Financial index data processing system and method based on artificial intelligence

Abstract

The invention relates to a financial index data processing system and method, and relates to the technical field of electric power. The system comprises a financial index data analysis platform which specifically comprises a multi-source financial data integration module for generating a financial index knowledge graph; the artificial intelligence model analysis module is used for generating a preset model and predicting an electricity benefit index corresponding to the current electricity price fluctuation index by adopting the preset model so as to obtain a current financial index; the intelligent reasoning engine module is used for extracting a plurality of target entities of entity characteristics in the current electricity benefit index and the current electricity price fluctuation index; reasoning the plurality of target entities according to associated paths in the financial index knowledge graph to obtain a plurality of first target paths; and the financial report generation module is used for generating a first financial analysis report according to the current financial index and the plurality of first target paths.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to the technical field of electric power financial data processing, in particular to a financial index data processing system and method based on artificial intelligence. BACKGROUND

[0002] With the continuous development of the power system, the collection and analysis of financial index data become increasingly important. However, the current power industry is facing the challenge of integrating and analyzing massive data in business management. The related technology relies on manual processing, which is low in efficiency and prone to errors. At present, a financial index data analysis platform based on simple static rules is gradually adopted to analyze financial index data. However, this financial index data analysis platform is difficult to adapt to complex and variable business scenarios, lacks the mining and analysis of the internal relations between various financial indexes, and thus leads to the generation of financial report data with single dimension and poor reasoning logic by the financial index data analysis platform. When financial users perform account supervision based on the above financial reports, the error rate is high and the supervision efficiency is low, resulting in poor user experience. SUMMARY

[0003] The present disclosure provides a financial index data processing system and method to at least solve the technical problems of high error rate and low supervision efficiency of the financial index data analysis platform in the related art. The technical solutions of the present disclosure are as follows:

[0004] According to a first aspect of the present invention, a financial indicator data processing system is provided. The system includes: a financial indicator data analysis platform; the financial indicator data analysis platform includes a multi-source financial data integration module, an artificial intelligence model analysis module, an intelligent inference engine module, and a financial report generation module. The multi-source financial data integration module is configured to: generate a financial indicator knowledge graph based on multi-dimensional financial indicator data from multiple databases; the financial indicator knowledge graph includes entities representing the characteristics of financial indicators, financial events representing the financial indicators, and the association paths between entities and financial events; the financial indicators include electricity price fluctuation indicators and electricity efficiency indicators; the artificial intelligence model analysis module is configured to: perform reinforcement training using historical electricity price fluctuation indicators, historical financial events, and historical electricity efficiency indicators as training samples to obtain a preset model. The system employs a preset model to predict the electricity benefit index corresponding to the current electricity price fluctuation index, thereby obtaining the current electricity benefit index. The preset model is a reinforcement learning model used to predict the electricity benefit index under the influence of corresponding financial events, where the corresponding financial events are those with a higher probability of occurrence than the preset probability under the corresponding electricity price fluctuation index. The intelligent reasoning engine module is configured to: extract multiple target entities with entity features from the current electricity benefit index and the current electricity price fluctuation index; and reason about the multiple target entities based on the association paths in the financial indicator knowledge graph to obtain multiple first target paths, which include multiple first target financial events corresponding to the multiple target entities. The financial report generation module is configured to generate a first financial analysis report based on the current financial index and the multiple first target paths. The first financial analysis report includes the correlation trend between the current electricity benefit index, the current electricity price fluctuation index, and the first target financial events.

[0005] The aforementioned multi-source financial data integration module, artificial intelligence model analysis module, intelligent inference engine module, and financial report generation module are interconnected via communication. The multi-source financial data integration module connects to multiple databases.

[0006] The preset probability is the probability of being located before the preset position, or the highest probability.

[0007] The system also includes: a financial analysis platform monitoring unit and a financial monitoring terminal.

[0008] The aforementioned databases include a front-end business management database, a financial database, and an electricity price database.

[0009] The electricity price fluctuation indicators include one or more of the following: statistical fluctuation indicators (e.g., price volatility rate), frequency and persistence indicators (price fluctuation frequency, high / low price fluctuation duration), spikes (spike factor representing the ratio of peak price to average price, value at risk, conditional value at risk), and risk indicators, and electricity price prediction error indicators (deviation of predicted electricity price from actual electricity price).

[0010] The electricity efficiency indicators include one or more of the following: efficiency indicators (unit output / production electricity consumption, unit area electricity consumption, equipment / system efficiency, power factor), economic indicators (unit electricity cost, electricity cost proportion, energy saving benefit, demand response benefit, investment payback period / net present value / internal rate of return), load characteristic indicators, i.e., indirectly reflecting benefits (load factor, peak-valley electricity consumption ratio), management and behavior indicators (demand response participation / potential, energy saving target completion rate), and environmental benefit indicators (unit output / production carbon emission, renewable energy consumption proportion).

[0011] The entities include physical entities (e.g., device entities such as transformer stations / units / lines, route entities), abstract entities (e.g., market entities such as electricity price types / policy documents, policy entities), and indicator entities (e.g., indicator parameters such as network loss rate / renewable energy penetration rate / load fluctuation coefficient).

[0012] In an implementation manner, the intelligent reasoning engine module is further configured to: display the plurality of first target paths on the front-end interface, so that the user account selects a second target path from the plurality of first target paths displayed on the front-end interface; receive the second target path returned by the front-end interface; according to the financial indicator knowledge graph, output the path reasoning process of the second target path in a visual form according to a reasoning chain; and the financial report generation module is further configured to: generate a second financial analysis report according to the second target path; and the second financial analysis report includes the correlation trend between the current electricity efficiency indicators, the current electricity price fluctuation indicators, and the second target financial event in the second target path.

[0013] In another implementation, the financial indicator data analysis platform further comprises an interactive response interface module in communication connection with other modules; the interactive response interface module is configured to perform an auditing operation on the first financial analysis report and / or the second financial analysis report; send the financial analysis report after the auditing operation to a target user terminal, so that a user of the target user terminal revises or confirms the financial analysis report after the auditing operation on the target user terminal, and sends the financial analysis report after the revision or confirmation operation to the financial report generation module; the financial report generation module is further configured to generate a target financial analysis report according to the financial analysis report after the revision or confirmation operation; wherein the auditing operation on the first financial analysis report and / or the second financial analysis report includes one or more of the following: format checking on the text format and / or picture format in the first financial analysis report and / or the second financial analysis report; sensitive word filtering and replacing operation on the preset sensitive words included in the first financial analysis report and / or the second financial analysis report; and marking the revised content in the first financial analysis report and / or the second financial analysis report that does not comply with the preset logic condition.

[0014] In another implementation, the system further comprises a financial analysis platform supervision unit in communication connection with the financial indicator data analysis platform; the financial analysis platform supervision unit is used to determine the running stability of the financial indicator data analysis platform; the financial analysis platform supervision unit comprises a security and stability evaluation module, a database docking monitoring and analysis module, a revision tracking analysis module and a request processing detection module.

