Civil aviation ticket number level account and subject balance checking system based on digital assets

Abstract

This invention discloses a ticket number-level accounting and account balance verification system for the civil aviation industry based on digital assets, comprising: a basic data processing module for collecting basic data for ticket-level accounting verification and generating a standardized record set of tickets; an accounting input construction module for generating a ticket-level accounting input set and establishing a ticket-level account system; an accounting penetration model construction module for constructing an accounting penetration Graph-Mamba model; a penetration graph construction module for generating an accounting penetration graph state based on the ticket-level accounting input set and the ticket-level account system; an account mapping and conversion module for generating mapped ticket-level debit and credit vouchers; a time-series accounting scanning module for generating monthly revenue accounting vouchers and core account balances based on the mapped ticket-level debit and credit vouchers; and a two-way verification and analysis module for generating ticket number-level accounting and account balance verification results. This invention improves verification accuracy, traceability, and processing efficiency.

Description

Technical Field

[0001] This invention relates to the field of civil aviation transportation settlement and revenue management, and in particular to a civil aviation industry ticket number-level accounting and account balance verification system based on digital assets. Background Technology

[0002] Revenue recognition in the civil aviation industry exhibits a distinct phased characteristic. Accounts receivable and contract liabilities are generated during the ticket sales phase, and revenue is only recognized after the transportation service is completed. Due to the mismatch between the sales month and the transportation month, and the existence of partial usage, refunds, changes, and cross-period operations of tickets, the correspondence between contract liabilities, accounts receivable, and transportation revenue is quite complex. In the existing business process, there has long been a problem of difficulty in accurately tracing the source of account balances.

[0003] Currently, civil aviation revenue settlement and financial accounting generally adopt a separate processing architecture. The civil aviation business system is responsible for processing sales and operation data, while the financial accounting system independently completes voucher generation and account posting. The connection between the two systems mainly relies on report export, voucher import, interface transmission, and manual verification. Although this approach can complete basic accounting processing, it still suffers from problems such as fragmented processing of settlement data and accounting vouchers, insufficient automation, and reliance on manual intervention for accounting conversion.

[0004] Furthermore, existing technologies typically process settlement business summary results or financial account summary results, resulting in inconsistent data granularity with financial reconciliation granularity. Once discrepancies in account balances occur, it is difficult to quickly locate specific invoice numbers, invoice serial numbers, and business status. At the same time, the data consistency guarantee capability between multiple systems is weak, the discrepancy handling process is cumbersome, and the real-time performance is insufficient, making it difficult to meet the application requirements of refined accounting, audit traceability, and dynamic financial management.

[0005] Therefore, how to provide a ticket number-level accounting and account balance reconciliation system for the civil aviation industry based on digital assets is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0006] One objective of this invention is to propose a ticket number-level accounting and account balance verification system for the civil aviation industry based on digital assets. This invention constructs a standardized ticket record set, a ticket-level accounting input set, and a ticket-level account system based on ticket-level accounting verification data. It then builds an accounting penetration Graph-Mamba model. Through a penetration graph construction layer, a rule coupling layer, a time-series scanning layer, and a two-way verification layer, it completes the generation of the accounting penetration graph status, the mapping and conversion of accounting accounts, the generation of monthly revenue accounting vouchers, the scanning of core account balances, and the output of ticket number-level accounting and account balance verification results. This system possesses advantages such as strong ticket number-level penetration verification capabilities, high accuracy in account balance tracing, and accurate discrepancy location.

[0007] According to an embodiment of the present invention, a civil aviation industry ticket number-level accounting and account balance reconciliation system based on digital assets includes: The basic data processing module is used to collect basic data for ticket-level accounting verification, perform field cleaning, format unification, ticket number verification, ticket serial number alignment, construct unique ticket identifiers, and generate standardized ticket record sets. The accounting input construction module is used to organize and generate a ticket-level accounting input set and establish a ticket-level subject system based on the standardized record set and unique ticket identifier. The accounting penetration model construction module is used to build an accounting penetration Graph-Mamba model, including a penetration graph construction layer, a rule coupling layer, a time-series scanning layer, and a two-way verification layer; The penetration diagram construction module is used to construct a three-level heterogeneous accounting penetration diagram in the penetration diagram construction layer based on the ticket-level accounting input set and ticket-level account system, establish inter-layer connections, and generate the accounting penetration diagram status. The account mapping and conversion module is used to determine the debit and credit accounts in the rule coupling layer based on the accounting penetration diagram status and the voucher-level accounting input set, generate the corresponding debit and credit records, and generate the mapped voucher-level debit and credit vouchers. The time-series accounting scanning module is used to perform status scanning in the time-series scanning layer based on the mapped voucher-level debit and credit vouchers, and generate monthly revenue accounting vouchers and core account balances; The two-way verification analysis module is used to perform consistency checks based on the two-way verification layer. It reverse-traces the monthly revenue accounting vouchers and mapped ticket-level debit and credit vouchers based on the core account balances in the civil aviation revenue accounting system, and calculates and generates ticket number-level accounting and account balance verification results.

[0008] Optionally, the basic data processing module includes: Collect ticket sales data, ticket carrier data, ticket business status data, and ticket-level account mapping rules; perform field integrity checks on ticket sales data, ticket carrier data, and ticket business status data; and generate ticket-level accounting reconciliation basic data. Based on the basic data for ticket-level accounting reconciliation, a unified format processing is performed; Based on the basic data for ticket-level accounting verification after the format has been standardized, ticket number verification is performed, and a unique ticket identifier is constructed according to a fixed combination of ticket number, ticket serial number, and carriage date index. Based on the unique ticket identifier, the ticket sales data, ticket carrier data, and ticket business status data are linked one by one. Based on each record after the association is completed, a standardized record set of tickets is output according to the unique ticket identifier. Each record in the standardized record set of tickets retains the ticket number, ticket serial number, amount item, business type, carrier status, customer identifier, and carrier date index.

[0009] Optionally, the accounting input construction module includes: Based on the standardized record set of tickets, the data records corresponding to the same ticket are merged and organized according to the unique ticket identifier; Based on the merged and organized data records, the amount items are split into ticket amount, tax amount, refund adjustment amount, and rescheduling adjustment amount, forming fixed-length field records. Based on fixed-length field records, output a ticket-level accounting input set according to the unique ticket identifier; Based on the document-level subject mapping rules, the business attributes and accounting subject correspondences corresponding to business types are extracted, and the mapping relationships between business types, carrier status and accounts receivable, contract liabilities, and transportation revenue are organized into a hierarchical subject index structure. A ticket-level subject system is established based on a hierarchical subject index structure.

[0010] Optionally, the accounting look-through model construction module includes: The original Graph-Mamba model is invoked, including the graph structure representation construction part, the graph structure serialization part, the Mamba state space scanning part, and the graph readout part; Based on the graph structure representation construction part in the original Graph-Mamba model, it is transformed into a penetrating graph construction layer, and the graph structure organization method in the graph structure representation construction part is transformed into a heterogeneous graph organization method oriented towards the state of the accounting penetrating graph. Based on the graph structure serialization part, a rule coupling layer is configured after the penetrating graph construction layer to transform the scanning input sequence organization method of the graph structure serialization part into a scanning input sequence organization method that includes rule constraint information; Based on the Mamba state space scanning part, it was modified into a time-series scanning layer; Based on the graph readout section, a bidirectional verification layer is configured after the timing scan layer, including a forward readout connection structure and a reverse readout connection structure; Based on the penetration graph construction layer, rule coupling layer, time-series scanning layer, and bidirectional verification layer, the accounting penetration Graph-Mamba model is constructed.

