Restaurant self-checkout method and system based on intelligent disk binding

Abstract

The invention relates to the technical field of self-checkout, in particular to a restaurant self-checkout method and system based on intelligent tray binding, and the method comprises the following steps: capturing a contact track through installing a sensor at the edge of a tray, judging the attribution of the track, marking the tray to be matched with a seat, obtaining a time period classification food mapping table through a time sequence group, and storing the time period classification food mapping table; and extracting a payment request and identifying delay to obtain an abnormal payment record list, updating the payment interface display structure and outputting an interface payment mode updating list. According to the invention, through tray induction identification, seat matching is realized, consumption sources are clearly identified, data confusion is avoided, a consumption sequence is constructed through track merging, data continuity is improved, time frame mapping is associated with food information, data structure linkage is enhanced, payment behaviors are identified and delayed, processing efficiency is improved, and payment interfaces are dynamically sorted according to abnormal frequency and call amount. The operation fluency and the system stability are optimized, and the settlement efficiency and the payment process stability in the peak period are comprehensively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of self-service checkout, and particularly to a restaurant self-service checkout method and system based on intelligent tray binding. Background Art

[0002] The technical field of self-service checkout includes related technologies in service industries such as retail and catering, where users independently complete the payment process for goods or services. The core content of this technical field lies in optimizing the efficiency of the payment link after consumption by consumers, reducing the dependence on manual services, and improving the overall operation efficiency. Self-service checkout technology covers processes such as identity recognition, product information entry, fee calculation, and payment execution, and involves specific means such as barcode recognition devices, payment integration, interface interaction design, information transmission, and verification mechanisms. In the catering industry, self-service checkout technology is gradually shifting from relying on waiters' operations in the past to a system where users independently complete it, covering forms such as mobile payment, scanning code to order, and electronic menus, forming a complete self-service consumption process system.

[0003] Among them, the restaurant self-service checkout method refers to the operation process in which users, during the dining process in a restaurant, view the consumption details and complete the payment by scanning the QR code on the table or using the restaurant's mobile application. This patent theme focuses on the operation method of the checkout link during the dining process, covering specific technical matters such as the encoding of the table QR code and the binding with the dining position, real-time synchronization of order information, display of order information on the user side, automatic generation of payment links, invocation of third-party payment platform interfaces, and feedback of payment status. The implementation method includes binding each table with a unique identifier to generate a dedicated QR code, and then automatically invoking the server-side order information after the user scans it, while integrating the payment channel to complete the fee settlement process.

[0004] In the prior art, QR codes are often used to bind seats and complete the payment process by users scanning the codes. However, the QR code binding lacks the ability to recognize dynamic behaviors, resulting in difficulties in accurately distinguishing consumption ownership in scenarios where multiple customers dine simultaneously or change seats, thus causing situations such as bill confusion and settlement errors. The consumption behavior record overly relies on the static order synchronization of the server. Once there is network latency or system update lag in the middle, it is easy to cause problems where the picking behavior is inconsistent with the order items, affecting the accuracy during settlement. The payment process is uniformly displayed in a standard interface, and the interface sorting is not optimized according to the response capabilities of different payment methods in actual operation, resulting in some payment methods frequently timing out but still being ranked in the priority display position, dragging down the overall payment efficiency. In terms of the processing of payment exception records, only the status returned by the payment interface is judged, lacking in-depth recognition means based on behavior trajectories and time nodes, making it difficult to timely discover and optimize the structural bottlenecks in the system. Especially during the peak dining period, the phenomenon of slow system response and frequent payment delays has a greater impact on the operation efficiency. Summary of the Invention

[0005] In order to solve the technical problems that in the prior art, a two-dimensional code is often used to bind seats and the user scans the code to complete the payment process, but the two-dimensional code binding lacks the ability to identify dynamic behaviors, resulting in difficulties in accurately distinguishing consumption attribution in scenarios where multiple customers dine at the same time or change seats, which in turn causes situations such as bill chaos and settlement errors. The consumption behavior record overly relies on the static order synchronization of the server. Once there is network latency or system update lag in the middle, it is easy to cause the problem that the picking behavior is inconsistent with the order items, affecting the accuracy during settlement. The payment process is uniformly displayed in a standard interface, and the interface sorting is not optimized according to the response capabilities of different payment methods in actual operation, resulting in some payment methods still being ranked in the priority display position despite frequent timeouts, dragging down the overall payment efficiency. In terms of the processing of payment exception records, only the status returned by the payment interface is judged, lacking in-depth identification means based on behavior trajectories and time nodes, making it difficult to discover and optimize the structural bottlenecks in the system in a timely manner. Especially during the peak dining period, the phenomenon of slow system response and frequent payment delays has a great impact on the operation efficiency. The embodiment of the present invention provides a restaurant self-checkout method and system based on intelligent tray binding. The technical solution is as follows:

[0006] On the one hand, a restaurant self-checkout method based on intelligent tray binding is provided, and the method includes:

[0007] S1: Install sensors on the tray edge and capture the contact trajectory of the sensor contacts. By judging the time difference and offset direction between the starting point of the trajectory and the trigger area of the tray edge, extract the continuously changing coordinates of the picking trajectory path, judge the trajectory attribution, and mark the unique matching result of the tray and the seat to obtain the tray binding seat recognition value;

[0008] S2: Based on the tray binding seat recognition value, collect the time nodes of each tray trigger and the corresponding picking events, screen the continuous records with an interval less than the standard dining period in chronological order, and merge them into a group of consumption behaviors to obtain a consumption behavior time series group;

[0009] S3: According to the consumption behavior time series group and the preset business hour table, classify each picking time into the corresponding time period, align the time nodes of the time period with the meal serving records, and mark the meal names and time period types to obtain a time period classified meal mapping table;

[0010] S4: According to the checkout items in the time period classified meal mapping table, extract the time points of payment request sending and return, identify the time difference and compare it with the payment delay threshold, and screen the payment record numbers that exceed the delay limit to obtain a list of abnormal payment records.

[0011] As a further solution of the present invention, the tray-bound seat identification value includes a path unique identifier, a trajectory stability factor, and a trigger position parameter. The consumption behavior time series group includes a clamping frequency distribution, a time interval sequence, and a behavior duration. The time period classification food mapping table includes a time period label, a set of food types, and a meal delivery record number. The abnormal payment record list includes a payment delay duration, a corresponding payment method code, and an abnormal identification label.

