined with Figure 3 The embodiments of the present invention will be described in detail below. Figure 3 This is a schematic diagram of the code for optimizing text direction according to an embodiment of the present invention.

[0061] Specifically, when using OCR technology to extract text from each image frame in screen video data, since the orientation classifier has a relatively high recognition accuracy in the two directions where the text is at 0 degrees and 180 degrees (i.e., the specific directions mentioned above), in order to improve the accuracy of text recognition and extraction, the following can be adopted: Figure 3 The optimized code shown performs orientation conversion on video frames with text orientation other than the specified orientation during the preprocessing stage, converting them to 0 degrees or 180 degrees, before extracting the text to improve the accuracy of text recognition.

[0062] It should be noted that the text orientation here refers to the orientation of the text as seen under normal visual conditions. In the specific recognition process, principal component analysis can be used to determine the main axis orientation of the text outline, and a pre-trained deep learning model can be used to classify the orientation of the text. The model has learned to distinguish between horizontal, vertical and slanted text to identify the text orientation. Of course, other methods can also be used to identify the text orientation, which will not be elaborated here.

[0063] According to the above embodiments of the present invention, before processing the screen text information, work task information, and current running status using the two-ticket generation model to generate work tickets and operation tickets, the intelligent generation method of the two tickets further includes: acquiring multiple sets of sample data, wherein each set of multiple sets of sample data includes: sample input data, sample work tickets and sample operation tickets corresponding to the sample input data, and the sample input data includes: sample screen text information, sample work task information, and sample current running status; dividing the multiple sets of sample data to obtain sample training data and sample test data, wherein the sample training data is used to train the two-ticket generation model, and the sample test data is used to test the two-ticket generation model; utilizing the multiple sets of sample training data... The AI-generated content model is iteratively trained until the iteration termination condition is met. The currently trained AI-generated content model is then identified as the initial two-vote generation model. This AI-generated content model is a deep learning model that automatically generates various types of content based on input information. These various content types include at least one of the following: text and images. The iteration termination condition includes at least one of the following: the number of iterations reaches a preset number of iterations, or the model output error is below an error threshold. The initial two-vote generation model is then validated using sample test data to obtain validation results. If the validation results indicate that the performance of the initial two-vote generation model meets the preset performance standard, the initial two-vote generation model is determined to be a two-vote generation model.

[0064] Specifically, during the training of the two-ticket generation model, a large amount of historical data can be acquired as sample training data for AIGC (Artificial Intelligence Generated Content Model). For example, the entire history of computer operations by maintenance personnel can be collected, including web browsing information and application dialog box content. This computer operation history can be made into a traceable and searchable database. Historical work task information and historical operating status information of the power system corresponding to this time can also be obtained as much as possible and used as training data to train AIGC, thus obtaining the two-ticket generation model. A portion of the sample training data is retained as test data to test the performance of the trained two-ticket generation model, ensuring its accuracy. Furthermore, human-computer interaction can be achieved through various modal signals (such as voice and text), ultimately making human-computer interaction as convenient and natural as human-to-human interaction. In addition, when using the two-ticket generation model to learn new tasks or data, it can quickly adapt and update. Its continuous learning capability allows the system to continuously evolve and update according to the needs of maintenance personnel, obtaining more personalized and relevant results.

[0065] According to the above embodiments of the present invention, the intelligent generation method for the two votes further includes: before iteratively training the artificial intelligence content generation model using multiple sets of sample training data, randomly initializing the model parameters of the artificial intelligence content generation model using an improved chaotic mapping, wherein the first expression of the improved chaotic mapping is: x i Let x represent the state value of the i-th generation population. i+1 Let represent the state value of the next generation of the i-th generation population; during the iterative training of the AI-generated content model using multiple sets of sample training data, the learning rate of the AI-generated content model is optimized using the first formula, where the first formula is: GP represents the learning rate, and n represents the current iteration number. max This indicates the maximum number of iterations, and the learning rate is used to control the step size for updating model parameters.

[0066] The following refers to Figures 4(a), 4(b), 5(a), and 5(b) and... Figure 6 The embodiments of the present invention will be described in detail below. Figure 4(a) is a scatter plot of the Circle chaotic map according to an embodiment of the present invention, Figure 4(b) is a scatter plot of the improved Circle chaotic map according to an embodiment of the present invention, Figure 5(a) is a distribution histogram of the Circle chaotic map according to an embodiment of the present invention, and Figure 5(b) is a distribution histogram of the improved Circle chaotic map according to an embodiment of the present invention. Figure 6 This is a schematic diagram illustrating the change of the learning rate parameter according to an embodiment of the present invention.

