Enterprise financial comprehensive management system based on intelligent technology

Abstract

The application provides an enterprise financial comprehensive management system based on intelligent technology, and relates to the technical field of financial management.The system comprises: an accounting intelligent subsystem for obtaining financial data, performing semantic analysis and conversion on the financial data through a preset rule and a multistage data mapping mechanism, generating standardized accounting business instructions and issuing the instructions to an automatic process engine, so as to automatically generate financial documents in accordance with accounting standards; a financial management subsystem for integrally realizing the execution of a to-do list management, the query of a compliance guidebook and the structured storage of financial data according to a preset standardized process; and a business service subsystem for providing tax-related business services, asset management services, fund management services and operation management service processing and guidance for enterprises through modular design.The application aims to improve operation efficiency, reduce error rate, break information islands, realize effective integration and utilization of resources and provide strong support for the decision-making of enterprises.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of financial management, and particularly relates to an enterprise financial comprehensive management system based on intelligent technology. BACKGROUND

[0002] In today's digital era, enterprise financial management faces increasingly complex business environments and massive data processing needs. Traditional financial management methods rely on a large number of manual operations, which are inefficient and prone to errors, and lack comprehensive automation support for business processes and centralized control of data. For example, in the process of expense accrual, bill generation, and invoice management, manual processing not only consumes time and labor costs, but also makes it difficult to ensure data accuracy and compliance. In addition, enterprises lack a standardized and intelligent service system for financial business support, such as tax-related business, asset management, and fund management, resulting in cumbersome business processes, paper accumulation, and high storage costs. Therefore, it is of great practical significance to develop an intelligent financial management system that can integrate various business processes, implement automated operations, traceability and query, and provide comprehensive financial business support. SUMMARY

[0003] In view of the above deficiencies of the prior art, the present application proposes an enterprise financial comprehensive management system based on intelligent technology by integrating accounting intelligence, financial management, and business service modules, aiming to improve operational efficiency, reduce error rates, break down information silos, and effectively integrate and utilize resources, providing strong support for enterprise decision-making.

[0004] The present application proposes an enterprise financial comprehensive management system based on intelligent technology, which includes:

[0005] An accounting intelligence subsystem, which integrates an ERP system and an automated process engine, is used to obtain financial data from the ERP system, perform semantic analysis and conversion of the financial data through preset rules and multi-level data mapping mechanisms, generate standardized accounting business instructions, and issue them to the automated process engine, thereby driving the automated process engine to automatically generate financial documents in accordance with accounting standards in the ERP system;

[0006] A financial management subsystem, which is used to implement pending item management, compliance guidebook query, and structured storage of financial data in an integrated manner according to preset standardized processes;

[0007] A business service subsystem, which is used to collaboratively handle tax-related business services, asset management services, fund management services, and operational management services through several integrated service modules.

[0008] Further, the accounting intelligence subsystem includes:

[0009] a checkout management module configured to perform data extraction on financial data obtained from an ERP system through a regular expression, perform integrity checking and semantic analysis on the extracted financial data, and generate standardized data fields; and generate financial documents in compliance with accounting standards by invoking a specified voucher template and filling in the standardized data fields;

[0010] an income management module configured to obtain unstructured financial data files, extract structured financial data from the financial data files through intelligent data analysis and rule mapping technology, and automatically generate standardized financial vouchers in compliance with accounting standards according to preset rules and the extracted financial data;

[0011] a treasurer application module configured to traverse each entity according to a preconfigured entity list, obtain financial management business data of each entity automatically, and perform a predefined financial management task based on the obtained financial management business data; the financial management task includes at least one of an intelligent supplier account management task, an uncleaned account book intelligent management task, a financial budget execution intelligent management task, or a fund allocation payment intelligent application task.

[0012] an auxiliary management module configured to obtain business data in different business scenarios automatically, and perform a predefined enterprise financial and management task based on the business data; the enterprise financial and management task includes at least one of an electronic invoice intelligent management task, a bank diary account intelligent query task, a financial project profit automatic extraction task, an invoice application single automatic generation task, or an operating index report automatic generation task.

[0013] Further, the checkout management module includes:

[0014] a welfare fund transfer automatic generation business unit configured to automatically download a subject balance table from an ERP system, match a preset keyword in a subject name column of the subject balance table line by line according to a preset regular expression, extract all matched line data as welfare-related items, perform semantic analysis and integrity checking on the welfare-related items, and generate standardized data fields; and invoke a voucher template and automatically fill in the standardized data fields to generate a standard format of a deduction list.

[0015] a safety production fee automatic deduction unit configured to invoke a pre-deduction expense calculation template, extract deduction data from the pre-deduction expense calculation template in combination with a predefined regular expression and semantic analysis, and further generate a safety production fee deduction list according to the deduction data by project dimension in batches.

[0016] The cost allocation company layer automatic generation business unit is used for automatically screening business detail data of different projects of each company from business data stored in a preconfigured data source according to screening conditions defined in a pre-set company sub-project adjustment template, performing data cleaning and format unification, and generating a standardized business detail data file; the parallel computing technology is adopted to match the business detail data file with preset cost allocation rules and a general ledger template, and perform matching and summary processing, so as to generate a cost allocation company general ledger;

[0017] The production cost settlement automatic generation business unit is used for obtaining subject balance detail data in a two-dimensional table structure by performing directional analysis on a subject balance table automatically downloaded from an ERP system; based on a preset configuration table and a dynamic correlation matrix, a target settlement voucher subject is acquired by taking a cost center and an accounting subject in the subject balance detail data as a query condition, and a debit amount of the subject balance detail data is automatically filled into a voucher template according to the target settlement voucher subject; a flow control mechanism is adopted to check data processing exceptions, skip abnormal data and record logs; and finally, a standardized cost settlement single conforming to accounting standards is automatically generated.

[0018] Further, the income management module comprises:

[0019] The sales income confirmation automatic generation unit is used for obtaining an income balance table from a specified external data source, performing parallel traversal of the income balance table by using a multithreading technology, extracting sales income data, and further automatically generating an electricity fee confirmation single according to the sales income data; meanwhile, an electricity quantity settlement single issued by a power grid company is automatically uploaded.

[0020] The sales income temporary estimate intelligent processing unit is used for obtaining a temporary estimate electricity fee income single from a specified external data source, and obtaining original data by analyzing the temporary estimate electricity fee income single; based on a multistage data mapping mechanism, the original data are dynamically converted into standard data recognizable by an ERP system; by using an automatic technology, a business detail page of the ERP system is automatically jumped to according to a preset menu path, and the standard data recognizable by the ERP system are automatically filled in, so that sales temporary estimate singles are batch generated.

[0021] The collection flow automatic generation unit is used for obtaining a collection flow detail file from a specified external data source, extracting collection flow key information from the collection flow detail file in combination with a pre-defined regular expression and semantic analysis; a total settlement receivable amount of a profit center is acquired by querying an income account book through SQL taking the profit center in the collection flow key information as a correlation key, and is verified; after verification, a collection flow record conforming to a rule is automatically generated according to the verified collection flow key information.

[0022] The collection flow key information includes a profit center, a collection account, a collection account opening bank name, an amount, a company code, a customer number, a collection method, a currency, and a transaction date.

[0023] Further, the treasurer application module comprises:

[0024] The intelligent supplier clearing account management unit is configured to traverse each company code according to a preconfigured company code list, automatically obtain the current account data of the company for each company code and perform data cleaning, and then perform clearing account operation on the cleaned data according to a preset clearing account rule, and automatically perform a skip operation and record a log for an abnormal company that cannot normally perform clearing account due to the nonexistence of the company code.

[0025] The uncleaned account intelligent management unit is configured to traverse each company code according to a preconfigured company code list, automatically configure a query parameter for each company code, load the accounts receivable and accounts payable of the company according to the query parameter using an intelligent waiting mechanism, and perform paging query on the loaded accounts receivable and accounts payable according to a configured subject and batch export the query results in a preset format.

[0026] The fund budget execution intelligent management unit is configured to traverse each company code according to a preconfigured company code list and automatically query the fund plan data of each company according to a preset multi-dimensional combined query condition, batch export the fund plan data of each company according to the company dimension to generate a standardized file, and verify, clean, and standardize the standardized file using a multi-level intelligent processing mechanism to generate structured fund plan data.

[0027] The fund allocation payment intelligent application unit is configured to traverse each company code according to a preconfigured company code list, query and load the fund allocation payment application information of each company using a multi-thread parallel processing technology, perform integrity verification on the fund allocation payment application information of each company, and batch submit the fund allocation payment application information that passes the verification for approval.

[0028] Further, the auxiliary management module comprises:

[0029] The electronic invoice intelligent management unit is configured to scan a specified folder and automatically identify invoice files through file extensions, establish a file fingerprint library based on a SHA-256 algorithm, filter out all identified duplicate files using the file fingerprint library, and batch upload the filtered invoice files to an invoice folder.

[0030] The bank journal account intelligent query unit is configured to automatically query a bank journal account and batch export according to a preset filtering condition.

[0031] The financial item profit automatic grabbing unit is used for acquiring a multi-level account balance table based on a company and a profit center in real time, dynamically checking the multi-level account balance table through a double closed loop checking mechanism based on a rule engine and a generative adversarial network, automatically identifying and extracting profit and loss type index data from the checked data through an intelligent semantic mapping engine, and automatically filling the profit and loss type index data into a preset intelligent analysis template to generate a financial item profit analysis report.

[0032] The billing application automatic generation unit is used for acquiring an income account book, performing line-by-line analysis according to a preset rule, extracting data fields respectively constituting basic business information, financial and tax information and invoice detail information from the income account book in combination with a regular expression and semantic analysis, and filling the extracted data fields into a tax system standard form through calling a billing application template engine to generate a billing application single.

[0033] The operating index completion condition report automatic generation unit is used for automatically collecting ERP financial data in a monthly cycle, and performing line-by-line analysis on the ERP financial data by using a missing value dynamic interpolation and derived field recalculation method based on a business rule to generate an operating analysis report.

[0034] Further, the specific method for dynamically checking the multi-level account balance table through the double closed loop checking mechanism based on the rule engine and the generative adversarial network is as follows:

[0035] The original data in the multi-level account balance table is cleaned through a standardized rule engine, and the cleaned data is converted into structured data with standardized semantics according to a preset conversion rule;

[0036] The structured data with business semantics is subjected to logical contradiction detection by using a pre-trained generative adversarial network; if the structured data with logical contradiction is detected, the standardized rule engine is fed back with abnormal information of the logical contradiction; the standardized rule engine locates the original data corresponding to the structured data according to the received abnormal information, modifies the original data by calling a predefined correction rule, and re-performs data cleaning and data conversion on the modified original data to generate new structured data with business semantics and re-perform logical contradiction detection, until the generative adversarial network cannot detect logical contradiction from the current structured data, and the checked data is obtained.