[0012] As a further solution of the present invention, the steps of the tray-bound seat identification value are specifically as follows:

[0013] S101: By installing a sensor on the tray edge and capturing the sensor contact point trajectory, record the time point when the contact point number contacts the trigger area, collect the contact point coordinate change sequence and moving direction within a continuous time, and judge the direction offset and contact interval between the contact point and the tray edge according to the coordinate change direction and time sequence, and generate an offset direction and a time delay value;

[0014] S102: Call the offset direction and time delay value, extract the spatial displacement sequence of the same contact point within a continuous time period, identify the change rate between adjacent coordinate points and the path curvature per unit time, and judge the path change stability according to the coherence of the change rate and the curvature fluctuation range to obtain a trajectory path persistence index group;

[0015] S103: Based on the trajectory path persistence index group, compare the number belonging of each group of trajectory features in the tray sensor sequence, and identify the matching relationship of the seat numbers according to the feature belonging result to generate a tray-bound seat identification value.

[0016] As a further solution of the present invention, the steps of the consumption behavior time series group are specifically as follows:

[0017] S201: Based on the tray-bound seat identification value, collect the tray identification number and time node of each tray trigger record, arrange them in order according to the time node field, and sequentially identify the time interval between adjacent records to obtain a tray time interval sequence;

[0018] S202: According to the tray time interval sequence and a preset standard dining cycle threshold, compare the time interval with the threshold, screen out continuous time periods with an interval less than the threshold, count the tray identification numbers and start and end times within each time period, and perform classification labeling and sequential integration to obtain a continuous record segment interval group;

[0019] S203: Call the tray identification number in the continuous record segment interval group, locate the corresponding clamping event record and extract the time node, and merge them into the start and end time node intervals of the segment for time sorting to obtain a consumption behavior time series group.

[0020] As a further solution of the present invention, for the consumption behavior time series group, the formula is used:

[0021]

[0022] where G represents the consumption behavior time series group, Q i represents the start time of the i-th time node, and E i represents the end time of the i-th time node, and n represents the total number of time nodes.

[0023] As a further solution of the present invention, the steps of the meal mapping table classified by time period are specifically as follows:

[0024] S301: According to the consumption behavior time series group, extract the clamping time, compare it with the start and end times of each time period in the preset business time period table, divide it into business time period types according to the time range, eliminate the data not within any time period, and generate a business time period interval type;

[0025] S302: Call the business time period interval type, compare the clamping time and the meal delivery record time in the short term, extract the meal names of the matching records, and combine them with the corresponding time periods to obtain the combination value of the meal and the business time period;

[0026] S303: Classify according to the time period according to the combination value of the meal and the business time period, sort out the corresponding meal names, count the frequencies after removing duplicates, and output in the order of time periods to obtain the meal mapping table classified by time period.

[0027] As a further solution of the present invention, the steps of the abnormal payment record list are specifically as follows:

[0028] S401: According to the settlement items in the meal mapping table classified by time period, extract the payment request sending time and the payment return time in each record, group them according to the record number, identify the time difference between the sending time and the return time within each group, and store them uniformly according to the record number to obtain the payment response time difference;

[0029] For the payment response time difference, the formula is used:

[0030]

[0031] where ΔT represents the payment response time difference, W j represents the payment request sending time in the j-th group of records, and R j represents the payment return time in the j-th group of records, and m represents the total number of record groups;

[0032] S402: Invoke the payment response time difference, select the time difference corresponding to the record number, compare it item by item with the set payment delay threshold, determine the record numbers greater than the threshold, extract and store the numbers to obtain a set of over-limit payment numbers;

[0033] S403: According to the set of over-limit payment numbers, match the corresponding records in the original checkout items, extract all fields, eliminate the non-abnormal items with time differences within the threshold, and sort them in ascending order of record number to obtain a list of abnormal payment records.

[0034] As a further solution of the present invention, the method further includes step S5:

[0035] S5: Based on the payment method numbers in the list of abnormal payment records, extract the current payment interface configuration items, move the abnormal numbers backward in the interface sorting, and arrange them in order according to the call frequency of non-abnormal items to update the interface display structure, and output a list of updated payment methods for the interface;

[0036] The list of updated payment methods for the interface includes the interface arrangement structure, the call frequency of payment methods, and the payment method priority label.

[0037] As a further solution of the present invention, the steps of the list of updated payment methods for the interface are specifically as follows:

[0038] S501: Based on the payment method numbers in the list of abnormal payment records, extract the number order and display structure of the payment methods from the payment interface configuration, screen the items matching the abnormal numbers, and compare them with the original order to generate the marked positions of abnormal payment methods;

[0039] S502: Invoke the marked positions of abnormal payment methods, move the matching numbers backward in the original configuration order, keep the corresponding order unchanged, extract the call times values of non-abnormal numbers, sort them according to the call times and insert them in front of the abnormal numbers to obtain a re-sorted structure of payment methods;

[0040] S503: According to the re-sorted structure of payment methods, update the display order field in the original configuration item, eliminate duplicate numbers and correct the sorting value range, keep the payment method numbers and sorting sequences, and output a list of updated payment methods for the interface.

[0041] On the other hand, the electric vehicle status monitoring system is used to execute the above-mentioned electric vehicle status monitoring method, and the system includes:

[0042] The trajectory recognition module installs sensors at the edge of the tray, extracts the start and end times of the restaurant slide rail contacts, judges the deviation direction and time difference, combines the coordinate continuity, matches the tray number and the seat number, and obtains the tray-seat pairing information;

[0043] Based on the pairing information of the tray and the seat, the behavior merging module extracts the picking time points under each tray number, arranges them in chronological order, groups the consecutive picking actions with a time interval less than the dining period into one group, and marks them as the same dining behavior according to the seat number to obtain a consumption sequence bound to the seat.

[0044] Based on the consumption sequence bound to the seat, the time period mapping module extracts each picking time, determines the business time period to which it belongs, matches the corresponding dish menu item for the time period, and determines the dish corresponding to each picking event by comparing the coincidence interval between the picking event and the dish serving time, to obtain a list of dishes corresponding to the time period.

[0045] Based on the settlement records in the list of dishes corresponding to the time period, the payment screening module extracts the payment sending time and return time of each record, screens the record numbers that exceed the preset response time, and obtains a set of abnormal payment numbers.