[0011] Optionally, the penetration map construction module includes: Based on the ticket-level accounting input set, records are merged according to the unique ticket identifier in the penetration graph construction layer to generate sales events and transportation events, and a ticket business layer is constructed to form sales event nodes and transportation event nodes. Each node in the ticket business layer includes a unique ticket identifier, business type, transportation status, and amount item index. Based on the amount items, business types, and carrier status in the ticket-level accounting input set, extract the corresponding amount item values ​​for each record, generate the amount value, determine the corresponding debit / credit direction mark and account index, and write the unique ticket identifier into the source ticket index. Based on the debit / credit direction mark, account index, amount value, and source ticket index, construct debit record nodes and credit record nodes to build a ticket-level debit / credit voucher layer. Based on the ticket-level account system, accounts receivable, contract liabilities, and transportation revenue are extracted to construct an account balance layer, forming accounts receivable nodes, contract liabilities nodes, and transportation revenue nodes. The beginning balance, current period change value, and ending balance are written to each node respectively. Based on the ticket business layer, ticket-level debit and credit voucher layer, and account balance layer, a three-level heterogeneous accounting penetration diagram is constructed. The nodes in the ticket business layer, ticket-level debit and credit voucher layer, and account balance layer are respectively encoded and organized to generate business node table, voucher node table, and balance node table. Based on the sales event nodes in the ticket business layer and the corresponding nodes in the ticket-level loan certificate layer, a sales formation relationship is constructed. Based on the carrier event nodes in the ticket business layer and the corresponding nodes in the ticket-level lending and borrowing voucher layer, a carrier transfer relationship is constructed. Based on the debit record nodes and credit record nodes in the voucher-level debit and credit voucher layer and the corresponding nodes in the account balance layer, an account mapping relationship is constructed; Based on the established account mapping relationship between the debit record node, credit record node and the corresponding node in the account balance layer, construct the balance accumulation relationship; Based on sales formation relationships, transportation transfer relationships, account mapping relationships, and balance accumulation relationships, establish inter-layer connections in a three-level heterogeneous accounting penetration diagram; Based on the business node table, voucher node table, balance node table, and inter-layer connections, an accounting penetration diagram is generated, which includes a set of voucher business layer nodes, a set of voucher-level debit and credit voucher layer nodes, a set of account balance layer nodes, and a set of inter-layer connections.

[0012] Optionally, the subject mapping conversion module includes: Based on the accounting penetration diagram status, the business type, transportation status, amount item index and unique ticket identifier are extracted from the ticket business layer node set in the rule coupling layer to form the rule query key value corresponding to the current ticket record. Based on the ticket-level subject mapping rules, the mapping condition field and subject output field are extracted. The rule query key value is compared with the mapping condition field item by item. Based on the ticket-level subject mapping rules that meet all comparison conditions, the corresponding debit subject index, credit subject index and amount item correspondence are extracted to generate the subject mapping result corresponding to the current ticket record. Based on the account mapping results, the business type, carrier status, debit account index, credit account index and amount item correspondence are organized into a rule embedding representation, and the rule embedding representation is updated in relation to the set of ticket business layer nodes in the accounting penetration diagram to generate a coupled mapping state. Based on the coupled mapping state and amount items, the debit record node and credit record node with the same unique ticket identifier corresponding to the source ticket index are located in the ticket-level debit and credit certificate layer. The mapping update process is performed on the debit record node and credit record node to generate the mapped updated debit record node and mapped updated credit record node. Based on the updated debit and credit record nodes, perform the corresponding association, organize them into debit and credit records, and establish an index correspondence between the debit and credit records and the corresponding nodes in the ticket business layer to complete the accounting subject mapping conversion; Based on the debit and credit records after the accounting subject mapping conversion is completed, a mapped voucher-level debit and credit voucher is generated.

[0013] Optionally, the time-series accounting scanning module includes: Based on the mapped ticket-level debit and credit vouchers and ticket carrier data, the mapped ticket-level debit and credit vouchers are sorted and organized in the time-series scanning layer according to the accounting period, unique ticket identifier and accounting order, arranged into a state sequence, and state scanning input is generated. Based on the state scan input, establish the account balance status that is synchronously updated with the state sequence, including the account balance status at the previous moment, the current voucher change value, and the updated account balance status; Based on the mapped voucher-level debit and credit vouchers with the business type of sales in the state sequence, the state update is performed according to the accounting causal order of sales forming contract liabilities, and the state update result of sales forming contract liabilities is generated. Based on the corresponding carrier records in the state sequence and the carrier status in the ticket carrier data, the status update is performed according to the accounting causal order of carrier offsetting contract liabilities, and the carrier offsetting contract liabilities status update result is generated. Based on the status update result of the carrier's reduction of contract liabilities, the status update is performed according to the accounting causal order of revenue recognition to generate the status update result of transportation revenue recognition; Based on the status update results of sales-generated contract liabilities, carrier-reduced contract liabilities, and transportation revenue recognition, balance accumulation processing is performed according to the order of the status sequence to generate a continuously changing core account balance trajectory during the accounting period. Based on the core account balance trajectory after the completion of balance accumulation processing, the summary organization is performed according to the accounting period. The debit and credit records corresponding to all status update results within the same accounting period are organized into monthly revenue accounting vouchers, and the balances of accounts receivable, contract liabilities, and transportation revenue at the end of the same accounting period are organized into core account balances.

[0014] Optionally, the two-way verification and analysis module includes: Based on the two-way verification layer, the balances of core accounts in the civil aviation revenue accounting system and the financial accounting system are extracted, and corresponding information is generated to produce the balance correspondence relationship under the same accounting period. Based on the balance correspondence, the core account balances in the civil aviation revenue accounting system are read forward through the forward read connection structure, and the core account balances in the financial accounting system are read backward through the reverse read connection structure. The core account balances in the civil aviation revenue accounting system and the core account balances in the financial accounting system are compared item by item, and the comparison results are organized into a consistency verification result. Based on the consistency verification results, core account balances with inconsistent amounts are filtered out. The corresponding monthly revenue accounting vouchers and mapped voucher-level debit and credit vouchers are traced back from the core account balances in the civil aviation revenue accounting system according to the source voucher index, and a corresponding relationship is established. Based on the monthly revenue accounting vouchers and the mapped voucher-level debit and credit vouchers after establishing the correspondence, the voucher number-level composition corresponding to the current account balance is statistically analyzed and verified against the amount value in the core account balance to determine the voucher number-level composition details of the current account balance. Based on the ticket number level composition and the mapped ticket level debit and credit vouchers, extract the sales accounting amount and the carrier transfer amount corresponding to the same ticket number, and calculate the untransferred amount; The total number of ticket copies and the number of ticket copies already shipped under the same ticket number are counted. The usage ratio is calculated by using the correspondence between the number of ticket copies already shipped and the total number of ticket copies. For ticket numbers where the amount items have been split and recorded by ticket copy, the total amount and the amount already shipped under the same ticket number are counted. The usage ratio is calculated by using the correspondence between the amount already shipped and the total amount. Based on the ticket business status data and the mapped ticket-level credit vouchers, extract the refund and reschedule records corresponding to the same ticket number, and calculate the refund and reschedule adjustment results. Based on the consistency verification results, the amount not carried over, the partial usage ratio, and the refund and rescheduling adjustment results, determine the type of discrepancy; Based on the consistency verification results, ticket number composition, uncarried amount, partial usage ratio, refund and change adjustment results, and difference type, generate ticket number-level accounting and account balance verification results.

[0015] The beneficial effects of this invention are: First, this invention establishes a penetrating verification path from the ticket business layer, the ticket-level accounting input set, the ticket-level subject system, the accounting penetration diagram status, the mapped ticket-level debit and credit vouchers, the monthly revenue accounting vouchers, and the core subject balances through a continuous processing relationship. This allows the core subject balances in the civil aviation revenue accounting system to not only complete the verification at the monetary level, but also to reverse-locate the corresponding ticket number-level composition, untransferred amounts, partial usage ratios, and refund / change adjustment results. This effectively solves the problem in existing technologies where contract liability account balances have long existed but their sources and compositions are unclear, and ticket number-level details are difficult to trace. It significantly improves the traceability and audit support capabilities of ticket number-level accounting and subject balance verification results.

[0016] Secondly, this invention utilizes the penetration graph construction layer, rule coupling layer, time-series scanning layer, and bidirectional verification layer in the accounting penetration Graph-Mamba model to uniformly process ticket sales data, ticket carrier data, ticket business status data, and ticket-level account mapping rules. It can generate monthly revenue accounting vouchers and core account balances according to the accounting causal sequence of sales forming contract liabilities, carrier reducing contract liabilities, revenue recognition, and balance accumulation. Combined with consistency verification, it achieves automatic identification of discrepancies. This changes the existing technology's separation of settlement system and financial accounting system, which relies on manual import of vouchers and manual comparison of discrepancies. It reduces efficiency losses and human errors caused by manual verification, and improves the integration of accounting processing and verification processing.

