Travel subsidy intelligent optimization method, system and device and medium

Abstract

The invention provides a travel subsidy intelligent optimization method, system and device and a medium, and belongs to the technical field of travel fee management. Receiving a business travel reimbursement request, collecting basic information and a matched voucher, and forming an original data set; the data is cleaned and verified, missing and abnormity are removed, and the data are converted into standardized data; verifying compliance through a rule engine, and screening out non-compliance data; historical data are extracted, a random forest model is utilized to analyze key influence factors, a cost matrix is constructed based on the model, and ant colony algorithm parameters are configured; and iteratively optimizing the cost combination scheme through an ant colony algorithm, selecting an optimal scheme, adjusting in combination with travel demands, and outputting a final subsidy scheme and a contrastive analysis chart. According to the invention, the business travel subsidy is calculated through the ant colony algorithm, the optimal business travel route is automatically selected, the final subsidy scheme is determined, and efficient and accurate business travel fee management is realized.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of travel expense management, and more particularly relates to a travel allowance intelligent optimization method, system, device and medium. BACKGROUND

[0002] In the operation and management of modern enterprises and institutions, the calculation and management of travel allowances are important links to protect the rights and interests of employees and control operating costs. As a cross-field of enterprise human resource management and financial management, the calculation accuracy of travel allowances directly affects the work enthusiasm of employees, and the management efficiency is related to the control level of enterprise operating costs, so it has become one of the core issues of enterprise management.

[0003] The traditional travel allowance calculation method mainly relies on manual auditing and manual accounting, which has significant efficiency bottlenecks and management shortcomings. In actual operation, employees need to manually fill out travel reimbursement forms and attach hundreds of pages of paper vouchers, and financial personnel need to check the itinerary information and expense details one by one, which not only occupies a large amount of human cost, but also is difficult to avoid the omissions caused by manual operation.

[0004] Currently, the related technology lacks effective intelligent support in dealing with complex travel scenarios, making it difficult to meet the needs of enterprise fine management. On the one hand, different industries and different sizes of enterprises have differentiated travel policies, involving city classification, transportation mode restrictions, accommodation standard fluctuations and other multiple rules, and traditional systems are difficult to achieve dynamic policy adaptation; on the other hand, travel expenses are affected by multiple factors such as seasonal fluctuations, market changes, and unexpected events, and fixed standard allowance calculation methods cannot accurately reflect the actual consumption differences, and may result in insufficient allowances or excessive reimbursement. Especially in dealing with complex scenarios such as multi-segment connecting travel and cross-departmental collaboration, the related technology often results in distorted calculation results due to rule hardening and data fragmentation.

[0005] With the deepening of enterprise digital transformation, the limitations of traditional travel allowance management mode are increasingly prominent. At the data level, reimbursement vouchers and travel information are stored in scattered places, making it difficult to form effective data assets for cost analysis and policy optimization; at the process level, there are many audit links and long approval cycles, resulting in delayed expense settlement, affecting employee experience and timeliness of financial data; at the decision-making level, there is a lack of deep mining of historical travel data, which cannot provide data support for enterprise travel policy adjustment, and it is difficult to achieve fine control of travel costs. These problems not only increase the management cost of enterprises, but also restrict the intelligent upgrading of travel management, so there is an urgent need for a travel allowance optimization scheme that integrates intelligent algorithms and management rules to improve calculation accuracy, management efficiency and decision-making scientificity. SUMMARY

[0006] In view of the above problems, the purpose of the present application is to provide a business travel allowance intelligent optimization method, system, device and medium, which calculates the business travel allowance through an ant colony algorithm, automatically selects the optimal business travel route, determines the final allowance scheme, and realizes the optimal allocation and management of financial resources.

[0007] To achieve the above purpose, the present application realizes the following technical solutions: In a first aspect, the present application provides a business travel allowance intelligent optimization method, comprising: Receiving a business expense claim request, collecting basic information and supporting vouchers of the business trip, and forming an original data set; Performing a cleaning operation on the original data set, checking the voucher information and basic information, identifying outliers, filling in missing data, and converting the processed data into standardized data; Calling a rule engine based on the standardized data, performing project compliance, standard compliance and voucher compliance verification according to the enterprise pre-set travel rule library, and generating compliant travel data; Extracting historical compliant data to build a data set, identifying key influence factors using a cost prediction model based on a random forest algorithm, building a cost matrix based on model output and configuring ant colony algorithm parameters; According to the compliant travel data, the ant colony algorithm is used to calculate the business travel allowance, the ant represents the cost combination scheme, the cost items are selected according to the probability formula, the scheme quality is evaluated by the objective function, and the path selection, cost calculation and pheromone update operations are iteratively executed until the preset iteration number is reached; Selecting the optimal scheme in the iteration process, adjusting according to the business travel requirements, generating the final allowance scheme, and outputting a comparative analysis chart.

[0008] In an optional embodiment, the receiving of the business expense claim request, the collection of the basic information and supporting vouchers of the business trip, and the formation of the original data set comprise: Obtaining the business expense claim request through a front-end interface and automatically triggering a data collection process; Obtaining the basic information of the business traveler, the core information of the trip and the cost items, and receiving the supporting vouchers through an encrypted upload interface as basic data; Synchronizing the basic data to a distributed database in real time to form an original data set, and automatically generating a unique verification code for the original data set for subsequent tracing; The basic information of the business traveler includes name, employee number and department; the core information of the trip includes business purpose, number of business travelers, position level, departure time, return time, departure city and arrival city; the cost items include transportation cost items, accommodation cost items and catering cost items; the supporting vouchers include PDF format electronic invoices, JPG format paper invoice scans and XML format electronic travel orders.

