Method and device for automatically generating a financial analysis report based on a power scenario

Abstract

The application discloses a financial analysis report automatic generation method and device based on a power scene, and the method comprises the following steps: collecting and sorting power grid financial operation related data, wherein the power grid financial operation related data comprises customer provided data and online open source data; according to the specific task and data characteristics of the power grid financial operation, a corresponding large model is matched for fine tuning and training; the architecture of an AIGC demonstration platform is designed, and a development tool is used for platform development; specific application scenarios are designed, and the developed AIGC demonstration platform is used to realize the decision support of the scenarios, and the effect is verified; knowledge retrieval and intelligent question answering are carried out based on user demand, a report is formed, and information extraction is carried out. The technical problem that the AIGC demonstration platform cannot be constructed to be applied to the knowledge retrieval task in the prior art, so that the power industry knowledge cannot be quickly retrieved and accurately matched is solved.

Description

Technical Field

[0001] This invention relates to the field of automatic report generation technology, specifically to a method and apparatus for automatically generating financial analysis reports based on the power industry scenario. Background Technology

[0002] New power system construction needs: With the transformation of energy structure and the growth of power load, the power system faces a more complex operating environment and higher safety requirements, requiring more intelligent technical means to ensure the safe and stable operation of the system; Digital transformation and intelligent development trend: The booming development of the digital economy and the rapid development of artificial intelligence technology have brought new development opportunities to the power industry. Digital transformation and intelligent development have become an inevitable trend in the development of the power industry.

[0003] However, existing technologies cannot form an intelligent response technology system architecture for specific scenarios in the field of digital intelligent operation decision-making for power grid enterprises, nor can they propose theoretical methods for constructing an intelligent response technology system. As a result, data related to power grid financial operations cannot be effectively collected and organized, leading to a decrease in the efficiency of automatic report generation. Furthermore, the AIGC demonstration platform cannot achieve rapid retrieval and accurate matching of knowledge in the power industry.

[0004] To address the aforementioned technical shortcomings, a solution is proposed. Summary of the Invention

[0005] The purpose of this invention is to solve the problems mentioned above by proposing a method and apparatus for automatically generating financial analysis reports based on power scenarios, which aims to overcome or at least partially overcome the shortcomings of the prior art.

[0006] The objective of this invention can be achieved through the following technical solutions:

[0007] Firstly, this application provides a method for automatically generating financial analysis reports based on power scenarios, including:

[0008] During the data preparation phase, relevant data on power grid financial operations are collected and organized, including customer-provided data and open-source data from the internet.

[0009] During the model training phase, the corresponding large model is matched for fine-tuning and training based on the specific tasks and data characteristics of power grid financial operations.

[0010] During the platform setup phase, the architecture of the AIGC demonstration platform was designed, and development tools were used for platform development.

[0011] In the application verification phase, specific application scenarios are designed, and the developed AIGC demonstration platform is used to implement decision support for the scenarios and verify their effectiveness.

[0012] During the application implementation phase, knowledge retrieval and intelligent question answering are performed based on user needs, generating reports and extracting information.

[0013] Optionally, in the above-mentioned method for automatically generating financial analysis reports based on the power scenario, during the data preparation stage, data cleaning is used to mark and monitor outlier handling, missing value filling, and data standardization to ensure data quality and accuracy.

[0014] For certain application scenarios that require model training, data annotation is performed to provide a foundation for model training.

[0015] Optionally, in the above-mentioned method for automatically generating financial analysis reports based on the power scenario, a large model with time series analysis capabilities will be selected for the electricity price forecasting task; and a large model with regression analysis capabilities will be selected for the cost control task.

[0016] During the model training phase, model fine-tuning will utilize data related to power grid financial operations to adaptively adjust the matched large model, which is a pre-trained large model. Furthermore, during the model training process, the convergence of the model will be continuously monitored, and the model parameters will be adjusted based on the evaluation results to optimize model performance.