[0017] Furthermore, this invention uses the core account balances in the civil aviation revenue accounting system and the financial accounting system as the objects of two-way reconciliation. Based on consistency verification, it further distinguishes between differences where accounting treatment has been completed but the operational status has not, and differences where the operational status has been completed but the accounting treatment has not. This allows for discrepancy identification to go beyond the level of total amount deviation, refining it down to the reconciliation result records corresponding to ticket number, ticket serial number, account name, accounting period, and amount value. This ensures consistent reconciliation standards even under multi-system parallel operation, enhances the data consistency guarantee capability between systems, and improves the accuracy of subsequent maintenance, regulatory verification, and anomaly handling. Attached Figure Description

[0018] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings: Figure 1 This is the overall flowchart of the civil aviation industry ticket number-level accounting and account balance verification system based on digital assets proposed in this invention; Figure 2This is a schematic diagram of the Graph-Mamba model structure for accounting penetration proposed in this invention; Figure 3 This is a schematic diagram of the three-level heterogeneous accounting penetration diagram and two-way verification analysis process proposed in this invention. Detailed Implementation

[0019] The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, illustrating only the basic structure of the invention, and therefore only show the components relevant to the invention.

[0020] refer to Figures 1-3 A civil aviation industry ticket number-level accounting and account balance reconciliation system based on digital assets, including: The basic data processing module is used to collect basic data for ticket-level accounting reconciliation. This basic data includes ticket sales data, ticket carrier data, ticket business status data, and ticket-level account mapping rules. The module performs field cleaning, format unification, ticket number verification, and ticket serial number alignment on the basic data. Based on the ticket number and ticket serial number, it constructs a unique ticket identifier and generates a standardized ticket record set. The accounting input construction module is used to organize ticket number, ticket serial number, amount item, customer identifier, carrier status, and business type into a ticket-level accounting input set based on the standardized record set of tickets and unique ticket identifiers, and to establish a ticket-level subject system based on ticket-level subject mapping rules. The accounting penetration model construction module is used to build an accounting penetration Graph-Mamba model. The accounting penetration Graph-Mamba model includes a penetration graph construction layer, a rule coupling layer, a time-series scanning layer, and a two-way verification layer. The penetration diagram construction module is used to construct a three-level heterogeneous accounting penetration diagram based on the ticket-level accounting input set and ticket-level account system. It includes a ticket business layer, a ticket-level debit / credit voucher layer, and an account balance layer. The ticket business layer is constructed by extracting sales events and transportation events from the ticket-level accounting input set. The ticket-level debit / credit voucher layer is constructed based on the amount items, business types, and transportation status in the ticket-level accounting input set. The account balance layer is constructed based on the accounts receivable, contract liabilities, and transportation revenue accounts in the ticket-level account system. The module establishes inter-layer connections based on sales formation relationships, transportation transfer relationships, account mapping relationships, and balance accumulation relationships to generate the accounting penetration diagram status. The subject mapping conversion module is used to extract business types and carrier status in the rule coupling layer based on the accounting penetration diagram status and the ticket-level accounting input set, determine the corresponding debit and credit accounts according to the ticket-level subject mapping rules, generate corresponding debit and credit records in the ticket-level debit and credit voucher layer based on the amount itemization, complete the accounting subject mapping conversion, and generate the mapped ticket-level debit and credit vouchers. The time-series accounting scanning module is used to perform status scanning in the time-series scanning layer based on the mapped ticket-level debit and credit vouchers and ticket carrier data, in accordance with the accounting causal sequence of sales forming contract liabilities, carrier reducing contract liabilities, revenue recognition, and balance accumulation, to generate monthly revenue accounting vouchers and core account balances. The core account balances include accounts receivable balances, contract liabilities balances, and transportation revenue balances. The two-way reconciliation analysis module is used to perform consistency verification between the core account balances in the civil aviation revenue accounting system and the core account balances in the financial accounting system based on the two-way reconciliation layer. Based on the core account balances in the civil aviation revenue accounting system, it traces back the monthly revenue accounting vouchers and mapped ticket-level debit and credit vouchers to calculate the ticket number-level composition, uncarried amount, partial usage ratio, and refund / change adjustment results corresponding to the current account balance. It identifies the differences between accounting treatments that have been formed but the carrier status has not been formed and the differences between carrier statuses that have been formed but the accounting treatment has not been formed, and generates ticket number-level accounting and account balance reconciliation results.

[0021] In this embodiment, the basic data processing module includes: The process involves collecting ticket sales data, ticket transport data, ticket business status data, and ticket-level account mapping rules. Field integrity checks are performed on the ticket number, ticket serial number, amount breakdown, customer identifier, and business type fields in the ticket sales data; on the ticket transport data, on the ticket number, ticket serial number, transport date, and transport status fields; and on the ticket business status data, on the ticket number, ticket serial number, and business status fields. Data that passes the field integrity checks is used as the foundation for ticket-level accounting reconciliation. The ticket-level account mapping rules are also included. It is a collection of ticket-level rule records established according to business type and carrier status. Each rule record includes a mapping condition field, a business attribute field, and an account output field. The mapping condition field includes business type and carrier status. The business attribute field is used to identify the business attribute corresponding to the current business type. The account output field includes debit account index, credit account index, and amount item correspondence. The debit account index and credit account index are used to identify the accounts receivable, contract liabilities, and transportation revenue accounts corresponding to the business attribute. The amount item correspondence is used to identify the correspondence between the amount item and the debit account index and credit account index. Based on the basic data for ticket-level accounting reconciliation, a unified format processing is performed. Specifically, the ticket number field is processed using a fixed-length character format, the ticket serial number field is processed using a numerical format, the amount item field is processed using a precise amount format, the customer identifier field is processed using a coding format, and the business type field, carrier status field, and business status field are processed using code values. Records with empty values, missing key fields, and duplicate records are deleted, while the ticket number, ticket serial number, amount item, customer identifier, business type, carrier status, business status, and carrier date are retained. Based on the ticket-level accounting reconciliation basic data after the format unification processing, ticket number verification is performed, including checking the check bit and length of the ticket number field, checking the serial number range of the ticket serial number field, performing sequential alignment processing on the ticket serial numbers corresponding to the same ticket number, extracting the transportation date index from the transportation date field in the transportation data of the ticket, and constructing a unique ticket identifier according to the fixed combination order of ticket number, ticket serial number, and transportation date index; Based on the unique ticket identifier, the ticket sales data, ticket carrier data, and ticket business status data are linked one by one. The ticket sales data, ticket carrier data, and ticket business status data corresponding to the same unique ticket identifier are merged into a single record. During the merging process, the ticket number, ticket serial number, amount item, business type, carrier status, customer identifier, and carrier date index are retained. Based on the individual records after each association is completed, a standardized record set of tickets is output according to the unique ticket identifier. Each record in the standardized record set of tickets retains the ticket number, ticket serial number, amount item, business type, carrier status, customer identifier, and carrier date index, which are used for the generation of ticket-level accounting input sets and the establishment of ticket-level subject systems.

[0022] In this embodiment, the accounting input construction module includes: Based on the ticket number, ticket serial number, amount item, customer identifier, carrier status, and business type in the standardized ticket record set, the data records corresponding to the same ticket are merged and organized according to the unique ticket identifier, and the unique ticket identifier is used as the organizational primary key of each record. Based on the merged and organized data records, the amount items are split into ticket amount, tax amount, refund adjustment amount, and rescheduling adjustment amount. The ticket number, ticket serial number, ticket amount, tax amount, refund adjustment amount, rescheduling adjustment amount, customer identifier, carrier status, and business type are written into the same record in a fixed field order to form a fixed-length field record. Based on fixed-length field records, output ticket-level accounting input sets according to unique ticket identifiers. Each record in the ticket-level accounting input set corresponds to a unique ticket identifier. Each record retains the ticket number, ticket serial number, ticket amount, tax amount, refund adjustment amount, rescheduling adjustment amount, customer identifier, carrier status, and business type. Based on the ticket-level subject mapping rules, the corresponding business attributes and accounting subjects of business types are extracted. The mapping relationship between business types, carrier status and accounts receivable, contract liabilities and transportation revenue is organized into a hierarchical subject index structure. The hierarchical subject index structure includes a business attribute index layer, a mapping relationship index layer and an accounting subject index layer. Based on the hierarchical subject index structure, a ticket-level subject system is established. The business attribute index layer records the business type and transportation status, the mapping relationship index layer records the corresponding debit accounts, credit accounts and amount items for the business type and transportation status, and the accounting subject index layer records accounts receivable, contract liabilities, and transportation revenue. The ticket-level subject system is used for constructing a three-level heterogeneous accounting penetration diagram.