[0009] In an optional embodiment, the cleaning operation is performed on the original data set, the voucher information and the basic information are checked, the abnormal values are identified, the missing data is filled, and the processed data is converted into standardized data, comprising: Based on the original data set, the key information of the supporting voucher is extracted by using the OCR technology of the CNN+LSTM deep learning model, and cross-checking is performed with the expense items to identify inconsistent data and mark them; the key information of the supporting voucher includes invoice amount, invoice date, and consumption items; The Z-score method is used to calculate the deviation of the expense items, and the values exceeding 3 times the average value of the same type of expenses in the same city are determined as abnormal values; for missing data, a filling model based on the K-nearest neighbor algorithm is called to interpolate and fill the missing data according to historical similar trip data; all processed data is converted into a standardized format to generate standardized data.

[0010] In an optional embodiment, the standardized data is called based on the rule engine, and the item compliance, standard compliance and voucher compliance are checked according to the enterprise pre-set travel rule library to generate compliant travel data, comprising: The rule engine based on the rete algorithm is called to load the enterprise pre-set travel rule library; The expense upper limit standards of reimbursable list, trip city level and trip person position level are extracted from the enterprise pre-set travel rule library; The expense items are compared with the reimbursable list one by one to exclude non-compliant items and record the corresponding non-compliant data; According to the trip city level and the trip person position level, the corresponding expense upper limit standards are matched to check whether the transportation expense items and the accommodation expense items are over-standard, and the corresponding non-compliant data is recorded; The authenticity of the PDF format electronic invoice and the JPG format paper invoice scan in the supporting voucher is verified by connecting the platform interface of the tax authorities, and the overlap degree of the consumption time and the trip time period is checked according to the electronic travel sheet whether it is greater than or equal to 80%, and the corresponding non-compliant data is recorded; All non-compliant data is excluded from the standardized data to generate compliant travel data; All non-compliant data is generated to include rule-based correction suggestions and real-time feedback.

[0011] In an optional embodiment, the historical compliant data is extracted to construct a data set, a cost prediction model based on the random forest algorithm is used to identify key influencing factors, a cost matrix is constructed based on the model output and the parameters of the ant colony algorithm are configured, comprising: Compliant travel data in a pre-set time period is extracted from the system database, and effective samples are selected according to city level, trip season, transportation mode, accommodation hotel star level, and daily average catering expense to construct a data set; The random forest algorithm is used to construct the cost prediction model, 200 decision trees are set, and the tree depth and node splitting threshold are optimized through 5-fold cross-validation; The data set is input into the cost prediction model, and the key influencing factors are identified to provide quantitative basis for the construction of the cost matrix of the ant colony algorithm; The key influencing factors include the correlation weight of city GDP and accommodation cost, and the influence coefficient of seasonal factors on transportation cost; Based on the key influencing factors, the compliance travel data in the preset time period is converted into an n*m cost matrix; Wherein, n is the number of travel links, and m is the number of optional schemes of travel links; Configure the ant colony algorithm parameters, set the number of ants to 50, set the initial pheromone concentration to 0.5, set the pheromone evaporation coefficient to 0.1, set the pheromone importance factor to 1.2, and set the iteration number to 100.

[0012] In an optional implementation, the ant colony algorithm is executed according to the compliance travel data to calculate the travel allowance, the ant represents the cost combination scheme, the cost item is selected according to the probability formula, the scheme quality is evaluated according to the objective function, and the path selection, cost calculation and pheromone updating operations are iteratively executed until the preset iteration number is reached, including: Initialize the ant colony population, and use each ant to represent a complete travel cost combination scheme; Each ant selects a cost item through the following probability formula:

[0013] Wherein, is the probability of selecting scheme j in travel link i, k is the scheme index, is the current pheromone concentration, A is the pheromone importance factor, is the heuristic information based on historical adoption rate and satisfaction, B is the heuristic information importance factor; After the scheme is constructed, the quality evaluation value F is calculated through the following objective function:

[0014] Wherein, is the total cost weight coefficient, is the weight coefficient of compliance deviation ; After each iteration, the pheromone updating process is executed as follows: According to the formula Global evaporation is performed; wherein, the pheromone evaporation coefficient R=0.1; The top 20% of the schemes ranked according to the quality evaluation value F are sorted according to the formula Perform pheromone enhancement; wherein the enhancement constant Q = 100; Repeat the above process until 100 iterations are completed.

[0015] In an optional embodiment, the optimal scheme in the iteration process is selected, adjusted according to the travel needs, a final subsidy scheme is generated, and a comparative analysis chart is output, including: After the iteration is completed, the occurrence frequencies of all feasible schemes are counted, and the top 3 high-frequency optimal schemes are selected; The high-frequency optimal schemes are adjusted in parameters in combination with preset needs of the current travel; A final subsidy scheme including cost composition details, calculation basis of each item, and rule matching degree score is generated; Based on the final subsidy scheme, a visual comparative analysis chart is generated and output.