[0017] Optionally, in the above-mentioned method for automatically generating financial analysis reports based on power scenarios, the platform architecture will adopt a modular design during the platform construction phase, dividing the platform into a data input module, a model processing module, and a result output module, and determining the functions and interaction methods of each module; and the platform development will support the readability, maintainability, and scalability of the code to facilitate subsequent platform maintenance and upgrades.

[0018] Optionally, in the above-mentioned method for automatically generating financial analysis reports based on the power industry scenario, the AIGC demonstration platform implements the following specific scenarios during the application verification phase:

[0019] Based on the professional knowledge and characteristics of the financial operations field, we will build a private sample set for the financial operations field and design a high-quality prompting system that fits the needs of the scenario.

[0020] Based on the self-built AIGC demonstration platform, the selected technical route is used to promote the realization of intelligent response tasks in the field of financial operations;

[0021] By combining the RLHF-based approach to enhance expert knowledge participation and deeply optimize the quality of model responses, an intelligent question-answering system is constructed.

[0022] Optionally, in the above-mentioned method for automatically generating financial analysis reports based on power scenarios, application scenarios can be designed on the AIGC demonstration platform according to actual business needs.

[0023] Several evaluation indicators are established, including at least one of the following: prediction accuracy, cost reduction rate, and return on investment.

[0024] Several verification methods are used for application verification, including at least one of experimental comparison method, case analysis method and expert evaluation method, to ensure the objectivity and reliability of the evaluation results.

[0025] Optionally, in the above-mentioned method for automatically generating financial analysis reports based on power scenarios, during the application implementation phase,

[0026] Knowledge Retrieval: The AIGC demonstration platform is applied to knowledge retrieval tasks, and natural language processing technology is used to achieve rapid retrieval and accurate matching of knowledge in the power industry;

[0027] Intelligent Question Answering: The AIGC demonstration platform is applied to an intelligent question answering system. Through semantic understanding, it can automatically identify and answer user questions, providing users with efficient and convenient consultation services.

[0028] Report Generation: The AIGC demonstration platform is applied to report generation tasks, automatically generating financial analysis reports and risk assessment reports through natural language generation technology;

[0029] Information Extraction: The AIGC demonstration platform is applied to information extraction tasks. Through named entity recognition and relation extraction, key information is extracted from massive amounts of data to provide users with data analysis and decision support.

[0030] Secondly, this application also provides an automatic financial analysis report generation device based on the power scenario, the device being as follows:

[0031] The data preparation unit is used to collect and organize data related to power grid financial operations, including customer-provided data and open-source data from the internet.

[0032] The model training unit is used to fine-tune and train the corresponding large model according to the specific tasks and data characteristics of power grid financial operations.

[0033] The platform construction unit was used to design the architecture of the AIGC demonstration platform and to develop the platform using development tools.

[0034] The application verification unit is used to design specific application scenarios, implement decision support for the scenarios using the developed AIGC demonstration platform, and verify its effectiveness.

[0035] The application implementation unit is used to perform knowledge retrieval and intelligent question answering based on user needs, generate reports, and extract information.

[0036] Thirdly, this application also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor executes the computer program to implement the steps of any of the above-mentioned methods for automatically generating financial analysis reports based on power scenarios.

[0037] Fourthly, this application also provides a computer-readable storage medium storing a computer program, characterized in that the computer program, when instructed by a processor, implements the steps of any of the above-mentioned methods for automatically generating financial analysis reports based on power scenarios.

[0038] Compared with the prior art, the beneficial effects of the present invention are:

[0039] 1. In this invention, platform performance evaluation: The performance of the AIGC demonstration platform is evaluated, including indicators such as model accuracy, efficiency, and scalability, in order to verify the platform's practicality and effectiveness.

[0040] Economic benefit analysis: Conduct an economic benefit analysis on the application effect of the AIGC demonstration platform, such as calculating the platform's input-output ratio and cost savings, in order to evaluate the platform's economic value.

[0041] Social benefit analysis: Conduct a social benefit analysis on the application effect of the AIGC demonstration platform, such as calculating the reduction of carbon emissions and energy savings of the platform, in order to assess the social value of the platform.