[0023] In this embodiment, the accounting look-through model construction module includes: The original Graph-Mamba model is invoked. The original Graph-Mamba model includes a graph structure representation construction part, a graph structure serialization part, a Mamba state space scanning part, and a graph readout part. The graph structure representation construction part is used to receive graph node representations and graph connection relationships. The graph structure serialization part is used to organize the graph structure representations into a scan input sequence according to a predetermined order. The Mamba state space scanning part is used to perform recursive state updates on the scan input sequence. The graph readout part is used to perform output organization on the recursive state update results. Based on the graph structure representation construction part in the original Graph-Mamba model, the graph structure representation construction part is transformed into a penetrating graph construction layer. In the penetrating graph construction layer, the graph node representation interface and the graph connection relationship interface are retained, and the graph node representation interface and the graph connection relationship interface are uniformly configured as the accounting penetrating graph state generation interface. The graph structure organization method in the graph structure representation construction part is transformed into a heterogeneous graph organization method oriented towards the accounting penetrating graph state. Based on the graph structure serialization part in the original Graph-Mamba model, a rule coupling layer is configured after the penetrating graph construction layer. The rule coupling layer is connected in series with the penetrating graph construction layer. The input end of the rule coupling layer is connected to the accounting penetrating graph state generation interface of the penetrating graph construction layer, and the output end of the rule coupling layer is connected to the graph structure serialization part. The scanning input sequence organization method of the graph structure serialization part is transformed into a scanning input sequence organization method that includes rule constraint information. Based on the Mamba state space scanning part in the original Graph-Mamba model, the Mamba state space scanning part is transformed into a temporal scanning layer. The input end of the temporal scanning layer is connected to the output end of the regular coupling layer. The temporal scanning layer retains the input projection structure, state recursion structure, and selective scanning structure. The input projection structure, state recursion structure, and selective scanning structure are connected in series in sequence. The scanning input sequence output by the regular coupling layer is written into the input projection structure. After processing by the state recursion structure and the selective scanning structure, a scanning state sequence is formed. Based on the graph readout part of the original Graph-Mamba model, a bidirectional check layer is configured after the temporal scan layer. The input of the bidirectional check layer is connected to the output of the temporal scan layer. The bidirectional check layer includes a forward readout connection structure and a reverse readout connection structure. The forward readout connection structure and the reverse readout connection structure are respectively connected to the scan state sequence. The bidirectional check layer organizes the execution results according to the forward and reverse arrangement directions of the scan state sequence. Based on the penetration graph construction layer, rule coupling layer, temporal scan layer, and bidirectional verification layer, the accounting penetration Graph-Mamba model is constructed. The output of the penetration graph construction layer is connected to the input of the rule coupling layer, the output of the rule coupling layer is connected to the input of the temporal scan layer, the output of the temporal scan layer is connected to the input of the bidirectional verification layer, the bidirectional verification layer retains the bidirectional verification output interface, and the accounting penetration Graph-Mamba model is used for accounting penetration graph state processing. During the training of the Graph-Mamba model for accounting penetration, historical ticket sales data, historical ticket carrier data, historical ticket business status data, historical ticket-level account mapping rules, and the balances of core accounts in the historical civil aviation revenue accounting system and the historical financial accounting system are collected. Following the processing standards for ticket sales data, ticket carrier data, and ticket business status data, historical data undergoes field cleaning, format standardization, ticket number verification, and ticket serial number alignment. A unique historical ticket identifier is constructed based on the ticket number and ticket serial number, and data corresponding to the same unique historical ticket identifier is associated as a single historical record, generating a standardized historical ticket record set. Then, based on the standardized historical ticket record set and the unique historical ticket identifier, a historical ticket-level accounting input set and a historical ticket-level account system are generated. Following the processing flow of the Graph-Mamba model for accounting penetration, historical accounting penetration graph status, historically mapped ticket-level debit and credit vouchers, historical monthly revenue accounting vouchers, and historical core account balances are generated. The historical monthly revenue accounting vouchers, historical core account balances, and the actual verification results from the historical financial accounting system are jointly organized into a historical training sample set. During training, the historical training sample set is sorted and organized according to the accounting period and the historical unique ticket identifier, and then input into the penetration graph construction layer, rule coupling layer, time-series scanning layer, and two-way verification layer respectively. Among them, the penetration graph construction layer outputs the graph node representation and graph connection relationship corresponding to the historical accounting penetration graph status; the rule coupling layer outputs the account mapping result constrained by the business type, carrier status, and amount item correspondence; the time-series scanning layer outputs the scan status sequence after scanning item by item; and the two-way verification layer outputs the verification results of the predicted monthly revenue accounting voucher, the predicted core account balance, and the predicted ticket number-level accounting and account balance. Then, the predicted monthly revenue accounting voucher is matched with the historical monthly revenue accounting voucher item by item, the predicted core account balance is matched with the historical core account balance item by item, and the verification results of the predicted ticket number-level accounting and account balance are matched with the actual verification results item by item, which serves as the basis for loss calculation. In the loss construction process, the total loss is divided into accounting voucher loss, account balance loss, and reconciliation result loss. The accounting voucher loss is calculated based on the differences in debit / credit direction markings, account index differences, and amount value differences between the predicted monthly revenue accounting vouchers and historical monthly revenue accounting vouchers. The account balance loss is calculated based on the balance difference between the predicted core account balance and the historical core account balance. The reconciliation result loss is calculated based on the differences in uncarried amounts, partial usage ratios, refund / refund adjustment results, and consistency of difference types between the predicted invoice number-level accounting and account balance reconciliation results and the actual reconciliation results. The accounting voucher loss, account balance loss, and reconciliation result loss are then combined into the total loss. The Adam optimizer is used to iteratively update the parameters of the accounting penetration Graph-Mamba model. After each training epoch, the validation loss is calculated based on the validation sample set. When the validation loss decreases below a preset threshold for multiple consecutive training epochs or the training epochs reach a preset upper limit, training is stopped, and the model parameters corresponding to the minimum validation loss are retained, resulting in the trained accounting penetration Graph-Mamba model.