[0016] In a second aspect, the embodiments of the present application also provide a travel subsidy intelligent optimization system, comprising: A travel data collection module is configured to receive a travel reimbursement request, collect basic information and supporting vouchers of the travel, and form an original data set; A data processing module is configured to perform a cleaning operation on the original data set, perform voucher information and basic information verification, abnormal value identification, missing data filling processing, and convert the processed data into standardized data; A rule verification module is configured to call a rule engine based on the standardized data, perform item compliance, standard compliance, and voucher compliance verification according to a preset travel rule library of an enterprise, and generate compliant travel data; An algorithm configuration module is configured to extract historical compliant data to construct a data set, identify key influence factors by using a cost prediction model based on a random forest algorithm, construct a cost matrix based on model output, and configure ant colony algorithm parameters; An optimization solving module is configured to perform an ant colony algorithm to calculate travel subsidies according to the compliant travel data, select cost items according to a probability formula through an ant representing a cost combination scheme, evaluate scheme quality through an objective function, and iteratively perform path selection, cost calculation, and pheromone update operations until a preset iteration number is reached; An optimal scheme output module is configured to select an optimal scheme in an iteration process, adjust the optimal scheme according to travel needs, generate a final subsidy scheme, and output a comparative analysis chart.

[0017] In a third aspect, the embodiments of the present application also provide an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the travel subsidy intelligent optimization method according to any one of the above embodiments when executing the program.

[0018] In a fourth aspect, the embodiments of the present application further provide a storage medium having a computer program stored thereon, wherein the computer program is executed by a processor to implement the steps of the intelligent travel allowance optimization method according to any one of the preceding aspects.

[0019] From the above technical solutions, the present application has the following advantages: In the intelligent travel allowance optimization method provided by the present application, the process of ants searching for the optimal path is simulated to optimize the allocation and calculation of travel expenses, ensuring the rationality and accuracy of travel allowances. This method not only considers different business trip needs and budget limitations, but also takes into account the combination of historical data and real-time information, thus more comprehensively reflecting the actual situation of business trips. Through repeated iteration and pheromone updating, the ant colony algorithm can quickly adapt to changes and ensure the best solution in a constantly changing environment. Ultimately, this method aims to improve the decision-making efficiency of enterprises in travel management, reduce unnecessary expenses, and provide scientific and reasonable travel allowances for employees, thereby achieving optimal allocation and management of financial resources and promoting the sustainable development of enterprises.

[0020] The present application replaces the traditional manual review and manual accounting mode by automatic data collection, intelligent preprocessing, and automatic compliance verification by rule engine, significantly shortens the travel reimbursement processing period, reduces repetitive labor, and reduces labor cost investment.

[0021] The present application relies on a multi-dimensional data quality control mechanism, including cross-verification of vouchers and reported data by OCR technology, accurate identification of outliers by Z-score method, interpolation and filling of missing values based on historical similar trip data, and three-level compliance verification (project, standard, and voucher) by rule engine, effectively avoiding human operation omissions and significantly reducing calculation error rate.

[0022] The present application mines key influence factors (such as the correlation between city GDP and accommodation expenses) from historical data through a random forest model, iteratively optimizes the fee combination scheme by combining the ant colony algorithm, and selects the optimal solution based on the objective function (total cost + compliance deviation weighting) to reduce unnecessary expenses while ensuring compliance, thereby improving the efficiency of financial resource allocation.

[0023] The rule engine based on the rete algorithm supports dynamic loading of enterprise travel policies, can flexibly match fee standards according to city level, position level, seasonal factors, and other multi-dimensional factors, and can adapt to complex travel scenarios through preset exception clauses, thus meeting the differentiated management needs of different enterprises.

[0024] The application identifies the cost influencing factors through historical data modeling, and can output visualized charts containing cost composition proportion, historical mean deviation rate and compliance score by combining iterative process high-frequency optimal scheme analysis, so as to provide data-driven decision basis for enterprise travel policy adjustment and cost control strategy optimization, and realize fine management upgrade. BRIEF DESCRIPTION OF DRAWINGS

[0025] In order to more clearly illustrate the technical solutions of the present application, the drawings needed to be used in the description will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor on the basis of these drawings.

[0026] Figure 1 The flowchart of the intelligent optimization method of travel allowance provided by the present application.

[0027] Figure 2 The structural schematic diagram of the intelligent optimization system of travel allowance provided by the present application.

[0028] Figure 3 The structural schematic diagram of the electronic device provided by the present application. DETAILED DESCRIPTION

[0029] In the following detailed description of the specific steps of the intelligent optimization method of travel allowance, various embodiments of the present disclosure will be described more fully. The present disclosure can have various embodiments, and adjustments and changes can be made therein. However, it should be understood that there is no intention to limit various embodiments of the present disclosure to the specific embodiments disclosed herein, but the present disclosure should be understood to cover all adjustments, equivalents and / or alternatives falling within the spirit and scope of various embodiments of the present disclosure.

[0030] In the following, the term "include" or "may include" used in various embodiments of the present disclosure indicates the presence of the disclosed functions, operations or elements, and does not limit the addition of one or more functions, operations or elements. In addition, as used in various embodiments of the present disclosure, the terms "include", "have" and their synonyms only mean to indicate the presence of a specific feature, number, step, operation, element, component or combination of the foregoing, and should not be understood as first excluding the presence or addition of one or more other features, numbers, steps, operations, elements, components or combinations of the foregoing or the possibility of one or more features, numbers, steps, operations, elements, components or combinations of the foregoing.