[0046] Based on the payment method numbers in the set of abnormal payment numbers, the interface update module counts the sorting positions and call frequencies of the numbers in the interface configuration, moves the sorted positions of the overtime numbers backward, rearranges the remaining numbers according to the frequencies, and outputs an updated list of interface payment methods.

[0047] The beneficial effects brought by the technical solution provided by the embodiment of the present invention at least include:

[0048] By setting sensors on the tray edge, using the starting contact time and direction offset of the contact track, extracting the continuous coordinate points of the picking path, and judging the corresponding relationship based on the track persistence and curve characteristics, the unique matching between the tray and the seat is realized, effectively distinguishing and identifying the consumption source, and reducing the possibility of confusion of consumption data of different seats. According to this identification value, record the corresponding time node for each picking, and merge the continuous operations according to the time interval rule to form a consumption behavior sequence, avoiding the problem that consumption records are scattered and difficult to merge in manual settlement. Connect the business time period table and the dish serving records according to the consumption behavior time sequence, match the picking time and dish data, and form a dish mapping result under the time period classification, improving the linkage between consumption data and dish content, and providing structured data support for subsequent analysis. Extract and compare the response time of payment records, screen out delayed payment behaviors and form a list of numbers, effectively capturing potential payment anomalies, achieving improvements in both accuracy and efficiency. Rearrange the display order according to the payment method numbers, move the frequently abnormal methods backward, and at the same time optimize the display structure in combination with historical call frequencies, enhancing the stability of the payment interface and the smoothness of user operations. Through accurate identification, data merging, behavior mapping and interactive sorting optimization, strengthen the integrity of the consumption data structure and the traceability of the behavior link, and enhance the adaptability of the checkout interface at the operation level, effectively improving the stable performance of the payment process and the overall settlement efficiency during peak periods. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 is a schematic diagram of the working process of the present invention;

[0050] Figure 2 is a detailed flowchart of S1 of the present invention;

[0051] Figure 3 is a detailed flowchart of S2 of the present invention;

[0052] Figure 4 is a detailed flowchart of S3 of the present invention;

[0053] Figure 5 is a detailed flowchart of S4 of the present invention;

[0054] Figure 6 is a detailed flowchart of S5 of the present invention;

[0055] Figure 7 is the system flowchart of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0056] The technical solutions in the present invention will be described below with reference to the accompanying drawings.

[0057] In the embodiments of the present invention, words such as "exemplarily" and "for example" are used to indicate examples, illustrations or explanations. Any embodiment or design solution described as an "example" in the present invention should not be construed as being more preferred or having more advantages than other embodiments or design solutions. Rather, the use of the word "example" is intended to present concepts in a specific manner. In addition, in the embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one of the two.

[0058] In the embodiments of the present invention, "image" and "picture" can sometimes be used interchangeably. It should be noted that when the difference is not emphasized, the meaning they express is the same. "Of", "corresponding", and "corresponding" can sometimes be used interchangeably. It should be noted that when the difference is not emphasized, the meaning they express is the same.

[0059] In the embodiments of the present invention, sometimes subscripts such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0060] To make the technical problems, technical solutions and advantages to be solved by the present invention clearer, the following will be described in detail with reference to the accompanying drawings and specific embodiments.

[0061] Please refer to Figure 1, an embodiment of the present invention provides a restaurant self-checkout method based on intelligent tray binding. The processing flow of this method may include the following steps:

[0062] S1: By installing sensors on the tray edge and capturing the contact track of the sensor contacts, by judging the contact time difference and offset direction between the track starting point and the tray edge trigger area, extracting the continuous change coordinates of the clamping track path, combining the path persistence and curve stability to judge the track attribution, and marking the unique matching result of the tray and the seat, the tray binding seat recognition value is obtained;

[0063] S2: Based on the tray binding seat recognition value, collect the time nodes of each tray trigger and the corresponding clamping events, screen the continuous records with an interval less than the standard dining cycle in chronological order, and merge them into a group of consumption behaviors to obtain a consumption behavior time series group;

[0064] S3: According to the consumption behavior time series group and the preset business time period table, classify each clamping time into the corresponding time period, align the time nodes of the time period with the meal serving records, and mark the meal names and time period types to obtain a time period classified meal mapping table;

[0065] S4: According to the checkout items in the time period classified meal mapping table, extract the payment request sending and return time points, identify the time difference and compare it with the payment delay threshold, and screen the payment record numbers that exceed the delay limit to obtain a list of abnormal payment records;

[0066] S5: Based on the payment method numbers in the list of abnormal payment records, extract the current payment interface configuration items, move the abnormal numbers to the back in the interface sorting, and arrange them in order according to the call frequency of non-abnormal items to update the interface display structure, and output the interface payment method update list.

[0067] The tray binding seat recognition value includes a path unique identifier, a track stability factor, and a trigger position parameter. The consumption behavior time series group includes a clamping frequency distribution, a time interval sequence, and a behavior duration. The time period classified meal mapping table includes a time period label, a set of meal types, and a meal serving record number. The list of abnormal payment records includes a payment delay duration, a corresponding payment method code, and an abnormal recognition label. The interface payment method update list includes an interface arrangement structure, a payment method call frequency, and a payment method priority label.

[0068] Specifically, as Figure 2 shown, the steps of the tray binding seat recognition value are specifically:

[0069] S101: By installing sensors at the tray edge, capturing the track of the sensor contacts, recording the time points when the contact numbers touch the trigger area, collecting the sequence of contact coordinate changes and the moving direction within a continuous time period, and judging the direction offset and contact interval between the contacts and the tray edge according to the coordinate change direction and the time sequence, generating the offset direction and the time delay value;

[0070] In a high-traffic restaurant environment, the waiter uses a tray equipped with sensors to carry meals. When the edge of the tray touches a table, chair, or object, the sensor is activated and records the coordinate sequence and moving direction of each contact. The process involves first initializing the sensor array to ensure the activity and response speed of each sensor, and then continuously monitoring the coordinate changes of the contacts. Whenever the contact moves, the new coordinate position is captured and the moving direction and speed are calculated immediately. If the contact moves in the same direction and at a similar speed several times in a row, it is judged that the contact is moving along the edge rather than simply jittering or being accidentally touched. This data is then used to analyze the moving pattern and path stability of the contact. For example, if a contact moves 10 cm to the right along the tray edge within 5 consecutive seconds and the moving rate remains at 2 cm per second, the system records the behavior pattern and predicts the future moving trend, generating the offset direction and the time delay value. The generated information can be used to evaluate the waiter's carrying skills and can also be used for further data analysis, such as optimizing the tray design to reduce false triggers.