[0024] In this embodiment, the penetration map construction module includes: Based on the ticket-level accounting input set, records are merged according to unique ticket identifiers in the penetration graph construction layer. The business type, transportation status, and amount item index corresponding to each unique ticket identifier are extracted. Sales events are generated based on the sales record registration content corresponding to the ticket sales data, and transportation events are generated based on the actual transportation record registration content corresponding to the ticket transportation data. The ticket business layer is constructed based on the sales events and transportation events, forming sales event nodes and transportation event nodes. Each node in the ticket business layer includes a unique ticket identifier, business type, transportation status, and amount item index. Among them, a sales event is a business record in which the ticket completes the sales record registration and forms an initial business amount correspondence, and a transportation event is a business record in which the ticket completes the actual transportation record registration and forms a transportation transfer correspondence. Based on the amount items, business types, and transportation status in the ticket-level accounting input set, the amount item values ​​corresponding to each record are extracted in the penetration diagram construction layer according to the unique ticket identifier. The amount value is generated based on the amount item values. The corresponding debit / credit direction mark is determined according to the business type and transportation status. The corresponding account index is determined according to the ticket-level account system. The unique ticket identifier is written into the source ticket index. Debit record nodes and credit record nodes are constructed based on the debit / credit direction mark, account index, amount value, and source ticket index. The ticket-level debit / credit voucher layer is constructed based on the debit record nodes and credit record nodes. Among them, the ticket-level debit / credit voucher layer node with the debit / credit direction mark is defined as the debit record node, and the ticket-level debit / credit voucher layer node with the credit / credit direction mark is defined as the credit record node. Based on the ticket-level account system, accounts receivable, contract liabilities, and transportation revenue are extracted in the penetration diagram construction layer. The account balance layer is constructed according to the account category, forming accounts receivable nodes, contract liabilities nodes, and transportation revenue nodes. The beginning balance, current change value, and ending balance are written to each node. The beginning balance is taken from the balance record of the corresponding account at the beginning of the current reconciliation period, the current change value is taken from the summary result of the amount mapped into the corresponding account within the current reconciliation period, and the ending balance is taken from the balance record of the corresponding account at the end of the current reconciliation period. Based on the ticket business layer, ticket-level debit and credit voucher layer, and account balance layer, a three-level heterogeneous accounting penetration diagram is constructed. In the penetration diagram construction layer, the nodes in the ticket business layer, ticket-level debit and credit voucher layer, and account balance layer are respectively encoded and organized to generate business node table, voucher node table, and balance node table. Based on the sales event nodes in the ticket business layer and the corresponding nodes in the ticket-level debit and credit voucher layer, a sales formation relationship is constructed. The starting point of the sales formation relationship is the sales event node, and the ending point of the sales formation relationship is the debit record node and credit record node whose source ticket index and the unique ticket identifier corresponding to the sales event node are consistent. The relationship constraints of the sales formation relationship are consistent unique ticket identifier, consistent amount item index, and consistent business type. Based on the carrier event node in the ticket business layer and the corresponding node in the ticket-level debit and credit voucher layer, a carrier transfer relationship is constructed. The starting point of the carrier transfer relationship is the carrier event node, and the ending point of the carrier transfer relationship is the debit record node and credit record node whose source ticket index and the unique ticket identifier corresponding to the carrier event node are consistent. The relationship constraints of the carrier transfer relationship are consistent unique ticket identifier, consistent carrier status, and consistent amount item index. Based on the debit and credit record nodes in the voucher-level debit and credit certificate layer and the corresponding nodes in the account balance layer, an account mapping relationship is constructed. The starting point of the account mapping relationship is the debit and credit record nodes, and the ending point of the account mapping relationship is the accounts receivable account node, contract liability account node, and transportation revenue account node corresponding to the account index. The relationship constraint of the account mapping relationship is that the debit and credit direction markers are consistent with the correspondence of the account index. Based on the established account mapping relationship between the debit and credit record nodes and the corresponding nodes in the account balance layer, a balance accumulation relationship is constructed. The starting point of the balance accumulation relationship is the established account mapping relationship between the debit and credit record nodes, and the ending point of the balance accumulation relationship is the corresponding accounts receivable node, contract liabilities node, and transportation revenue node. The relationship constraint of the balance accumulation relationship is that the current period change value is updated by cumulative addition according to the corresponding amount value, and the ending balance is updated by balance update according to the beginning balance and the current period change value. Based on sales formation relationships, transportation transfer relationships, account mapping relationships, and balance accumulation relationships, a three-level heterogeneous accounting penetration diagram is established to establish inter-layer connections. Inter-layer connections include business connections between the ticket business layer and the ticket-level debit and credit voucher layer, and account connections between the ticket-level debit and credit voucher layer and the account balance layer. Based on the business node table, voucher node table, balance node table, and inter-layer connections, the sales event nodes and transportation event nodes in the business node table, the debit record nodes and credit record nodes in the voucher node table, and the accounts receivable nodes, contract liability nodes, and transportation revenue nodes in the balance node table are associated according to the unique voucher identifier, source voucher index, and account index. The sales formation relationship, transportation transfer relationship, account mapping relationship, and balance accumulation relationship are written into the connection records between the corresponding nodes, generating an accounting penetration diagram state that includes a set of voucher business layer nodes, a set of voucher-level debit and credit voucher layer nodes, a set of account balance layer nodes, and a set of inter-layer connections.

[0025] In this embodiment, the subject mapping conversion module includes: Based on the accounting penetration diagram status, the business type, transportation status, amount item index and unique ticket identifier are extracted from the ticket business layer node set in the rule coupling layer. The amount items in the ticket-level accounting input set are associated with the unique ticket identifier to form the rule query key value corresponding to the current ticket record. The rule query key value consists of business type, transportation status and amount item index. Based on the ticket-level subject mapping rules, the mapping condition field and subject output field are extracted in the rule coupling layer. The rule query key value is compared with the mapping condition field item by item. Based on the ticket-level subject mapping rules that meet all comparison conditions, the corresponding debit subject index, credit subject index and amount item correspondence are extracted to generate the subject mapping result corresponding to the current ticket record. Among them, the amount item correspondence is used to determine the correspondence between the amount item and the debit subject index and the credit subject index. Based on the account mapping results, in the rule coupling layer, the business type, carrier status, debit account index, credit account index and amount item correspondence are organized into a rule embedding representation. The rule embedding representation is then updated in relation to the set of ticket business layer nodes in the accounting penetration diagram to generate a coupled mapping state containing business status information and account information. Based on the coupled mapping state and amount items, the debit and credit record nodes with the same unique ticket identifier corresponding to the source ticket index are located in the ticket-level debit and credit voucher layer. The mapping update process is performed on the debit and credit record nodes. The corresponding account index is written to the debit record node according to the debit account index, and the corresponding account index is written to the credit record node according to the credit account index. The corresponding amount value is extracted from the amount item according to the amount item correspondence and written to the debit and credit record nodes. The source ticket index is corrected according to the unique ticket identifier, and the mapped updated debit and credit record nodes are generated. Based on the updated debit and credit record nodes, corresponding associations are performed in the voucher-level debit and credit voucher layer according to the unique voucher identifier, amount sub-index, and business type. The updated debit and credit record nodes corresponding to the same voucher record are organized into debit and credit records, and an index correspondence is established between the debit and credit records and the corresponding nodes in the voucher business layer to complete the accounting subject mapping conversion. Based on the debit and credit records after the accounting subject mapping conversion is completed, the records are merged according to the unique ticket identifier, and the debit and credit direction mark, subject index, amount value and source ticket index are organized into the corresponding voucher fields to generate the mapped ticket-level debit and credit vouchers.

[0026] In this embodiment, the time-series accounting scanning module includes: Based on the mapped ticket-level debit and credit vouchers and ticket-carrying data, the mapped ticket-level debit and credit vouchers are sorted and organized in the time-series scanning layer according to the accounting period, unique ticket identifier, and accounting order. The mapped ticket-level debit and credit vouchers corresponding to the same unique ticket identifier within the same accounting period are arranged into a state sequence, and the carrying date and carrying status in the ticket-carrying data are written into the corresponding records in the state sequence to generate state scanning input. Based on the state scan input, the debit / credit direction marker, account index, amount value, source document index, transportation date, and transportation status corresponding to each record in the state sequence are extracted. In the time-series scan layer, an account balance status that is synchronously updated with the state sequence is established. The account balance status includes the account balance status at the previous moment, the current voucher change value, and the updated account balance status. The account balance status at the previous moment is used to record the accounts receivable balance, contract liabilities balance, and transportation revenue balance before the current record is executed for state scan. The current voucher change value is used to record the change in the amount value corresponding to the current record for the current account. The updated account balance status is used to record the accounts receivable balance, contract liabilities balance, and transportation revenue balance after the current record is executed for state scan. Based on the mapped voucher-level debit and credit vouchers with sales business type in the state sequence, the state update is performed in the time-series scanning layer according to the accounting causal order of sales forming contract liabilities. The corresponding amount value is written into the current voucher change value. The contract liability account balance is increased according to the debit / credit direction mark and account index. The account receivable account balance is synchronously posted. The correspondence between the contract liability account balance before and after the update is retained, and the sales forming contract liability state update result is generated. Based on the corresponding carrier records in the state sequence and the carrier status in the ticket carrier data, the status update is performed in the time-series scanning layer according to the accounting causal order of carrier offsetting contract liabilities. The amount value in the corresponding carrier record is written into the current voucher change value. The contract liability account balance is offset according to the debit / credit direction mark and account index. The correspondence between the contract liability account balance before and after the update is retained, and the carrier offsetting contract liability status update result is generated. Based on the status update result of the carrier's reduction of contract liabilities, the status update is performed in the time-series scanning layer according to the accounting causal order of revenue recognition. The amount value of the corresponding transportation revenue account in the corresponding carrier record is written into the current voucher change value. The balance of the transportation revenue account is confirmed according to the debit / credit direction mark and account index, and the correspondence between the transportation revenue account balance before and after the update is retained, generating the transportation revenue recognition status update result. Based on the status update results of sales forming contract liabilities, carrier offsetting contract liabilities, and transportation revenue recognition, balance accumulation processing is performed in the time-series scanning layer according to the order of the status sequence. The current voucher change value corresponding to each record is written into the update process of the accounts receivable account balance, contract liabilities account balance, and transportation revenue account balance, generating the core account balance trajectory that changes continuously within the accounting period. The correspondence between the account balance status at the previous moment, the current voucher change value, and the updated account balance status is retained at each record in the status sequence. Based on the core account balance trajectory after the balance accumulation process is completed, the summary organization is performed according to the accounting period. The debit and credit records corresponding to all status update results within the same accounting period are organized into monthly revenue accounting vouchers. The accounts receivable account balance, contract liability account balance, and transportation revenue account balance at the end of the same accounting period are organized into core account balances, generating monthly revenue accounting vouchers and core account balances.