[0067] Specifically, in the AIGC standard model, pseudo-random initialization is used during the initialization phase. This can easily lead to clustering of initial values, while chaotic initialization can effectively improve this phenomenon. There are many methods of chaotic mapping, among which Circle chaotic mapping produces a high coverage of chaotic values, but the chaotic values ​​still exhibit local uneven distribution, so it needs to be improved. The standard Circle chaotic mapping formula is: The improved Circle chaotic mapping formula is: In the formula, i represents the dimension of the solution. The improved effect can be visually demonstrated by the scatter plots shown in Figure 4(a) and Figure 4(b) and the distribution histograms shown in Figure 5(a) and Figure 5(b). i can be taken as 2000 (this is only an example and can be adjusted according to the actual situation; no specific restrictions are imposed here). As shown in Figures 4(a), 4(b), 5(a), and 5(b), it can be intuitively found by comparison that the distribution of the improved Circle chaotic mapping values ​​is more uniform. Therefore, using the improved Circle chaotic mapping method to initialize the population can enhance the diversity of the population and thus enhance the optimization algorithm's search capability.

[0068] Furthermore, the learning rate is a crucial parameter controlling the speed of model parameter updates. An excessively large learning rate can lead to model oscillations and instability. A fixed learning rate is typically used, which is the simplest method for adjusting the learning rate—keeping it constant throughout the training process. While this method is simple and intuitive, it may not adapt well to different training stages, leading to instability or slow convergence. Therefore, adaptive algorithms are needed for parameter tuning, as shown below: GP represents the learning rate, and n represents the current iteration number. max This indicates the maximum number of iterations.

[0069] Testing revealed that a GP range of [0.095, 0.1] allows the algorithm to achieve better global search capabilities while maintaining local exploration capabilities (if other GP ranges are found to achieve better performance in practice, adjustments can be made accordingly; no specific restrictions are imposed here). Figure 6 As shown, the generation probability is small in the early stage of the iteration process. As the generation probability increases, it eventually reaches a value of 0.1. The global search capability of the algorithm is continuously enhanced in the optimization process, preventing the algorithm from getting stuck in local optima in the later stage of the optimization process.

[0070] According to the above embodiments of the present invention, the intelligent generation method of the two tickets further includes: during the process of generating work tickets and operation tickets using the two-ticket generation model, dynamically adjusting the regularization parameter of the two-ticket generation model using a second formula to adaptively optimize the two-ticket generation model, wherein the regularization parameter, the second formula, is: 'a' represents the regularization parameter, which determines the degree of penalty applied to the two-vote generation model.

[0071] The following is combined with Figure 7 The embodiments of the present invention will be described in detail below. Figure 7 This is a schematic diagram illustrating the variation of regularization parameters according to an embodiment of the present invention.

[0072] Specifically, the regularization parameter is used to control the complexity of the model and avoid overfitting. By adjusting the regularization parameter, the model's fitting ability and generalization ability can be balanced. Generally, a larger regularization parameter reduces model complexity and is suitable for handling high variance problems; while a smaller regularization parameter is suitable for handling high bias problems. To balance the algorithm's fitting ability and generalization ability, the regularization parameter is tuned using the sine and cosine principle, as shown below: like Figure 7 As shown, a increases non-linearly from around 0.35 to 0.5 with the number of iterations. As a changes, the algorithm gradually enhances its local exploration ability, enabling it to find particles that are infinitely close to the optimal solution within the range of the global optimal solution, thus improving the algorithm's optimization ability.

[0073] According to the above embodiments of the present invention, after processing screen text information, work task information, and current operating status using a two-ticket generation model to generate work tickets and operation tickets, the intelligent generation method of the two tickets further includes: sending the work tickets and operation tickets to a front-end page for visual presentation to review the work tickets and operation tickets; upon receiving a two-ticket adjustment request, parsing the two-ticket adjustment request to obtain an adjustment strategy for adjusting the work tickets and / or operation tickets, and receiving update information, wherein the two-ticket adjustment request is an adjustment request generated by the adjustment component when the front-end page senses that the work tickets and operation tickets have failed the review, and the update information is information that needs to be presented on the work tickets and / or operation tickets; adjusting the work tickets and / or operation tickets according to the adjustment strategy based on the update information to obtain target work tickets and / or target operation tickets.