[0042] Through the above expansion, the research objectives and content will be clearer and more specific, the research methods and technical routes will be more complete, and the research results and presentations will be more comprehensive, thus better reflecting the depth and breadth of the research.

[0043] 2. This invention establishes an intelligent response technology system architecture for specific scenarios in the field of digital intelligent operation decision-making for power grid enterprises, proposes a theoretical method for constructing the intelligent response technology system, provides a technical route for the intelligent response layer for the integrated application of generative AI technology in the financial operation decision-making field, and lays a theoretical foundation for providing scenario application services for the establishment of an AI technology platform; it also establishes an application paradigm for intelligent response technology for specific scenarios in the field of digital intelligent operation decision-making for power grid enterprises, and provides a key intelligent response application implementation path for specific scenarios in the operation decision-making field.

[0044] 3. In this invention, the model comparison analysis involves comparing and analyzing different types of large models, such as language models and knowledge graph models, to evaluate their applicability and advantages / disadvantages in power grid financial operation decision-making; model fine-tuning involves fine-tuning the large models according to the specific tasks and data characteristics of power grid financial operations to improve their performance and accuracy in this field; and model compression involves researching model compression techniques to reduce model size, lower computational resource requirements, and improve platform efficiency and scalability.

[0045] Modular design: The platform is divided into data input modules, model processing modules, result output modules, etc., to achieve modular design and improve the platform's flexibility and maintainability; Scalability design: Distributed architecture and microservice architecture technologies are adopted to improve the platform's scalability and enable it to adapt to the future development and changes in the power industry; Security design: Data encryption, access control and other technologies are used to ensure the platform's security and privacy protection.

[0046] 4. Knowledge Retrieval: The AIGC demonstration platform has been successfully applied to knowledge retrieval tasks. Through natural language processing technology, it enables rapid retrieval and accurate matching of knowledge in the power industry, providing users with convenient knowledge acquisition channels and improving work efficiency.

[0047] Intelligent Question Answering: The AIGC demonstration platform has been successfully applied to an intelligent question answering system. Through semantic understanding, it can automatically identify and answer user questions, providing users with efficient and convenient consultation services and reducing the cost of manual customer service.

[0048] Report Generation: The AIGC demonstration platform has been successfully applied to report generation tasks. Through natural language generation technology, it automatically generates financial analysis reports, risk assessment reports, etc., improving the efficiency and accuracy of report generation and reducing the workload of manual report writing.

[0049] Information Extraction: The AIGC demonstration platform was successfully applied to information extraction tasks. Through technologies such as named entity recognition and relation extraction, it extracted key information from massive amounts of data, such as equipment failure information and market information, to provide users with data analysis and decision support. Attached Figure Description

[0050] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings.

[0051] Figure 1 This is a flowchart illustrating an embodiment of the present invention of an automatic generation method for financial analysis reports based on power scenarios.

[0052] Figure 2 This is a statistical chart showing the performance improvement of a large model after training, according to one embodiment of the present invention.

[0053] Figure 3 This is a schematic diagram of a modular design according to an embodiment of the present invention;

[0054] Figure 4 This is a diagram illustrating the improved effect of one embodiment of the present invention;

[0055] Figure 5 This is a technical roadmap for intelligent question answering according to an embodiment of the present invention;

[0056] Figure 6 This is a schematic diagram of the structure of an automatic financial analysis report generation device based on a power scenario, according to an embodiment of the present invention.

[0057] Figure 7 A schematic diagram of the structure of a computer device according to an embodiment of the present invention is shown. Detailed Implementation

[0058] 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 are within the scope of protection of the present invention.

[0059] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of the invention. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0060] This research aims to explore the architecture of a large-scale AIGC demonstration platform and its application in power grid financial operation decision-making, thereby promoting the intelligent and efficient development of the power industry. The research employs multiple methods, including customer scenario research, model selection and optimization, experimental verification, case studies, and expert interviews, to ensure comprehensiveness and depth. The research technical roadmap covers data preparation, model training, platform construction, and application verification. Please refer to [link / reference]. Figures 1-5 As shown in one embodiment provided in this application, for details please refer to... Figure 1 The method for automatically generating financial analysis reports based on the power industry scenario is as follows:

[0061] First, the data preparation stage

[0062] Collect and organize data related to the financial operations of the power grid, including customer-provided data and open-source data from the internet.