[0027] In this embodiment, the two-way verification and analysis module includes: Based on the two-way verification layer, the balances of core accounts in the civil aviation revenue accounting system and the financial accounting system are extracted. The balances of accounts receivable, contract liabilities, and transportation revenue in the core accounts are then sorted according to account name, accounting period, source document index, and amount value to generate the balance correspondence under the same accounting period. Based on the balance correspondence, in the two-way verification layer, the core account balances in the civil aviation revenue accounting system are read forward through the forward read connection structure, and the core account balances in the financial accounting system are read backward through the reverse read connection structure. The core account balances in the civil aviation revenue accounting system and the core account balances in the financial accounting system are compared item by item. For the receivables account balance, contract liabilities account balance, and transportation revenue account balance, it is determined whether the account name, accounting period, source document index, and amount are consistent. The comparison results are then organized into a consistency verification result. Based on the consistency verification results, core account balances with inconsistent amounts are filtered out. The corresponding monthly revenue accounting vouchers are traced back from the core account balances in the civil aviation revenue accounting system according to the source ticket index. The corresponding mapped ticket-level debit and credit vouchers are traced back from the source ticket index in the monthly revenue accounting vouchers. A corresponding relationship is established between the monthly revenue accounting vouchers and the mapped ticket-level debit and credit vouchers corresponding to the same source ticket index. Based on the monthly revenue accounting vouchers and the mapped voucher-level debit and credit vouchers after establishing the correspondence, the voucher number-level composition corresponding to the current account balance is statistically analyzed. The voucher number-level composition includes the voucher number, voucher serial number, debit / credit direction mark, account index, and amount value that participate in forming the current account balance. The voucher number-level composition is matched with the amount value in the core account balance to determine the voucher number-level composition details of the current account balance. Based on the ticket number level structure and the mapped ticket level debit and credit vouchers, the sales accounting amount and the transportation carry-forward amount corresponding to the same ticket number are extracted. The sales accounting amount is subtracted from the transportation carry-forward amount to calculate the uncarried amount. Among them, the sales accounting amount is the cumulative amount corresponding to the contract liability formed by the same ticket number in the mapped ticket level debit and credit vouchers, and the transportation carry-forward amount is the cumulative amount corresponding to the transportation reduction executed in the mapped ticket level debit and credit vouchers for the same ticket number. Based on the ticket number structure, ticket serial number, and ticket carrying data, the total number of ticket copies and the number of ticket copies already carried under the same ticket number are counted. The usage ratio is calculated using the correspondence between the number of ticket copies already carried and the total number of ticket copies. For ticket numbers where the amount items have been split and recorded by ticket copy, the total amount and the amount already carried under the same ticket number are counted. The usage ratio is calculated using the correspondence between the amount already carried and the total amount. Based on the ticket business status data and the mapped ticket-level debit and credit vouchers, the refund records and rescheduling records corresponding to the same ticket number are extracted. The amount to be reduced corresponding to the refund record and the amount to be recalculated corresponding to the rescheduling record are written into the mapped ticket-level debit and credit vouchers corresponding to the original ticket number and the mapped ticket-level debit and credit vouchers corresponding to the adjusted ticket number, respectively. The refund and rescheduling adjustment results are calculated based on the changes in the amount after reduction and the changes in the amount after recalculation. Based on the consistency verification results, uncarried amount, partial usage ratio, and refund / change adjustment results, the type of difference is determined. For ticket numbers where the monthly revenue accounting voucher in the civil aviation revenue accounting system already has a corresponding record but the corresponding carrying status has not been formed in the ticket carrier data, the difference is identified as the difference where the accounting treatment has been formed but the carrying status has not been formed. For ticket numbers where the corresponding carrying status has been formed in the ticket carrier data but the monthly revenue accounting voucher in the civil aviation revenue accounting system does not have a corresponding record, the difference is identified as the difference where the carrying status has been formed but the accounting treatment has not been formed. Based on the consistency verification results, ticket number-level composition, unsettled amount, partial usage ratio, refund / change / refund adjustment results, and difference type, a verification result record is constructed according to ticket number, ticket copy number, account name, accounting period, and amount value. The consistency verification results, ticket number-level composition, unsettled amount, partial usage ratio, refund / change / refund adjustment results, and difference type are written into the corresponding verification result record to generate ticket number-level accounting and account balance verification results.

[0028] Example 1: To verify the feasibility of this invention in practice, it was applied to a civil aviation enterprise's ticket-level revenue settlement and account balance reconciliation scenario. This enterprise has long operated its civil aviation revenue accounting system and financial accounting system simultaneously, handling domestic passenger tickets, connecting tickets, refunds, rebookings, and some carrier tickets daily. Because the revenue recognition chain—from sales forming contract liabilities to transportation revenue forming transportation revenue—spans different accounting periods, month-end reconciliation frequently encounters problems such as long-term outstanding contract liability balances, inconsistent transportation revenue carry-over criteria, scattered refund and rebooking adjustment records, and difficulty in quickly tracing discrepancies in ticket numbers. Currently, civil aviation revenue settlement and financial accounting generally adopt a separate architecture, with settlement data, accounting vouchers, and account balances processed separately in different systems. This results in a large workload for manual import, comparison, and tracing. Once discrepancies in account balances occur, it is difficult to accurately pinpoint the specific ticket details, easily leading to audit pressure and financial risks.

[0029] In this embodiment, ticket sales data, ticket-related transportation data, ticket business status data, and ticket-level account mapping rules are first collected. Field cleaning, format standardization, ticket number verification, and ticket-related serial number alignment are then performed on all types of basic data. A unique ticket identifier is constructed based on the ticket number and ticket-related serial number, forming a standardized ticket record set. Subsequently, a ticket-level accounting input set is organized based on this standardized record set, and a ticket-level account system is established, incorporating ticket number, ticket-related serial number, amount itemization, customer identifier, transportation status, and business type into a unified processing standard. After completing the input construction, an accounting penetration Graph-Mamba model is constructed. In the penetration graph construction layer, sales events, transportation events, ticket-level debit / credit vouchers, and account balances are organized into a three-level heterogeneous accounting penetration graph, enabling continuous transmission relationships between the ticket business layer, the ticket-level debit / credit voucher layer, and the account balance layer.

[0030] During the rule-coupled processing, the business type, transportation status, and amount items in the ticket-level accounting input set are coupled with the ticket-level account mapping rules to automatically generate mapped ticket-level debit and credit vouchers. For records of sales forming contract liabilities, the system writes corresponding debit and credit records according to the ticket-level account system; for records of transportation revenue formed by transportation transfer, the system triggers corresponding transfers based on the ticket's transportation data; for adjustment records formed by ticket refunds or rescheduling, the system writes the reduction amount and recalculation amount based on the ticket business status data. This processing method ensures that every accounting record can be traced back to the specific ticket number and ticket serial number, avoiding the previous problem of only seeing the summary amount of the account without being able to see the source composition.