[0037] Further, the specific content of the method for performing line-by-line analysis on the ERP financial data by using the missing value dynamic interpolation and derived field recalculation method based on the business rule to generate the operating analysis report is as follows:

[0038] A mapping relationship table of subject types and precision rules is constructed by analyzing a data dictionary of an ERP system;

[0039] For any data in the ERP financial data, traverse each field in the data, and obtain the metadata label of each field by querying the data dictionary of the ERP system; wherein the metadata label is any one of a basic field and a derived field;

[0040] For each basic field, detect whether the subject is missing, if not, according to the mapping relationship table of subject type and precision rule, through dynamic matching of the pre-set precision rule library, the basic field is converted into the precision conforming to the business specification; if so, according to the mapping relationship table of subject type and precision rule, the business attribute of the basic field is automatically identified, and the moving weighted average method is used to dynamically interpolate the basic field to generate the missing value of the basic field and convert it into the precision conforming to the business specification;

[0041] For each derived field, the calculation formula of the derived field is parsed from the data dictionary of the ERP system, the variable current effective value of each variable in the calculation formula is obtained, and all variable current effective values of the variables are substituted into the calculation formula to recalculate, to obtain the substitute value of the derived field and convert it into the precision conforming to the business specification;

[0042] After the precision conversion of all fields in the data is completed, the cleaned data is obtained, and the data cleaning of the ERP financial data is completed to obtain the standardized financial data set;

[0043] The standardized financial analysis template is called and the standardized financial data set is automatically filled in to generate the operation analysis report.

[0044] Further, the financial management subsystem comprises:

[0045] The to-do list management module is used to obtain the to-do list and visually display the to-do list by using a density-sensitive dynamic time axis compression algorithm; the visual display content includes real-time synchronous display of file download and completion state of the to-do list, display of the to-do list through a monthly calendar view, and automatic push of a reminder message according to the deadline of the to-do list;

[0046] The guide manual query module is used to query the financial management business compliance guide manual in the built-in standardized compliance guide system according to the keywords or categories provided by the user;

[0047] The financial electronic database is used to obtain financial data related to funds, budgets, accounting, taxes and assets from a specified data source, and construct a multidimensional index according to the financial data; the financial data is stored in a structured manner according to the multidimensional index.

[0048] Further, the business service subsystem comprises:

[0049] The tax-related business service module is configured to store various tax-related business operation guide files based on tax regulations, receive a user's access request for a specified tax-related business, and provide the user with an operation guide file corresponding to the specified tax-related business according to the access request, wherein the specified tax-related business is at least one of a land occupation tax business, a personal income tax withholding business, an enterprise income tax declaration business, and a value-added tax input tax deduction business.

[0050] The asset management service module is configured to store standardized operation guide files of various asset management services, receive a user's access request for a specified asset management service, and provide the user with a standardized operation guide file corresponding to the specified asset management service according to the access request, wherein the specified asset management service is at least one of a production preparation service, a solid conversion management service, a completion settlement service, an asset temporary estimation service, and an asset inventory service.

[0051] The fund management service module is configured to store standardized operation guide files of various fund management services, receive a user's access request for a specified fund management service, and provide the user with a standardized operation guide file corresponding to the specified fund management service according to the access request, wherein the specified fund management service is at least one of a fund plan preparation service, a fund plan execution service, a special account use service, a special fund payment service, and a capital management service.

[0052] The operation management service module is configured to store standardized operation guide files of various operation management services, receive a user's access request for a specified operation management service, and provide the user with a standardized operation guide file corresponding to the specified operation management service according to the access request, wherein the specified operation management service is at least one of a current account management service, a budget control service, and a dividend payment management service.

[0053] The beneficial effects of the above technical solutions are as follows:

[0054] 1. Efficiently improving financial management efficiency and significantly reducing operating costs: The system of the present application greatly reduces manual intervention through full-process automation, and significantly improves processing efficiency in core links such as expense accrual and bill generation. Not only does it effectively reduce labor costs, but it also reduces operation error rates through intelligent verification mechanisms, achieving triple optimization of efficiency, cost, and accuracy.

[0055] 2. Building a full-link data traceability system to strengthen decision support capabilities: The system of the present application realizes full-life cycle traceability and query of financial data from collection, processing to archiving through standardized data storage and intelligent indexing technology. It supports multi-dimensional dynamic data analysis and provides real-time and accurate data support.

[0056] 3. Establish a standardized intelligent management system to build a risk prevention and control barrier: The system relies on an automated process engine and a compliance rule library to build a standardized management system covering all fields such as accounting, fund management, and tax processing. By automatic compliance verification, compliance management requirements are embedded in business processes, improving management efficiency while strengthening compliance and risk prevention and control capabilities.

[0057] 4. Integrate the full-scenario financial service ecosystem to promote digital green transformation: The system integrates professional service modules in multiple fields such as tax-related business, asset management, and fund scheduling to provide full-life cycle support for enterprises from the business front end to the financial back end. Through electronic document flow and intelligent archive management, paper material use is reduced, storage costs are significantly reduced, and the green office concept and enterprise digital transformation strategy are deeply integrated.

[0058] In summary, the system deeply integrates intelligent and digital technologies to reshape the financial management mode with systematic integration thinking. Through full-process automation and data-driven intelligent decision-making mechanisms, not only does it significantly reduce the use of paper documents and reduce carbon emissions in the enterprise operation process, but it also effectively promotes green and low-carbon transformation. With cross-module data linkage and precise analysis capabilities, it realizes the fine allocation and efficient use of enterprise resources. The system has achieved remarkable results in improving financial management efficiency, eliminating information barriers, reducing operating costs, and enhancing market competitiveness. At the same time, through standardized compliance management and full-cycle business support, it has significantly improved customer service satisfaction, providing solid support for enterprises to achieve coordinated growth of economic benefits and social value and move towards sustainable development. BRIEF DESCRIPTION OF DRAWINGS

[0059] Figure 1 Figure 1 is an architectural diagram of an enterprise financial comprehensive management system based on intelligent technology in the present embodiment. DETAILED DESCRIPTION

[0060] To facilitate understanding of the present application, the specific embodiments of the present application are further described in detail below in conjunction with the drawings and embodiments. The following embodiments are used to illustrate the present application, but are not used to limit the scope of the present application. On the contrary, the purpose of providing these embodiments is to make the disclosure of the present application more thorough and comprehensive.

[0061] Embodiment 1:

[0062] An enterprise financial comprehensive management system based on intelligent technology in the present embodiment, as shown in Figure 1, includes an accounting intelligent subsystem, a financial management subsystem, and a business service subsystem. The specific embodiments of each subsystem are as follows: Figure 1

[0063] ​The accounting intelligent subsystem integrates the ERP system and the automatic process engine, obtains financial data from the ERP system, performs semantic analysis and conversion on the financial data through preset rules and a multi-level data mapping mechanism, generates standardized accounting business instructions, and issues the instructions to the automatic process engine, so that the automatic process engine automatically generates financial documents in compliance with accounting standards in the ERP system.

[0064] In the present embodiment, the accounting intelligent subsystem integrates the external automatic enterprise resource planning (ERP) system and the automatic process engine, realizes automatic processing of the whole process of financial business, reduces manual intervention, and improves data accuracy. The subsystem covers multiple business processes such as welfare fee accrual, allocation of company expenses management, production cost transfer, electricity fee confirmation, and invoice management. Through preset rules and data mapping mechanisms, the subsystem realizes closed-loop automation from data collection to document generation.

[0065] It should be noted that the system described in the present embodiment establishes an efficient, bidirectional, and secure data interaction channel with the ERP system, realizes full-link automation from data collection to processing, and its interaction mechanism is as follows:

[0066] (1) Intelligent data calling mechanism:

[0067] In order to realize comprehensive and efficient acquisition of massive financial data (such as subject balance table) in the ERP system, the present embodiment adopts a diversified hybrid access strategy, including parallel data traversal technology, intelligent semantic analysis and conversion technology, standardized file interface adaptation technology, and visual interface intelligent collection technology. The four parts support the data acquisition needs of modules such as the checkout management module and the income management module. The specific implementation method is as follows:

[0068] Parallel data traversal technology: when the ERP system receives a data acquisition request (such as the welfare fee transfer automatic generation business unit needs to download the subject balance table) from the internal business unit of the system described in the present embodiment, a SQL query is constructed at the database level according to the data acquisition request, which is used to pre-filter the to-be-traversed data set (such as the latest subject balance table) related to the data acquisition request from the database. Due to the huge amount of subject balance table data, a number of data query subtasks are usually constructed according to the dimensions of the company or organizational structure. The task scheduler assigns the above subtasks to multiple worker threads, so that multiple data segments are extracted by performing multi-threaded parallel queries on the to-be-traversed data set. All data segments are aggregated to restore a complete subject balance table that meets the data acquisition request. Taking the production cost transfer automatic generation business unit as an example, this technology provides core support for its rapid acquisition of subject balance detail data.

[0069] Intelligent semantic analysis and conversion technology: In this embodiment, regular expressions and multi-level data mapping technology are integrated to perform semantic analysis and cleaning on the obtained unstructured or semi-structured data, and dynamically convert it into fully structured data conforming to system standards. Specifically, this embodiment maintains an extensible regular expression pattern library, for example, for welfare fee identification, pre-installs patterns such as "payable employee compensation-welfare fee" and "management expenses-welfare expenditure" to cover different expressions. The welfare fee transfer automatic business unit uses this technology to match keywords such as "payable employee compensation-welfare fee" in the subject balance table through regular expressions, to complete the accurate extraction and standardized conversion of welfare fee related items.

[0070] Standardized file interface adaptation technology: This embodiment realizes standardized file interface adaptation by constructing a file adapter component, thereby establishing a data exchange channel with the ERP system and supporting the parsing of Excel, CSV and other standard format export files. By integrating high-performance open source libraries such as Apache POI (for parsing Excel format) and OpenCSV (for parsing CSV format) as the core parsing engine, automatic recognition, semantic extraction and format standardization conversion are realized. This technology is the basis for the split company layer expense automatic business unit to obtain standardized business detail data files from pre-configured data sources.

[0071] Visual interface intelligent collection technology: This embodiment integrates process automation robot RPA engine, uses XPath and CSS selector-based UI element precise positioning technology to simulate authorized user operations, automatically completes system login, page navigation and target element positioning, and uses dynamic content grabbing and structure reorganization technology (such as dynamic DOM analysis and data grabbing script), finally converts interface information into usable structured business data through rule engine and template mapping, realizes automatic collection and conversion of interface data. For example, the sales revenue provisional estimate intelligent processing unit uses this technology to automatically locate and fill in data in the business detail page of the ERP system.

[0072] (2) Automatic data feedback and writing mechanism:

[0073] The processing results generated by the system described in this embodiment are accurately and reliably fed back to the ERP system in the following ways:

[0074] Standardized instruction feedback and execution: After the business logic processing is completed, the system integrates an automatic process engine to return the generated standardized accounting instructions and associated data to the ERP system via a secure interface, driving it to automatically generate financial documents that meet the accounting standards. For example, the checkout management module fills the generated standardized data fields into the voucher template, and finally automatically generates financial documents such as accruals and transfer documents in the ERP.