[0063] During the data preparation phase, data cleaning can be performed to handle outliers, imputate missing values, and standardize data. This process is then monitored to ensure data quality and accuracy. Furthermore, for application scenarios requiring model training, data annotation is conducted to provide a foundation for model training.

[0064] Second, the model training phase

[0065] Based on the specific tasks and data characteristics of power grid financial operations, a corresponding large model is matched for fine-tuning and training.

[0066] For example, for electricity price forecasting tasks, a large model with time series analysis capabilities will be selected; for cost control tasks, a large model with regression analysis capabilities will be selected.

[0067] Model fine-tuning refers to adaptively adjusting a matched large model using data related to power grid financial operations to improve its performance and accuracy in this field. In some embodiments of this application, it is recommended to use a pre-trained large model. During model training, the convergence of the large model is continuously monitored, and model parameters are adjusted based on evaluation results to optimize model performance.

[0068] from Figure 2 It can be seen that the large models trained using the method described in this application show improved performance across different tasks.

[0069] Third, the platform construction phase

[0070] The architecture of the AIGC demonstration platform was designed, and development tools such as Python were used for platform development. The previously trained large model was built into the AIGC demonstration platform for subsequent use.

[0071] The platform architecture will adopt a modular design, dividing the platform into data input modules, model processing modules, result output modules, etc., and defining the functions and interaction methods of each module. Furthermore, the platform development will focus on code readability, maintainability, and scalability to facilitate subsequent platform maintenance and upgrades.

[0072] from Figure 3As can be seen, the AIGC demonstration platform of this application adopts a modular design concept, dividing the platform into data input modules, model processing modules, and result output modules, thereby improving the platform's flexibility and maintainability; scalability design: further, it employs technologies such as distributed architecture and microservice architecture to enhance the platform's scalability, enabling it to adapt to the future development and changes in the power industry; security design: it uses technologies such as data encryption and access control to ensure the platform's security and privacy protection.

[0073] Fourth, application verification phase

[0074] Design specific application scenarios, such as knowledge Q&A, policy retrieval, and document collaboration, and use the developed demonstration platform to implement decision support for these scenarios and verify their effectiveness.

[0075] Furthermore, we will study a research and development roadmap for large-scale, highly adaptable intelligent responses, so that the model can provide multiple inference chains for comparison when facing complex inference tasks, thereby improving the adaptability of downstream tasks.

[0076] Meanwhile, based on the specific scenario requirements in the field of intelligent operation decision-making, we will analyze the development trend and application prospects of intelligent response technology in specific scenarios of digital intelligent operation decision-making for power grid enterprises; based on the professional knowledge characteristics of the financial operation field, we will study the real-world needs of common question-and-answer scenarios in the financial operation field, construct a private sample set in the financial operation field, and design a high-quality prompting project that fits the scenario requirements.

[0077] Based on a self-built AIGC demonstration platform, this project leverages the selected technical approach to drive the implementation of intelligent question-answering tasks in the financial operations field. Simultaneously, by combining this with RLHF-based methods to enhance expert knowledge participation and deeply optimize model response quality, it constructs an efficient, high-quality, and high-value-added intelligent question-answering system.

[0078] Please refer to Figure 5 , Figure 5 This document illustrates a technical roadmap for intelligent question answering based on an embodiment of the present invention. In this embodiment, the AIGC demonstration platform is successfully applied in an intelligent question answering system. First, for local documents, data preprocessing, intelligent slicing, and content vectorization are performed sequentially, and the processed data is stored in a vector database. When a user asks a question, the system first vectorizes the question and then performs vector similarity matching in the vector database. Next, through steps such as model hints, hint engineering, and reordering, the model accurately understands the question and provides an answer.

