Search

CN-122019819-A - Non-invasive invoice data acquisition and continuous number violation intelligent identification and early warning system and method

CN122019819ACN 122019819 ACN122019819 ACN 122019819ACN-122019819-A

Abstract

The application provides an audit scene-oriented non-invasive invoice data acquisition and continuous number violation intelligent identification and early warning system and method, and belongs to the field of invoice data processing; the system comprises a data acquisition module for acquiring invoice raw data on an enterprise financial system based on an RPA (remote processing architecture) robot, a data storage and verification module for cleaning and storing the acquired data, a continuous number comparison module for judging the illegal invoice, an RPA robot for regularly reading the verified current-year invoice data from a database and traversing the invoice numbers one by one, a basic rule base for calling the RPA robot to calculate adjacent differences of the traversed invoice numbers one by one, marking the illegal data in real time and generating a unique temporary identifier for each illegal data, and an electronic image acquisition module for realizing accurate acquisition and standard storage of the illegal invoice images in linkage with the continuous number comparison module.

Inventors

  • XUE YIFANG

Assignees

  • 中国能源建设集团山西省电力勘测设计院有限公司

Dates

Publication Date
20260512
Application Date
20251231

Claims (10)

  1. 1. The non-invasive invoice data acquisition and continuous number violation intelligent recognition and early warning system is characterized by comprising the following components: The data acquisition module is used for acquiring invoice original data on an enterprise financial system based on the RPA robot; The data storage and verification module is used for storing the invoice original data acquired by the data acquisition module into a database, and cleaning the data by adopting an RPA built-in verification rule in the storage process; the continuous number comparison module comprises a real-time traversing unit, a rule configuration unit and a result marking unit, wherein: The timing traversing unit is used for reading the current-year invoice data after verification from the database through the RPA robot at regular time and traversing the invoice numbers one by one; the rule configuration unit comprises a basic rule base, wherein a general audit rule is arranged in the basic rule base, and the RPA robot can call the basic rule base to calculate adjacent difference values of the traversed invoice numbers pair by pair and mark violation data in real time; The result marking unit is used for synchronizing the violation data marked by the real-time traversing unit to a result database, generating a unique temporary identifier for each piece of violation data, and associating the identifier with invoice core information as a positioning index for the subsequent electronic image acquisition; And the electronic image acquisition module is linked with the continuous number comparison module to realize accurate acquisition and standard storage of the illegal invoice image.
  2. 2. The system for non-invasive invoice data collection and continuous number violation intelligent recognition and early warning according to claim 1, further comprising a result pushing module, wherein the result pushing module is used for pushing a result document and an image file to audit responsibility at regular time, violation data is stored in the result document, and violation invoice images associated with the violation data are stored in the influence file.
  3. 3. The system for non-invasive invoice data collection and continuous number violation intelligent recognition and early warning according to claim 1 or 2, wherein the electronic image acquisition module comprises: The RPA robot reads the unique temporary identifier of the violation data in the result database to obtain the combined inquiry condition of the bill number and the invoice number, and automatically returns to the corresponding enterprise financial system; The RPA robot simulates the operation of manually clicking a 'download electronic image' button, acquires the electronic image file in PDF or JPG format, automatically names according to the standardized naming rule of 'year-month-invoice number-violation type', stores the electronic image file in an intranet designated path, establishes the association index of the image file storage path and violation data in a result database, and realizes one-key association retrieval of the image file and violation information; The abnormal processing unit is used for setting a download state real-time monitoring mechanism, automatically marking the 'image missing' state of the invoice in a result database if abnormal conditions such as image file missing, download overtime or file damage are monitored, and sending an abnormal prompt containing invoice information to auditors through a preset channel.
  4. 4. The non-invasive invoice data acquisition and continuous number violation intelligent recognition and early warning system according to claim 3, wherein a visual configuration interface is further arranged in the rule configuration unit, supports the code-free operation of auditors, and can directly adjust continuous number judgment standards and expand comparison dimensions.
  5. 5. The non-invasive invoice data acquisition and continuous number violation intelligent recognition and early warning system according to claim 4, wherein a continuous number comparison algorithm is built in a basic rule base, and the automatic recognition of batch continuous number violation invoices is realized through the continuous number comparison algorithm, and is specifically as follows: aiming at the structural invoice data of the date after cleaning belonging to the target examination year, ascending sorting is carried out according to the invoice numbers, and a global ordered invoice number sequence is generated: ; Wherein the method comprises the steps of Is the ith invoice number, and The date of the manufacture is not distinguished; Acquiring a currently effective number connection judgment threshold t and a number difference threshold d from a rule configuration unit; For the global ordered sequence S, calculating the adjacent number difference ; Traversing a sequence of differences Identifying subsequences with continuous t-1 difference values equal to d, and recording corresponding invoice number ranges; marking all invoices in the subsequence meeting the condition as 'batch number connection violation'; And writing the violation records into a annual violation result table, and establishing an association relation between the temporary identifier and the bill number and invoice number for subsequent image positioning.
  6. 6. The non-invasive invoice data acquisition and continuous number violation intelligent identification and early warning method is characterized by adopting the non-invasive invoice data acquisition and continuous number violation intelligent identification and early warning system as claimed in any one of claims 1-5, and comprising the following steps: 1. Initializing and configuring a system; 2. invoice data is automatically collected and checked; 3. automatically comparing invoice data; 4. and automatically acquiring electronic images of the illegal invoices.
  7. 7. The non-invasive invoice data acquisition and continuous number violation intelligent identification and early warning method according to claim 6, further comprising the steps of: 5. And automatically pushing the violation results.
  8. 8. The non-invasive invoice data acquisition and continuous number violation intelligent identification and early warning method according to claim 6 or 7 is characterized in that the invoice data automatic comparison in the third step specifically comprises the following steps: S7, the data acquisition module transmits the original invoice data list to the data storage and verification module in real time to clean and store the data; s8, after the daily data acquisition is finished, a built-in continuous number comparison algorithm is regularly called, and automatic judgment of continuous number relation is carried out on all the extracted invoice numbers; S9, after traversing and reading invoice data, calling a basic rule base in a rule configuration unit, sorting invoice numbers according to ascending order of open date, and calculating the difference value of adjacent invoice numbers; And S10, the result marking unit carries out information complementation on the recognized violation data, adds fields such as violation type, examination time, rule number and the like, generates a unique temporary identifier for each piece of violation data, synchronously writes the violation data into a violation result database, establishes an association relation between the temporary identifier and a bill number and an invoice number, and takes the association relation as a positioning index for subsequent image acquisition, and simultaneously generates an examination log and uploads the examination log to the RPA control center.
  9. 9. A computer device comprising a memory, a processor and a computer program stored on the memory, wherein the processor executes the computer program to perform the steps of the method of claim 8.
  10. 10. A computer-readable storage medium, on which a computer program/instruction is stored which, when executed by a processor, implements the steps of the method as claimed in claim 8.