[0015] The security and stability evaluation module is configured to detect target running data in the financial indicator data analysis platform according to a preset detection frequency, fit the target running data to obtain a target fitting curve, and determine a stability running result according to the curve trend of the target fitting curve, wherein the stability running result comprises a security and stability normal signal or a security and stability alarm signal.

[0016] The database docking monitoring and analysis module is configured to, when the security and stability normal signal is received, monitor and analyze the docking status of the multi-source financial data integration module and each database in the detection period, and generate a docking qualified signal or a docking abnormal signal through analysis;

[0017] The revision tracking analysis module is configured to, when the docking qualified signal is received, analyze the interaction efficiency of the user on the financial indicator data analysis platform in the detection period to generate an interaction abnormal signal or an interaction normal signal;

[0018] The request processing detection module is configured to, when the interaction normal signal is received, analyze the query request processing load condition of the financial indicator data analysis platform in the detection period to generate a processing high load signal or a processing low load signal.

[0019] In another implementation, the security and stability evaluation module is further configured to obtain a total number of times of network attacks on the financial indicator data analysis platform in a detection period, and mark the total number of times of network attacks as a network feature value; and obtain a total number of times of crashes of the financial indicator data analysis platform in the detection period, and mark the total number of times of crashes as a paralysis feature value; if the network feature value is greater than or equal to a first preset threshold value and / or the paralysis feature value is greater than or equal to a second preset threshold value, a security and stability alarm signal is generated; if the network feature value is less than the first preset threshold value and the paralysis feature value is less than the second preset threshold value, the network feature value, the paralysis feature value, the intervention time value, and the intervention risk value are weighted and summed to obtain a security and stability evaluation coefficient; if the security and stability evaluation coefficient is greater than or equal to a preset security and stability evaluation coefficient threshold value, the security and stability alarm signal is generated; and if the security and stability evaluation coefficient is less than the preset security and stability evaluation coefficient threshold value, a security and stability normal signal is generated.

[0020] In another implementation, the database connection monitoring and analysis module is specifically configured to, if the total disconnection time value or the total disconnection frequency value of any database in the plurality of databases exceeds the corresponding preset disconnection threshold value, mark the any database as a stable and different database; if there is a stable and different database in the plurality of databases in the detection period, generate a connection exception signal; if there is no stable and different database in the plurality of databases in the detection period, weighted sum calculation is performed on the disconnection time measurement value and the disconnection frequency measurement value of the stable and different database to obtain a disconnection monitoring evaluation value, and if the disconnection monitoring evaluation value is greater than or equal to a preset disconnection monitoring evaluation threshold value, the connection exception signal is generated; and if the disconnection monitoring evaluation value is less than the preset disconnection monitoring evaluation threshold value, a connection qualified signal is generated.

[0021] In another implementation, the revision tracking analysis module is specifically configured to obtain an interaction number of unqualified time lengths in the detection period, and perform ratio calculation on the interaction number and the total number of analysis time lengths to obtain an unqualified time frequency value; mark a ratio of each interaction analysis time length to a corresponding preset analysis time length threshold value as an analysis time length; and perform mean value calculation on all analysis time lengths in the detection period to obtain an analysis evaluation value; if the unqualified time frequency value is greater than or equal to a time frequency threshold value, and / or the analysis evaluation value is greater than or equal to a preset evaluation threshold value, a revision exception signal is generated; otherwise, a revision normal signal is generated.

[0022] In another implementation, the request processing detection module is specifically configured to obtain a total number of query requests received by the financial indicator data analysis platform in a detection period, and mark the total number of query requests as a load detection value; if the load detection value is greater than or equal to a preset load detection threshold, a high-load processing signal is generated; if the load detection value is less than the preset load detection threshold, a load evaluation value is obtained by performing weighted summation calculation on the load detection value, the delay detection value, and the waiting condition value; if the load evaluation value is greater than or equal to a preset load evaluation threshold, the high-load processing signal is generated; and if the load evaluation value is less than the preset load evaluation threshold, a low-load processing signal is generated.

[0023] According to a second aspect of an embodiment of the present application, a financial indicator data processing method is provided, applied to a financial indicator data processing system, the system including a financial indicator knowledge graph and a preset model; the financial indicator knowledge graph including entities representing characteristics of financial indicators, financial events affecting the financial indicators, and associated paths between the entities and the financial events; the financial indicators including an electricity price fluctuation indicator and an electricity benefit indicator; the preset model being a reinforcement learning model, used to predict the electricity benefit indicator under the influence of the corresponding financial event of the electricity price fluctuation indicator, the corresponding financial event being a financial event with a probability of occurrence higher than a preset probability under the corresponding electricity price fluctuation indicator; the financial indicator knowledge graph being generated according to multi-dimensional financial indicator data in multiple databases; the method including: obtaining a current electricity price fluctuation indicator, and inputting the current electricity price fluctuation indicator into the preset model to obtain a current electricity benefit indicator; extracting a plurality of target entities of entity characteristics in the current electricity benefit indicator and the current electricity price fluctuation indicator; reasoning the plurality of target entities according to the associated paths in the financial indicator knowledge graph to obtain a plurality of first target paths, the plurality of first target paths including a plurality of first target financial events corresponding to the plurality of target entities; generating a first financial analysis report according to the current financial indicator and the plurality of first target paths; the first financial analysis report including an associated trend between the current electricity benefit indicator, the current electricity price fluctuation indicator, and the first target financial events.

[0024] The plurality of first target paths include the plurality of target entities and a plurality of first target financial events corresponding to the plurality of target entities.

[0025] In an implementation, the method further includes determining the stability of the financial indicator data analysis platform, obtaining multi-dimensional financial indicator data, obtaining training samples, and obtaining the current electricity price fluctuation indicator when the financial indicator data analysis platform is determined to be in a stable running state; the training samples including historical electricity price fluctuation indicators, historical financial events, and historical electricity benefit indicators.

[0026] The determination of the stable operation state of the financial indicator data analysis platform can be achieved by one or more of the following modules: a safety and stability evaluation module, a database docking monitoring and analysis module, a revision tracking analysis module, and a request processing detection module.

[0027] According to a third aspect of the embodiments of the present application, a computer readable storage medium is provided, and the computer readable storage medium stores instructions, when the instructions in the computer readable storage medium are executed by a processor of an electronic device, the electronic device can perform the financial indicator data processing method according to the first aspect and any possible implementation manner thereof.

[0028] According to a fourth aspect of the embodiments of the present application, a computer program product is provided, and the computer program product includes computer instructions, when the computer instructions are run on an electronic device, the electronic device performs the financial indicator data processing method according to the first aspect and any possible implementation manner thereof.