[0079] This technological approach, leveraging semantic understanding, enables the automatic identification and response to user questions, significantly improving the efficiency and convenience of consultation services. It allows users to quickly obtain the information they need while substantially reducing the cost of human customer service, propelling intelligent question-and-answer systems to a higher level. Application scenario design will fully consider actual business needs and establish reasonable evaluation metrics, such as prediction accuracy, cost reduction rate, and return on investment.

[0080] Application validation will employ a variety of methods, such as experimental comparison, case analysis, and expert evaluation, to ensure the objectivity and reliability of the evaluation results.

[0081] In the research process, this invention focuses on the following technical issues:

[0082] Model evaluation: Evaluate the performance of the trained model, such as accuracy, recall, F1 score, etc., and adjust the model parameters based on the evaluation results to optimize model performance.

[0083] Model compression: Research model compression techniques, such as model pruning and model quantization, to reduce model size, lower computing resource requirements, and improve platform efficiency and scalability.

[0084] Model interpretability: Researching model interpretability techniques, such as attention mechanisms and gradient-weighted class activation mapping, to explain the model's decision-making process and improve the model's credibility.

[0085] Platform security: Research platform security technologies, such as data encryption and access control, to ensure platform security and privacy protection.

[0086] Platform Deployment: Deploy the AIGC demonstration platform into the actual production environment and perform operation and management.

[0087] Platform monitoring: Monitor the platform's operational status, such as resource utilization and model performance, and promptly identify and resolve any issues that arise.

[0088] Platform maintenance: Regularly maintain the platform, such as updating models and fixing vulnerabilities, to ensure the platform's stability and security.

[0089] Fifth, application implementation stage

[0090] Knowledge Retrieval: The AIGC demonstration platform has been successfully applied to knowledge retrieval tasks. Through natural language processing technology, it enables rapid retrieval and accurate matching of knowledge in the power industry, providing users with a convenient channel for knowledge acquisition and improving work efficiency.

[0091] Intelligent Question Answering: The AIGC demonstration platform has been successfully applied to an intelligent question answering system. Through semantic understanding, it can automatically identify and answer user questions, providing users with efficient and convenient consultation services and reducing the cost of manual customer service.

[0092] Report Generation: The AIGC demonstration platform has been successfully applied to report generation tasks. Through natural language generation technology, it automatically generates financial analysis reports, risk assessment reports, etc., improving the efficiency and accuracy of report generation and reducing the workload of manual report writing. Please refer to [link / reference]. Figure 4 , Figure 4 This is a diagram illustrating the improved effect of one embodiment of the present invention. Figure 4 As can be seen, after being processed by the AIGC demonstration platform, the report is more reasonable in terms of form, format, and language. This shows that the AIGC platform has been successfully applied to report generation tasks. Through natural language generation technology, it automatically generates financial analysis reports, risk assessment reports, etc., improving the efficiency and accuracy of report generation and reducing the workload of manual report writing.

[0093] Information Extraction: The AIGC demonstration platform was successfully applied to information extraction tasks. Through technologies such as named entity recognition and relation extraction, it extracted key information from massive amounts of data, such as equipment failure information and market information, to provide users with data analysis and decision support.

[0094] In the application of this invention, the data preparation stage involves collecting and organizing relevant data on power grid financial operations, including customer-provided data and data collected from open-source websites. The model training stage involves selecting a suitable large-scale model for fine-tuning and training based on the specific tasks and data characteristics of power grid financial operations. The platform construction stage involves designing the architecture of the AIGC demonstration platform and developing the platform using development tools. The application verification stage involves designing specific application scenarios and using the developed demonstration platform to implement decision support for these scenarios, thus verifying their effectiveness.

[0095] Figure 6 This is a schematic diagram of the structure of an automatic financial analysis report generation device based on a power scenario, according to an embodiment of the present invention. Figure 6 It can be seen that the automatic financial analysis report generation device 600 based on the power scenario includes:

[0096] Data preparation unit 610 is used to collect and organize data related to power grid financial operations, including customer-provided data and online open-source data.