[0031] With reference to the drawings of the embodiments of the present application, the technical solutions in the embodiments of the present application will be clearly and completely described. Obviously, the described embodiments are only a part of the embodiments of the present application, but not all the embodiments of the present application. Based on the embodiments of the present application, all the other embodiments obtained by a person of ordinary skill in the art without creative work are within the protection scope of the present application.

[0032] Please refer to Figure 1 As shown in FIG. 1, which is a flow chart of a method for intelligent optimization of travel allowance in an embodiment, the method comprises: S1: receiving a travel reimbursement request, collecting basic information and supporting vouchers of the travel, and forming an original data set.

[0033] In the specific embodiment, after the document initiator submits the travel reimbursement request through the front-end page of the enterprise travel management system, the system automatically triggers the data collection process. First, the user identity verification module is used to obtain the basic information of the person going out, including the name, work number and department to which the person belongs; then the travel information filling interface is loaded to guide the user to enter the core information of the trip, which specifically covers the purpose of the trip (such as business negotiation, training and participation in meetings, etc.), the number of people going out, the position level of the person going out (management layer / ordinary employee), the departure time and return time accurate to the hour, the departure city and the arrival city name. At the same time, the cost item entry area is used to read the transportation cost item (including transportation mode, ticket price, etc.), the accommodation cost item (including hotel name, accommodation days, cost amount, etc.), and the catering cost item (including daily catering budget, etc.). Then, the supporting vouchers uploaded by the user are received through the encrypted HTTPS interface, supporting three formats: the electronic invoice in PDF format needs to contain the invoice code, the invoice date, the amount and other key fields; the scanned paper invoice in JPG format requires a resolution of not less than 300 dpi and clear legible text; the electronic itinerary in XML format needs to comply with the standard format of the transportation department. All the collected basic data are synchronized in real time to the distributed database (such as HBase), and a unique 32-bit check code is automatically generated for each original data set, which contains a timestamp and a user ID hash value, for subsequent data tracing and checking in the whole process.

[0034] S2: performing cleaning operation on the original data set, performing voucher information and basic information checking, abnormal value identification, missing data filling processing, and converting the processed data into standardized data.

[0035] In the specific embodiment, first, the voucher information verification process is started: the OCR technology of CNN+LSTM deep learning model is used to recognize the text of the electronic invoice and the scanned copy of the paper invoice, extract the invoice amount, invoice date, consumption project and other key information, and perform cross comparison with the expense item data filled by the user, and when the deviation is more than 5%, it is marked as inconsistent data and triggers the manual review prompt.

[0036] Then the outlier identification link is executed, the Z-score method is used to calculate the deviation of each expense item from the average value of the same type of expense in the same city, the specific formula is Z=(X-μ) / σ, where X is the expense item value, μ is the average value of the same city, and σ is the standard deviation. When |Z|>3, it is determined as an outlier and marked. For missing data (such as the dining expense of part of the trip not filled in), call the filling model based on K-nearest neighbor (K=5) algorithm, and interpolate and fill in the average dining expense of the historical same trip (same city level, season, position level).

[0037] Finally, the data standardization processing is executed: the date is uniformly converted to YYYY-MM-DD format, the time is accurate to HH:MM; the currency unit is unified to RMB yuan and two decimal places; the city name is matched with the standard library of national administrative division, and the city level label (first-tier city / second-tier city / third-tier city and below) is supplemented, forming structured standardized data.

[0038] S3: Based on the standardized data, call the rule engine, execute the project compliance, standard compliance and voucher compliance verification according to the pre-set business travel rule library, and generate compliant travel data.

[0039] In the specific embodiment, first, call the rule engine based on rete algorithm, load the pre-set travel rule library from the enterprise travel policy database, the rule library includes the list of reimbursable items (such as transportation expenses, accommodation expenses, and dining expenses are reimbursable items, entertainment expenses and luxury goods consumption are non-reimbursable items), and the expense limit standard table divided by city level and position level (such as the accommodation expense limit for managers in first-tier cities is 800 yuan per day). In the project compliance verification, the rule engine compares the expense items with the reimbursable list one by one, filters out non-compliant items such as KTV consumption and high-end gift purchase, and records the non-compliant reasons. In the standard compliance verification link, according to the city level and user position level marked in step S2, the corresponding expense limit standard is matched, for example, whether the accommodation expense of an ordinary employee in a first-tier city exceeds 500 yuan per day, and whether the transportation expense chooses a standard-exceeding transportation tool (such as unapproved first class). In the compliance verification of the certificate, the authenticity of the electronic invoice is verified by connecting to the electronic invoice inspection platform interface, and the invoice code and number of the scanned paper invoice are extracted for network verification. At the same time, according to the time information of the electronic travel order, the overlap of the consumption time and the business trip time period is calculated, and the formula is: Overlap = (consumption time intersection business trip time) / business trip total time x 100 When the overlap is less than 80%, it is marked as non-compliant. All non-compliant data generates a correction suggestion containing specific rule clause basis, which is pushed to the user terminal through the system message, and the user corrects and re-submits after correction. The data that passes the verification forms compliant travel data and enters the next link.