[0074] Specifically, after initially generating work tickets and operation tickets using the two-ticket generation model, the initially generated work tickets and operation tickets can be shown to operations and maintenance personnel for review. If the operations and maintenance personnel approve the work tickets, they can directly trigger the corresponding buttons to download and save them. The system will also respond and save the work tickets and operation tickets. If the operations and maintenance personnel fail to approve the work tickets, they can trigger the corresponding adjustment buttons and simultaneously upload the information that needs to be updated. While responding to the two-ticket adjustment request, the system will parse the request to determine which part of the work ticket and / or operation ticket needs to be adjusted. Based on the update information submitted by the operations and maintenance personnel, the system will make the corresponding adjustments and regenerate the work tickets and / or operation tickets to be shown to the operations and maintenance personnel again.

[0075] The technical solutions provided by the above embodiments of the present invention also achieve the following technical effects: 1) It can achieve a more natural and fluent human-computer dialogue experience. By learning a large amount of text data, it can generate fluent and coherent language expressions, better understand the user's intentions, and has long-term memory capabilities. In addition, its sustainable learning ability can enable it to continuously evolve and update with user needs, obtain more personalized and relevant results, and improve the human-computer interaction experience; 2) It can use more advanced incremental learning algorithms and more flexible model structures to adapt to different tasks and data, or more intelligent ensemble learning methods to integrate the prediction results of multiple models, thereby improving the system's generalization ability and adaptability.

[0076] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to this application. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to this application.

[0077] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods according to the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0078] According to embodiments of the present invention, an intelligent two-vote generation apparatus for implementing the above-described intelligent two-vote generation method is also provided. Figure 8 This is a schematic diagram of an intelligent two-vote generation device according to an embodiment of the present invention, as shown below. Figure 8 As shown, the device includes: a first generation unit 81, a first acquisition unit 83, a second acquisition unit 85, a third acquisition unit 87, and a second generation unit 89. The intelligent generation device for these two votes will be described in detail below.

[0079] The first generation unit 81 is used to respond to the two ticket generation operations applied to the front-end page to generate screen recording instructions and information acquisition instructions. The two tickets include a work ticket and an operation ticket. The work ticket is used to record maintenance tasks for power equipment, and the operation ticket is used to record operation tasks for power equipment.

[0080] The first acquisition unit 83 is used to control the screen recording component to perform screen recording operations according to the screen recording instruction, and obtain screen video data. The screen recording operation refers to controlling the screen recording component to record the operations sensed by the front-end page.

[0081] The second acquisition unit 85 is used to extract text from the screen video data to obtain screen text information in the screen video data.

[0082] The third acquisition unit 87 is used to acquire the current operating status and work task information of the power system based on the information acquisition instruction. The work task information refers to the information of the work tasks that need to be recorded in the work order and operation order.

[0083] The second generation unit 89 is used to process screen text information, work task information and current running status using the two-ticket generation model to generate work tickets and operation tickets. The two-ticket generation model is trained using multiple sets of training data through deep learning. Each set of training data includes: sample input data, sample work tickets and sample operation tickets corresponding to the sample input data. The sample input data includes: sample screen text information, sample work task information and sample current running status.

[0084] It should be noted that the first generation unit 81, the first acquisition unit 83, the second acquisition unit 85, the third acquisition unit 87 and the second generation unit 89 mentioned above correspond to steps S202 to S210 in the above embodiments. The three modules and the corresponding steps implement the same instances and application scenarios, but are not limited to the content disclosed in the above embodiments.

[0085] As can be seen from the above, in the solution described in the above embodiments of the present invention, the first generation unit can respond to the two ticket generation operations acting on the front-end page to generate screen recording instructions and information acquisition instructions. The two tickets include a work ticket and an operation ticket. The work ticket is used to record maintenance tasks for power equipment, and the operation ticket is used to record operation tasks for power equipment. Then, the first acquisition unit controls the screen recording component to perform screen recording operations according to the screen recording instructions to obtain screen video data. The screen recording operation refers to controlling the screen recording component to record operations sensed by the front-end page. Next, the second acquisition unit extracts text from the screen video data to obtain screen text information. Then, the third acquisition unit uses the information acquisition instructions to obtain the current operating status and work task information of the power system. The work task information refers to the information of the work tasks that need to be recorded in the work ticket and operation ticket. Finally, the second generation unit uses the two... The ticket generation model processes screen text information, work task information, and current operating status to generate work tickets and operation tickets. The model is trained using deep learning with multiple sets of training data. Each set includes sample input data, a corresponding sample work ticket, and a sample operation ticket. The sample input data includes sample screen text information, sample work task information, and sample current operating status. This model achieves the goal of recording operations performed by maintenance personnel at the front end, extracting text, obtaining the submitted work task information and the current operating status of the power system, and then using the trained ticket generation model to process the information to generate work tickets and operation tickets. This achieves a more comprehensive acquisition of the content needed to generate the two tickets from multiple dimensions, improving the efficiency and accuracy of ticket generation and simplifying the ticket processing procedure.