[0031] After the mapping is completed and ticket-level debit / credit vouchers are generated, the time-series accounting scanning module performs a status scan according to the accounting causal sequence of sales forming contract liabilities, transportation reducing contract liabilities, revenue recognition, and balance accumulation, generating monthly revenue accounting vouchers and core account balances. Core account balances include accounts receivable balances, contract liabilities balances, and transportation revenue balances. Subsequently, in the two-way reconciliation analysis module, consistency checks are performed between the core account balances in the civil aviation revenue accounting system and the core account balances in the financial accounting system. For records with inconsistent amounts, the system traces back along the source ticket index to the monthly revenue accounting voucher, and then further traces back to the mapped ticket-level debit / credit vouchers, statistically analyzing the ticket number composition, calculating uncarried amounts, partial usage ratios, and refund / change adjustment results, and identifying the types of discrepancies.

[0032] This embodiment selects the actual business structure within three consecutive accounting periods as the verification basis, incorporating a total of 128,640 ticket records and covering 241,380 ticket copy records. These include 198,426 normal sales and transportation records, 18,742 partial transportation records, 9,315 refund records, 14,897 rebooking records, and 36,208 cross-month carry-over records. The verification focuses on three core accounts: accounts receivable, contract liabilities, and transportation revenue. To ensure the comparison is reliable, traditional method one is a manual report verification method, traditional method two is an intermediate database summary verification method, and the method of this invention is a civil aviation industry ticket number-level accounting and account balance verification method based on digital assets. The three methods are tested under the same data range, the same accounting period, and the same verification criteria, recording six indicators: verification accuracy, difference location completion rate, monthly closing verification cycle, average ticket number backtracking time, manual review ratio, and accuracy rate of uncarried-over amount identification.

[0033] Table 1 Comparison of the Reconciliation Results of Ticket Number-Level Accounts and Account Balances

[0034] As shown in Table 1, the method of this invention achieves an accuracy rate of 91.8%, which is 9.2 percentage points higher than the traditional manual report verification method and 4.4 percentage points higher than the intermediate database summary verification method. This result indicates that this invention does not stop at comparing summary amounts, but rather integrates the document business layer, document-level debit / credit voucher layer, and account balance layer, shifting the verification basis from "whether the account amounts are consistent" to "whether the document number composition is consistent." Therefore, it can reduce the bias caused by manual judgment in complex carry-over, partial transportation, and refund / rescheduling scenarios.

[0035] The discrepancy location completion rate reached 90.7%, with a relatively stable improvement. This is because the method of this invention establishes a continuous traceability chain in the two-way verification analysis stage, starting from the core account balance and tracing back to the monthly income accounting vouchers, and then tracing back to the mapped voucher-level debit and credit vouchers. Traditional manual report verification methods often require cross-system retrieval of ticket numbers, ticket copies, and voucher numbers after discovering discrepancies, resulting in a fragmented processing process. While intermediate database summary verification methods can shorten some retrieval paths, they still mainly rely on summary table structures and cannot fully encompass the accounting composition of each ticket number. The method of this invention processes the source ticket index, ticket number, ticket copy serial number, and amount value under a unified structure, thus achieving more complete discrepancy location.

[0036] The monthly reconciliation cycle was shortened from 8.6 days to 2.4 days, the average time for ticket number backtracking was reduced from 27.8 minutes to 4.3 minutes, and the proportion of manual review decreased from 61.5% to 16.9%. These three results collectively demonstrate that the method of this invention directly improves the actual business process. The performance improvement is due to the continuous generation path adopted between the standardized ticket record set, ticket-level accounting input set, mapped ticket-level debit and credit vouchers, monthly revenue accounting vouchers, and core account balances. The system can complete data cleaning, account mapping, time-series scanning, balance verification, and difference tracing within the same process, eliminating the need for financial personnel to repeatedly switch manually between settlement reports, voucher files, and general ledger balance sheets.

[0037] The accuracy rate of identifying unreconciled amounts reached 90.9%, indicating that the present invention is relatively stable in cross-month transportation scenarios. Civil aviation revenue recognition inherently involves discrepancies between the sales month and the transportation month. This business characteristic leads to long-term outstanding contract liabilities and makes it difficult to accurately trace ticket details. The method of this invention constructs a unique ticket identifier based on the ticket number and ticket serial number, and then performs continuous scanning by combining the transportation reduction path and the balance accumulation path. This eliminates the need for manual estimation of unreconciled amounts; instead, it directly calculates them based on the correspondence between the sales accounting amount at the ticket number level and the transportation carry-forward amount, thus resulting in more stable identification results.

[0038] The above are merely preferred embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A civil aviation industry ticket number-level accounting and account balance reconciliation system based on digital assets, characterized in that: include: The basic data processing module is used to collect basic data for ticket-level accounting verification, perform field cleaning, format unification, ticket number verification, ticket serial number alignment, construct unique ticket identifiers, and generate standardized ticket record sets. The accounting input construction module is used to organize and generate a ticket-level accounting input set and establish a ticket-level subject system based on the standardized record set and unique ticket identifier. The accounting penetration model construction module is used to build an accounting penetration Graph-Mamba model, including a penetration graph construction layer, a rule coupling layer, a time-series scanning layer, and a two-way verification layer; The penetration diagram construction module is used to construct a three-level heterogeneous accounting penetration diagram in the penetration diagram construction layer based on the ticket-level accounting input set and ticket-level account system, establish inter-layer connections, and generate the accounting penetration diagram status. The account mapping and conversion module is used to determine the debit and credit accounts in the rule coupling layer based on the accounting penetration diagram status and the voucher-level accounting input set, generate the corresponding debit and credit records, and generate the mapped voucher-level debit and credit vouchers. The time-series accounting scanning module is used to perform status scanning in the time-series scanning layer based on the mapped voucher-level debit and credit vouchers, and generate monthly revenue accounting vouchers and core account balances; The two-way verification analysis module is used to perform consistency checks based on the two-way verification layer. It reverse-traces the monthly revenue accounting vouchers and mapped ticket-level debit and credit vouchers based on the core account balances in the civil aviation revenue accounting system, and calculates and generates ticket number-level accounting and account balance verification results.

2. The civil aviation industry ticket number-level accounting and account balance reconciliation system based on digital assets according to claim 1, characterized in that, The basic data processing module includes: Collect ticket sales data, ticket carrier data, ticket business status data, and ticket-level account mapping rules; perform field integrity checks on ticket sales data, ticket carrier data, and ticket business status data; and generate ticket-level accounting reconciliation basic data. Based on the basic data for ticket-level accounting reconciliation, a unified format processing is performed; Based on the basic data for ticket-level accounting verification after the format has been standardized, ticket number verification is performed, and a unique ticket identifier is constructed according to a fixed combination of ticket number, ticket serial number, and carriage date index. Based on the unique ticket identifier, the ticket sales data, ticket carrier data, and ticket business status data are linked one by one. Based on each record after the association is completed, a standardized record set of tickets is output according to the unique ticket identifier. Each record in the standardized record set of tickets retains the ticket number, ticket serial number, amount item, business type, carrier status, customer identifier, and carrier date index.

3. The civil aviation industry ticket number-level accounting and account balance reconciliation system based on digital assets according to claim 1, characterized in that, The accounting input construction module includes: Based on the standardized record set of tickets, the data records corresponding to the same ticket are merged and organized according to the unique ticket identifier; Based on the merged and organized data records, the amount items are split into ticket amount, tax amount, refund adjustment amount, and rescheduling adjustment amount, forming fixed-length field records. Based on fixed-length field records, output a ticket-level accounting input set according to the unique ticket identifier; Based on the document-level subject mapping rules, the business attributes and accounting subject correspondences corresponding to business types are extracted, and the mapping relationships between business types, carrier status and accounts receivable, contract liabilities, and transportation revenue are organized into a hierarchical subject index structure. A ticket-level subject system is established based on a hierarchical subject index structure.

4. The civil aviation industry ticket number-level accounting and account balance reconciliation system based on digital assets according to claim 1, characterized in that, The accounting transparency model construction module includes: The original Graph-Mamba model is invoked, including the graph structure representation construction part, the graph structure serialization part, the Mamba state space scanning part, and the graph readout part; Based on the graph structure representation construction part in the original Graph-Mamba model, it is transformed into a penetrating graph construction layer, and the graph structure organization method in the graph structure representation construction part is transformed into a heterogeneous graph organization method oriented towards the state of the accounting penetrating graph. Based on the graph structure serialization part, a rule coupling layer is configured after the penetrating graph construction layer to transform the scanning input sequence organization method of the graph structure serialization part into a scanning input sequence organization method that includes rule constraint information; Based on the Mamba state space scanning part, it was modified into a time-series scanning layer; Based on the graph readout section, a bidirectional verification layer is configured after the timing scan layer, including a forward readout connection structure and a reverse readout connection structure; Based on the penetration graph construction layer, rule coupling layer, time-series scanning layer, and bidirectional verification layer, the accounting penetration Graph-Mamba model is constructed.