[0075] Precise positioning and batch filling: Support automatic jump to specified business nodes (such as voucher entry, account management page) in ERP system, and fill the data after system verification and conversion into target fields in batch mode, forming a full-process automation closed loop from data collection to business operation. For example, the batch generation of sales provisional estimate and the batch submission of payment application in the sales provisional estimate intelligent processing unit and the fund allocation payment intelligent application unit.

[0076] Robust exception handling mechanism: During data writing, the system has built-in verification and fault tolerance logic. When detecting data format exceptions, verification failures, or system response timeouts, etc., the exception handling process is automatically triggered: record detailed logs, skip the current exception entry and continue to execute subsequent tasks, so as to ensure the high availability and continuity of core business processes. This technology is reflected in the production cost transfer automatic business unit and the intelligent supplier account management unit, effectively ensuring the stable operation of the automated process.

[0077] Further, the accounting intelligentization subsystem comprises:

[0078] The checkout management module is configured to perform data extraction on the financial data obtained from the ERP system through a regular expression, perform integrity verification and semantic analysis on the extracted financial data, and generate standardized data fields; by calling a specified voucher template and filling in the standardized data fields, a financial document conforming to the accounting standards is generated.

[0079] The checkout management module comprises:

[0080] The welfare expense transfer automatic business unit is configured to automatically download the subject balance table from the ERP system, match the preset keywords in the subject name column of the subject balance table line by line according to the preset regular expression, extract all matched line data as welfare expense related entries, perform semantic analysis and integrity verification on the welfare expense related entries, and generate standardized data fields; call the voucher template and automatically fill in the standardized data fields to generate a standard format of the accrual.

[0081] In the present embodiment, the welfare fund transfer automatic generation business unit automatically downloads the latest subject balance table file by integrating with the ERP system, and parses the subject balance table using regular expressions combined with semantics, i.e., defining regular expressions according to preset keywords, such as "wages payable- welfare fund", "management expenses- welfare expenditure", etc. The regular expressions are used to match the preset keywords in the subject name column of the subject balance table line by line, and if the matching is successful, all the data of the line are extracted as welfare-related items. The accuracy and integrity of the data are further confirmed by semantic analysis of the matched items. According to the defined data integrity check rules, the integrity of the parsed data is checked. At the same time, the voucher template engine of the ERP system is called to automatically fill in the summary, subject and amount, etc. related information, and generate a standard format of the accrual.

[0082] For example, in the welfare fund transfer automatic generation business unit, after the data extraction is completed, the integrity check is performed based on the preset keyword rules, and the subject balance table is subjected to multi-dimensional semantic analysis.

[0083] The safety production fee automatic accrual unit is used to call the pre-withdrawal expense calculation template, extract the accrual data from the pre-withdrawal expense calculation template by combining the predefined regular expressions and semantic analysis, and then generate safety production fee accrual sheets in batches according to the accrual data by project dimension.

[0084] In the present embodiment, by parsing the pre-withdrawal expense calculation template, the key data such as project, cost center, amount, etc. are extracted using regular expressions combined with semantic analysis, and by integrating with the ERP system, the key data are grouped by project dimension and the accrual sheets are generated, i.e., the safety production fee accrual sheets are automatically generated by using automatic programs, to ensure the accuracy and timeliness of the safety production fee accrual.

[0085] The allocated company layer fee automatic generation business unit is used to automatically filter out the business detail data of different projects of each company from the business data stored in the preconfigured data source according to the filtering conditions defined in the pre-set company project adjustment template, and to perform data cleaning and format unification to generate standardized business detail data files; using parallel computing technology, the business detail data files are matched and summarized with the preset fee allocation rules and the general ledger template to generate the allocated company fee general ledger.

[0086] In the embodiment, each company sub-project adjustment template is preset, and according to the screening conditions in the template, the business detail data of different projects of each company is automatically screened from the massive relevant business data stored in the pre-configured data source, that is, the memory and the preset dictionary, and data cleaning and format unification are performed to generate a standardized business detail data file. The business detail data file contains project name, amount, cost center and other key fields. The generated business detail data file is matched with the preset cost allocation rule and the general ledger template, wherein the preset cost allocation rule and the general ledger template contain the mapping relationship between the project and the cost center, the allocation logic rule, such as subject code 40 representing debit and 50 representing credit, and the summary field definition. Parallel computing technology is used to accelerate the above data matching process, and the cost allocation company general ledger is automatically summarized to realize automatic management of the cost.

[0087] For example, when the cost allocation company layer cost is generated, the system checks the screened business detail data of each company, specifically: through format standardization processing, it is ensured that the data meets the requirements of the preset template.

[0088] The production cost settlement automatic generation business unit is used to obtain the subject balance detail data in the form of a two-dimensional table structure by directional analysis of the subject balance table automatically downloaded from the ERP system; based on the preset configuration table and the dynamic association matrix, the cost center and the accounting subject in the subject balance detail data are used as query conditions to obtain the target settlement voucher subject, and the debit amount of the subject balance detail data is automatically filled into the voucher template according to the target settlement voucher subject; a flow control mechanism is used to check data processing abnormalities and record logs; and finally a standardized cost settlement single meeting the accounting standards is automatically generated.

[0089] In this embodiment, for the subject balance table obtained from the ERP system, the Excel file exported by the ERP system is parsed by pandas, that is, the unstructured Excel data exported by the ERP system is converted into a standardized two-dimensional relational data structure by directional parsing, so as to realize the conversion from semi-structured financial data to structured data. Based on the preset rules, that is, by configuring a table and a dynamic association matrix such as an external Excel mapping table, the cost center and the accounting subject are accurately mapped. Therefore, the debit amount of the subject balance table is automatically filled in the voucher field according to the preset rules, so as to realize the automatic generation of the standardized cost transfer sheet in accordance with the accounting standards, and improve the efficiency and accuracy of the cost processing. At the same time, in order to ensure the integrity and reliability of the above operation process, the flow control mechanism is used to check the data processing exception in this embodiment. If the check is passed, the data processing process is continued; if the check finds an exception, the current abnormal data is skipped and a log is recorded, and then the data processing process is continued, so as to avoid interruption of the process, that is, the unit contains exception handling, for example, skipping the current loop when the data is not found, but the user needs to ensure that all dependencies are correct to avoid runtime errors. Finally, the standardized cost transfer sheet in accordance with the accounting standards is generated.

[0090] The income management module is used for obtaining unstructured financial data files, and extracting structured financial data from the financial data files by intelligent data parsing and rule mapping technology, and automatically generating standardized financial vouchers in accordance with the preset rules and the extracted financial data.

[0091] In this embodiment, the income management module is used for enterprise income accounting related processes, specifically including: sales income confirmation automatic generation process, sales income temporary estimation intelligent processing process and collection flow automatic generation process. It should be noted that the unstructured wind power financial data files obtained by the income management module are all from specified external data sources, such as obtaining the wind power income balance table and the power settlement sheet from the power grid company, obtaining the temporary electricity fee income sheet from the enterprise marketing management system or the financial sharing platform, and obtaining the collection flow detail file from the enterprise online banking system or the fund management platform. The standardized financial vouchers in accordance with the accounting standards include: electricity fee confirmation sheet, sales temporary estimation sheet and collection flow record.

[0092] The income management module includes:

[0093] The sales income confirmation automatic generation unit is used for obtaining the income balance table from the specified external data source, using multi-threading technology to traverse the income balance table in parallel, extracting sales income data, and then automatically generating the electricity fee confirmation sheet according to the sales income data; at the same time, the power settlement sheet issued by the power grid company is automatically uploaded.

[0094] In the embodiment, the sales revenue confirmation automatic generation unit obtains the intelligent association relationship between the power consumption data and the financial subjects by reading the wind power revenue balance table, so as to realize the automatic generation of the sales revenue confirmation, simplify the electricity fee confirmation process, and reduce manual intervention. Specifically, the sales revenue confirmation automatic generation unit automatically obtains the wind power revenue balance table and uses the multi-thread technology to traverse the data in the table in parallel, extracts the sales revenue data including the power, the price, the amount and the like, and automatically generates the sales revenue confirmation, i.e., the electricity fee confirmation sheet. Meanwhile, the electricity grid company can automatically upload the power balance sheet issued by the electricity grid company, so as to realize the digitization of the whole process of the electricity fee confirmation.

[0095] The sales revenue provisional estimation intelligent processing unit is used to obtain the provisional electricity fee revenue sheet from a specified external data source, and obtain the original data by analyzing the provisional electricity fee revenue sheet. Based on the multi-level data mapping mechanism, the original data is dynamically converted into the standard data recognizable by the ERP system. Through the automatic technology, the business detail page of the ERP system is automatically jumped according to the preset menu path, and the standard data recognizable by the ERP system is automatically filled in, so as to batch generate the sales provisional estimation sheet.

[0096] In the embodiment, the sales revenue provisional estimation intelligent processing unit dynamically generates the price and power data and the like according to the configuration table and the information of the corresponding business detail page, so as to improve the intelligent level of the sales revenue provisional estimation processing. Specifically, the provisional electricity fee revenue sheet is analyzed based on the Python related library to obtain the original data including the company information, the profit center information, the business type, the power data, the price data, the amount data and the like. The original data is mapped through the multi-level data mapping mechanism, so as to realize the accurate association and management of the multi-level data. The multi-level data mapping mechanism includes that the first dimension covers the company code, the profit center code and the like, the second dimension includes the business type, and the original data analyzed from the provisional electricity fee revenue sheet is mapped through the three dimensions of “company+profit center+business type” to generate the mapped structured data. At the data processing level, the price and power data are dynamically matched according to the three dimensions of “company+profit center+business type”, so that the configuration table information can be accurately matched according to the dimensions, and the profit center abbreviation and full name can be automatically associated through the fuzzy matching function, so as to ensure the dynamic and accuracy of the data filling. In addition, the unit also has the data batch filling capacity, integrates the UiPath automatic robot and the ERP system, automatically jumps and completes the automatic operation of the business detail page through the preset menu path, that is, the batch filling of the company name, the profit center, the power, the price and the like is realized, and the sales provisional estimation sheet is generated based on the filled data, so as to realize the whole process automation from the data processing to the sheet generation.

[0097] The collection flow automatic generation unit is configured to obtain a collection flow detail file from a designated external data source, extract key information of the collection flow from the collection flow detail file by combining a predefined regular expression and semantic analysis, take a profit center in the key information of the collection flow as a correlation key, query an income account book by SQL to obtain a total settlement receivable amount of the profit center and verify the total settlement receivable amount, and automatically generate a collection flow record in accordance with the verified key information of the collection flow and in accordance with rules.

[0098] The key information of the collection flow includes a profit center, a collection party account, a collection party bank name, an amount, a company code, a customer number, a collection method, a currency, and a transaction date.