[0071] S102: Call the offset direction and the time delay value, extract the spatial displacement sequence of the same contact within a continuous time period, identify the change rate between adjacent coordinate points and the path curvature per unit time, judge the path change stability according to the coherence of the change rate and the curvature fluctuation range, and obtain the trajectory path persistence index group;

[0072] Use a tray with sensors to carry food. By analyzing the contact displacement sequence within a continuous time period, stable and unstable carrying trajectories are identified. The process includes first extracting the spatial displacement sequence of each contact from the stored data, and then calculating the displacement rate and path curvature between adjacent coordinate points. For example, if the contact moves from coordinate (10, 10) to (15, 10) within 2 seconds, the moving rate is 2.5 cm per second. Then, the coherence of the rate is evaluated. If the subsequent contact rate drops significantly to 1 cm per second, the system marks the trajectory as unstable. Based on the data, the path persistence index of each contact is further calculated, such as the average rate, the rate change coefficient, the maximum and minimum curvatures. The index set is used to judge the stability of the entire moving process and finally obtain the trajectory path persistence index group. This data helps the management optimize the work process of the packaging line and ensure the accuracy and safety of food handling.

[0073] S103: Based on the trajectory path persistence index group, compare the serial number attribution of each group of trajectory features in the tray sensor sequence. According to the feature attribution result, identify the matching relationship of seat numbers and generate the tray-bound seat recognition value.

[0074] By analyzing the trajectory path persistence indicators captured by each sensor on the tray, the trajectory data is matched with the preset seat numbers on the tray. This process involves using an algorithm to compare each trajectory indicator with the known tray seat standard template. For example, if the average speed and curvature of a certain trajectory match the preset trajectory features of a specific seat, then the trajectory is marked as belonging to that seat. In this way, the tray seat corresponding to each item can be accurately identified. This identification is not based on a single trajectory feature, but on a comprehensive judgment based on a series of persistence indicators, and finally generates the tray-bound seat recognition value. This value is crucial for ensuring the correct sorting and efficient processing of items in the logistics automation configuration.

[0075] Specifically, as Figure 3 shown, the steps of the consumer behavior time series group are specifically as follows:

[0076] S201: Based on the tray-bound seat recognition value, collect the tray identification number and time node recorded each time the tray is triggered, arrange them in order according to the time node field, and sequentially identify the time interval between adjacent records to obtain the tray time interval sequence.

[0077] In a restaurant environment, whenever a waiter or customer takes a meal from the cashier using a tray equipped with a sensor, the identification number of the tray and the meal-taking time are recorded. Thereafter, each time the tray triggers a sensor (such as when it is put down or picked up again), a new time node is generated. This recording method helps the restaurant track the fluidity and usage frequency of the tray. The data is arranged in the order of the time nodes. This arrangement process involves data pulling, time formatting, and sequence generation. For example, a tray numbered 001 is taken at 12:00 and triggers the sensor at different tables at 12:15 and 12:45 respectively. Then these three time nodes are recorded and arranged in sequence. Subsequently, the interval between adjacent time nodes is identified and recorded. Calculating the interval helps analyze the tray usage cycle and the length of the customer's dining time. The generation of the time interval involves a simple subtraction operation between two time points. For example, the time interval from 12:00 to 12:15 is 15 minutes, and the time interval from 12:15 to 12:45 is 30 minutes, obtaining the tray time interval sequence. This sequence provides important data on the restaurant's operation efficiency and can be used to further optimize customer service and resource allocation.

[0078] S202: Compare the time intervals with the threshold according to the tray time interval sequence and the preset standard dining cycle threshold, screen out the continuous time periods with intervals less than the threshold, count the tray identification numbers and start and end times within each time period, and perform classification labeling and sequential integration to obtain a continuous record segment interval group;

[0079] The restaurant management sets the standard dining cycle to 30 minutes. This threshold is obtained through statistical analysis based on average dining time data. The threshold setting process includes collecting historical data, calculating the average value and standard deviation, and determining the optimal threshold to adapt to the dining habits of most customers. After the tray time interval sequence is generated, each time interval is compared with the 30-minute threshold. This comparison operation is a basic conditional judgment process. For example, if an interval is 20 minutes, which is less than the 30-minute threshold, then this interval is recorded as a fast dining. Screen out all time intervals less than the threshold. The intervals indicate faster dining transitions or very busy periods. The screening operation involves traversing the time interval sequence and applying conditional statements. Count the tray identification numbers and start and end times within each time period, and perform classification labeling and sequential integration. This statistical and integration process involves data classification and time period determination, such as determining all fast dining time periods and the corresponding tray numbers, and then integrating the information in chronological order to obtain a continuous record segment interval group. This interval group provides the restaurant with insights into peak dining periods and customer dining behavior patterns, helping to further adjust the restaurant layout and service strategies to meet customer needs.

[0080] S203: Call the tray identification numbers in the continuous record segment interval group, locate the corresponding picking event records and extract the time nodes, and import them into the start and end time node intervals of the segment for time sorting and merging to obtain a consumption behavior time series group;

[0081] For the consumption behavior time series group, use the formula:

[0082]

[0083] where G represents the consumption behavior time series group, Q i represents the start time of the i-th time node, E i represents the end time of the i-th time node, and n represents the total number of time nodes;

[0084] The "consumer behavior time series group" is formed by sorting multiple picking records generated during the use of trays in a restaurant in chronological order, screening consecutive records with a time interval less than the "standard dining cycle", and then merging them. The specific process includes: First, identify the tray number and time node to form a tray time interval sequence; then, combine consecutive segments with a time interval less than the threshold (i.e., the standard dining cycle) into a group of consumer behaviors; finally, extract the picking events in these segments and integrate them into the final "consumer behavior time series group". From these steps, it can be seen that the scope of the "accumulation process" is defined based on whether the time interval is less than the standard dining cycle. Any consecutive picking behavior that meets this condition will be regarded as a consecutive dining behavior and included in the calculation scope. Therefore, this process does not accumulate all picking records without discrimination, but has a threshold limit for behavior merging, that is, the boundary condition of the "dining cycle". This method ensures the time coherence and behavior integrity of the consumption sequence, which is a conditional "restricted accumulation";