[0029] The technical scheme provided by the embodiments of the present application at least brings the following beneficial effects: in the financial indicator data processing system of the present application, a financial indicator knowledge graph of the correlation relationship between multi-dimensional financial indicators is first constructed, and a reinforcement learning model of the correlation relationship between the electricity price fluctuation indicator and the electricity benefit indicator, i.e., a preset model, is also trained. Based on the preset model, the current electricity benefit indicator of the current electricity price fluctuation indicator can be intelligently and accurately predicted. Based on the financial indicator knowledge graph, the financial events related to the target entities in the current electricity price fluctuation indicator and the current electricity benefit indicator are reasoned to form a first target path. Based on the current financial indicators and the first target path, a first financial analysis report representing the mutual influence between various financial indicator influence factors is generated, so that the mutual influence trend between different financial indicator influence factors can be accurately reflected in the financial analysis report. Therefore, the financial user can more effectively supervise the power financial status based on the financial report, and the financial supervision efficiency is improved.

[0030] It should be understood that the above general description and the following detailed description are only exemplary and explanatory, and cannot limit the present disclosure. BRIEF DESCRIPTION OF DRAWINGS

[0031] The accompanying drawings, which are incorporated into and form part of the specification, illustrate embodiments consistent with the present disclosure and, together with the specification, serve to explain the principles of the present disclosure, and do not constitute an improper limitation on the present disclosure.

[0032] Figure 1 is a financial indicator data processing system block diagram according to an exemplary embodiment;

[0033] Figure 2Fig. 1 is a flow chart of a financial indicator data processing method according to an exemplary embodiment. DETAILED DESCRIPTION

[0034] In order for those skilled in the art to better understand the technical solutions of the present disclosure, the technical solutions in the embodiments of the present disclosure will be clearly and completely described below in conjunction with the drawings.

[0035] It should be noted that the terms "first", "second", etc. in the specification and claims of the present disclosure and the above drawings are used to distinguish similar objects, and do not necessarily indicate a specific order or sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments of the present disclosure described herein can be implemented in an order other than that illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present disclosure. Rather, they are merely examples of devices and methods consistent with some aspects of the present disclosure as detailed in the appended claims.

[0036] As shown in Figure 1 Fig. 1, the financial indicator data processing system includes a financial indicator data analysis platform 1; the financial indicator data analysis platform 1 includes a multi-source financial data integration module 11, an artificial intelligence model analysis module 12, an intelligent reasoning engine module 13, and a financial report generation module 14; the multi-source financial data integration module 11 is connected with a plurality of databases; the multi-source financial data integration module 11, the artificial intelligence model analysis module 12, the intelligent reasoning engine module 13, and the financial report generation module 14 are connected through communication.

[0037] In some embodiments, the above-mentioned financial indicator data processing system is also referred to as an artificial intelligence-based financial indicator data multi-dimensional analysis system. The system includes a financial indicator data analysis platform 1, a financial analysis platform supervision unit 2, and a financial supervision terminal 3, wherein the financial indicator data analysis platform 1 includes a multi-source financial data integration module 11, an artificial intelligence model analysis module 12, an intelligent reasoning engine module 13, a financial report generation module 14, and an interactive response interface module 15.

[0038] The multi-source financial data integration module 11 is configured to generate a financial indicator knowledge graph according to multi-dimensional financial indicator data in a plurality of databases; the financial indicator knowledge graph includes entities representing financial indicator characteristics, financial events affecting financial indicators, and associated paths between entities and financial events; the financial indicators include electricity price fluctuation indicators and electricity efficiency indicators.

[0039] Exemplarily, the multi-source financial data integration module 11 interfaces multiple databases including front-end business management, financial and electricity price databases, extracts structured data through entity extraction, financial event extraction technology, and constructs a financial indicator knowledge graph; the artificial intelligence model analysis module 12 trains a special model for multiple scenarios including electricity price fluctuation indicators and electricity benefit indicators based on a pre-trained large model (such as the GPT series) and combined with user feedback reinforcement learning algorithm (RLHF) and prompt learning algorithm (Prompt Learning), receives the financial indicator knowledge graph and user query information (such as analyzing the reasons for electricity price fluctuation indicators), and generates preliminary prediction results (such as “supply chain cost increase” and “policy adjustment impact”) through the trained model.

[0040] The artificial intelligence model analysis module 12 is configured to: perform reinforcement training on historical electricity price fluctuation indicators, historical financial events and historical electricity benefit indicators as training samples to obtain a preset model; predict the current electricity benefit indicator using the preset model based on the current electricity price fluctuation indicator to obtain the current electricity benefit indicator; the preset model is a reinforcement learning model for predicting the electricity benefit indicator under the influence of the corresponding financial event of the electricity price fluctuation indicator, and the corresponding financial event is a financial event with a probability higher than a preset probability under the corresponding electricity price fluctuation indicator.

[0041] The intelligent reasoning engine module 13 is configured to: extract a plurality of target entities of entity features in the current electricity benefit indicator and the current electricity price fluctuation indicator; reason the plurality of target entities according to the association path in the financial indicator knowledge graph to obtain a plurality of first target paths, and the plurality of first target paths include a plurality of first target financial events corresponding to the plurality of target entities.

[0042] The financial report generation module 14 is configured to generate a first financial analysis report according to the current financial indicator and the plurality of first target paths; the first financial analysis report includes the association trend between the current electricity benefit indicator, the current electricity price fluctuation indicator and the first target financial event.

[0043] The preset probability is a probability before a preset bit sequence or the highest probability.

[0044] The system further includes a financial analysis platform supervision unit 2 and a financial supervision terminal 3.

[0045] The above-mentioned multiple databases include front-end business management databases, financial databases and electricity price databases.

[0046] The electricity price fluctuation indicators include one or more of the following: statistical fluctuation indicators (e.g., price fluctuation rate), frequency and persistence indicators (price fluctuation frequency, high / low price fluctuation duration), spike (spike factor representing the ratio of peak price to average price, value at risk, conditional value at risk), and risk indicators, and electricity price prediction error indicators (deviation of predicted electricity price from actual electricity price).

[0047] The electricity efficiency indicators include one or more of the following: efficiency indicators (unit output / production electricity consumption, unit area electricity consumption, equipment / system efficiency, power factor), economic indicators (unit electricity cost, electricity cost proportion, energy saving benefit, demand response benefit, investment payback period / net present value / internal rate of return), load characteristic indicators, i.e., indirectly reflecting benefits (load factor, peak-valley electricity consumption ratio), management and behavior indicators (demand response participation / potential, energy saving target completion rate), and environmental benefit indicators (unit output / production carbon emissions, renewable energy consumption proportion).

[0048] The entities include physical entities (e.g., device entities such as substations / units / lines, route entities), abstract entities (e.g., market entities such as electricity price types / policy documents, policy entities), and indicator entities (e.g., indicator parameters such as network loss rate / renewable energy penetration rate / load fluctuation coefficient).