[0040] S4: Extract historical compliant data to construct a dataset, identify key impact factors using a random forest algorithm-based cost prediction model, and configure ant colony algorithm parameters based on model output to construct a cost matrix.

[0041] In the specific embodiment, first, the compliant travel data of the enterprise in the past 3 years that passes the verification through step S3 is extracted from the database, and is divided into a training set and a test set in a ratio of 8:2. The data selection link retains valid samples containing city level, business trip season (divided by quarter), transportation mode (airplane / high-speed rail / car, etc.), hotel star rating, daily average catering cost, etc., and eliminates samples with a data missing rate of more than 10%, to construct a structured dataset. The cost prediction model is constructed using a random forest algorithm, 200 CART decision trees are set, and the parameters are optimized through 5-fold cross-validation: the tree depth limit is 10-15 layers, the minimum sample number for node splitting is set to 20, and the minimum sample number for leaf node is set to 5. The training set is input into the model for training, and the model performance is verified through the test set, so that the total cost prediction error is controlled within ±8%. The output layer of the model identifies the key impact factors through the feature importance evaluation module, wherein the correlation weight of city GDP and accommodation cost is 0.72, and the influence coefficient of seasonal factors on transportation cost is 1.2 in peak season (holiday / peak season) and 0.8 in off-season. Based on the above model output, the historical compliant data is converted into an n x m cost matrix (n = 3, corresponding to transportation / accommodation / catering three travel links; m is the number of optional schemes for each link, such as 5 kinds of optional transportation modes in the transportation link). At the same time, the ant colony algorithm parameters are configured: the number of ants N = 50 (based on the dynamic adjustment of historical data volume), the initial pheromone concentration T0 = 0.5, the pheromone volatilization coefficient R = 0.1, the pheromone importance factor A = 1.2, the heuristic information importance factor B = 1.0, and the iteration number K = 100.

[0042] S5: Calculate travel allowances using ant colony algorithm based on compliant travel data. Use ants to represent cost combination schemes, select cost items according to probability formulas, evaluate the quality of the scheme through objective function, and iteratively execute path selection, cost calculation and pheromone update operations until the preset number of iterations is reached.

[0043] In a specific implementation, when initializing the ant colony, each ant is assigned a unique identifier representing a complete travel expense combination (e.g., "second-class high-speed rail seat + three-star hotel + daily meals costing 80 yuan"). Each ant selects an expense item based on a probability formula:

[0044] in, Let k be the probability of choosing option j in travel segment i, and k be the option index. Let A be the current pheromone concentration and A be the pheromone importance factor. B represents the importance factor of the heuristic information, which is based on historical adoption rates and satisfaction levels (calculated by weighting historical solution adoption rates at 60% and user satisfaction scores at 40%).

[0045] After the solution is constructed, the quality assessment value F is calculated using the objective function:

[0046] In the formula, =0.6 is the total cost The weighting coefficients, =0.4 is the compliance deviation. The weighting coefficients (the lower the deviation, the better the solution).

[0047] After each iteration, the pheromone update process is executed: first, according to the formula... Perform global evaporation; where the pheromone evaporation coefficient R = 0.1. Then, for the top 20% of schemes ranked by quantity evaluation value F, apply the formula... Pheromones are enhanced to prioritize high-quality solutions in subsequent iterations. The enhancement constant Q = 100. Repeat the path selection, cost calculation, and pheromone update process until 100 iterations are completed.

[0048] S6: Select the optimal solution during the iteration process, adjust it according to travel needs, generate the final subsidy plan, and output a comparative analysis chart.

[0049] In a specific implementation, after the iteration is completed, the frequency of occurrence of all feasible solutions is counted, and the top 3 most frequent optimal solutions (each occurring ≥15 times) are selected. Parameters are then fine-tuned based on the specific preset requirements of the current business trip. For example, a 20% transportation subsidy is added to itineraries including travel after 22:00, and accommodation costs in remote areas (such as county-level cities) can be increased by 20%. In the final subsidy scheme generation link, the cost composition details (transportation / accommodation / meal expense proportions), the calculation basis of each item (historical average, rule standard, and algorithm optimization result), and the rule matching degree score (100 points in total, including item compliance, standard compliance, and voucher compliance) are automatically integrated to generate the final subsidy scheme; Finally, based on the final subsidy scheme, visual comparison and analysis charts are generated and output. The cost composition proportion pie chart is used to intuitively display the proportions of each cost item; the deviation rate column chart of the average of similar trips in the past half year is used to mark the cost saving / overspending amplitude; the rule compliance radar chart clearly presents the compliance of each item, and all results can be fed back to the user and the financial department through the system interface.

[0050] In this embodiment, through the application of intelligent technology throughout the whole process, multi-dimensional upgrading of travel expense reimbursement management is achieved. Specifically, in the data collection link, automation and encrypted storage are used to ensure information integrity and security; in the preprocessing stage, OCR verification and algorithm filling are used to greatly reduce data error rate; the three-level compliance verification of the rule engine is used to effectively intercept non-compliant expenses and provide real-time feedback and correction suggestions; the random forest modeling and ant colony algorithm iteration optimization are used to achieve accurate matching of expenses and rules; finally, through high-frequency scheme screening, special demand adjustment, and visual analysis, the reimbursement processing efficiency is improved, the cost control and reasonable needs of employees are balanced, data-driven travel policy optimization basis is provided for enterprises, and the accuracy, compliance, and decision-making scientificity of travel management are comprehensively improved.