[0086] Therefore, the technical solution provided by the above embodiments of the present invention solves the technical problem in the related art that the operation tickets and work tickets are usually generated by manual handwriting or by a single, inflexible intelligent algorithm, which leads to a cumbersome filling and approval process and low processing efficiency.

[0087] Optionally, the intelligent generation device for the two tickets further includes: a fourth acquisition unit, used to acquire the sensing habits of the front-end page based on the sensing identifier of the two ticket generation operation sensed by the front-end page after responding to the two ticket generation operation acting on the front-end page to generate a screen recording instruction and an information acquisition instruction, wherein the sensing habits refer to the habits when the two ticket generation operation acts on the front-end page; a fifth acquisition unit, used to optimize the screen recording instruction based on the sensing habits to obtain an optimized screen recording instruction; a sixth acquisition unit, used to acquire the sensing time of the two ticket generation operation sensed by the front-end page; a first determination unit, used to determine the operating habits of the power system based on the time period in which the sensing time is located, wherein the operating habits refer to the operating state of the power system in different time periods; and a seventh acquisition unit, used to optimize the information acquisition instruction based on the operating habits to obtain an optimized information acquisition instruction.

[0088] Optionally, the second acquisition unit includes: a first acquisition module, used to decompose the screen video data to obtain multiple video frame images; a second acquisition module, used to identify the text direction of the text in each video frame image to obtain an identification result, wherein the text direction refers to the direction of the center line of the text; a first determination module, used to determine the video frame image as a target video frame image when there is no offset between the text direction indicated by the identification result and the specific direction; a second determination module, used to correct the video frame image when there is an offset between the text direction indicated by the identification result and the specific direction, until the text direction of the text in the video frame image is the same as the specific direction, and then determine the video frame image as a target video frame image, wherein the specific direction refers to the vertical direction; and a third acquisition module, used to extract text from each target video frame image to obtain screen text information.

[0089] Optionally, the intelligent generation device for the two tickets further includes: an eighth acquisition unit, used to acquire multiple sets of sample data before processing screen text information, work task information, and current running status using the two-ticket generation model to generate work tickets and operation tickets, wherein each set of sample data includes: sample input data, sample work tickets and sample operation tickets corresponding to the sample input data, and the sample input data includes: sample screen text information, sample work task information, and sample current running status; a ninth acquisition unit, used to divide the multiple sets of sample data to obtain sample training data and sample test data, wherein the sample training data is used to train the two-ticket generation model, and the sample test data is used to test the two-ticket generation model; and a second determination unit, used to utilize the multiple sets of sample training data... The AI-generated content model is iteratively trained until the iteration termination condition is met. The resulting AI-generated content model is then identified as the initial two-vote generation model. This AI-generated content model is a deep learning model that automatically generates various types of content based on input information. These various content types include at least one of the following: text and images. The iteration termination condition includes at least one of the following: the number of iterations reaches a preset number of iterations, or the model output error is below an error threshold. The tenth acquisition unit is used to verify the performance of the initial two-vote generation model using sample test data and obtain the verification result. The third determination unit is used to determine the initial two-vote generation model as a two-vote generation model if the verification result indicates that the performance of the initial two-vote generation model meets a preset performance standard.

[0090] Optionally, the intelligent generation device for the two votes further includes: an initialization unit, used to randomly initialize the model parameters of the AI-generated content model using an improved chaotic mapping before iteratively training the AI-generated content model using multiple sets of sample training data, wherein the first expression of the improved chaotic mapping is: x i Let x represent the state value of the i-th generation population. i+1 This represents the state value of the next generation of the i-th generation population; the first optimization unit is used to optimize the learning rate of the AI-generated content model using a first formula during the iterative training of the AI-generated content model using multiple sets of sample training data, wherein the first formula is: GP represents the learning rate, and n represents the current iteration number. max This indicates the maximum number of iterations, and the learning rate is used to control the step size for updating model parameters.