[0049] For example, the intelligent reasoning engine module 13 performs multi-path reasoning on the preliminary prediction results (generates 3-5 reasoning chains), selects the optimal path (e.g., “policy adjustment as the main cause”) in response to multiple user selection results, preferably uses the self-consistency method to generate multiple reasoning chains and performs majority voting, and outputs the reasoning chains in a visual form (e.g., a flowchart). The financial report generation module 14 calls a dynamic template library, automatically fills in the indicator data and the reasoning results, generates initial analysis results (i.e., the first financial analysis report) including trend charts and attribution analysis, and performs format checking and sensitive word filtering on the initial analysis results through the verification engine and marks parts that need to be revised manually. After review, the final draft is generated and output. User revisions to the report (e.g., adjusting chart types) are fed back through the interactive response interface module 15, which receives user instructions, triggers model reanalysis or extended query range, and thus realizes dynamic updating of the report. In this way, through the integration of multi-source data, adaptive models, and self-consistent reasoning technology, intelligent processing of electricity price fluctuation indicator attribution and electricity efficiency indicator correlation analysis is realized, and the use of dynamic templates and multiple rounds of review mechanism improves the efficiency and accuracy of report generation, providing technical support for digital transformation of the power industry, i.e., through the integration of generative large models, self-consistent reasoning chains, and multi-source data extraction technology, the problems of scattered data, single analysis dimension, and low report generation efficiency in the power industry are solved.

[0050] Through the above embodiments, in the financial indicator data processing system of the present application, the financial indicator knowledge graph of the correlation relationship between the multi-dimensional financial indicators is first constructed, and the reinforcement learning model of the correlation relationship between the electricity price fluctuation indicator and the electricity benefit indicator, i.e., the preset model, is also trained. Based on the preset model, the current electricity benefit indicator of the current electricity price fluctuation indicator can be intelligently and accurately predicted. The financial events associated with the target entities in the current electricity price fluctuation indicator and the current electricity benefit indicator in the financial indicator knowledge graph are reasoned to form a first target path. Based on the current financial indicators and the first target path, a first financial analysis report representing the mutual influence between various financial indicator influence factors is generated, so that the mutual influence trend between different financial indicator influence factors is accurately reflected in the financial analysis report. Therefore, the financial user can more effectively supervise the power financial situation based on the financial report, and improve the efficiency of financial supervision.

[0051] In an embodiment, the intelligent reasoning engine module 13 is further configured to: display the plurality of first target paths on the front-end interface, so that the user account selects a second target path from the plurality of first target paths displayed on the front-end interface; receive the second target path returned by the front-end interface; output the path reasoning process of the second target path in a visual form according to the reasoning chain based on the financial indicator knowledge graph; and the financial report generation module 14 is further configured to: generate a second financial analysis report according to the second target path; and the second financial analysis report includes the correlation trend between the current electricity benefit indicator, the current electricity price fluctuation indicator and the second target financial event in the second target path.

[0052] In another embodiment, the financial indicator data analysis platform 1 further comprises an interactive response interface module 15 communicatively connected with other modules; the interactive response interface module 15 is configured to: perform an auditing operation on the first financial analysis report and / or the second financial analysis report; send the financial analysis report after the auditing operation to a target user terminal, so that the user of the target user terminal performs a revision operation or a confirmation operation on the financial analysis report after the auditing operation on the target user terminal, and sends the financial analysis report after the revision or confirmation operation to the financial report generation module 14; and the financial report generation module 14 is further configured to: generate a target financial analysis report according to the financial analysis report after the revision or confirmation operation; and wherein the auditing operation on the first financial analysis report and / or the second financial analysis report includes one or more of the following: format checking on the text format and / or picture format in the first financial analysis report and / or the second financial analysis report; sensitive word filtering and replacement operation on the preset sensitive words included in the first financial analysis report and / or the second financial analysis report; and marking the revision content in the first financial analysis report and / or the second financial analysis report that does not meet the preset logical conditions.

[0053] In another implementation, the system further comprises a financial analysis platform supervision unit 2 connected in communication with the financial indicator data analysis platform 1; the financial analysis platform supervision unit 2 is configured to determine the operational stability of the financial indicator data analysis platform 1.

[0054] As shown in Figure 1 , the system can further be provided with a financial analysis platform supervision unit 2, which comprises one or more of the following: a security and stability evaluation module 21, a database interfacing monitoring and analysis module 22, a revision tracking analysis module 23, and a request processing detection module 24.

[0055] The financial analysis platform supervision unit 2 is configured to ensure the stability of the financial indicator data analysis platform 1, so as to ensure that the multi-dimensional financial indicator data is obtained under various normal conditions.

[0056] The security and stability evaluation module 21 is configured to detect target operation data in the financial indicator data analysis platform 1 at a preset detection frequency, fit the target operation data to obtain a target fitting curve, and determine a stability operation result according to the curve trend of the target fitting curve, wherein the stability operation result comprises a security and stability normal signal or a security and stability alarm signal.

[0057] The database interfacing monitoring and analysis module 22 is configured to, upon receiving the security and stability normal signal, monitor and analyze the interfacing status of the multi-source financial data integration module 11 with each database within the detection period, and generate an interfacing qualified signal or an interfacing abnormal signal through analysis.

[0058] The revision tracking analysis module 23 is configured to, upon receiving the interfacing qualified signal, analyze the interaction efficiency of the user with the financial indicator data analysis platform 1 within the detection period, and generate an interaction abnormal signal or an interaction normal signal.

[0059] The request processing detection module 24 is configured to, upon receiving the interaction normal signal, analyze the query request processing load status of the financial indicator data analysis platform 1 within the detection period, and generate a processing high load signal or a processing low load signal.

[0060] As an implementation, as shown in Figure 1 , the system can further be provided with a financial supervision terminal 3.

[0061] The financial analysis platform supervision unit 2 comprehensively monitors the operation of the financial index data analysis platform 1, and the financial analysis platform supervision unit 2 comprises a safety and stability evaluation module 21 and a database docking monitoring analysis module 22; wherein the safety and stability evaluation module 21 detects and evaluates the safety and stability of the operation of the financial index data analysis platform 1 in the detection period (preferably, the detection period is twenty-five days), and generates a safety and stability normal signal or a safety and stability alarm signal accordingly. Moreover, the safety and stability alarm signal is sent to the financial supervision terminal 3, and the financial supervision terminal 3 sends out a corresponding early warning when receiving the safety and stability alarm signal, so as to remind the supervisor to take corresponding safety optimization and improvement measures for the financial index data analysis platform 1, to strengthen the subsequent operation supervision of the financial index data analysis platform 1, and to ensure the subsequent safe and stable operation of the financial index data analysis platform 1.