[0085] This formula is used to sort a set of time nodes to obtain the final consumer behavior time series. The acquisition and processing methods of each parameter in the formula are as follows:

[0086] G: The final sorted time series group, representing the set of time nodes of all consumer behaviors after sorting. This result represents the time nodes arranged in chronological order, reflecting the chronological relationship of each consumer behavior;

[0087] Q i : The start time of the i-th time node. This refers to the start time corresponding to a certain consumer behavior within the entire time interval. The acquisition of this parameter needs to be automatically recorded by the system and is collected through induction devices or Internet of Things sensors. For example, through the time stamp of barcode scanning or time marking methods. During actual data collection, the unit of Q i is seconds, and the time stamp during collection can be obtained through the time interface of the system and measured in seconds;

[0088] E i : The end time of the i-th time node, representing the end time of the corresponding consumer behavior, similar to Q i , this data is collected through the sensors of the automation system, with the unit of seconds. The data quantization process is the same as that of Q i , and the end time of each consumer behavior is also collected through recording devices;

[0089] n: The total number of time nodes, representing the number of all collected time nodes, obtained by counting the consumer behavior records within a certain time interval. In the calculation, this parameter represents the total number of all consumer event records;

[0090] Qi and E i Collection of: The time parameters can be automatically recorded through IoT sensors, barcode scanners, timestamps of mobile devices, etc. Whenever a consumption behavior occurs, its start and end times are recorded, with the unit being seconds, representing the time since a certain fixed time point (such as epoch time);

[0091] Calculation of n: n is obtained by counting the start and end times of each consumption event. During a certain period, count the total number of start and end time points of consumption behaviors and determine the value of n based on the number of time points;

[0092] Unification of time units: All time parameters (Q i and E i ) in the formula are in seconds to ensure consistency. If the data source uses different time units (such as milliseconds or minutes), appropriate conversion is required to make the units consistent;

[0093] Suppose there are 5 consumption behavior events collected within a certain period, and their corresponding start times and end times are as follows:

[0094] Q 1 = 10 seconds, E 1 = 15 seconds, Q 2 = 25 seconds, E 2 = 35 seconds, Q 3 = 50 seconds, E 3 = 55 seconds, Q 4 = 70 seconds, E 4 = 75 seconds, Q 5 = 90 seconds, E 5 = 100 seconds;

[0095] Step 1: Calculate the absolute time difference for each time period:

[0096] |Q 1 - E 1 | = |10 - 15| = 5 seconds;

[0097] |Q 2 - E 2 | = |25 - 35| = 10 seconds;

[0098] |Q 3 - E 3 | = |50 - 55| = 5 seconds;

[0099] |Q 4 - E 4 | = |70 - 75| = 5 seconds;

[0100] |Q 5 - E5 |=|90-100|=10 seconds;

[0101] Compute the square root term:

[0102]

[0103] Calculate the time node sorting results:

[0104]

[0105] The final G value is 4.24, which represents the weighted sum of the total time series combination after time sorting. The value 4.24 indicates the sum of the weighted time differences within the specified time interval. This value can be further used in subsequent analysis to reflect the time characteristics of each consumer behavior event in the entire sequence.

[0106] Specifically, if Figure 4 As shown, the steps of the time period classification meal mapping table are as follows:

[0107] S301: Extract the gripping time according to the consumption behavior time series group, compare it with the start and end time of each time period in the preset business period table, divide it into business period types according to the time range, remove the data that is not in any time period, and generate the business period interval type;

[0108] Extract the picking time and compare it with the start and end time of each time period in the preset business time table. In the actual operation of the restaurant, the picking time is extracted from each customer's buffet picking record. For example, a customer picks at 12:05 and compares this time with the preset business hours of the restaurant. The business time table sets the lunch period from 11:00 to 14:00 and the dinner period from 17:00 to 20:00. First, compare the time point of 12:05 with the start and end time of each business period one by one to determine whether it is within the preset time period. If it is picked up at 12:05, the time point is compared with the start and end time of each business period. The time 12:05 is between 11:00 and 14:00, so this time point is classified as "lunch period". Then all the picking times are processed in the same way, classified by time period, and those data that are not in any time period are eliminated. For example, if a picking time is 10:30, the system will eliminate it because it does not meet any preset business hours. Finally, the data of the business hour interval type is generated, in which each picking time is accurately classified into the corresponding time period interval, for example, 12:05 is classified as "lunch period" and 10:30 is eliminated.

[0109] S302: calling the business hour interval type, comparing the short-term pick-up time with the meal delivery record time, extracting the matching recorded meal names, and combining them with the corresponding time periods to obtain a meal and business hour combination value;

[0110] In the automated ordering system of a restaurant, customers select and pick up dishes through self-service devices, record the time of each pick-up, and then compare it with the dish preparation records in the kitchen. For example, if a customer picks up Dish A at 12:05, check if there is a matching dish preparation completion time. Suppose the kitchen finishes making Dish A and is ready to serve it at 12:07. If the time difference between the two is small, it is a pick-up and dish preparation record within a short period. At this time, the system will extract the dish name in this record, such as Dish A, and then combine this dish with its "lunch period" to generate a combined value of the dish and the business period. For example, Dish A and the "lunch period" are combined into one value. This process is repeated for each pick-up time and dish preparation record, and finally, the combination of the dish and its corresponding period is obtained. The combined value will help the restaurant analyze the most popular dishes during each period and optimize the dish supply strategy.

[0111] S303: Classify according to the combined value of the dish and the business period, sort out the corresponding dish names, remove duplicates, count the frequencies, and output in the order of the periods to obtain the mapping table of the classified dishes by period;

[0112] When processing the combined value of the dish and the business period, classify and sort according to the dish combinations within each period, group all dishes by the business period. For example, put all the dishes in the "lunch period" together and all the dishes in the "dinner period" together. Then remove the duplicate dish names to ensure that each dish is only counted once within the period. Count the frequency of each dish within each period. For example, assume that Dish A appears 5 times and Dish B appears 3 times during the lunch period. Count the frequencies and output the results in the order of the periods. The dishes and their frequency data within all periods will generate the mapping table of the classified dishes by period in the order of the periods. In the example, during the lunch period, Dish A appears 5 times and Dish B appears 3 times, and during the dinner period, Dish C appears 8 times and Dish D appears 4 times, etc. Finally, generate the mapping table of the classified dishes by period, which can provide specific data support for the restaurant's operation decision-making.