[0099] In the embodiment, the collection flow detail file is parsed by using a pandas library of Python, and the key information of the collection flow is extracted by using a regular expression and combining semantic analysis. In the extraction process, the profit center field contains both “structured” and “unstructured” characteristics, so the embodiment proposes a method of supporting both coding rule matching (regular expression) and fuzzy matching (semantic analysis) to extract the profit center. Specifically, a predefined regular expression is used to identify a standardized profit center code, semantic analysis is used to identify an irregular profit center name, and the profit center name is converted into a standard profit center code by calculating a similarity. Based on the extracted profit center, an income account book of wind power is queried by SQL to obtain a corresponding total settlement receivable amount. Subsequently, a record of the collection flow in accordance with rules is automatically generated by integrating with an ERP system, and full-process automation from data parsing to flow generation is realized.

[0100] The treasurer application module is configured to traverse each entity according to a preconfigured entity list, for each entity, automatically obtain fund management business data of the entity, and execute a predefined fund management task based on the obtained fund management business data; wherein the fund management task includes at least one of an intelligent supplier account clearing management task, an un-cleared account intelligent management task, a fund budget execution intelligent management task, or a fund allocation payment intelligent application task.

[0101] In the embodiment, the treasurer application module assists fund management work from different links around the treasurer business, and specifically includes an intelligent supplier account clearing management process, an intelligent supplier account clearing management process, a fund budget execution intelligent management process, and a fund budget execution intelligent management process.

[0102] The treasurer application module includes:

[0103] The intelligent supplier account clearing management unit is configured to traverse each company code according to a preconfigured company code list, automatically obtain the account data of each company and perform data cleaning, and then perform account clearing on the cleaned data according to a preset account clearing rule, and automatically perform a skip operation and record a log for an abnormal company that cannot be normally cleared due to the nonexistence of the company code.

[0104] In this embodiment, each company is traversed in sequence through the preconfigured company code list. During the traversal process, the intelligent supplier account clearing management unit supports abnormal skipping, that is, when a company code does not exist in the ERP system, the company code is automatically skipped and a log is recorded. Based on the UiPath automation robot framework, the report module is automatically operated by automatically navigating through the menu path of the ERP system. The automatic operation is to obtain the account data of each company and perform a data cleaning process, and to uniformly convert the fields such as amount and date into numerical or date types to avoid calculation errors caused by text formats. Based on the cleaned data, the supplier account clearing operation is automatically performed according to the built-in account clearing strategy, and the efficiency of fund management is improved.

[0105] The un-cleared account intelligent management unit is configured to traverse each company code according to a preconfigured company code list, automatically configure query parameters for each company code, load the accounts receivable and accounts payable of each company according to the query parameters by using an intelligent waiting mechanism, and perform a paging query on the loaded accounts receivable and accounts payable according to a configured subject and batch export the query results in a preset format.

[0106] In this embodiment, the companies are traversed in sequence based on the preconfigured company code list by integrating with the treasurer system. Based on the UiPath automation robot framework, the query parameters such as company code and accounting period are automatically configured, and the data loading is processed by using an intelligent waiting mechanism. After the data loading, the accounts receivable and accounts payable are queried in pages, batch export is supported, and the format is standardized. Finally, the accounts receivable and accounts payable are automatically queried and batch exported, which facilitates the management and analysis of the un-cleared accounts by enterprises.

[0107] The fund budget execution intelligent management unit is configured to traverse each company code according to a preconfigured company code list, and automatically query the fund plan data of each company according to a preset multi-dimensional combined query condition. The fund plan data of each company is batch exported in the company dimension to generate a standardized file. The standardized file is verified, cleaned and standardized integrated by using a multi-level intelligent processing mechanism to generate structured fund plan data.

[0108] In the embodiment, the fund budget execution intelligent management unit supports automatic real-time query, data cleaning and integration functions of the fund plan of each company, and realizes efficient management of the fund budget execution. Specifically, the fund budget execution intelligent management unit integrates the ERP system and the treasurer system. According to the preset rules, the codes of each branch company are sequentially traversed. The unit supports multi-dimensional combination query function, that is, users can accurately filter the required fund plan data based on multi-dimensional combination query conditions including accounting period, fund nature, project type and other information. At the same time, the batch check function of the unit supports one-key export of standardized Excel or CSV files according to company dimension, which greatly improves the data acquisition efficiency. For the exported fund plan data, that is, the standardized file, a multi-level intelligent processing mechanism is immediately started: on the one hand, through the data verification engine, the preset verification rules are used to automatically identify missing values and logical conflict values in the data; on the other hand, the fund plan fields are standardized and integrated, so as to form structured fund plan data, providing standardized and high-quality data support for subsequent dynamic fund scheduling.

[0109] The fund allocation payment intelligent application unit is used for traversing each company code according to a preconfigured company code list, querying and loading fund allocation payment application information of each company by using multi-thread parallel processing technology, performing integrity check on the fund allocation payment application information of each company, and batch submitting the fund allocation payment application information that passes the check for approval.

[0110] In the embodiment, the fund allocation payment intelligent application unit integrates the ERP system and the treasurer management system, automatically traverses each company code based on the preconfigured company code list by using multi-thread parallel processing technology, and executes intelligent query of fund allocation payment application information by using UiPath automation robot. In the batch submission link, the fund allocation payment application information is automatically checked for data integrity, and after passing the check, the fund allocation payment application information that meets the conditions is automatically checked and batch submitted for approval, realizing automatic flow of fund payment application and improving the efficiency and standardization of fund payment application.

[0111] The auxiliary management module is used for automatically obtaining business data under different business scenarios and executing a predefined enterprise financial and management task based on the business data; wherein the enterprise financial and management task includes at least one of an electronic invoice intelligent management task, a bank diary account intelligent query task, a financial project profit automatic grabbing task, an invoice application single automatic generation task or an operating index report automatic generation task.

[0112] In the embodiment, the auxiliary management module assists the enterprise financial and management work from different dimensions, specifically including: electronic invoice intelligent management process, bank diary account intelligent query process, financial project profit automatic extraction process, invoice application automatic generation process and operation index completion report automatic generation process.

[0113] The auxiliary management module comprises:

[0114] The electronic invoice intelligent management unit is configured to scan a specified folder and automatically identify invoice files through file extensions; a file fingerprint library is established based on a SHA-256 algorithm, and all identified duplicate files are filtered out by using the file fingerprint library, and then the filtered invoice files are uploaded to an invoice folder in batches.

[0115] In the embodiment, the unit supports automatic uploading of invoices, improving the invoice processing efficiency. Specifically, a specified folder is scanned, and invoice files are automatically identified through file extensions including.pdf / .ofd / .jpg. The unit constructs a file fingerprint library based on a SHA-256 hash algorithm, generates a unique hash value fingerprint for each invoice file, and compares it with the fingerprint of the scanned invoice file, thereby realizing efficient and accurate duplicate file filtering, ensuring the uniqueness of the files uploaded to the ERP invoice folder, and avoiding duplicate uploading. An automatic tool is used to simulate an operation of opening a file dialog box, i.e. a pop-up window, and the filtered invoice files are uploaded to the invoice folder in the ERP system in batches.

[0116] The bank diary account intelligent query unit is configured to automatically query the bank diary account and batch export according to preset filtering conditions.

[0117] In the embodiment, the bank diary account intelligent query unit is integrated with the ERP system, and the query is triggered according to the user's preset filtering conditions. The filtering conditions include: company code, accounting year, accounting period, bank account number and other core filtering conditions. The query results are batch exported in Excel / PDF format, meeting the data application needs of different scenarios and facilitating the enterprise's monitoring of bank account fund flow. It should be noted that the unit supports multiple company batch queries by simultaneously inputting multiple company codes in one query operation; the accounting period can be a natural month or a custom period; for bank account number query, the unit provides two modes of fuzzy query and accurate query.

[0118] The financial item profit automatic grabbing unit is used for acquiring a multi-level account balance table based on a company and a profit center in real time, dynamically checking the multi-level account balance table through a double closed loop checking mechanism based on a rule engine and a generative adversarial network, automatically identifying and extracting profit and loss type index data from the checked data through an intelligent semantic mapping engine, and automatically filling the profit and loss type index data into a preset intelligent analysis template to generate a financial item profit analysis report.

[0119] In the embodiment, the multi-level account balance table is acquired from an ERP system in real time in a two-level architecture sequence of "company -> profit center" according to a preconfigured profit center list. It should be noted that the multi-level account balance table has intelligent fault tolerance capability when being acquired automatically, that is, when the profit center code is invalid, the acquisition is automatically skipped and a log is recorded, and when the data volume exceeds 100,000 rows or the set file size, a block processing mechanism is automatically triggered to guarantee the efficiency and stability of data processing.

[0120] The specific method for dynamically checking the multi-level account balance table through the double closed loop checking mechanism based on the rule engine and the generative adversarial network is as follows:

[0121] The original data in the multi-level account balance table is cleaned through a standardized rule engine, and the cleaned data is converted into structured data with business semantics according to a preset conversion rule.

[0122] The conversion rule includes but is not limited to a subject mapping rule, an entity association rule and a data polarity conversion rule. The subject mapping rule is used for realizing accurate mapping from a subject code to a business index, the entity association rule is used for realizing intelligent association from a profit center to a project level, and the data polarity conversion rule is used for realizing dynamic conversion from a debit-credit direction to an index polarity. These conversion rules are realized through a multi-level condition judgment and a dynamic query mechanism, are used for completing semantic mapping conversion of a business index, and ensure accurate expression of business semantics. The converted structured data is a data set containing standardized fields, mainly includes a project, an index type, a numerical value and the like, and will be used as an input of a subsequent generative adversarial network checking process.

[0123] The pre-trained adversarial generative network is used for logical contradiction detection on structured data with business semantics; if structured data with logical contradiction is detected, the standardization rule engine is fed with abnormal information of the logical contradiction; the standardization rule engine locates the original data corresponding to the structured data according to the received abnormal information, corrects the original data by calling a predefined correction rule, re-performs data cleaning and data conversion on the corrected original data, generates new structured data with business semantics and re-performs logical contradiction detection, until the adversarial generative network cannot detect logical contradiction from the current structured data, and obtains verified data.

[0124] The abnormal information includes but is not limited to: logical contradiction type, data value causing logical contradiction and unique identifier of the located original data item.