[0091] Optionally, the intelligent generation device for the two tickets further includes: a second optimization unit, used to dynamically adjust the regularization parameter of the two-ticket generation model using a second formula during the process of generating work tickets and operation tickets using the two-ticket generation model, so as to adaptively optimize the two-ticket generation model, wherein the regularization parameter, the second formula is: 'a' represents the regularization parameter, which determines the degree of penalty applied to the two-vote generation model.

[0092] Optionally, the intelligent generation device for the two tickets further includes: a sending unit, used to process screen text information, work task information, and current operating status using the two-ticket generation model to generate work tickets and operation tickets, and then send the work tickets and operation tickets to the front-end page for visual presentation to review the work tickets and operation tickets; a receiving unit, used to parse the two-ticket adjustment request upon receiving it, obtain the adjustment strategy for adjusting the work tickets and / or operation tickets, and receive update information, wherein the two-ticket adjustment request is the adjustment request generated by the adjustment component when the front-end page senses that the work tickets and operation tickets have failed the review, and the update information is the information that needs to be presented on the work tickets and / or operation tickets; and an eleventh obtaining unit, used to adjust the work tickets and / or operation tickets according to the adjustment strategy based on the update information to obtain the target work tickets and / or target operation tickets.

[0093] According to another aspect of the present invention, an intelligent two-vote generation system is also provided, which uses any of the above-described intelligent two-vote generation methods.

[0094] According to another aspect of the present invention, a computer-readable storage medium is also provided, the computer-readable storage medium including a stored program, wherein the program executes any of the above-described intelligent two-vote generation methods.

[0095] Optionally, in this embodiment, the computer-readable storage medium may be located in any computer terminal in a group of computer terminals in a computer network, or in any communication device in a group of communication devices.

[0096] Optionally, in this embodiment, the computer-readable storage medium is configured to store program code for performing the following steps: responding to a two-ticket generation operation acting on the front-end page to generate a screen recording instruction and an information retrieval instruction, wherein the two tickets include: a work ticket and an operation ticket, the work ticket being used to record maintenance tasks for power equipment, and the operation ticket being used to record operation tasks for performing operations on the power equipment; controlling the screen recording component to perform a screen recording operation according to the screen recording instruction to obtain screen video data, wherein the screen recording operation refers to controlling the screen recording component to record the operation sensed by the front-end page; and extracting text from the screen video data to obtain the screen in the screen video data. Text information; based on information acquisition instructions, the current operating status and work task information of the power system are obtained. Among them, the work task information refers to the information of the work tasks that need to be recorded in the work order and operation order. The two-ticket generation model is used to process the screen text information, work task information and current operating status to generate work orders and operation orders. The two-ticket generation model is trained by deep learning using multiple sets of training data. Each set of training data includes: sample input data, sample work orders and sample operation orders corresponding to the sample input data. The sample input data includes: sample screen text information, sample work task information and sample current operating status.

[0097] Optionally, in this embodiment, the computer-readable storage medium is configured to store program code for performing the following steps: obtaining the sensing habits of the front-end page based on the sensing identifier of the two-ticket generation operation, wherein the sensing habits refer to the habits when the two-ticket generation operation is applied to the front-end page; optimizing the screen recording instruction based on the sensing habits to obtain an optimized screen recording instruction; obtaining the sensing time of the two-ticket generation operation detected by the front-end page; determining the operating habits of the power system based on the time period in which the sensing time is located, wherein the operating habits refer to the operating state of the power system in different time periods; and optimizing the information acquisition instruction based on the operating habits to obtain an optimized information acquisition instruction.

[0098] Optionally, in this embodiment, the computer-readable storage medium is configured to store program code for performing the following steps: decomposing screen video data to obtain multiple video frame images; identifying the text direction of the text in each video frame image to obtain an identification result, wherein the text direction refers to the direction of the center line of the text; determining the video frame image as a target video frame image when the identification result indicates that there is no offset between the text direction and the specific direction; correcting the video frame image until the text direction of the text in the video frame image is the same as the specific direction, wherein the specific direction refers to the vertical direction; and extracting text from each target video frame image to obtain screen text information.