[0097] Model training unit 620 is used to fine-tune and train the corresponding large model according to the specific tasks and data characteristics of power grid financial operations.

[0098] Platform building unit 630 was used to design the architecture of the AIGC demonstration platform and to develop the platform using development tools.

[0099] Application verification unit 640 is used to design specific application scenarios, implement decision support for the scenarios using the developed AIGC demonstration platform, and verify its effectiveness.

[0100] Application implementation unit 650 is used for knowledge retrieval and intelligent question answering based on user needs, generating reports and extracting information.

[0101] It should be noted that the aforementioned automatic financial analysis report generation device 600 based on the power scenario can realize the aforementioned automatic financial analysis report generation method based on the power scenario, which will not be elaborated further.

[0102] Figure 7 This application shows a schematic diagram of the structure of a computer device according to one embodiment of the present application. Figure 7 As shown, the internal structure of this computer device may include a processor, memory, network interface, and database connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile and / or volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and database. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The network interface is used for communication with external devices via a network connection. When the computer program is executed by the processor, it implements the functions or steps of a method for automatically generating financial analysis reports based on power scenarios.

[0103] In one embodiment, the computer device provided in this application includes a memory and a processor. The memory stores a database and a computer program that can run on the processor. When the processor executes the computer program, it implements the steps of an automatic generation method for financial analysis reports based on power scenarios.

[0104] In one embodiment, a computer-readable storage medium is also provided, on which a computer program is stored, which, when executed by a processor, implements the steps of a method for automatically generating financial analysis reports based on power scenarios.

[0105] It should be noted that the functions or steps that can be implemented by the computer device or computer-readable storage medium described above can be referred to the relevant descriptions in the foregoing method embodiments. To avoid repetition, they will not be described one by one here.

[0106] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0107] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0108] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

[0109] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to any specific implementation. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A method for automatically generating financial analysis reports based on power scenarios, characterized in that: include: During the data preparation phase, relevant data on power grid financial operations are collected and organized, including customer-provided data and open-source data from the internet. During the model training phase, the corresponding large model is matched for fine-tuning and training based on the specific tasks and data characteristics of power grid financial operations. During the platform setup phase, the architecture of the AIGC demonstration platform was designed, and development tools were used for platform development. In the application verification phase, specific application scenarios are designed, and the developed AIGC demonstration platform is used to implement decision support for the scenarios and verify their effectiveness. During the application implementation phase, knowledge retrieval and intelligent question answering are performed based on user needs, resulting in reports and information extraction. For the electricity price forecasting task, a large model with time series analysis capabilities will be selected; for the cost control task, a large model with regression analysis capabilities will be selected. During the model training phase, model fine-tuning will utilize data related to power grid financial operations to adaptively adjust the matched large model, which is a pre-trained large model. Furthermore, during the model training process, the convergence of the model will be continuously monitored, and the model parameters will be adjusted based on the evaluation results to optimize model performance. During the platform construction phase, the platform architecture will adopt a modular design, dividing the platform into a data input module, a model processing module, and a result output module, and defining the functions and interaction methods of each module; in addition, the platform development will support code readability, maintainability, and scalability to facilitate subsequent platform maintenance and upgrades; In the application verification phase, the AIGC demonstration platform implements the following specific scenarios: Based on the professional knowledge and characteristics of the financial operations field, we will build a private sample set for the financial operations field and design a high-quality prompting system that fits the needs of the scenario. Based on the self-built AIGC demonstration platform, the selected technical route is used to promote the realization of intelligent response tasks in the field of financial operations; By combining the RLHF-based method to enhance expert knowledge participation and deeply optimize the model's response quality, an intelligent question-answering system can be constructed. In the application implementation phase, Knowledge Retrieval: The AIGC demonstration platform is applied to knowledge retrieval tasks, and natural language processing technology is used to achieve rapid retrieval and accurate matching of knowledge in the power industry; Intelligent Question Answering: The AIGC demonstration platform is applied to an intelligent question answering system. Through semantic understanding, it can automatically identify and answer user questions, providing users with efficient and convenient consultation services. Report Generation: The AIGC demonstration platform is applied to report generation tasks, automatically generating financial analysis reports and risk assessment reports through natural language generation technology; Information Extraction: The AIGC demonstration platform is applied to information extraction tasks. Through named entity recognition and relation extraction, key information is extracted from massive amounts of data to provide users with data analysis and decision support.