[0062] In another embodiment, the safety and stability evaluation module 21 is further configured to obtain the total number of network attacks on the financial index data analysis platform 1 in the detection period, and mark the total number of network attacks as a network feature value; and obtain the total number of crashes of the financial index data analysis platform 1 in the detection period, and mark the total number of crashes as a paralysis feature value; if the network feature value is greater than or equal to a first preset threshold value and / or the paralysis feature value is greater than or equal to a second preset threshold value, a safety and stability alarm signal is generated; if the network feature value is less than the first preset threshold value and the paralysis feature value is less than the second preset threshold value, the network feature value, the paralysis feature value, the intervention time value and the intervention risk value are weighted and summed to obtain a safety and stability evaluation coefficient; if the safety and stability evaluation coefficient is greater than or equal to a preset safety and stability evaluation coefficient threshold value, a safety and stability alarm signal is generated; and if the safety and stability evaluation coefficient is less than the preset safety and stability evaluation coefficient threshold value, a safety and stability normal signal is generated.

[0063] In this embodiment, the total number of network attacks on the financial index data analysis platform 1 in the detection period is obtained and marked as a network feature value, and the total number of crashes of the financial index data analysis platform 1 in the detection period is obtained and marked as a paralysis feature value. The network feature value and the paralysis feature value are compared with the preset network feature threshold value and the preset paralysis feature threshold value respectively, and if the network feature value or the paralysis feature value exceeds the corresponding preset threshold value, it indicates that the safety and stability of the financial index data analysis platform 1 in the detection period is poor, and a safety and stability alarm signal is generated.

[0064] In another embodiment, the database connection monitoring analysis module 22 is specifically configured to mark any database as a stable abnormal database if the total disconnection time value or the total disconnection frequency value of any database in the plurality of databases exceeds the corresponding preset disconnection threshold value; generate a connection abnormal signal if there is a stable abnormal database in the plurality of databases during the detection period; if there is no stable abnormal database in the plurality of databases during the detection period, calculate the weighted sum of the disconnection time value and the disconnection frequency value of the stable abnormal database to obtain a disconnection monitoring evaluation value, and generate a connection abnormal signal if the disconnection monitoring evaluation value is greater than or equal to a preset disconnection monitoring evaluation threshold value; and generate a connection qualified signal if the disconnection monitoring evaluation value is less than the preset disconnection monitoring evaluation threshold value.

[0065] In this embodiment, if neither the network characteristic value nor the paralysis characteristic value exceeds the corresponding preset threshold value, the detection period is divided into a plurality of monitoring time periods, and the duration of all monitoring time periods is the same; if the financial indicator data analysis platform 1 is attacked by a network attack or collapses and paralyzes in a corresponding monitoring time period, it indicates that the operation of the financial indicator data analysis platform 1 in the corresponding monitoring time period is hindered, and the corresponding monitoring time period is marked as an emergency intervention time period; if the financial indicator data analysis platform 1 is not attacked by a network attack and does not collapse and paralyze in a corresponding monitoring time period, it indicates that the operation of the financial indicator data analysis platform 1 in the corresponding monitoring time period is relatively smooth, and the corresponding monitoring time period is marked as a safe and stable time period.

[0066] In the first embodiment, the financial analysis platform supervision unit 2 includes a safe and stable evaluation module 21 and a database connection monitoring analysis module 22, and the stability of the analysis platform is specifically implemented.

[0067] Firstly, the number of emergency intervention time periods in the detection period is obtained, and the ratio of the number of emergency intervention time periods to the total number of monitoring time periods is calculated to obtain an intervention time occupancy value. The number of emergency intervention time periods between adjacent two safe and stable time periods is marked as an intervention duration value. The intervention duration value is compared with a preset intervention duration threshold value. If the intervention duration value exceeds the preset intervention duration threshold value, the corresponding intervention duration value is marked as an intervention abnormal value.

[0068] Secondly, the number of intervention abnormal values in the detection period is obtained and marked as an intervention risk value. A safe and stable evaluation coefficient is calculated by weighted sum of the network characteristic value, the paralysis characteristic value, the intervention time occupancy value, and the intervention risk value. That is, the network characteristic value, the paralysis characteristic value, the intervention time occupancy value, and the intervention risk value are respectively assigned corresponding preset weight coefficients, and the network characteristic value, the paralysis characteristic value, the intervention time occupancy value, and the intervention risk value are respectively multiplied by the corresponding preset weight coefficients. The sum of the four product results is marked as the safe and stable evaluation coefficient. Moreover, the larger the value of the safe and stable evaluation coefficient, the worse the comprehensive safety and stability of the financial indicator data analysis platform 1 in the detection period.

[0069] Thirdly, the safety and stability evaluation coefficient is compared with the preset safety and stability evaluation coefficient threshold value. If the safety and stability evaluation coefficient exceeds the preset safety and stability evaluation coefficient threshold value, it indicates that the safety and stability of the financial index data analysis platform 1 in the detection period is poor in general, and a safety and stability alarm signal is generated. If the safety and stability evaluation coefficient does not exceed the preset safety and stability evaluation coefficient threshold value, it indicates that the safety and stability of the financial index data analysis platform 1 in the detection period is good in general, and a safety and stability normal signal is generated.

[0070] Thirdly, the safety and stability evaluation module 21 sends the safety and stability normal signal to the database docking monitoring analysis module 22. When the database docking monitoring analysis module 22 receives the safety and stability normal signal, it monitors and analyzes the docking status of the multi-source financial data integration module 11 and each database in the detection period, and generates a docking qualified signal or a docking abnormal signal through analysis.

[0071] Fourthly, the docking abnormal signal is sent to the financial supervision terminal 3. When the financial supervision terminal 3 receives the docking abnormal signal, it issues a corresponding early warning to timely strengthen the docking supervision of all databases, ensure the continuous and stable docking of the financial index data analysis platform 1 and all databases, avoid the analysis process from stopping due to the inability to obtain data, and further ensure the smoothness of the operation of the financial index data analysis platform 1.

[0072] The specific analysis process of the database docking monitoring analysis module 22 is as follows:

[0073] All databases docked by the multi-source financial data integration module 11 are obtained, and the corresponding database is marked as i, and i is a natural number greater than 1. When the docking between the financial index data analysis platform 1 and the database i is disconnected (i.e. the financial index data analysis platform 1 cannot obtain data from the database i), the disconnection duration is collected, and the sum of all disconnection durations of the database i in the detection period is calculated to obtain the total disconnection time value, and the total number of disconnections of the database i in the detection period is marked as the total disconnection frequency value.

[0074] The total disconnection time value and the total disconnection frequency value are compared with the preset total disconnection time threshold value and the preset total disconnection frequency threshold value respectively. If the total disconnection time value or the total disconnection frequency value exceeds the corresponding preset threshold value, it indicates that the docking with the database i is unstable in the detection period, and the database i is marked as a stable database. If there is a stable database in the detection period, it indicates that the docking of the database in the detection period is poor, and a docking abnormal signal is generated.

[0075] Further, if there is no stable database in the detection period, the total disconnection time value of all databases is calculated to obtain the disconnection time measurement value, and the total disconnection frequency value of all databases is calculated to obtain the disconnection frequency measurement value.