[0113] Specifically, as Figure 5 shown, the steps of the list of abnormal payment records are specifically as follows:

[0114] S401: According to the settlement items in the mapping table of the classified dishes by period, extract the payment request sending time and the payment return time in each record, group them by the record number, identify the time difference between the sending time and the return time within each group, and store them uniformly by the record number to obtain the payment response time difference;

[0115] The payment response time difference uses the formula:

[0116]

[0117] where, ΔT represents the payment response time difference, Wj represents the payment request sending time in the j-th group of records, R j represents the payment return time in the j-th group of records, and m represents the total number of record groups;

[0118] The logic for obtaining the "payment response time difference" is to extract the time point of sending the payment request and the time point of payment return confirmation for each payment record, and calculate the difference between these two time points, that is, the time consumed from the payment request to the payment response. This time difference is used to evaluate the response efficiency of the payment system, and can reveal whether there are problems with slow processing for certain payment methods, thus providing data basis for optimizing the checkout experience;

[0119] The operation method of "dividing the difference by the total number of record groups" mentioned in the text is not simply to obtain the difference of a single record, but to uniformly process the response time differences of multiple records, that is, by calculating the sum of the response time differences of all payment records and then dividing by the total number of records, so as to obtain an "average payment response time difference". Such a calculation method is not to calculate the difference between two values within a certain group, but to reflect the overall response performance of the entire payment system during a certain period or for a certain payment method;

[0120] The reason for adopting such an operation method is to achieve the purpose of data standardization and trend recognition. By averaging multiple payment response time differences, the interference caused by individual extreme values can be avoided, and at the same time, the system can macroscopically judge the overall trend of payment performance. If the average response time of a certain payment method is significantly higher than the preset delay threshold of the system, then it can be determined that this method is abnormal, and then it can be sorted and moved backward or marked as low priority in the interface. This method not only improves the rationality of the payment method display, but also plays an optimization role in terms of user experience and system stability;

[0121] In this formula, it is mainly used to calculate the response time difference between the payment request sending time and the payment return time in each group of records. The acquisition and quantification process of each specific parameter is as follows:

[0122] W j : represents the payment request sending time in the j-th group of records. This data is obtained by real-time monitoring and recording in the payment system, with the unit of seconds. The payment request sending time is the moment when the payment request is sent after the user performs the payment operation, and is obtained through server logs or payment platform APIs. For example, assume that in a certain group of records, W j = 12:05:01 (i.e., 12 hours, 05 minutes, and 01 second);

[0123] R j: Represents the payment return time in the j-th group of records. This time is recorded from the confirmation information returned by the payment platform, with the unit of seconds, indicating the response time after the payment request is sent. It is obtained by recording the confirmation response of the payment platform or the return time of the payment system. For example, assume that in a certain group of records, R j = 12:05:10 (i.e., 12 hours, 05 minutes, and 10 seconds);

[0124] |W j -R j |: Calculate the absolute difference between the sending time and the return time, with the unit of seconds. This value represents the response time difference for each payment operation, reflecting the time delay in payment processing. For example, the payment request sending time for the first group of records is 12:05:01, and the payment return time is 12:05:10. The response time difference is:

[0125] |W j -R j | = |12:05:10 - 12:05:01| = 9;

[0126] Represents the sum of the squares of the response time differences in all records. Here, m is the number of record groups. After calculating the absolute value of the response time difference for each group of records and then squaring it, the squared values of all records are accumulated to obtain the total sum. Assume there are 3 groups of records, and the calculated response time differences are 9 seconds, 7 seconds, and 5 seconds respectively;

[0127]

[0128] m: The total number of record groups, representing the number of payment request groups currently being counted. For the above example, assume there are 3 groups of records, so m = 3;

[0129] By substituting the data, the average value of the payment response time difference can be calculated;

[0130] The specific calculation is as follows:

[0131] The calculation result shows that the average response time difference for the 3 groups of payment records counted is 51.67 seconds. This result reflects the average delay time between the payment request and the payment return. Through this value, the delay in the payment system during the processing of requests and responses can be evaluated, and further optimization of the system performance can be carried out.

[0132] S402: Invoke the payment response time difference, select the time difference corresponding to the record number, compare it item by item with the set payment delay threshold, determine the record numbers greater than the threshold, extract and store the numbers to obtain the set of over-limit payment numbers;

[0133] In the self-checkout of a smart restaurant, the payment response time difference is the time difference between when a customer completes a payment operation and when the payment system confirms the payment status. First, based on the time nodes recorded for each payment event, calculate the payment response time difference. For example, assume a customer completes a payment at 12:10:05 and the payment confirmation time is 12:10:30. Then the payment response time difference is 25 seconds. Compare the time difference of each payment record with a preset payment delay threshold. If the set payment delay threshold is 20 seconds, then 25 seconds exceeds the threshold, and record the payment operation number as abnormal. Payment events that exceed the threshold indicate a slower payment confirmation speed, resulting in a poor customer experience. Filter out the record numbers of all payment responses with a time greater than the set threshold, extract and store the numbers of abnormal payment events, and finally obtain a set of over-limit payment numbers. For example, payment events numbered 101, 102, and 105 are recorded as the set of over-limit payment numbers, which provides basic data for subsequent exception handling.

[0134] S403: According to the set of over-limit payment numbers, match the corresponding records in the original checkout items, extract all fields, eliminate non-abnormal items with a time difference within the threshold, and sort them in ascending order by record number to obtain a list of abnormal payment records;

[0135] By comparing and matching the set of over-limit payment numbers with the payment records in the original checkout record form, each payment record includes fields such as payment number, payment time, customer number, and payment amount. According to the extracted over-limit payment numbers, find the corresponding complete payment records and extract all relevant fields from them. For example, assume the record with payment number 101 involves customer A paying 100 yuan at 12:10:05. Extract the payment information of customer A from the original checkout record form, including all fields such as payment time and payment amount. Further eliminate the normal payment records with a payment response time within the preset threshold, that is, remove the data with a payment response time less than or equal to the set threshold, which can be achieved through filtering operations in the database. All eligible abnormal payment records will be extracted and sorted in ascending order by record number for subsequent review and processing, obtaining a list of abnormal payment records sorted by time.