2. The method for automatically generating financial analysis reports based on power scenarios according to claim 1, characterized in that, During the data preparation phase, data cleaning is employed to mark and monitor outlier handling, missing value imputation, and data standardization work to ensure data quality and accuracy. For certain application scenarios that require model training, data annotation is performed to provide a foundation for model training.

3. The method for automatically generating financial analysis reports based on power scenarios according to claim 1, characterized in that, On the AIGC demonstration platform, application scenarios are designed based on actual business needs; Several evaluation indicators are established, including at least one of the following: prediction accuracy, cost reduction rate, and return on investment. Several verification methods are used for application verification, including at least one of experimental comparison method, case analysis method and expert evaluation method, to ensure the objectivity and reliability of the evaluation results.

4. A device for automatically generating financial analysis reports based on power scenarios, characterized in that, The device is as follows: The data preparation unit is used to collect and organize data related to power grid financial operations, including customer-provided data and open-source data from the internet. The model training unit is used to fine-tune and train the corresponding large model according to the specific tasks and data characteristics of power grid financial operations. The platform construction unit was used to design the architecture of the AIGC demonstration platform and to develop the platform using development tools. The application verification unit is used to design specific application scenarios, implement decision support for the scenarios using the developed AIGC demonstration platform, and verify its effectiveness. The application implementation unit is used to perform knowledge retrieval and intelligent question answering based on user needs, generate reports, and extract information. The model training unit is used for: For electricity price forecasting, a large model with time series analysis capabilities will be selected; for cost control, a large model with regression analysis capabilities will be selected. Model fine-tuning will utilize data related to power grid financial operations to adaptively adjust the matched large model, which is a pre-trained large model; and during the model training process, the convergence of the model will be continuously monitored, and the model parameters will be adjusted based on the evaluation results to optimize the model performance. The platform building unit is used for: The platform architecture adopts a modular design, dividing the platform into a data input module, a model processing module, and a result output module, and defining the functions and interaction methods of each module; the platform development supports code readability, maintainability, and scalability to facilitate subsequent platform maintenance and upgrades; The application verification unit is used for: The AIGC demonstration platform will enable the following specific scenarios: Based on the professional knowledge and characteristics of the financial operations field, we will build a private sample set for the financial operations field and design a high-quality prompting system that fits the needs of the scenario. Based on the self-built AIGC demonstration platform, the selected technical route is used to promote the realization of intelligent response tasks in the field of financial operations; By combining the RLHF-based method to enhance expert knowledge participation and deeply optimize the model's response quality, an intelligent question-answering system can be constructed. The application implementation unit is used for: Knowledge Retrieval: The AIGC demonstration platform is applied to knowledge retrieval tasks, and natural language processing technology is used to achieve rapid retrieval and accurate matching of knowledge in the power industry; Intelligent Question Answering: The AIGC demonstration platform is applied to an intelligent question answering system. Through semantic understanding, it can automatically identify and answer user questions, providing users with efficient and convenient consultation services. Report Generation: The AIGC demonstration platform is applied to report generation tasks, automatically generating financial analysis reports and risk assessment reports through natural language generation technology; Information Extraction: The AIGC demonstration platform is applied to information extraction tasks. Through named entity recognition and relation extraction, key information is extracted from massive amounts of data to provide users with data analysis and decision support.

5. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, The processor executes a computer program to implement the steps of the automatic generation method for financial analysis reports based on power scenarios as described in any one of claims 1 to 3.

6. A computer-readable storage medium storing a computer program, characterized in that, The computer program, when instructed by the processor, implements the steps of the automatic generation method for financial analysis reports based on power scenarios as described in any one of claims 1 to 3.

Images