[0125] In the embodiment, the automatic financial item profit extraction process refers to intelligent matching and extraction of profit and loss indicators according to the subject balance table automatically obtained from the ERP system, automatic filling of the analysis template, and provision of accurate data support for financial decision-making. In this process, in view of the defect that the prior art is difficult to identify implicit logical contradiction, the embodiment proposes an intelligent financial data verification system based on an adversarial generative network, realizes multi-level abnormality detection through a "rule engine + adversarial verification" double closed loop mechanism, and ensures the logical consistency of profit data. Specifically, in the data intelligent processing stage, the adversarial generative network technology is introduced, the original data is first cleaned and the semantic mapping conversion of business indicators is completed through the standardization rule engine, and then the dynamic verification process driven by the adversarial generative network is started, the logical contradiction is detected and fed back to the cleaning module for iterative correction, thereby forming a double closed loop mechanism of "rule engine first level verification + adversarial generative network adversarial second level verification". Since the traditional rule has the problem of covering the out-of-range value, the embodiment uses the adversarial generative network mechanism to accurately capture the implicit conflict such as balance sheet abnormality, thereby improving the reliability of the data. Finally, the loss and benefit indicator data is automatically identified and extracted through the intelligent semantic mapping engine, and then automatically filled into the preset intelligent analysis template to generate a financial item profit analysis report, realizing end-to-end automation closed loop from data collection to decision support.

[0126] The invoice application automatic generation unit is used for acquiring the income account book and performing line-by-line analysis according to the preset rule, extracting data fields respectively constituting basic business information, financial and tax information and invoice detail information from the income account book in combination with regular expressions and semantic analysis; the extracted data fields are filled into the tax system standard form by calling the invoice application template engine to generate an invoice application form.

[0127] In this embodiment, the wind power income account is parsed line by line by the openpyxl library of Python, and non-business rows are skipped according to preset rules, regular expressions and semantic analysis are combined to locate key data fields, and data fields corresponding to company code, profit center, project name, business type, unit on-grid settlement power, payable power purchase fee and other information are extracted. Based on the extracted data field, the data is mapped to the standard form of the tax system by automatically calling the invoice application template engine, and the invoice application form is intelligently generated, simplifying the invoice application process and supporting progress visualization monitoring throughout the process.

[0128] The operating indicator completion condition report automatic generation unit is used to automatically collect ERP financial data in a monthly cycle, and to parse the ERP financial data piece by piece by using a missing value dynamic interpolation method based on business rules and a derived field recalculation method to generate an operating analysis report.

[0129] In this embodiment, the operating indicator completion condition report automatic generation process refers to triggering a data collection task in a preset monthly cycle, automatically collecting ERP financial data through integration with an ERP system, and the ERP financial data includes structured data of a profit table, an income detail table, a strategic emerging industry income statistical table, and a new energy power generation industry operating condition table. After data collection is completed, a context-aware ETL processing engine is started immediately, and based on a preset subject type-precision rule mapping table, the precision requirements of different financial indicators are automatically adapted, such as income retaining 2 decimal places and power generation taking integer.

[0130] In this embodiment, the context-aware ETL processing engine automatically identifies data characteristics and matches the best processing strategy by dynamically analyzing the meta-information of ERP data, such as subject type, business scenario, and period label. The engine starts immediately after data collection, first analyzes the data context, for example, judges whether the field belongs to a currency subject, a derived indicator or a statistical dimension, and then calls the scenario processing scheme in the intelligent business rule engine, such as using moving weighted average interpolation for currency missing values, and recalculating the derived fields by reversing the analysis formula. The operating indicator completion condition report automatic generation unit realizes efficient and accurate conversion from raw data to analysis-ready data through this context-driven and rule-coordinated way, providing high-quality input for subsequent report generation. Finally, by calling the template document generation engine, a high-precision operating analysis report is automatically generated, supporting Word standardization one-key export and key indicator customization, and providing an efficient and accurate intelligent financial reporting solution for enterprises.

[0131] The specific content of the method of using a missing value dynamic interpolation method based on business rules and a derived field recalculation method to parse the ERP financial data piece by piece to generate an operating analysis report is as follows:

[0132] A mapping relationship table of subject type and precision rule is constructed by analyzing the data dictionary of the ERP system.

[0133] In the embodiment, a mapping relationship table of subject type-precision rule is constructed by analyzing the metadata tags of the ERP data dictionary, so as to break through the limitation of the traditional ETL unified precision processing.

[0134] For any data in the ERP financial data, each field in the data is traversed, and the metadata tag of each field is obtained by querying the data dictionary of the ERP system; wherein the metadata tag is any one of a basic field or a derived field.

[0135] For each basic field, it is detected whether the subject is missing, if not, the basic field is converted into precision conforming to the business specification according to the mapping relationship table of subject type and precision rule and by dynamically matching the preset precision rule library; if yes, the business attribute of the basic field is automatically identified according to the mapping relationship table of subject type and precision rule, and the basic field is dynamically interpolated by using the moving weighted average method to generate the missing value of the basic field and convert it into precision conforming to the business specification.

[0136] In the embodiment, the interpolation algorithm is selected according to the business context of the field, and the missing value is interpolated in a scenario. When the subject is detected to be missing, the moving weighted average method is used for dynamic interpolation according to the precision requirement in the mapping relationship table, which is expressed as:

[0137] ;

[0138] Wherein is the missing value; is the weight coefficient; is a constant dynamically set according to the business rule, which is used to represent the number of historical periods used to calculate the average value; represents the index; represents the time point before the known historical observation value.

[0139] In the embodiment, the field business attributes are automatically identified from the ERP financial data according to the mapping relationship table of subject types and precision rules, and the fields extracted from the ERP system are dynamically matched with the precision rule library. The precision rule library is usually pre-constructed according to business requirements, and the rules in the precision rule library include but are not limited to: decimal place requirements of currency subjects, time series fields and static reference fields; wherein the decimal place requirements of the currency subjects are, for example, income data is accurate to ten thousand yuan, cost data is accurate to a penny, and the rest is kept to two decimal places; the time series field is a periodical aggregation label, such as monthly or quarterly periodical aggregation; and the static reference field is a constant that cannot be modified.

[0140] For each derived field, the calculation formula of the derived field is inversely analyzed from the data dictionary of the ERP system, the variable current effective values of variables in the calculation formula are obtained, and all the variable current effective values of variables are substituted into the calculation formula for recalculation to obtain the substitute value of the derived field and convert it into precision conforming to business specifications.

[0141] The calculation formula of the derived field is:

[0142] ;

[0143] Wherein represents a derived field; represents a predefined calculation formula; are variables in the predefined calculation formula.

[0144] ;

[0145] Wherein are variable current effective values of , respectively. It should be noted that the variable current effective value is the latest, compliant and logically consistent data dynamically selected from the data obtained by the ERP, combined with business rules, and if the original value is missing, it is filled by weighted average to ensure the accuracy of the calculation of the derived field.

[0146] After the precision conversion of all fields in the data is completed, the cleaned data is obtained, and the data cleaning of the ERP financial data is completed to obtain a standardized financial data set.

[0147] The standardized financial analysis template is called and the standardized financial data set is automatically filled in to generate an operating analysis report.

[0148] In this embodiment, based on the cleaned high-quality data obtained through the above data processing process, the financial report materials containing subject precision adapted core index analysis are automatically generated by calling the modularly designed intelligent document generation engine, and the Word format standardized export is supported. In the example of full-process automatic processing, if the automatic generation of a process fails, the system will automatically display the failure reason, and optimize the automatic process.

[0149] The financial management subsystem is used to integratively implement the to-do list management, compliance guidebook query and financial data structured storage according to the preset standardized process.

[0150] In this embodiment, the centralized mode is used to plan the financial data management and control, covering the to-do list management, compliance guidance and electronic database three functions, implementing the compliance requirements and strengthening the risk prevention and control. Specifically, through the standardized process and real-time data tracking, the financial management collaboration efficiency and decision support capability are improved, and the risk prevention and control capability is comprehensively strengthened. Through the built-in standardized compliance guidance system, the compliance requirements in various fields such as accounting, fund management and tax treatment are covered, which facilitates user query and learning, ensures the compliance of enterprise financial management business, and realizes the standardized, structured storage and efficient retrieval of financial data, providing data support for enterprise financial management.

[0151] The financial management subsystem comprises:

[0152] The to-do list management module is used to obtain the to-do list and visually display the to-do list by using the density-sensitive dynamic time axis compression algorithm; the content of the visual display includes: real-time synchronous display of file download and completion state of the to-do list, display of the to-do list through a monthly calendar view, and automatic push of a reminder message according to the deadline of the to-do list.

[0153] In this embodiment, the to-do list management module proposes to dynamically present the to-do list by using an intelligent clustering algorithm, displays the to-do task in a time axis view, supports multi-dimensional visualization and one-key operation, facilitates users to view the to-do list, download the to-do list related files, save the completion state of the to-do list, view all to-do lists in the month, automatically push a reminder message near the deadline, support uploading the completion record, and realize the association and traceability of the document version and the task state, and comprehensively improve the financial collaboration efficiency. Through the combination of the intelligent interaction layer and the data processing layer, the visual efficiency and version tracking accuracy of the financial to-do list management are significantly improved.

[0154] The specific content of the visual display of the to-do list by using the density-sensitive dynamic time axis compression algorithm is:

[0155] Set a time window , real-time statistics of the total number of to-do items in the current time window and the weight coefficient of each to-do item; wherein is the time interval span; is the display area pixel width.

[0156] Calculate the item distribution density of the current time window , expressed as:

[0157] ;

[0158] wherein is the weight coefficient of the th to-do item.

[0159] In this embodiment, the total number of to-do items m in the current time window is counted by a predefined query, such as the query statement SELECT COUNT(*) FROM tasks WHERE deadline BETWEEN [T-W] AND [T+W]. If there is a weight coefficient, the weight system of each item is also obtained synchronously. When the user adds, completes or modifies a to-do item, an event listener is triggered to update the item distribution density D in the cache.

[0160] In this embodiment, when , the multi-dimensional clustering engine based on the improved K-means algorithm is started, the time axis equal division priority strategy is used to initialize the clustering center, and the key node set is preserved during the iteration process. When the density threshold exceeds the preset value, the items are aggregated according to the two-dimensional feature matrix of business type and urgency, and the summary node with quantity label is generated. Through the preset rules, the core items are always exposed, and important information is avoided from being folded. The clustering state on the time axis is synchronized to the calendar view in real time, and the month view shows the item density in the date dimension.

[0161] Compare the item distribution density of the current time window with the preset item distribution density threshold , when , the improved K-means algorithm is used to cluster the to-do items in the current window, and the specific content is as follows:

[0162] According to the business type and urgency of two dimensions, all to-do items in the current time window are divided into several groups;

[0163] For each group of to-do items, the time interval to which all to-do items in the group belong is divided into several equal length subintervals using the time axis equal division priority strategy, and the midpoint of each subinterval is taken as an initial clustering center of a cluster.

[0164] For any to-do item in each group , calculate the weighted Euclidean distance between the to-do item and all initial cluster centers, and assign the to-do item to the initial cluster center with the minimum weighted Euclidean distance.

[0165] The weighted Euclidean distance is defined as follows:

[0166] ;

[0167] wherein represents the feature vector of the to-do item; represents the feature vector of the cluster center; and are both business type codes; and are both urgency scores; and are both preset weight coefficients, and , .