[0136] Specifically, as Figure 6 shown, the steps for updating the list of interface payment methods are specifically as follows:

[0137] S501: Based on the payment method numbers in the list of abnormal payment records, extract the number order and display structure of the payment methods from the payment interface configuration, filter out the items that match the abnormal numbers, and compare them with the original order to generate the marked positions of abnormal payment methods;

[0138] According to the payment method number in the list of abnormal payment records, extract the order and display structure of the payment method from the payment interface configuration. Assume that in the payment interface, the configuration order of the payment methods is: credit card, WeChat Pay, Alipay, cash. First, according to the payment method number in the abnormal payment record (for example, a record with the number "2" represents WeChat Pay), extract the position and display order of WeChat Pay from the configuration table. Then, by comparing the number order of each payment method, filter out the items that match the abnormal payment record number. Assume that the payment method number "3" (i.e., Alipay payment) appears in the abnormal payment record, and it will match Alipay in the third position in the display structure. And according to its order in the configuration table, compare it with the original order to further determine which payment methods do not meet the display order requirements, generate the marked position of the abnormal payment method, and record the abnormal position of the payment method. For example, if Alipay payment should be displayed in the second position but is actually displayed in the third position, then mark this payment method as abnormal.

[0139] S502: Invoke the marked position of the abnormal payment method, move the matching number backward in the original configuration order, and keep the corresponding order unchanged. Extract the call count values of the non-abnormal numbers, sort them by the call count, and insert them in front of the abnormal number to obtain the reordered structure of the payment methods.

[0140] Move the matching payment method number backward in the original configuration order. Assume that WeChat Pay is marked as abnormal and its position in the payment method display order should be the second. Adjust the payment method number of WeChat Pay to the end of the display order, keeping the relative order of the payment methods unchanged. For example, if the order of Alipay payment and cash payment was originally the first and the fourth, keep the relative positions of these two payment methods unchanged in the updated configuration, just move WeChat Pay to the last fifth position, ensuring that the payment method order is adjusted, but the relative order of the payment methods is not affected. Extract the call count values of the non-abnormal numbers. For example, a certain payment method may be called 10 times, and another payment method is only called 2 times. The system sorts the payment methods according to the call count, gives priority to arranging the payment method with a higher call count in the front, and inserts the payment method with a higher call count in front of the abnormal number according to the sorting result to obtain the reordered structure of the payment methods.

[0141] S503: According to the reordered structure of the payment methods, update the display order field in the original configuration item, remove duplicate numbers, correct the sorting value range, keep the payment method number and the sorting sequence, and output the updated list of the payment methods on the interface.

[0142] Reorder the structure according to the payment method, update the display order field in the original payment configuration item. Assume that there are payment method numbers "1" to "5" in the original payment method configuration table. Rearrange the display order of the payment methods, eliminate duplicate number records, ensure that each payment method appears only once, correct the sorting value range of the payment methods. If the number of a certain payment method is "1", ensure that it is displayed at the front. If the payment method with the number "5" is moved to the end, update its sorting position in the configuration table according to the new sorting logic to ensure the display order is accurate. Retain the payment method number and its new sorting sequence, and finally output the interface payment method update list. The content in the list will include the number of each payment method, the updated order, and the sorting value range, ensuring that the payment interface can display the payment methods according to the updated configuration so that customers can more conveniently select payment methods during self-checkout.

[0143] As Figure 7 shown, a restaurant self-checkout system based on intelligent tray binding, the system includes:

[0144] The trajectory recognition module installs sensors on the edge of the tray, extracts the start and end times of the contacts on the restaurant slide rail, judges the deviation direction and time difference, combines the coordinate continuity, matches the tray number with the seat number, and obtains the tray-seat pairing information;

[0145] The behavior merging module, based on the tray-seat pairing information, extracts the picking time points under each tray number, arranges them in chronological order, groups the continuous picking actions with a time interval less than the dining period into one group, and marks them as the same dining behavior according to the seat number to obtain the consumption sequence of the bound seat;

[0146] The time period mapping module, based on the consumption sequence of the bound seat, extracts each picking time, judges the business time period to which it belongs, matches the corresponding dish menu items for the time period, and determines the corresponding dishes for each picking event by comparing the coincidence interval between the picking event and the dish serving time to obtain the dish list corresponding to the time period;

[0147] The payment screening module, based on the checkout records in the dish list corresponding to the time period, extracts the payment sending time and return time of each record, screens the record numbers that exceed the preset response time to obtain a set of abnormal payment numbers;

[0148] The interface update module, based on the payment method numbers in the set of abnormal payment numbers, counts the sorting positions and call frequencies of the numbers in the interface configuration, moves the timed-out numbers to the back after sorting, rearranges the remaining numbers according to the frequencies, and outputs the interface payment method update list.

[0149] The above are only the specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention can easily think of changes or substitutions, which should all be covered by the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

Claims

1. A restaurant self-service checkout method based on smart binding disk, characterized in that: The following steps are involved: S1: Install a sensor on the edge of the pallet and capture the sensor contact trajectory. By determining the contact time difference and offset direction between the starting point of the trajectory and the trigger area on the edge of the pallet, extract the continuously changing coordinates of the gripping trajectory path, determine the trajectory ownership, mark the unique matching result between the pallet and the seat, and obtain the pallet-bound seat identification value. S2: Based on the tray-bound seat identification value, the time node of each tray triggering and the corresponding gripping event are collected, and the continuous records with an interval less than the standard dining cycle are filtered in chronological order, and merged into a group of consumption behaviors to obtain a consumption behavior time series group; S3: According to the consumption behavior time series group and the preset business period table, each gripping time is classified into the corresponding time period, and the time node of the time period is aligned with the meal delivery record, and the meal name and time period type are marked to obtain a time period classification meal mapping table; S4: According to the checkout items in the meal mapping table classified by the time period, the time points of sending and returning the payment request are extracted, the time difference is identified and compared with the payment delay threshold, and the payment record numbers exceeding the delay limit are screened to obtain a list of abnormal payment records.

2. The restaurant self-service checkout method based on smart disk binding according to claim 1 is characterized in that: The tray-bound seat identification value includes a unique path identifier, a trajectory stability factor, and a trigger position parameter; the consumption behavior time series group includes a gripping frequency distribution, a time interval sequence, and a behavior duration; the time period classification meal mapping table includes a time period label, a meal type set, and a meal delivery record number; the abnormal payment record list includes a payment delay duration, a corresponding payment method code, and an abnormal identification label.