[0051] As shown in Figure 2 The following is an embodiment of a travel expense reimbursement intelligent optimization system provided by the embodiments of the present disclosure. The system and the travel expense reimbursement intelligent optimization method described above belong to the same inventive concept. Details not described in the embodiment of the travel expense reimbursement intelligent optimization system can be referred to the embodiment of the travel expense reimbursement intelligent optimization method.

[0052] A travel expense reimbursement intelligent optimization system, comprising: A travel data collection module is configured to receive a travel reimbursement request, collect basic information and supporting vouchers of the travel, and form an original data set.

[0053] A data processing module is configured to perform cleaning operations on the original data set, including voucher information and basic information verification, outlier identification, missing data filling processing, and converting the processed data into standardized data.

[0054] A rule verification module is configured to call a rule engine based on the standardized data, perform item compliance, standard compliance, and voucher compliance verification according to a pre-set travel rule library of an enterprise, and generate compliant travel data.

[0055] An algorithm configuration module is configured to extract historical compliance data to construct a data set, identify key influence factors by using a cost prediction model based on a random forest algorithm, construct a cost matrix based on model output, and configure ant colony algorithm parameters.

[0056] An optimization solution module is configured to perform an ant colony algorithm to calculate travel allowance according to compliance travel data, select a cost item according to a probability formula through an ant representing a cost combination scheme, evaluate a scheme quality through an objective function, and iteratively perform path selection, cost calculation, and pheromone update operations until a preset iteration number is reached.

[0057] An optimal scheme output module is configured to select an optimal scheme in an iteration process, generate a final allowance scheme after adjustment according to travel requirements, and output a comparative analysis chart.

[0058] The travel allowance intelligent optimization system provided in the embodiment can automatically collect, clean, and verify data, analyze key influence factors by using a random forest algorithm, construct a cost matrix, iteratively optimize a cost combination scheme by using an ant colony algorithm, and finally generate an efficient, compliant, and cost-optimal allowance scheme, thereby significantly improving the scientificity, accuracy, and execution efficiency of enterprise travel cost management.

[0059] Figure 3 A hardware structure schematic diagram of an electronic device for implementing various embodiments of the present application.

[0060] The travel allowance intelligent optimization method provided in the embodiment can be applied to an electronic device. Those skilled in the art can understand that the electronic device structure involved in the embodiment of the present application does not constitute a limitation on the electronic device, and the electronic device can include more or fewer components than the illustration, or combine certain components, or different component arrangements. In the embodiment of the present application, the electronic device includes but is not limited to a laptop computer, a desktop computer, a workstation, a personal digital assistant, a server, a blade server, a mainframe computer, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as a personal digital processor, a cellular phone, a smart phone, a wearable device, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the embodiments of the present application described herein and / or claimed.

[0061] The electronic device can include a processor, an external memory interface, an internal memory, a universal serial bus (USB) interface, a charge management module, a power management module, a battery, a wireless communication module, an audio module, a speaker, a microphone, a sensor module, a key, a camera, a display screen, and a SIM card interface, etc.

[0062] The processor can include one or more processing units, such as: the processor can include a central processing unit (CPU), an application processor (AP), a modem processor, a graphics processing unit (GPU), an image signal processor (ISP), a controller, a memory, a video code, a digital signal processor (DSP), a baseband processor, and / or a neural-network processing unit (NPU), etc. Among them, different processing units can be independent devices, or can be integrated in one or more processors.

[0063] The processor can be the nerve center and command center of the electronic device. The controller can generate operation control signals according to instruction operation codes and timing signals, and complete the control of fetching and executing instructions.

[0064] The processor can also be provided with a memory for storing instructions and data. In some embodiments, the memory in the processor is a cache memory. The memory can save instructions or data that the processor has just used or repeatedly uses. If the processor needs to use the instruction or data again, it can be directly called from the memory. Avoiding repeated access reduces the waiting time of the processor, thus improving the efficiency of the system.

[0065] The external memory interface can be used to connect an external storage card, such as a MicroSD card, to realize the expansion of the storage capacity of the electronic device. The external storage card communicates with the processor through the external memory interface to realize the data storage function. For example, files such as music and video are saved in the external storage card.

[0066] The internal memory can be used to store computer executable program codes, which include instructions. The processor executes various functional applications and data processing of the electronic device by running the instructions stored in the internal memory. The internal memory can include a program storage area and a data storage area. The internal memory can include a high-speed random access memory, and can also include a non-volatile memory, such as at least one magnetic disk storage device, a flash memory device, a universal flash storage (UFS), etc.

[0067] The wireless communication function of the electronic device can be realized through an antenna, a wireless communication module, a modem processor, and a baseband processor, etc.

[0068] The wireless communication module can provide a wireless communication solution applied to the electronic device, including wireless local area networks (WLAN) (such as a wireless fidelity (Wi-Fi) network), Bluetooth (BT), a global navigation satellite system (GNSS), frequency modulation (FM), near field communication (NFC), infrared technology (IR), and the like.

[0069] The electronic device can realize an audio function and the like through an audio module, a speaker, a receiver, a microphone, a headset interface, an application processor, and the like.