5. The civil aviation industry ticket number-level accounting and account balance reconciliation system based on digital assets according to claim 1, characterized in that, The penetration map construction module includes: Based on the ticket-level accounting input set, records are merged according to the unique ticket identifier in the penetration graph construction layer to generate sales events and transportation events, and a ticket business layer is constructed to form sales event nodes and transportation event nodes. Each node in the ticket business layer includes a unique ticket identifier, business type, transportation status, and amount item index. Based on the amount items, business types, and carrier status in the ticket-level accounting input set, extract the corresponding amount item values ​​for each record, generate the amount value, determine the corresponding debit / credit direction mark and account index, and write the unique ticket identifier into the source ticket index. Based on the debit / credit direction mark, account index, amount value, and source ticket index, construct debit record nodes and credit record nodes to build a ticket-level debit / credit voucher layer. Based on the ticket-level account system, accounts receivable, contract liabilities, and transportation revenue are extracted to construct an account balance layer, forming accounts receivable nodes, contract liabilities nodes, and transportation revenue nodes. The beginning balance, current period change value, and ending balance are written to each node respectively. Based on the ticket business layer, ticket-level debit and credit voucher layer, and account balance layer, a three-level heterogeneous accounting penetration diagram is constructed. The nodes in the ticket business layer, ticket-level debit and credit voucher layer, and account balance layer are respectively encoded and organized to generate business node table, voucher node table, and balance node table. Based on the sales event nodes in the ticket business layer and the corresponding nodes in the ticket-level loan certificate layer, a sales formation relationship is constructed. Based on the carrier event nodes in the ticket business layer and the corresponding nodes in the ticket-level lending and borrowing voucher layer, a carrier transfer relationship is constructed. Based on the debit record nodes and credit record nodes in the voucher-level debit and credit voucher layer and the corresponding nodes in the account balance layer, an account mapping relationship is constructed; Based on the established account mapping relationship between the debit record node, credit record node and the corresponding node in the account balance layer, construct the balance accumulation relationship; Based on sales formation relationships, transportation transfer relationships, account mapping relationships, and balance accumulation relationships, establish inter-layer connections in a three-level heterogeneous accounting penetration diagram; Based on the business node table, voucher node table, balance node table, and inter-layer connections, an accounting penetration diagram is generated, which includes a set of voucher business layer nodes, a set of voucher-level debit and credit voucher layer nodes, a set of account balance layer nodes, and a set of inter-layer connections.

6. The civil aviation industry ticket number-level accounting and account balance reconciliation system based on digital assets according to claim 1, characterized in that, The subject mapping conversion module includes: Based on the accounting penetration diagram status, the business type, transportation status, amount item index and unique ticket identifier are extracted from the ticket business layer node set in the rule coupling layer to form the rule query key value corresponding to the current ticket record. Based on the ticket-level subject mapping rules, the mapping condition field and subject output field are extracted. The rule query key value is compared with the mapping condition field item by item. Based on the ticket-level subject mapping rules that meet all comparison conditions, the corresponding debit subject index, credit subject index and amount item correspondence are extracted to generate the subject mapping result corresponding to the current ticket record. Based on the account mapping results, the business type, carrier status, debit account index, credit account index and amount item correspondence are organized into a rule embedding representation, and the rule embedding representation is updated in relation to the set of ticket business layer nodes in the accounting penetration diagram to generate a coupled mapping state. Based on the coupled mapping state and amount items, the debit record node and credit record node with the same unique ticket identifier corresponding to the source ticket index are located in the ticket-level debit and credit certificate layer. The mapping update process is performed on the debit record node and credit record node to generate the mapped updated debit record node and mapped updated credit record node. Based on the updated debit and credit record nodes, perform the corresponding association, organize them into debit and credit records, and establish an index correspondence between the debit and credit records and the corresponding nodes in the ticket business layer to complete the accounting subject mapping conversion; Based on the debit and credit records after the accounting subject mapping conversion is completed, a mapped voucher-level debit and credit voucher is generated.

7. The civil aviation industry ticket number-level accounting and account balance reconciliation system based on digital assets according to claim 1, characterized in that, The time-series accounting scanning module includes: Based on the mapped ticket-level debit and credit vouchers and ticket carrier data, the mapped ticket-level debit and credit vouchers are sorted and organized in the time-series scanning layer according to the accounting period, unique ticket identifier and accounting order, arranged into a state sequence, and state scanning input is generated. Based on the state scan input, establish the account balance status that is updated synchronously with the state sequence, including the account balance status at the previous moment, the current voucher change value, and the updated account balance status; Based on the mapped voucher-level debit and credit vouchers with the business type of sales in the state sequence, the state update is performed according to the accounting causal order of sales forming contract liabilities, and the state update result of sales forming contract liabilities is generated. Based on the corresponding carrier records in the state sequence and the carrier status in the ticket carrier data, the status update is performed according to the accounting causal order of carrier offsetting contract liabilities, and the carrier offsetting contract liabilities status update result is generated. Based on the status update result of the carrier's reduction of contract liabilities, the status update is performed according to the accounting causal order of revenue recognition to generate the status update result of transportation revenue recognition; Based on the status update results of sales-generated contract liabilities, carrier-reduced contract liabilities, and transportation revenue recognition, balance accumulation processing is performed according to the order of the status sequence to generate a continuously changing core account balance trajectory during the accounting period. Based on the core account balance trajectory after the completion of balance accumulation processing, the summary organization is performed according to the accounting period. The debit and credit records corresponding to all status update results within the same accounting period are organized into monthly revenue accounting vouchers, and the balances of accounts receivable, contract liabilities, and transportation revenue at the end of the same accounting period are organized into core account balances.

8. The civil aviation industry ticket number-level accounting and account balance reconciliation system based on digital assets according to claim 1, characterized in that, The two-way verification and analysis module includes: Based on the two-way verification layer, the balances of core accounts in the civil aviation revenue accounting system and the financial accounting system are extracted, and corresponding information is generated to produce the balance correspondence relationship under the same accounting period. Based on the balance correspondence, the core account balances in the civil aviation revenue accounting system are read forward through the forward read connection structure, and the core account balances in the financial accounting system are read backward through the reverse read connection structure. The core account balances in the civil aviation revenue accounting system and the core account balances in the financial accounting system are compared item by item, and the comparison results are organized into a consistency verification result. Based on the consistency verification results, core account balances with inconsistent amounts are filtered out. The corresponding monthly revenue accounting vouchers and mapped voucher-level debit and credit vouchers are traced back from the core account balances in the civil aviation revenue accounting system according to the source voucher index, and a corresponding relationship is established. Based on the monthly revenue accounting vouchers and the mapped voucher-level debit and credit vouchers after establishing the correspondence, the voucher number-level composition corresponding to the current account balance is statistically analyzed and verified against the amount value in the core account balance to determine the voucher number-level composition details of the current account balance. Based on the ticket number level composition and the mapped ticket level debit and credit vouchers, extract the sales accounting amount and the carrier transfer amount corresponding to the same ticket number, and calculate the untransferred amount. The total number of ticket copies and the number of ticket copies already shipped under the same ticket number are counted. The usage ratio is calculated by using the correspondence between the number of ticket copies already shipped and the total number of ticket copies. For ticket numbers where the amount items have been split and recorded by ticket copy, the total amount and the amount already shipped under the same ticket number are counted. The usage ratio is calculated by using the correspondence between the amount already shipped and the total amount. Based on the ticket business status data and the mapped ticket-level credit vouchers, extract the refund and reschedule records corresponding to the same ticket number, and calculate the refund and reschedule adjustment results. Based on the consistency verification results, the amount not carried over, the partial usage ratio, and the refund and rescheduling adjustment results, determine the type of discrepancy; Based on the consistency verification results, ticket number-level composition, uncarried amount, partial usage ratio, refund and change adjustment results, and difference type, generate ticket number-level accounting and account balance reconciliation results.

Images