3. The restaurant self-service checkout method based on smart disk binding according to claim 1 is characterized in that: The steps of binding the tray to the seat identification value are specifically as follows: S101: Installing a sensor on the edge of the tray and capturing the sensor contact trajectory, recording the time point when the contact number contacts the trigger area, collecting the contact coordinate change sequence and movement direction in continuous time, judging the direction offset and contact interval between the contact and the tray edge according to the coordinate change direction and time sequence, and generating the offset direction and time delay value; S102: calling the offset direction and time delay value, extracting the spatial displacement sequence of the same contact point in a continuous time period, identifying the change rate between adjacent coordinate points and the path curvature per unit time, judging the path change stability according to the coherence of the change rate and the curvature fluctuation range, and obtaining a trajectory path continuity index group; S103: Based on the trajectory path continuity index group, compare the number attribution of each group of trajectory features in the pallet sensor sequence, identify the matching relationship of the seat number according to the feature attribution result, and generate a pallet-bound seat identification value.

4. The restaurant self-service checkout method based on smart disk binding according to claim 1 is characterized in that: The steps of the consumption behavior time series group are specifically as follows: S201: Based on the pallet-bound seat identification value, the pallet identification number and time node of each pallet triggering record are collected, arranged in sequence according to the time node field, and the time intervals between adjacent records are identified in sequence to obtain a pallet time interval sequence; S202: According to the tray time interval sequence and the preset standard dining cycle threshold, the time interval is compared with the threshold, the continuous time period whose interval is less than the threshold is selected, the tray identification number and the start and end time in each time period are counted, and the classification and labeling and sequential integration are performed to obtain a continuous record segment interval group; S203: calling the tray identification number in the continuous record segment interval group, locating the corresponding gripping event record and extracting the time node, importing it into the start and end time node interval of the segment to perform time sorting and merging, and obtaining the consumption behavior time series group.

5. The restaurant self-service checkout method based on smart disk binding according to claim 4 is characterized in that: The consumption behavior time series group adopts the formula: Among them, G represents the consumption behavior time series group, Q i represents the starting time of the i-th time node, E i represents the end time of the i-th time node, and n represents the total number of time nodes.

6. The restaurant self-service checkout method based on smart disk binding according to claim 1 is characterized in that: The steps of the time period classification meal mapping table are specifically as follows: S301: extracting the gripping time according to the consumption behavior time series group, comparing it with the start and end time of each time period in the preset business period table, dividing it into business period types according to the time range, eliminating data that is not in any time period, and generating business period interval types; S302: calling the business period interval type, comparing the short-term pick-up time with the meal delivery record time, extracting the names of the matching records, and combining them with the time period to obtain a combination value of the meal and the business period; S303: Classify by time period according to the combination value of the meal and the business hour, sort out the corresponding meal names, count the frequencies after removing duplicates, and output them in order of time periods to obtain a time period classified meal mapping table.

7. The restaurant self-service checkout method based on smart disk binding according to claim 1 is characterized in that: The steps of listing abnormal payment records are as follows: S401: extract the payment request sending time and payment return time in each record according to the checkout items in the time period classification meal mapping table, group them by record number, identify the time difference between the sending time and the return time in each group, store them uniformly by record number, and obtain the payment response time difference; The payment response time difference is calculated using the formula: Where ΔT represents the payment response time difference, W j represents the time when the payment request in the jth group of records is sent, R j represents the payment return time in the jth group of records, and m represents the total number of record groups; S402: calling the payment response time difference, selecting the time difference corresponding to the record number, comparing it item by item with the set payment delay threshold, determining the record number greater than the threshold, extracting and storing the number, and obtaining an over-limit payment number set; S403: According to the set of over-limit payment numbers, the corresponding records are matched in the original checkout items, all fields are extracted, non-abnormal items with time differences within the threshold are eliminated, and the records are sorted in ascending order by record number to obtain an abnormal payment record list.

8. The restaurant self-service checkout method based on smart disk binding according to claim 1 is characterized in that: The method further comprises step S5: S5: Based on the payment method number in the abnormal payment record list, extract the current payment interface configuration item, move the abnormal number backward in the interface sorting, and arrange the updated interface display structure in order according to the call frequency of non-abnormal items, and output the interface payment method update list; The interface payment method update list includes an interface arrangement structure, a payment method call frequency, and a payment method priority label.

9. The restaurant self-service checkout method based on smart disk binding according to claim 8 is characterized in that: The steps of updating the payment method list on the interface are specifically as follows: S501: Based on the payment method number in the abnormal payment record list, extract the number sequence and display structure of the payment method from the payment interface configuration, filter out items matching the abnormal number, and compare them with the original sequence to generate an abnormal payment method mark position; S502: calling the abnormal payment method mark position, moving the matching number backward in the original configuration order, and keeping the corresponding order unchanged, extracting the call count value of the non-abnormal number, sorting by the call count and inserting it in front of the abnormal number, and obtaining the payment method reordering structure; S503: According to the payment method reordering structure, the display order field in the original configuration item is updated, duplicate numbers are removed and the sorting value interval is corrected, the payment method number and sorting sequence are retained, and the payment method update list is output.

10. A restaurant self-service checkout system based on smart tray, characterized in that: According to any one of claims 1 to 9, the restaurant self-service checkout method based on the smart binding disk comprises: The trajectory recognition module installs a sensor on the edge of the tray to extract the start and end time of the restaurant slide rail contact, determine the offset direction and time difference, combine the coordinate continuity, match the tray number and the seat number, and obtain the tray and seat pairing information; The behavior merging module extracts the gripping time points under each tray number based on the tray and seat pairing information, arranges them in chronological order, groups the continuous gripping actions with a time interval less than the dining cycle into one group, and marks them as the same dining behavior according to the seat number, thereby obtaining the consumption sequence bound to the seat; The time period mapping module extracts each gripping time based on the consumption sequence of the bound seat, determines the business time period to which it belongs, and matches the meal delivery menu items corresponding to the time period. By comparing the overlapping intervals of the gripping events and the meal delivery time, the meal corresponding to each gripping event is determined, and a list of meals corresponding to the time period is obtained; The payment screening module extracts the payment sending time and return time of each record based on the checkout records in the meal list corresponding to the time period, and screens the record numbers that exceed the preset response time to obtain a set of abnormal payment numbers; The interface update module counts the payment method numbers in the abnormal payment number set, the ranking position and call frequency of the numbers in the interface configuration, sorts the timed numbers backward, rearranges the remaining numbers according to the frequency, and outputs the interface payment method update list.

Images