[0099] Optionally, in this embodiment, the computer-readable storage medium is configured to store program code for performing the following steps: acquiring multiple sets of sample data, wherein each set of sample data includes: sample input data, a sample work ticket and a sample operation ticket corresponding to the sample input data, and the sample input data includes: sample screen text information, sample work task information and the current running status of the sample; dividing the multiple sets of sample data to obtain sample training data and sample test data, wherein the sample training data is used to train the two-ticket generation model, and the sample test data is used to test the two-ticket generation model; and iteratively training the artificial intelligence content generation model using the multiple sets of sample training data. The process continues until the iteration termination condition is met. The currently trained AI-generated content model is then identified as the initial two-vote generation model. This AI-generated content model is a deep learning model that automatically generates various types of content based on input information. These various content types include at least one of the following: text and images. The iteration termination condition includes at least one of the following: the number of iterations reaches a preset number of iterations; the model output error is below an error threshold. The initial two-vote generation model is then validated using sample test data to obtain validation results. If the validation results indicate that the performance of the initial two-vote generation model meets the preset performance standard, the initial two-vote generation model is determined to be a two-vote generation model.

[0100] Optionally, in this embodiment, the computer-readable storage medium is configured to store program code for performing the following steps: before iteratively training the AI-generated content model using multiple sets of sample training data, the model parameters of the AI-generated content model are randomly initialized using an improved chaotic mapping, wherein the first expression of the improved chaotic mapping is: x i Let x represent the state value of the i-th generation population. i+1Let represent the state value of the next generation of the i-th generation population; during the iterative training of the AI-generated content model using multiple sets of sample training data, the learning rate of the AI-generated content model is optimized using the first formula, where the first formula is: GP represents the learning rate, and n represents the current iteration number. max This indicates the maximum number of iterations, and the learning rate is used to control the step size for updating model parameters.

[0101] Optionally, in this embodiment, the computer-readable storage medium is configured to store program code for performing the following steps: during the process of generating work tickets and operation tickets using the two-ticket generation model, the regularization parameter of the two-ticket generation model is dynamically adjusted using a second formula to adaptively optimize the two-ticket generation model, wherein the regularization parameter, the second formula, is: 'a' represents the regularization parameter, which determines the degree of penalty applied to the two-vote generation model.

[0102] Optionally, in this embodiment, the computer-readable storage medium is configured to store program code for performing the following steps: sending work orders and operation tickets to a front-end page for visual presentation to review the work orders and operation tickets; upon receiving two adjustment requests, parsing the two adjustment requests to obtain an adjustment strategy for adjusting the work orders and / or operation tickets, and receiving update information, wherein the two adjustment requests are adjustment requests generated by the adjustment component when the front-end page senses that the work orders and operation tickets have failed review, and the update information is information that needs to be presented on the work orders and / or operation tickets; adjusting the work orders and / or operation tickets according to the adjustment strategy based on the update information to obtain target work orders and / or target operation tickets.

[0103] According to another aspect of the present invention, a processor is also provided, which is used to run a program, wherein the program executes any of the above-described intelligent two-vote generation methods during runtime.

[0104] According to another aspect of the present invention, a computer program product is also provided, including computer instructions, which, when executed by a processor, perform an intelligent method for generating two votes as described above.

[0105] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0106] In the above embodiments of the present invention, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0107] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units can be a logical functional division, and in actual implementation, there may be other division methods. For instance, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling, direct coupling, or communication connection may be through some interfaces; the indirect coupling or communication connection between units or modules may be electrical or other forms.