[0070] The electronic device can realize a photographing function through an ISP, a camera, a video codec, a GPU, a display screen, an application processor, and the like.

[0071] The electronic device can realize a display function through a GPU, a display screen, an application processor, and the like.

[0072] The GPU is a microprocessor for image processing, connected to the display screen and the application processor. The GPU is used to perform mathematical and geometric calculations for graphics rendering. The processor can include one or more GPUs that execute program instructions to generate or change display information.

[0073] The display screen is used to display images, videos, and the like. The display screen includes a display panel.

[0074] The electronic device realizes the travel allowance intelligent optimization method through automatic data processing and intelligent algorithm optimization, and achieves the beneficial effect of efficiently generating a travel allowance scheme that is compliant and cost-optimal.

[0075] In the storage medium provided in the present application, a program product capable of realizing the travel allowance intelligent optimization method is stored.

[0076] The travel allowance intelligent optimization method comprises: receiving a travel reimbursement request, collecting basic information and supporting vouchers of the travel, and forming an original data set; performing a cleaning operation on the original data set, performing voucher information and basic information verification, abnormal value identification, missing data filling processing, and converting the processed data into standardized data; calling a rule engine based on the standardized data, performing project compliance, standard compliance, and voucher compliance verification according to the enterprise pre-set travel rule library, and generating compliant travel data; The historical compliance data is extracted to build a dataset, a cost prediction model based on a random forest algorithm is used to identify key influence factors, a cost matrix is built based on the model output and parameters of an ant colony algorithm are configured; The ant colony algorithm is executed according to the compliance travel data to calculate travel allowances, an ant represents a cost combination scheme, a cost item is selected according to a probability formula, a scheme quality is evaluated through an objective function, path selection, cost calculation and pheromone updating operations are iteratively executed until a preset iteration number is reached; An optimal scheme in the iteration process is selected, a final allowance scheme is generated after adjustment according to travel requirements, and a comparison analysis chart is output.

[0077] In some possible implementation manners, the travel allowance intelligent optimization method of the present disclosure can be implemented in the form of a program product, which includes program code for causing a terminal device to execute the steps according to various exemplary embodiments of the present disclosure described in the “Exemplary Method” section of the present specification when the program product is run on the terminal device.

[0078] The storage medium of the present disclosure can adopt any combination of one or more readable media. The readable medium can be a readable signal medium or a readable storage medium. The readable storage medium may, for example, be but is not limited to an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or component, or any combination of the above. More specific examples (non-exhaustive list) of readable storage media include an electrical connection having one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0079] The above description of the disclosed embodiments enables a person skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to these embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for intelligent optimization of travel allowance, characterized in that, The method comprises the following steps: receiving a travel expense claim request, collecting basic information and supporting documents of the travel, and forming an original data set; performing a cleaning operation on the original data set, checking the supporting document information and the basic information, identifying abnormal values, filling in missing data, and converting the processed data into standardized data; calling a rule engine based on the standardized data, performing project compliance, standard compliance, and document compliance verification according to the enterprise pre-set travel rule library, and generating compliant travel data; extracting historical compliance data to construct a data set, identifying key influence factors using a random forest algorithm-based cost prediction model, constructing a cost matrix based on the model output, and configuring ant colony algorithm parameters; performing ant colony algorithm calculation on the compliant travel data to calculate travel allowances, selecting cost items according to a probability formula through an ant representative cost combination scheme, evaluating the scheme quality through an objective function, and iteratively performing path selection, cost calculation, and pheromone updating operations until a preset iteration number is reached; selecting the optimal scheme in the iteration process, adjusting it according to the travel requirements, generating a final subsidy scheme, and outputting comparison analysis charts.

2. The method of claim 1, wherein, The method comprises the following steps: obtaining a travel expense claim request through a front-end interface and automatically triggering a data collection process; obtaining the basic information of the traveler, the core information of the trip, and the cost items, and receiving supporting documents through an encrypted upload interface as basic data; synchronizing the basic data to a distributed database in real time to form an original data set, and automatically generating a unique verification code for the original data set for subsequent tracing; The basic information of the traveler includes name, employee number, and department; the core information of the trip includes the purpose of the trip, the number of travelers, the position level, the departure time, the return time, the departure city, and the arrival city; the cost items include transportation costs, accommodation costs, and catering costs; the supporting documents include PDF format electronic invoices, JPG format paper invoice scans, and XML format electronic travel orders.

3. The method of claim 2, wherein, The method comprises the following steps: based on the original data set, using the OCR technology of the CNN+LSTM deep learning model to extract the key information of the supporting documents, and performing cross-checking with the cost items to identify inconsistent data and mark them; the key information of the supporting documents includes invoice amount, invoice date, and consumption items; using the Z-score method to calculate the deviation of the cost items, and determining that the value exceeding 3 times the average value of the same type of cost in the same city as an abnormal value; for missing data, call the filling model based on the K-nearest neighbor algorithm to interpolate and fill in the missing data; convert all processed data into a standardized format to generate standardized data.