[0076] The disconnection monitoring evaluation value is calculated by weighted summation of the disconnection time measurement value and the disconnection frequency measurement value; that is, the disconnection time measurement value and the disconnection frequency measurement value are respectively assigned with corresponding preset weight coefficients, and the disconnection time measurement value and the disconnection frequency measurement value are respectively multiplied by the corresponding preset weight coefficients, and the sum of the two groups of product results is marked as the disconnection monitoring evaluation value; and the larger the value of the disconnection monitoring evaluation value, the worse the overall performance of the database connection in the detection period.

[0077] The disconnection monitoring evaluation value is compared with the preset disconnection monitoring evaluation threshold value, if the disconnection monitoring evaluation value exceeds the preset disconnection monitoring evaluation threshold value, it indicates that the overall performance of the database connection in the detection period is poor, and then a connection abnormal signal is generated; if the disconnection monitoring evaluation value does not exceed the preset disconnection monitoring evaluation threshold value, it indicates that the overall performance of the database connection in the detection period is good, and then a connection qualified signal is generated.

[0078] In another embodiment, the revision tracking analysis module 23 is specifically configured to obtain the number of interactions of unqualified time length in the detection period, and calculate the ratio of the number of interactions to the total number of analysis time length to obtain the unqualified time frequency value; the ratio of each interaction analysis time length to the corresponding preset analysis time length threshold value is marked as the analysis time length; and the average of all analysis time lengths in the detection period is calculated to obtain the analysis evaluation value; if the unqualified time frequency value is greater than or equal to the time frequency threshold value, and / or the analysis evaluation value is greater than or equal to the preset evaluation threshold value, a revision abnormal signal is generated; otherwise, a revision normal signal is generated.

[0079] In another embodiment, the request processing detection module 24 is specifically configured to obtain the total number of query requests received by the financial index data analysis platform 1 in the detection period, and mark the total number of query requests as the load detection value, if the load detection value is greater than or equal to the preset load detection threshold value, a high load processing signal is generated; if the load detection value is less than the preset load detection threshold value, the load detection value, the delay detection value and the waiting condition value are calculated by weighted summation to obtain the load evaluation value, if the load evaluation value is greater than or equal to the preset load evaluation threshold value, a high load processing signal is generated; if the load evaluation value is less than the preset load evaluation threshold value, a low load processing signal is generated.

[0080] In example two, the difference between this example and example one is that the financial analysis platform supervision unit 2 further includes a revision tracking analysis module 23, and the database connection monitoring analysis module 22 sends the connection qualified signal to the revision tracking analysis module 23, and the revision tracking analysis module 23 analyzes the response efficiency of the financial index data analysis platform 1 to user revision opinions in the detection period when receiving the connection qualified signal, and generates a revision abnormal signal or a revision normal signal accordingly.

[0081] And send the revised abnormal signal to the financial supervision terminal 3, when the financial supervision terminal 3 receives the revised abnormal signal, the corresponding early warning is sent out to remind the supervisor to take corresponding optimization improvement measures in time, ensure the timeliness of the financial index data analysis platform 1 in response to user modification opinions, and improve the operation efficiency of the financial index data analysis platform 1.

[0082] The specific analysis process of the revision tracking analysis module 23 is as follows.

[0083] Firstly, the time when the financial index data analysis platform 1 receives the revision opinion is obtained and marked as the receiving time, and the generation and output time of the revised report is marked as the output time, the interval time between the output time and the corresponding receiving time is marked as the analysis time, the analysis time is compared with the corresponding preset analysis time threshold value, if the analysis time exceeds the corresponding preset analysis time threshold value, the corresponding analysis time is marked as unqualified time.

[0084] Secondly, the number of unqualified time in the detection period is obtained and compared with the total number of analysis time to obtain the unqualified time frequency value, and the ratio of analysis time to the corresponding preset analysis time threshold value is marked as the analysis time length, and the average of all analysis time lengths in the detection period is calculated to obtain the analysis evaluation value.

[0085] Thirdly, the unqualified time frequency value and the analysis evaluation value are compared with the preset unqualified time frequency threshold value and the preset analysis evaluation threshold value respectively, if the unqualified time frequency value or the analysis evaluation value exceeds the corresponding preset threshold value, the revision abnormal signal is generated; if the unqualified time frequency value and the analysis evaluation value do not exceed the corresponding preset threshold value, the revision normal signal is generated.

[0086] Embodiment three: the difference between this embodiment and embodiment one and embodiment two is that the financial analysis platform supervision unit 2 further comprises a request processing detection module 24, the revision tracking analysis module 23 sends the revision normal signal to the request processing detection module 24, the request processing detection module 24 receives the revision normal signal and analyzes the query request processing load condition of the financial index data analysis platform 1 in the detection period, and generates a high processing load signal or a low processing load signal through analysis;

[0087] And send the high processing load signal to the financial supervision terminal 3, when the financial supervision terminal 3 receives the high processing load signal, the corresponding early warning is sent out to remind the supervisor to strengthen the subsequent operation supervision of the financial index data analysis platform 1 and improve its operation load bearing capacity, so as to ensure its operation performance and user satisfaction.

[0088] The specific analysis process of the request processing detection module 24 is as follows.

[0089] The total number of query requests received by the financial index data analysis platform 1 in the detection period is obtained and marked as a load detection value. The load detection value is compared with a preset load detection threshold value. If the load detection value exceeds the preset load detection threshold value, it indicates that the request processing load of the detection period is initially high. A high load processing signal is generated.

[0090] If the load detection value does not exceed the preset load detection threshold value, the waiting processing time length of the corresponding query request is collected. The waiting processing time length is compared with a preset waiting processing time length threshold value. The number of query requests with a waiting processing time length exceeding the preset waiting processing time length threshold value in the detection period is marked as a delay detection value. The waiting processing time length of all query requests in the detection period is averaged to obtain a waiting condition value.

[0091] The load detection value, the delay detection value, and the waiting condition value are weighted and summed to obtain a load evaluation value. The load detection value, the delay detection value, and the waiting condition value are respectively assigned corresponding preset weight coefficients. The load detection value, the delay detection value, and the waiting condition value are respectively multiplied by the corresponding preset weight coefficients. The sum of the three groups of product results is marked as the load evaluation value. The larger the value of the load evaluation value, the higher the comprehensive request processing load of the detection period.

[0092] The load evaluation value is compared with a preset load evaluation threshold value. If the load evaluation value exceeds the preset load evaluation threshold value, it indicates that the comprehensive request processing load of the detection period is high. A high load processing signal is generated. If the load evaluation value does not exceed the preset load evaluation threshold value, it indicates that the comprehensive request processing load of the detection period is low. A low load processing signal is generated.