[0168] After all to-do items are assigned, the cluster centers of each cluster are updated, and each to-do item is assigned to a new cluster center by recalculating the weighted Euclidean distance between each to-do item and each cluster center until a preset maximum number of iterations is reached.

[0169] For the generated final clusters, a summary node with a quantity label is generated for each final cluster, and core items in the final cluster are selected according to a preset rule; the generated final clusters are displayed in the form of a summary node and a core item image in the calendar view, and the distribution density of items in the date dimension is displayed through the month view .

[0170] In this embodiment, the preset rule ensures that the core items are always exposed, avoiding important information being folded. The clustering state on the timeline is synchronized to the calendar view in real time, and the item density in the date dimension is displayed through the month view.

[0171] When there is a financial document in the file of the to-do item, the financial document is subjected to structured semantic analysis to extract business entities, business entity attributes, and relationships between business entities in the financial document.

[0172] The extracted business entity attributes are taken as leaf nodes, and the business entities and relationships between business entities are taken as non-leaf nodes to construct a semantic dependency tree.

[0173] ​​In the embodiment, by performing structured semantic analysis on the financial document, key business entities and their attribute relationships are extracted, thereby constructing a semantic dependency tree with hierarchical association characteristics; wherein the leaf nodes correspond to specific data fields, and the non-leaf nodes represent business logic association.

[0174] After each modification of the financial document, a new semantic dependency tree is constructed according to the new version of the financial document, and by comparing the semantic dependency trees before and after the modification of the financial document, incremental changes at the business entity level are extracted and a version modification chain with a time stamp is constructed.

[0175] By matching the semantic dependency tree and the predefined to-do list state rule, a bidirectional association relationship between the version modification chain and the to-do list state change is established.

[0176] Based on the bidirectional association relationship between the version modification chain and the to-do list state change, the current completion state of the to-do list is obtained and saved, and a revision traceability graph supporting interactive operation is generated.

[0177] For version management of the financial document, to break through the technical limitations of traditional binary difference comparison, the embodiment realizes intelligent version control through the following technical solutions: first, a business entity depth analysis module: by performing structured semantic analysis on the financial document, key business entities and their attribute relationships are extracted, thereby constructing a semantic dependency tree with hierarchical association characteristics. Second, a version-state mapping engine: based on the semantic dependency tree, a bidirectional association relationship between the document version modification chain and the to-do list state change is automatically established, the business entity level change and the to-do list state rule are dynamically bound, the bidirectional association index is established using a graph database, and the automatic synchronization of version and state is realized. Finally, a visual traceability interface: it generates a revision traceability graph supporting interactive operation, and synchronously displays the document version evolution path and the completion state of the associated to-do list in the form of a topological graph, wherein the node size dynamically reflects the modification impact range. Based on the above technical solutions, document version comparison enriches business semantic understanding, and at the same time breaks the problem of low collaboration efficiency caused by data isolation between views.

[0178] For example, in to-do list management, if the deadline of a to-do list is approaching, i.e. the time between the current date and the deadline of the to-do list is less than a preset threshold, a message is sent to the relevant personnel for reminding, so as to avoid overdue tasks.

[0179] The guide manual query module is used to query the financial management business compliance guide manual in the built-in standardized compliance guide system according to the keywords or categories provided by the user.

[0180] In the embodiment, the built-in standardized compliance guide system covers compliance requirements in various fields such as accounting, fund management, and tax treatment. Through keyword search and classified browsing, the user can conveniently query and learn the financial management business compliance guide manual required by the user, and ensure the compliance of the enterprise financial management business. For example, when the financial personnel performs accounting operation, the user can quickly query the relevant compliance standards through the keyword search function of the compliance guide manual, and ensure that the operation conforms to the regulations.

[0181] The financial electronic database is configured to obtain financial data related to funds, budgets, accounting, taxes, and assets from a specified data source, and build a multidimensional index according to the financial data; and the financial data is stored in a structured manner according to the multidimensional index.

[0182] In the embodiment, the financial electronic database realizes the standardized, structured storage and efficient retrieval of financial data, and provides data support for enterprise financial management.

[0183] The business service subsystem is configured to provide a plurality of integrated service modules for the user, and to call a standardized operation guide file corresponding to the service module from a preset business document library according to an access request of the user to any service module.

[0184] In the embodiment, the business service subsystem covers tax-related business services, asset management services, fund management services, and operation management services. By providing standardized operation guide files for the user, the user can be guided to complete the corresponding business processing flow, and the enterprise can be provided with tax-related business support, improved asset management efficiency, and efficient fund management. At the same time, through modular design, seamless connection from front-end business to back-end finance is realized, digital transformation and green office are promoted, and all-round operation management and financial support are provided for the enterprise.

[0185] The business service subsystem comprises:

[0186] The tax-related business service module is configured to store various tax-related business operation guide files formulated based on tax laws; receive an access request of a user to a specified tax-related business; call an operation guide file corresponding to the specified tax-related business according to the access request and provide the operation guide file to the user; and the specified tax-related business is at least one of a land occupation tax business, a personal income tax withholding business, an enterprise income tax declaration business, and a value-added tax input tax deduction business.

[0187] In this embodiment, the tax-related business service provides professional tax-related business support for enterprises, covering farmland occupation tax business, personal income tax business guidance, enterprise income tax business guidance, value-added tax input tax deduction, etc. Through this service, tax policies can be more accurately grasped, tax operation processes can be standardized, tax risks can be effectively reduced, tax incentives can be reasonably utilized, tax burdens can be reduced, and the smooth development of tax-related business can be ensured.

[0188] Specifically, the farmland occupation tax business is to divide the land use cost according to the actual situation during project planning, guide the user to divide the land class according to the land use approval document, clarify the tax scope and tax time of various land and forest land related approval documents, calculate the tax payable, and complete the tax payment and data archiving within 30 days. The personal income tax withholding service supports monthly collection of employee salary data, checking of special additional deduction information, and completion of wage and salary, labor remuneration, and annual one-time bonus reporting and withholding through the electronic tax bureau of natural persons. The personal income tax service clarifies the scope of company welfare expenditure related to personal income tax and the annual personal income tax settlement time node. The enterprise income tax declaration business clarifies the impact of the signing subject of the company contract on enterprise income tax, and performs quarterly prepayment and annual settlement of enterprise income tax, including: checking the total profit, adjusting the tax difference, and completing tax declaration on time. The value-added tax input tax deduction business is to complete invoice selection authentication, check the deduction details, and ensure accurate reporting of input tax within the tax period each month; standardize invoice acquisition and stamping requirements, and clarify the input tax transfer-out situation of collective welfare, etc. All the above tax-related businesses need to strictly follow the tax law, complete tax declaration on time, and properly keep relevant tax file data.

[0189] The asset management service module is configured to store standardized operation guideline files of various asset management services, receive a user's access request for a specified asset management service, and provide the user with a standardized operation guideline file corresponding to the specified asset management service according to the access request. The specified asset management service can be at least one of a commissioning preparation service, a solid conversion management service, a completion settlement service, a temporary asset valuation service, and an asset inventory service.

[0190] In this embodiment, the asset management service module is committed to improving asset management efficiency by managing various assets throughout their life cycles, tracking asset status in real time, ensuring that asset accounts are consistent, optimizing asset allocation, improving asset utilization, and reducing asset idle and loss risks. The life cycle management includes a commissioning preparation service, a solid conversion management service, a completion settlement service, a temporary asset valuation service, and an asset inventory service.

[0191] Specifically, the commissioning preparation service is the acceptance preparation work before the project is put into operation, including collecting the commissioning application form, acceptance certificate and operation data, etc. materials, determining the conditions for the equipment to reach the predetermined usable state, and providing compliance guidance for the formal operation of the capital construction project. The standardization operation guide document of this service requires that the financial department needs to cooperate with the engineering department to audit the completion final accounts data, establish the fixed asset card file, and determine the fixed point according to the grid connection acceptance report.

[0192] The fixed asset management service standardizes the fixed asset transfer into formal management process, covering technical improvement engineering and capital construction project transfer, and clearly stating that technical improvement engineering needs to complete transfer before the closing date, and capital construction project needs to complete temporary estimated transfer in the month when the asset is available. A tracking mechanism is established to standardize the allocation processing of temporary estimated transfer and formal transfer, ensure the accuracy of asset original value and depreciation accounting, and protect the compliance of asset management. The standardization operation guide document of this service requires that the financial personnel complete the technical improvement engineering entry processing, perform the capital construction project temporary estimated transfer operation, and ensure that the asset classification and value allocation comply with the accounting standards.

[0193] The completion final accounts service standardizes the capital construction project completion final accounts work, clearly stating that the final accounts preparation needs to be completed within 3 months after the project is completed, and no more than 6 months in special cases. The standardization operation guide document of this service requires that the financial department leads the preparation of complete financial final accounts report, and the project site provides asset basic data; the audit department is responsible for the audit process, the engineering department provides acceptance data to ensure the accurate determination of asset value, and provides legal basis for the transfer of property rights.

[0194] The asset temporary estimation service standardizes the capital construction project asset temporary estimation transfer process. When the asset reaches the predetermined usable state, the engineering department prepares a temporary estimation sheet according to the unsettled engineering cost (excluding tax), and after the financial department's audit, the temporary estimation transfer is completed according to the asset category, ensuring that the asset is timely entered into the management. The standardization operation guide document of this service guides users to create asset inventory, update asset value, and associate and match accounting vouchers with project WBS elements.

[0195] The asset inventory service refers to the organization of inventory and inspection from the asset physical management department, the joint completion of inventory work by relevant business departments, the provision of business standards and processes according to the inventory requirements, and the final provision of inventory report, inventory table and inventory results. The standardization operation guide document of this service guides users to organize annual fixed asset inventory, prepare inventory detail table, and issue inventory report and implement rectification measures.

[0196] All the above asset management services must strictly follow the internal control specifications to ensure the accuracy and completeness of asset information and the compliance and effectiveness of management processes.

[0197] The fund management service module is configured to store standardized operation instruction files of various fund management services, receive a user's access request for a specified fund management service, and call and provide the user with a standardized operation instruction file corresponding to the specified fund management service according to the access request. The specified fund management service is at least one of a fund plan preparation service, a fund plan execution service, a special account use service, a special fund payment service, and a capital management service.