4. The method of claim 3, wherein, The method comprises the following steps: calling a rule engine based on the rete algorithm, loading the enterprise pre-set travel rule library, and performing project compliance, standard compliance, and document compliance verification according to the enterprise pre-set travel rule library to generate compliant travel data. Extract the reimbursable list, the expense ceiling standard of the trip city level and the trip person position level from the enterprise pre-set travel rule library; Compare the expense items with the reimbursable list one by one, filter out non-compliant items, and record the corresponding non-compliant data; According to the trip city level and the trip person position level, match the corresponding expense ceiling standard, check whether the transportation expense item and the accommodation expense item exceed the standard, and record the corresponding non-compliant data; Verify the authenticity of the PDF format electronic invoice and the JPG format paper invoice scan in the supporting voucher through the platform interface of the tax authority, and check whether the overlap degree of the consumption time and the trip time period is greater than or equal to 80% according to the electronic itinerary, and record the corresponding non-compliant data; Filter out all non-compliant data from the standardized data to generate compliant travel data; Generate correction suggestions including rule basis for all non-compliant data and real-time feedback.

5. The method of claim 4, wherein, The extracted historical compliant data constructs a data set, adopts a cost prediction model based on a random forest algorithm to identify key influencing factors, constructs a cost matrix based on model output and configures ant colony algorithm parameters, including: Extract compliant travel data in a preset time period from the system database, filter out effective samples according to city level, trip season, transportation mode, accommodation hotel star level, and daily average catering expense to construct a data set; Use a random forest algorithm to construct a cost prediction model, set 200 decision trees, and optimize tree depth and node split threshold through 5-fold cross-validation; Input the data set into the cost prediction model to identify key influencing factors, which provide quantitative basis for the construction of the cost matrix of the ant colony algorithm; the key influencing factors include the correlation weight of city GDP and accommodation expense, and the influence coefficient of seasonal factors on transportation expense; Based on the key influencing factors, convert the compliant travel data in the preset time period into an n×m cost matrix; where n is the number of travel links and m is the number of selectable schemes for the travel links; Configure ant colony algorithm parameters, set the number of ants to 50, the initial pheromone concentration to 0.5, the pheromone evaporation coefficient to 0.1, the pheromone importance factor to 1.2, and the iteration number to 100.

6. The method of claim 5, wherein, The ant colony algorithm is executed according to the compliant travel data to calculate travel allowances, ants represent cost combination schemes, cost items are selected according to a probability formula, scheme quality is evaluated by an objective function, and path selection, cost calculation and pheromone updating operations are iteratively performed until a preset iteration number is reached, including: Initialize the ant colony population, with each ant representing a complete travel cost combination scheme; Each ant selects a cost item through the following probability formula: wherein, is the probability of selecting the scheme j in the trip segment i, k is the scheme index, is the current pheromone concentration, A is the pheromone importance factor, is the heuristic information based on the historical adoption rate and satisfaction, B is the heuristic information importance factor; After the scheme is constructed, the quality evaluation value F is calculated by the following objective function: wherein, is a weight coefficient for total cost is a weight coefficient for compliance deviation is a weight coefficient for compliance deviation is a weight coefficient for compliance deviation After each iteration, the pheromone updating process is performed as follows: According to the formula Global pheromone evaporation is performed; wherein the pheromone evaporation coefficient R = 0.1; The 20% of the solutions ranked in front of the evaluation value F are sorted according to the formula Perform pheromone reinforcement; wherein the reinforcement constant Q = 100; Repeat the above process until 100 iterations are completed.

7. The method of claim 6, wherein, Select the optimal scheme in the iteration process, adjust it according to the travel requirements to generate the final allowance scheme, and output comparative analysis charts, including: After the iteration is completed, the appearance frequency of all feasible schemes is counted, and the top 3 high-frequency optimal schemes are selected; Adjusting parameters of the high-frequency optimal solution according to preset requirements of the current trip; Generating a final subsidy solution including details of cost composition, calculation basis of each item, and matching degree score of rules; Based on the final subsidy solution, generating and outputting visual comparative analysis charts.

8. A travel reimbursement intelligent optimization system, comprising: The system adopts the intelligent optimization method for business trip subsidies according to any one of claims 1 to 7; The system comprises: A business trip data collection module for receiving a business trip reimbursement request, collecting basic information and supporting vouchers of the business trip, and forming an original data set; A data processing module for performing cleaning operations on the original data set, checking voucher information and basic information, identifying abnormal values, filling in missing data, and converting the processed data into standardized data; A rule verification module for calling a rule engine based on the standardized data, performing item compliance, standard compliance, and voucher compliance verification according to a pre-set business trip rule library, and generating compliant business trip data; An algorithm configuration module for extracting historical compliant data to construct a data set, identifying key influence factors using a cost prediction model based on a random forest algorithm, constructing a cost matrix based on model output, and configuring parameters of an ant colony algorithm; An optimization solving module for calculating business trip subsidies according to the compliant business trip data, selecting cost items according to a probability formula through an ant representing a cost combination solution, evaluating the quality of the solution through an objective function, and iteratively performing path selection, cost calculation, and pheromone update operations until a preset number of iterations is reached; An optimal solution output module for selecting the optimal solution in the iteration process, adjusting it according to the business trip requirements, generating a final subsidy solution, and outputting comparative analysis charts.

9. An electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, The processor implements the steps of the intelligent optimization method for business trip subsidies according to any one of claims 1 to 7 when executing the program.

10. A storage medium having stored thereon a computer program, characterized in that The computer program implements the steps of the intelligent optimization method for business trip subsidies according to any one of claims 1 to 7 when executed by the processor.

Images