[0093] In the above embodiments of the present application, when used, the multi-source financial data integration module 11 interfaces with multiple databases and constructs a financial indicator knowledge graph, the artificial intelligence model analysis module 12 receives the financial indicator knowledge graph and user query information and generates a preliminary prediction result through the trained model, the intelligent reasoning engine module 13 generates multiple reasoning chains and performs majority voting, the financial report generation module 14 generates an initial analysis result including a trend chart and attribution analysis, and generates a final draft after review and outputs it, the user's revision comments are fed back through the interactive response interface module 15 to trigger model reanalysis or expand the query range, through the integration of the generative large language model, self-consistent reasoning chain and multi-source data extraction technology, the problems of scattered data, single analysis dimension and low report generation efficiency in the power industry are solved, and through the financial analysis platform supervision unit 2, the safety and stability of the financial indicator data analysis platform 1, the database interface performance, the revision comment response efficiency status and the request processing load state are progressively and accurately analyzed, the overall monitoring of the financial indicator data analysis platform 1 is realized and timely alarm is given, which significantly reduces the operation supervision difficulty of the financial indicator data analysis platform 1 and ensures its operation performance.

[0094] As shown in Figure 2 , the financial indicator data processing method provided by the present application is specifically introduced.

[0095] S21, obtaining a current electricity price fluctuation indicator, and inputting the current electricity price fluctuation indicator into a preset model to obtain a current electricity benefit indicator.

[0096] S22, extracting a plurality of target entities of entity features in the current electricity benefit indicator and the current electricity price fluctuation indicator.

[0097] S23, reasoning the plurality of target entities according to the associated paths in the financial indicator knowledge graph to obtain a plurality of first target paths.

[0098] The plurality of first target paths include a plurality of first target financial events corresponding to the plurality of target entities.

[0099] S24, generating a first financial analysis report according to the current financial indicator and the plurality of first target paths.

[0100] The first financial analysis report includes the associated trends between the current electricity benefit indicator, the current electricity price fluctuation indicator and the first target financial events.

[0101] The plurality of first target paths include a plurality of target entities and a plurality of first target financial events corresponding to the plurality of target entities.

[0102] The financial indicator knowledge graph includes entities representing characteristics of the financial indicators, financial events affecting the financial indicators, and association paths between the entities and the financial events; the financial indicators include an electricity price fluctuation indicator and an electricity benefit indicator; the preset model is a reinforcement learning model, used to predict the electricity benefit indicator under the influence of the corresponding financial event on the electricity price fluctuation indicator, and the corresponding financial event is a financial event with a probability of occurrence higher than a preset probability under the corresponding electricity price fluctuation indicator; and the financial indicator knowledge graph is generated according to multi-dimensional financial indicator data in multiple databases.

[0103] In an embodiment, before or during the implementation of the above method steps, the following steps are performed to determine the stability of the financial indicator data analysis platform, so that the data on which the report is based is generated by the financial indicator data analysis platform under stable processing.

[0104] First, the stability of the financial indicator data analysis platform is determined.

[0105] Second, the multi-dimensional financial indicator data is obtained, the training samples are obtained, and the current electricity price fluctuation indicator is obtained under the condition that the financial indicator data analysis platform is in a stable running state.

[0106] The training samples include historical electricity price fluctuation indicators, historical financial events, and historical electricity benefit indicators.

[0107] The determination of the stable running state of the financial indicator data analysis platform can be achieved by one or more of the following modules: a safety and stability evaluation module 21, a database docking monitoring and analysis module 22, a revision tracking analysis module 23, and a request processing detection module 24.

[0108] The embodiments of the present application also provide a computer-readable storage medium, when the instructions in the computer-readable storage medium are executed by the processor of the financial indicator data processing device or the power system equipment, the financial indicator data processing device or the power system equipment can execute the financial indicator data processing method of any one of the possible embodiments as described above. And can achieve the same technical effect, to avoid repetition, here will not repeat.

[0109] The embodiments of the present application also provide a computer program product, including a computer program or instructions, the computer program or instructions are executed by the processor to execute the financial indicator data processing method of any one of the possible embodiments as described above. And can achieve the same technical effect, to avoid repetition, here will not repeat.

[0110] Other embodiments of the application will be apparent to those skilled in the art from consideration of the specification and practice of the application disclosed herein. It is intended that the specification and examples be considered as exemplary only, with the true scope and spirit of the application being indicated by the following claims.

[0111] It is to be understood that the application is not limited to the precise construction herein described and as shown in the attached drawings, and that various modifications and changes can be made by those skilled in the art without departing from the scope of the application. The scope of the application is to be indicated by the appended claims, rather than the description.

Claims

1. A financial indicator data processing system based on artificial intelligence, characterized in that, The system is used for intelligent generation of financial statements, and includes: a financial indicator data analysis platform; the financial indicator data analysis platform includes: The multi-source financial data integration module is configured to generate a financial indicator knowledge graph based on multi-dimensional financial indicator data from the multiple databases. The financial indicator knowledge graph includes entities representing financial indicators, financial events representing financial indicators, and the association paths between entities and financial events. The financial indicators include electricity price fluctuation indicators and electricity efficiency indicators. The multiple databases include a front-end business management database, a financial database, and an electricity price database. The artificial intelligence model analysis module is configured to: perform reinforcement learning training based on historical electricity price fluctuation indicators, historical financial events, and historical electricity benefit indicators as training samples to obtain a preset model; use the preset model to predict the electricity benefit indicator corresponding to the current electricity price fluctuation indicator to obtain the current electricity benefit indicator; the preset model is a reinforcement learning model used to predict the electricity benefit indicator under the influence of the corresponding financial event, wherein the corresponding financial event is a financial event with a higher probability of occurrence than a preset probability under the corresponding electricity price fluctuation indicator; The intelligent reasoning engine module is configured to: extract multiple target entities with entity features from the current electricity efficiency index and the current electricity price fluctuation index; and reason about the multiple target entities based on the association paths in the financial indicator knowledge graph to obtain multiple first target paths, wherein the multiple first target paths include multiple first target financial events corresponding to the multiple target entities. The financial report generation module is configured to generate a first financial analysis report based on current financial indicators and the multiple first target paths; the first financial analysis report includes the current electricity efficiency indicators, the current electricity price fluctuation indicators, and the correlation between the first target financial events.

2. The financial indicator data processing system according to claim 1, characterized in that, The intelligent inference engine module is also configured as follows: The plurality of first target paths are displayed on the front-end interface so that the user account can select a second target path from the plurality of first target paths displayed on the front-end interface; Receive the second target path returned by the front-end interface; Based on the financial indicator knowledge graph, the path reasoning process of the second target path is output in a visual form according to the reasoning chain; The financial report generation module is further configured to: generate a second financial analysis report based on the current financial indicators and the second target path; the second financial analysis report includes the correlation and influence trend between the current electricity benefit indicator, the current electricity price fluctuation indicator, and the second target financial event in the second target path; the current financial indicators include the current electricity benefit indicator and the current electricity price fluctuation indicator.