[0108] The units described as separate components may or may not be physically separate. 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 units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0109] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0110] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0111] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for intelligently generating two-tickets, characterized in that, include: The system responds to the two-ticket generation operation applied to the front-end page to generate a screen recording instruction and an information retrieval instruction. The two tickets include a work ticket and an operation ticket. The work ticket is used to record maintenance tasks for power equipment, and the operation ticket is used to record operation tasks for performing operations on the power equipment. The screen recording component is controlled to perform a screen recording operation according to the screen recording instruction to obtain screen video data. The screen recording operation refers to controlling the screen recording component to record the operations sensed by the front-end page. Text extraction is performed on the screen video data to obtain screen text information from the screen video data; Based on the information acquisition instruction, the current operating status and work task information of the power system are acquired, wherein the work task information refers to the information of the work task that needs to be recorded in the work ticket and the operation ticket; A two-ticket generation model is used to process the screen text information, the task information, and the current running status to generate the work ticket and the operation ticket. The two-ticket generation model is trained using deep learning with multiple sets of training data. Each set of training data includes: sample input data, a sample work ticket, and a sample operation ticket corresponding to the sample input data. The sample input data includes: sample screen text information, sample task information, and sample current running status. After responding to the two-ticket generation operation applied to the front-end page to generate a screen recording instruction and an information retrieval instruction, the method further includes: obtaining the front-end page's sensing habits based on the sensing identifier of the two-ticket generation operation, wherein the sensing habits refer to the habits when the two-ticket generation operation is applied to the front-end page; optimizing the screen recording instruction based on the sensing habits to obtain an optimized screen recording instruction; obtaining the sensing time of the front-end page sensing the two-ticket generation operation; determining the power system's operating habits based on the time period of the sensing time, wherein the operating habits refer to the operating state of the power system in different time periods; and optimizing the information retrieval instruction based on the operating habits to obtain an optimized information retrieval instruction. Before processing the screen text information, the task information, and the current running status using the two-ticket generation model to generate the work ticket and the operation ticket, the method further includes: acquiring multiple sets of sample data, wherein each set of sample data includes: sample input data, a sample work ticket and a sample operation ticket corresponding to the sample input data, wherein the sample input data includes: sample screen text information, sample task information, and sample current running status; dividing the multiple sets of sample data to obtain sample training data and sample test data, wherein the sample training data is used to train the two-ticket generation model, and the sample test data is used to test the two-ticket generation model; and using the multiple sets of sample training data to generate artificial intelligence. The content model undergoes iterative training until an iteration termination condition is met. The currently trained AI-generated content model is then identified as the initial two-vote generation model. This AI-generated content model is a deep learning model that automatically generates various types of content based on input information. These various content types include at least one of the following: text and images. The iteration termination condition includes at least one of the following: the number of iterations reaches a preset number of iterations, and the model output error is below an error threshold. The initial two-vote generation model is then used to perform performance verification using the sample test data, and verification results are obtained. If the verification results indicate that the performance of the initial two-vote generation model meets a preset performance standard, the initial two-vote generation model is identified as the two-vote generation model. The method further includes: before iteratively training the AI-generated content model using multiple sets of sample training data, randomly initializing the model parameters of the AI-generated content model using an improved chaotic mapping, wherein the first expression of the improved chaotic mapping is: , This represents the state value of the i-th generation of the population. The state value represents the next generation population of the i-th generation population; during the iterative training of the AI-generated content model using multiple sets of sample training data, the learning rate of the AI-generated content model is optimized using a first formula, wherein the first formula is: GP represents the learning rate, and n represents the current iteration number. The maximum number of iterations is indicated, and the learning rate is used to control the step size for updating the model parameters.

2. The intelligent generation method for two tickets according to claim 1, characterized in that, Text extraction is performed on the screen video data to obtain screen text information from the screen video data, including: The screen video data is decomposed to obtain multiple video frame images; The text direction of the text in each of the video frame images is identified to obtain the identification result, wherein the text direction refers to the direction of the center line of the text; If the recognition result indicates that there is no offset between the text direction and the specific direction, the video frame image is determined to be the target video frame image; If the recognition result indicates an offset between the text direction and the specific direction, the video frame image is corrected until the text direction in the video frame image is the same as the specific direction. Then, the video frame image is determined to be the target video frame image, where the specific direction refers to the vertical direction. Text is extracted from each of the target video frame images to obtain the screen text information.

3. The intelligent generation method for two tickets according to claim 1, characterized in that, Also includes: During the process of generating the work order and the operation order using the two-ticket generation model, the regularization parameter of the two-ticket generation model is dynamically adjusted using a second formula to adaptively optimize the model. The regularization parameter, as defined in the second formula, is: , where 'a' represents the regularization parameter, which is used to determine the degree of penalty for the two-vote generation model.

4. The intelligent generation method for two tickets according to claim 1, characterized in that, After processing the screen text information, the task information, and the current running status using the two-ticket generation model to generate the work ticket and the operation ticket, the method further includes: The work order and the operation ticket are sent to the front-end page for visual presentation in order to review the work order and the operation ticket; Upon receiving two adjustment requests, the two adjustment requests are parsed to obtain an adjustment strategy for adjusting the work ticket and / or the operation ticket, and update information is received. The two adjustment requests are adjustment requests generated by the adjustment component when the front-end page senses that the work ticket and the operation ticket have failed the review. The update information is information that needs to be presented on the work ticket and / or the operation ticket. Based on the updated information, the work order and / or operation order are adjusted according to the adjustment strategy to obtain the target work order and / or target operation order.