[0198] In the present embodiment, the fund management service module performs fund plan preparation, fund plan execution and monitoring, fund plan adjustment, special fund payment, and capital management. Specifically, the fund plan preparation service specifies the time limit requirement for reporting the fund plan online and offline each month, and dynamically adjusts the fund plan and re-performs the approval process in a timely manner in the event of major business adjustment or unexpected situation. The standardized operation instruction file of the service specifies that the time requirement and reporting mode requirement are provided to the business department from the online and offline modes, the business department is required to submit the fund demand on time, the financial department compiles the monthly fund plan after summarizing and performs the approval process, and the fund plan is filled in, summarized, and audited by relying on the treasury system online. The fund plan execution service provides the execution and monitoring requirement for the business department from the business department's audit of fund expenditure to the financial department's monitoring of fund flow direction and progress through the treasury system. The standardized operation instruction file of the service requires that the business department strictly audits each fund expenditure to ensure that it meets the plan arrangement, and the financial department monitors the fund flow direction and progress in real time through the system and timely warns of abnormal situations. When there is a business adjustment or unexpected situation, the fund plan needs to be dynamically adjusted and re-approved. In the special account use service, the special bank account is a bank settlement account opened according to the requirements of laws and regulations and limited in use to meet the needs of financing and other management. Except for policy provisions and regulatory restrictions, no more than one special bank account of the same category should be opened in the same financial institution. The special fund payment service covers enterprise special project funds and public welfare and charity funds, and the payment needs to strictly follow the approval and payment process and strictly prohibit the misappropriation of funds. The standardized operation instruction file of the service specifies the approval process and information traceability of special project funds and public welfare and charity funds to ensure the use of special funds and perform the information disclosure obligation. The capital management covers the registration of property rights, the use of supervision, and the management of changes, and the standardized operation instruction file of the service requires that the financial property rights department notifies the comprehensive department to issue a resolution of the shareholders' meeting according to the capital fund arrival reminder, handles the business registration procedures, and completes the registration and monitoring of property rights in the financial property rights department. The service automatically supervises and executes the registration, use, and change processes of the capital fund according to the pre-set laws and regulations, ensures that the process strictly complies with the laws and regulations such as the Company Law, standardizes the procedures of internal decision-making, external announcement, and business registration, and ensures the legality and compliance of the use of capital.

[0199] The operation and management service module is configured to store standardized operation guide files of various operation and management services, receive a user's access request for a specified operation and management service, and call and provide the user with a standardized operation guide file corresponding to the specified operation and management service according to the access request. The specified operation and management service is at least one of a current account management service, a budget control service, and a dividend payment management service.

[0200] In this embodiment, the operation and management service module provides comprehensive operation and management financial support, deeply integrates financial data with business operations, analyzes, monitors, and warns of financial indicators in the operation process, provides data support for operation decisions, discovers problems in the operation in a timely manner, and takes measures to improve. Meanwhile, the operation and management service module covers aspects such as financial accounting and tax-related matter management of a newly-established company, investment project pre-feasibility cost accounting and payment management, budget preparation and control points, economic business compliance management points, current account management points, wind power project life cycle management points, photovoltaic project life cycle management points, and wind power project large-to-small management, thereby improving the operation and management level and realizing sustainable development.

[0201] Specifically, the current account management service is a whole-process control of accounts receivable, accounts payable, and prepayments, which covers core contents such as document auditing, account management, periodic reconciliation, account age monitoring, and bad debt handling. The service establishes a risk warning mechanism through standardizing special processes such as business travel reimbursement and prepayment cancellation, and ensures the compliance of fund circulation and the prevention of financial risks. The standardized operation guide file of the service requires the finance department to audit business contract terms, perform accounting subject confirmation, periodically clean up accounts receivable and accounts payable, and complete the national capital supervision data reporting on time. The budget control service covers the decomposition of the cost budget to departments, the clear standard and process of monthly forecast assessment, quarterly difference analysis, and annual budget adjustment. The standardized operation guide file of the service guides the user to prepare a monthly rolling budget report, carry out budget execution deviation analysis, organize a quarterly economic activity analysis meeting, and complete annual budget preparation and decomposition. The dividend payment management standardizes the company's profit distribution and dividend payment process, including the required nodes of each work in the process and the division of responsibilities of different departments involved in the business. The standardized operation guide file of the service guides the user to develop a profit distribution plan, perform a party committee review procedure, complete an OA system approval process, and accurately handle dividend payment accounting. All operation and management businesses must strictly follow the enterprise accounting standards and the company's financial system to ensure standardized accounting, process compliance, and accurate data.

[0202] In summary, the embodiment deeply integrates information technology in operation and management, remolds the traditional financial management mode through automation, data integration, and intelligent decision-making, and the specific implementation manner includes:

[0203] (1) Whole-process automation and intelligent processing:

[0204] Data-driven automatic generation: By integrating with ERP systems, automatically download subject balance sheets, use regular expressions and semantic analysis techniques to accurately locate data such as "payroll-welfare fees" items, and generate financial documents such as accruals and settlement documents with one click, reducing more than 90% of manual input work.

[0205] Parallel computing and batch processing: In the context of allocating company expenses and confirming electricity charges, use multi-threading and parallel computing to speed up data matching and improve generation efficiency.

[0206] Intelligent exception handling: In the process of production cost transfer and account clearing management, the system automatically identifies data missing or format errors such as unstructured data conversion failures, skips abnormal items and records logs to avoid process interruption.

[0207] (2) Cross-system data integration and sharing:

[0208] The system is linked with ERP systems and treasurers: Through the UiPath robot framework, the system realizes real-time synchronization of data between ERP systems and treasurers, such as automatically retrieving bank statements and account balances when querying financial plans to ensure data consistency.

[0209] Multi-dimensional data mapping: In the sales provisional processing, through the "company + profit center + business type" three-dimensional dynamic matching of data, fuzzy matching is also supported, such as automatically associating profit center abbreviations and full names to improve data integration efficiency.

[0210] Standardized data storage: The financial electronic database uses structured storage methods to convert unstructured data into standardized fields, supporting fast retrieval and cross-module calls.

[0211] (3) Full-link data traceability and automated intelligent query:

[0212] Financial data traceability: Sales revenue confirmation can be traced back to the original power generation data of the wind power income balance sheet, realizing "business-finance" full-process traceability.

[0213] Multi-condition combination query: Bank journal, uncleaned account and other modules support multi-dimensional filtering such as company, period, amount, etc., such as automatically batch exporting the uncleaned accounts of a certain company, improving query efficiency.

[0214] (4) Decision support and compliance management:

[0215] Automatic financial index analysis: Extract profit and loss indicators from ERP systems and automatically fill them into analysis templates, support custom dimensions such as project and branch, generate financial reporting materials, and shorten report preparation time.

[0216] Embodiment 2

[0217] The electronic device can be a mobile phone, a computer, a tablet computer, or the like, and includes a memory and a processor. The memory stores a computer program, and the computer program, when executed by the processor, implements the functions of the enterprise financial comprehensive management system based on intelligent technology as described in the embodiments. It can be understood that the electronic device can further include an input / output (I / O) interface and a communication component.

[0218] The processor is configured to execute the operations of all or part of the functional modules of the enterprise financial comprehensive management system based on intelligent technology as described in the above embodiments. The memory is configured to store various types of data, which can include, for example, instructions of any application program or method in the electronic device, and application program related data.

[0219] The processor can be an Application Specific Integrated Cricuit (ASIC), a Digital Signal Processor (DSP), a Programmable Logic Device (PLD), a Field Programmable Gate Array (FPGA), a controller, a microcontroller, a microprocessor, or other electronic elements, and is configured to execute the functions of the enterprise financial comprehensive management system based on intelligent technology as described in the above embodiments.

[0220] The processor can be an Application Specific Integrated Cricuit (ASIC), a Digital Signal Processor (DSP), a Programmable Logic Device (PLD), a Field Programmable Gate Array (FPGA), a controller, a microcontroller, a microprocessor, or other electronic elements, and is configured to execute the functions of the enterprise financial comprehensive management system based on intelligent technology as described in the above embodiments.

[0221] Embodiment 3

[0222] The computer readable storage medium stores executable instructions, which, when executed, can be stored in a computer readable storage medium if implemented in the form of a software functional unit and sold or used as an independent product.

[0223] The computer software product is stored in a storage medium, and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the operations of all or part of the functional modules of the enterprise financial comprehensive management system based on intelligent technology as described in the embodiments.

[0224] The aforementioned storage medium includes a flash memory, a hard disk, a multimedia card, a card-type memory (e.g., an SD (Secure Digital Memory Card) or a DX (an abbreviation of Memory Data Register, MDR) memory, etc.), a random access memory (RAM), a static random access memory (SRAM), a read-only memory (ROM), an electrically erasable programmable read-only memory (EEPROM), a programmable read-only memory (PROM), a magnetic memory, a magnetic disk, an optical disk, a server, an APP (an abbreviation of Application) application store, and the like, which can store a program check code, and stores a computer program thereon, which, when executed by a processor, can implement the enterprise financial comprehensive management system based on intelligent technology.

[0225] Embodiment 4

[0226] The embodiment provides a computer program product, which includes a computer program or instructions, and the computer program or instructions, when executed by a processor, implement the enterprise financial comprehensive management system based on intelligent technology.

[0227] Based on such understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art, or the part of the technical solution can be embodied in the form of a computer program product.

[0228] Each of the embodiments in the present application is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other. Each embodiment mainly describes the difference from other embodiments.

[0229] The scope of protection of the present application is not limited to the above-described embodiments. Obviously, those skilled in the art can make various modifications and changes to the present disclosure without departing from the scope and spirit of the present disclosure. If these modifications and changes belong to the scope of the present disclosure and its equivalent technology, the present disclosure also includes these modifications and changes.

Claims

1. A comprehensive management system for enterprise finance based on intelligent technology, characterized in that, The system comprises: An accounting intelligent subsystem, which is configured to obtain financial data from an ERP system, perform semantic analysis and conversion on the financial data according to preset rules and a multi-level data mapping mechanism, generate standardized accounting business instructions, and deliver the standardized accounting business instructions to an automated process engine, so as to drive the automated process engine to automatically generate financial documents in compliance with accounting standards in the ERP system; The accounting intelligent subsystem comprises: A checkout management module, which is configured to perform data extraction on the financial data obtained from the ERP system through a regular expression, perform integrity checking and semantic analysis on the extracted financial data, and generate standardized data fields; and generate financial documents in compliance with accounting standards by calling a specified voucher template and filling in the standardized data fields; An income management module, which is configured to obtain an unstructured financial data file, extract structured financial data from the financial data file through intelligent data analysis and rule mapping technology, and automatically generate standardized financial vouchers in compliance with accounting standards according to preset rules and the extracted financial data; A treasurer application module, which is configured to traverse each entity according to a preconfigured entity list; for each entity, automatically obtain the financial management business data of the entity, and perform a pre-defined financial management task based on the obtained financial management business data; wherein the financial management task comprises at least one of an intelligent supplier account management task, an un-cleared account intelligent management task, a financial budget execution intelligent management task, or a financial allocation payment intelligent application task; An auxiliary management module, which is configured to automatically obtain business data under different business scenarios, and perform a pre-defined enterprise financial and management task based on the business data; wherein the enterprise financial and management task comprises at least one of an electronic invoice intelligent management task, a bank diary account intelligent query task, a financial project profit automatic extraction task, an invoice application single automatic generation task, or an operating index report automatic generation task; A financial management subsystem, which is configured to integratively implement a to-do list management, a compliance guide manual query, and a financial data structured storage according to a preset standardized process; A business service subsystem, which is configured to provide a user with a plurality of integrated service modules, and call a standard operation instruction file corresponding to the service module from a pre-stored business document library according to a user access request for any service module.