Description

Non-invasive invoice data acquisition and continuous number violation intelligent identification and early warning system and method Technical Field The application relates to the technical field of automatic invoice data processing, in particular to an audit scene-oriented noninvasive intelligent invoice data acquisition and continuous number violation recognition and early warning system and method. Background In the daily financial audit work of enterprises, invoice compliance examination is a key link for guaranteeing the reality and legality of financial data, wherein invoice serial number examination is important content for preventing false reimbursement and illegal reimbursement. Traditional invoice compliance examination relies on manual operation, financial staff needs to manually log in an enterprise financial system, current-year invoice data are derived one by one, and then invoice numbers are manually screened through Excel and other tools and are compared in serial number. This traditional mode of operation exposes limitations in practical applications: 1. when the enterprise invoices are large in number, a large amount of time is consumed for manual data collection and comparison and analysis, and especially in the financial settlement peak period of the end of the month, the end of the year and the like, the inspection work is easy to delay. 2. The accuracy is difficult to ensure, the problems of data missing, number comparison errors and the like easily occur in the manual operation process, and the compliance risk brought by the continuous number invoice cannot be effectively identified. 3. When the response is not timely, after the manual inspection is finished, the offending information needs to be manually arranged and a specialized person is informed, so that the optimal opportunity for risk management and control can be missed. In addition, in the prior art, some enterprises try to realize simple data screening through built-in functions of the financial system, and a communication interface between systems is required to be developed or a special examination system is required to be constructed, so that the development cost is high, the period is long, the stability of the existing enterprise financial system is possibly influenced, a closed loop of data acquisition, compliance examination and risk reminding is not formed, all links still need to be manually intervened and connected, and the pain point of manual examination cannot be fundamentally solved. Therefore, a method for realizing non-invasive invoice data collection, continuous number examination and risk reminding is needed to improve the management level of enterprise invoice compliance. Disclosure of Invention The application aims to provide an audit scene-oriented noninvasive invoice data acquisition and continuous number violation intelligent recognition and early warning system and method, which solve the problems that the efficiency is low, the accuracy is easy to be interfered by human factors, the inspection period is long and the like in the traditional manual invoice compliance inspection mode, and the prior art scheme needs to develop an intersystem communication interface or construct a special inspection system, has high development cost and long period, can influence the stability of the traditional enterprise financial system, and meanwhile, cannot realize the full-flow automation of data acquisition, compliance inspection and risk reminding, still needs to manually intervene in connecting links, and does not fundamentally solve the pain point of manual inspection. The technical scheme adopted by the application is that the non-invasive invoice data acquisition and continuous number violation intelligent identification and early warning system comprises: The data acquisition module is used for acquiring invoice original data on an enterprise financial system based on the RPA robot; The data storage and verification module is used for storing the invoice original data acquired by the data acquisition module into a database, and cleaning the data by adopting an RPA built-in verification rule in the storage process; the continuous number comparison module comprises a real-time traversing unit, a rule configuration unit and a result marking unit, wherein: The timing traversing unit is used for reading the current-year invoice data after verification from the database through the RPA robot at regular time and traversing the invoice numbers one by one; the rule configuration unit comprises a basic rule base, wherein a general audit rule is arranged in the basic rule base, and the RPA robot can call the basic rule base to calculate adjacent difference values of the traversed invoice numbers pair by pair and mark violation data in real time; The result marking unit is used for synchronizing the violation data marked by the real-time traversing unit to a result database, generating a unique temporary identifier for e