3. The financial indicator data processing system according to claim 2, characterized in that, The financial indicator data analysis platform also includes an interactive response interface module that communicates and connects with other modules. The interactive response interface module is configured to: review the first financial analysis report and / or the second financial analysis report; send the reviewed financial analysis report to the target user terminal so that the user of the target user terminal can revise or confirm the reviewed financial analysis report on the target user terminal; and send the revised or confirmed financial analysis report to the financial report generation module. The financial report generation module is also configured to generate a target financial analysis report based on the revised or confirmed financial analysis report. The auditing operation of the first financial analysis report and / or the second financial analysis report includes one or more of the following: performing format verification on the text format and / or image format of the first financial analysis report and / or the second financial analysis report; performing sensitive word filtering and replacement operations on the preset sensitive words included in the first financial analysis report and / or the second financial analysis report; and marking the revised content in the first financial analysis report and / or the second financial analysis report that does not conform to the preset logical conditions.

4. The financial indicator data processing system according to any one of claims 1 to 3, characterized in that, The system also includes a financial analysis platform monitoring unit that is communicatively connected to the financial indicator data analysis platform; the financial analysis platform monitoring unit is used to determine the operational stability of the financial indicator data analysis platform; the financial analysis platform monitoring unit includes a security and stability assessment module, a database connection monitoring and analysis module, a revision tracking and analysis module, and a request processing detection module; The safety and stability assessment module is configured to detect target operating data within the financial indicator data analysis platform at a preset detection frequency, fit the target operating data to obtain a target fitting curve, and determine the stability operating result based on the curve trend of the target fitting curve, wherein the stability operating result includes a safety and stability normal signal or a safety and stability alarm signal. The database docking monitoring and analysis module is configured to monitor and analyze the docking status of the multi-source financial data integration module with each database during the detection period when the safe, stable and normal signal is received, and generate a docking qualified signal or a docking abnormal signal through analysis. The revision tracking and analysis module is configured to analyze the user's interaction efficiency with the financial indicator data analysis platform during the detection period when the docking qualification signal is received, so as to generate an interaction abnormal signal or an interaction normal signal. The request processing detection module is configured to analyze the query request processing load of the financial indicator data analysis platform during the detection period when it receives a normal interaction signal, so as to generate a high-load signal or a low-load signal.

5. The financial indicator data processing system according to any one of claims 1 to 4, characterized in that, The security and stability assessment module is also configured to obtain the total number of cyberattacks suffered by the financial indicator data analysis platform during the detection period, and mark the total number of cyberattacks suffered as a network feature value. In addition, the total number of times the financial indicator data analysis platform crashed or became paralyzed during the detection period is obtained and the total number of crashes or becoming paralyzed is marked as a paralysis feature value; If the network feature value is greater than or equal to the first preset threshold and / or the paralysis feature value is greater than or equal to the second preset threshold, then the safety and stability alarm signal is generated. If the network feature value is less than the first preset threshold and the paralysis feature value is less than the second preset threshold, the network feature value, the paralysis feature value, the intervention time value, and the intervention risk value are weighted and summed to obtain the safety and stability evaluation coefficient. If the safety and stability evaluation coefficient is greater than or equal to the preset safety and stability evaluation coefficient threshold, then the safety and stability alarm signal is generated. If the safety and stability evaluation coefficient is less than the preset safety and stability evaluation coefficient threshold, the safety and stability normal signal is generated.

6. The financial indicator data processing system according to any one of claims 1 to 4, characterized in that, The database docking monitoring and analysis module is specifically configured to: mark any database as a stable abnormal database if the total disconnection time value or total disconnection frequency value of any database in the plurality of databases exceeds the corresponding preset disconnection threshold; generate a docking anomaly signal if a stable abnormal database exists in the plurality of databases during the detection period; generate a disconnection anomaly signal if no stable abnormal database exists in the plurality of databases during the detection period; calculate a weighted sum of the disconnection time measurement value and disconnection frequency measurement value of the stable abnormal database to obtain a disconnection monitoring evaluation value; generate a docking anomaly signal if the disconnection monitoring evaluation value is greater than or equal to the preset disconnection monitoring evaluation threshold; and generate a docking qualified signal if the disconnection monitoring evaluation value is less than the preset disconnection monitoring evaluation threshold.

7. The financial indicator data processing system according to any one of claims 1 to 4, characterized in that, The revision tracking and analysis module is specifically configured to obtain the number of interactions with non-compliant durations during the detection period, and calculate the ratio of the number of interactions to the total number of analysis durations to obtain the non-compliant time frequency value. The ratio of the command parsing duration of each interaction to the corresponding preset command parsing duration threshold is marked as the command duration. The average of all the specified durations within the detection period is calculated to obtain the duration evaluation value; if the non-compliance time-frequency value is greater than or equal to the time-frequency threshold, and / or the duration evaluation value is greater than or equal to the preset evaluation threshold, a revision anomaly signal is generated; Otherwise, a normal revision signal is generated.

8. The financial indicator data processing system according to any one of claims 1 to 4, characterized in that, The request processing detection module is specifically configured to obtain the total number of query requests received by the financial indicator data analysis platform during the detection period, and mark the total number of query requests as a load detection value. If the load detection value is greater than or equal to a preset load detection threshold, a high load signal is generated. If the load detection value is less than the preset load detection threshold, a weighted summation is performed on the load detection value, the delayed processing value, and the waiting status value to obtain a load assessment value. If the load assessment value is greater than or equal to a preset load assessment threshold, a high load signal is generated. If the load assessment value is less than the preset load assessment threshold, a low load signal is generated.

9. A method for processing financial indicator data, characterized in that, This system is applied to a financial indicator data processing system. The system includes a financial indicator knowledge graph and a preset model. The financial indicator knowledge graph includes entities that characterize the features of financial indicators, financial events that affect financial indicators, and the association paths between entities and financial events. The financial indicators include electricity price fluctuation indicators and electricity efficiency indicators. The preset model is a reinforcement learning model used to predict the electricity benefit index under the influence of corresponding financial events, where the corresponding financial events are financial events with a higher probability of occurrence than the preset probability under the corresponding electricity price fluctuation index. The financial indicator knowledge graph is generated based on multi-dimensional financial indicator data from the multiple databases; the method includes: Obtain the current electricity price fluctuation index and input the current electricity price fluctuation index into the preset model to obtain the current electricity benefit index; Extract multiple target entities from the entity features of the current electricity efficiency index and the current electricity price fluctuation index; based on the association paths in the financial indicator knowledge graph, reason about the multiple target entities to obtain multiple first target paths, the multiple first target paths including multiple first target financial events corresponding to the multiple target entities; Based on the multiple first target paths, a first financial analysis report is generated; the first financial analysis report includes the current electricity efficiency indicator, the current electricity price fluctuation indicator, and the correlation trend between the first target financial events.

10. The financial indicator data processing method according to claim 9, characterized in that, The financial indicator data processing system also includes a financial indicator data analysis platform, and the method further includes: Determine the stability of the financial indicator data analysis platform; Once the financial indicator data analysis platform is confirmed to be operating stably, multi-dimensional financial indicator data is acquired, training samples are obtained, and the current electricity price fluctuation indicator is acquired. The training samples include historical electricity price fluctuation indicators, historical financial events, and historical electricity benefit indicators.

Images