2. The enterprise financial comprehensive management system based on intelligent technology according to claim 1, characterized in that, The checkout management module comprises: A welfare fee transfer automatic generation business unit, which is configured to automatically download a subject balance table from the ERP system, match a preset keyword in a subject name column of the subject balance table line by line according to a preset regular expression, extract all matched line data as welfare fee related items, perform semantic analysis and integrity checking on the welfare fee related items, and generate standardized data fields; call a voucher template and automatically fill in the standardized data fields to generate a standard format of a deduction list; The safety production fee automatic deduction unit is used for calling a pre-deduction fee calculation template, extracting deduction data from the pre-deduction fee calculation template in combination with a predefined regular expression and semantic analysis, and further generating a safety production fee deduction list according to the deduction data in a project dimension; The allocated company layer fee automatic generation business unit is used for automatically screening business detail data of different projects of each company from business data stored in a preconfigured data source according to a screening condition defined in a pre-set company project adjustment template, and performing data cleaning and format unification to generate a standardized business detail data file; the business detail data file and a pre-set fee allocation rule and a general ledger template are matched and summarized by using parallel computing technology to generate an allocated company fee general ledger; The production cost settlement automatic generation business unit is used for obtaining subject balance detail data in a two-dimensional table structure by performing directional analysis on a subject balance table automatically downloaded from an ERP system; target settlement voucher subjects are acquired based on a pre-set configuration table and a dynamic correlation matrix with the cost center and the accounting subject in the subject balance detail data as the query condition, and the debit amount of the subject balance detail data is automatically filled into a voucher template according to the target settlement voucher subjects; a flow control mechanism is used to check data processing exceptions, skip abnormal data and record logs; and finally a standardized cost settlement single meeting the accounting standards is automatically generated.

3. The enterprise financial comprehensive management system based on intelligent technology according to claim 2, characterized in that, The income management module comprises: The sales income confirmation automatic generation unit is used for obtaining an income balance table from a specified external data source, extracting sales income data by parallel traversing the income balance table by using a multi-thread technology, and further automatically generating an electricity fee confirmation single according to the sales income data; meanwhile, an electricity quantity settlement single issued by a power grid company is automatically uploaded; The sales income temporary estimate intelligent processing unit is used for obtaining a temporary estimate electricity fee income single from a specified external data source, and obtaining original data by analyzing the temporary estimate electricity fee income single; the original data are dynamically converted into standard data recognizable by an ERP system based on a multi-level data mapping mechanism; the ERP system business detail page is automatically jumped to by following a pre-set menu path by using an automatic technology, and the standard data recognizable by the ERP system are automatically filled in, so that a sales temporary estimate single is batch generated; The collection flow automatic generation unit is used for obtaining a collection flow detail file from a specified external data source, extracting collection flow key information from the collection flow detail file in combination with a predefined regular expression and semantic analysis; the total settlement receivable amount of a profit center is obtained by querying an income account table through SQL with the profit center in the collection flow key information as the correlation key, and is verified; after the verification, a collection flow record meeting the rules is automatically generated according to the verified collection flow key information; The collection flow key information comprises a profit center, a collection party account, a collection party bank name, an amount, a company code, a customer number, a collection mode, a currency and a transaction date.

4. The enterprise financial comprehensive management system based on intelligent technology according to claim 3, characterized in that, The treasurer application module comprises: The intelligent supplier account management unit is configured to traverse each company code according to a preconfigured company code list, automatically obtain the account data of each company and perform data cleaning, and then perform account clearing operation on the cleaned data according to a preset account clearing rule; and automatically perform skip operation and record log for an abnormal company that cannot be normally cleared due to nonexistence of the company code. The un-cleared account intelligent management unit is configured to traverse each company code according to a preconfigured company code list, automatically configure query parameters for each company code, load the un-cleared accounts receivable and the un-cleared accounts payable of the company according to the query parameters by using an intelligent waiting mechanism, and perform paging query on the loaded un-cleared accounts receivable and un-cleared accounts payable according to a configured subject and batch export the query results in a preset format. The fund budget execution intelligent management unit is configured to traverse each company code according to a preconfigured company code list, automatically query the fund plan data of each company according to a preset multi-dimensional combined query condition, batch export the fund plan data of each company in the company dimension to generate a standardized file, and verify, clean and standardize the standardized file by using a multi-level intelligent processing mechanism to generate structured fund plan data. The fund allocation payment intelligent application unit is configured to traverse each company code according to a preconfigured company code list, query and load the fund allocation payment application information of each company by using a multi-thread parallel processing technology, perform integrity verification on the fund allocation payment application information of each company, and batch submit the fund allocation payment application information that passes the verification for approval.

5. The enterprise financial comprehensive management system based on intelligent technology according to claim 4, characterized in that, The auxiliary management module comprises: The electronic invoice intelligent management unit is configured to scan a specified folder and automatically identify invoice files through file extensions, establish a file fingerprint library based on the SHA-256 algorithm, filter out all identified duplicate files by using the file fingerprint library, and batch upload the filtered invoice files to an invoice folder. The bank journal account intelligent query unit is configured to automatically query the bank journal account and batch export according to a preset filtering condition. The financial project profit automatic extraction unit is configured to acquire a multi-level subject balance table based on a company and a profit center in real time, perform dynamic verification on the multi-level subject balance table by using a double closed loop verification mechanism based on a rule engine and a generative adversarial network, automatically identify and extract profit and loss type index data from the verified data by using an intelligent semantic mapping engine, and automatically fill the profit and loss type index data into a preset intelligent analysis template to generate a financial project profit analysis report. The invoice application automatic generation unit is configured to acquire an income account and perform line-by-line analysis according to a preset rule, extract data fields constituting basic business information, financial and tax information and invoice detail information from the income account by combining regular expressions and semantic analysis, and fill the extracted data fields into a tax system standard form by calling an invoice application template engine to generate an invoice application form. The operation index completion condition report automatic generation unit is used for automatically collecting ERP financial data in a monthly cycle, and performing piece-by-piece analysis on the ERP financial data by using a missing value dynamic interpolation and derived field recalculation method based on business rules to generate an operation analysis report.

6. The enterprise financial comprehensive management system based on intelligent technology according to claim 5, characterized in that, The specific method for dynamically checking the multi-level subject balance table through the double closed loop checking mechanism based on the rule engine and the generative adversarial network is: The original data in the multi-level subject balance table is cleaned through the standardized rule engine, and the cleaned data is converted into structured data with standardized semantics according to a preset conversion rule; The pre-trained generative adversarial network is used to detect logical contradictions in the structured data with business semantics; if the structured data with business semantics is detected to have logical contradictions, the standardized rule engine feeds back abnormal information of the logical contradictions to the standardized rule engine; The standardized rule engine locates the original data corresponding to the structured data according to the received abnormal information, modifies the original data by calling a predefined correction rule, and re-performs data cleaning and data conversion on the modified original data to generate new structured data with business semantics and re-perform logical contradiction detection, until the generative adversarial network cannot detect logical contradictions from the current structured data, and the checked data is obtained.

7. The enterprise financial comprehensive management system based on intelligent technology according to claim 6, characterized in that, The specific content of the method for performing piece-by-piece analysis on the ERP financial data by using the missing value dynamic interpolation and derived field recalculation method based on business rules to generate the operation analysis report is: A mapping relationship table of subject types and precision rules is constructed by analyzing the data dictionary of the ERP system; For any one data in the ERP financial data, each field in the data is traversed, and the metadata tag of each field is obtained by querying the data dictionary of the ERP system; wherein the metadata tag is any one of a basic field and a derived field; For each basic field, it is detected whether the basic field has a subject missing, if not, the basic field is converted into precision conforming to business specifications by dynamically matching the preconfigured precision rule library according to the mapping relationship table of subject types and precision rules; if so, the business attribute of the basic field is automatically identified according to the mapping relationship table of subject types and precision rules, and the missing value of the basic field is generated and converted into precision conforming to business specifications by using the moving weighted average method to dynamically interpolate the basic field; For each derived field, the calculation formula of the derived field is inversely analyzed from the data dictionary of the ERP system to obtain the variable current effective value of each variable in the calculation formula, and the variable current effective value of all variables is substituted into the calculation formula to recalculate the alternative value of the derived field and convert it into precision conforming to business specifications; After the precision conversion of all fields in the data is completed, the cleaned data is obtained, and the data cleaning of the ERP financial data is completed to obtain a standardized financial data set; The standardized financial analysis template is called and the standardized financial data set is automatically filled in to generate the operation analysis report.

8. The enterprise financial comprehensive management system based on intelligent technology according to claim 1, characterized in that, The financial management subsystem comprises: The to-do list management module is configured to acquire to-do lists and visualize the to-do lists by using a density-sensitive dynamic time axis compression algorithm; the visualized content includes real-time synchronization of file download and completion status of the to-do lists, display of the to-do lists through a monthly calendar view, and automatic push of a reminder message according to an expiration date of the to-do lists; The guidebook query module is configured to query a financial management business compliance guidebook from a built-in standardized compliance guide system according to a keyword or a category provided by a user; The financial electronic database is configured to acquire financial data related to funds, budgets, accounting, tax, and assets from a specified data source, and construct a multi-dimensional index according to the financial data; and the financial data is stored in a structured manner according to the multi-dimensional index.

9. The enterprise financial comprehensive management system based on intelligent technology according to claim 1, characterized in that, The business service subsystem includes: The tax-related business service module is configured to store various tax-related business operation guide files formulated based on tax laws; receive an access request of a user for a specified tax-related business; and call and provide the user with an operation guide file corresponding to the specified tax-related business according to the access request; wherein the specified tax-related business is at least one of a cultivated land occupation tax business, a personal income tax withholding business, an enterprise income tax declaration business, and a value-added tax input tax deduction business; The asset management service module is configured to store standardized operation guide files of various asset management services; receive an access request of a user for a specified asset management service; and call and provide the user with a standardized operation guide file corresponding to the specified asset management service according to the access request; wherein the specified asset management service is at least one of a production preparation service, a solid conversion management service, a completion settlement service, an asset temporary estimation service, and an asset inventory service; The fund management service module is configured to store standardized operation guide files of various fund management services; receive an access request of a user for a specified fund management service; and call and provide the user with a standardized operation guide file corresponding to the specified fund management service according to the access request; wherein the specified fund management service is at least one of a fund plan preparation service, a fund plan execution service, a special account use service, a special fund payment service, and a capital management service; The business management service module is configured to store standardized operation guide files of various business management services; receive an access request of a user for a specified business management service; and call and provide the user with a standardized operation guide file corresponding to the specified business management service according to the access request; wherein the specified business management service is at least one of a current account management service, a budget control service, and a dividend payment management service.

Images