5. An intelligent two-ticket generation device, characterized in that, include: The first generation unit is used to respond to the two ticket generation operations applied to the front-end page to generate screen recording instructions and information acquisition instructions. The two tickets include a work ticket and an operation ticket. The work ticket is used to record maintenance tasks for power equipment, and the operation ticket is used to record operation tasks for performing operations on the power equipment. The first acquisition unit is used to control the screen recording component to perform a screen recording operation according to the screen recording instruction to obtain screen video data, wherein the screen recording operation refers to controlling the screen recording component to record the operation sensed by the front-end page; The second acquisition unit is used to extract text from the screen video data to obtain screen text information in the screen video data. The third acquisition unit is used to acquire the current operating status and work task information of the power system based on the information acquisition instruction, wherein the work task information refers to the information of the work task that needs to be recorded in the work ticket and the operation ticket; The second generation unit is used to process the screen text information, the task information, and the current running status using a two-ticket generation model to generate the work ticket and the operation ticket. The two-ticket generation model is trained using deep learning with multiple sets of training data. Each set of training data includes: sample input data, a sample work ticket, and a sample operation ticket corresponding to the sample input data. The sample input data includes: sample screen text information, sample task information, and sample current running status. After responding to the two-ticket generation operation applied to the front-end page to generate a screen recording instruction and an information retrieval instruction, the method further includes: obtaining the front-end page's sensing habits based on the sensing identifier of the two-ticket generation operation, wherein the sensing habits refer to the habits when the two-ticket generation operation is applied to the front-end page; optimizing the screen recording instruction based on the sensing habits to obtain an optimized screen recording instruction; obtaining the sensing time of the front-end page sensing the two-ticket generation operation; determining the power system's operating habits based on the time period of the sensing time, wherein the operating habits refer to the operating state of the power system in different time periods; and optimizing the information retrieval instruction based on the operating habits to obtain an optimized information retrieval instruction. The intelligent two-ticket generation device further includes: an eighth acquisition unit, used to acquire multiple sets of sample data, wherein each set of sample data includes: sample input data, a sample work ticket and a sample operation ticket corresponding to the sample input data, wherein the sample input data includes: sample screen text information, sample work task information and sample current running status; a ninth acquisition unit, used to divide the multiple sets of sample data to obtain sample training data and sample test data, wherein the sample training data is used to train the two-ticket generation model and the sample test data is used to test the two-ticket generation model; and a second determination unit, used to iteratively train the artificial intelligence generated content model using the multiple sets of sample training data until the iteration reaches the final value. When the termination condition is met, the currently trained AI-generated content model is determined to be the initial two-vote generation model. The AI-generated content model is a deep learning model that automatically generates multiple types of content based on input information. These multiple types of content include at least one of the following: text and images. The iteration termination condition includes at least one of the following: the number of iterations reaches a preset number of iterations, and the model output error is lower than an error threshold. The tenth acquisition unit is used to perform performance verification on the initial two-vote generation model using the sample test data and obtain verification results. The third determination unit is used to determine the initial two-vote generation model as the two-vote generation model when the verification result indicates that the performance of the initial two-vote generation model meets a preset performance standard. The intelligent generation device for the two votes further includes: an initialization unit, used to randomly initialize the model parameters of the AI-generated content model using an improved chaotic mapping before iteratively training the AI-generated content model using multiple sets of sample training data, wherein the first expression of the improved chaotic mapping is: , This represents the state value of the i-th generation of the population. The state value represents the next generation population of the i-th generation population; the first optimization unit is used to optimize the learning rate of the AI-generated content model using a first formula during the iterative training of the AI-generated content model using multiple sets of sample training data, wherein the first formula is: GP represents the learning rate, and n represents the current iteration number. The maximum number of iterations is indicated, and the learning rate is used to control the step size for updating the model parameters.

6. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored program, wherein the program executes the intelligent two-vote generation method according to any one of claims 1 to 4.

7. A computer program product comprising computer instructions, characterized in that, When the computer instructions are executed by the processor, the intelligent two-vote generation method described in any one of claims 1 to 4 is performed.

Images