Hotel self-service check-in terminal based on multi-language interaction and intelligent room selection function

Abstract

The invention relates to the technical field of smart hotels, in particular to a hotel self-service check-in terminal based on multi-language interaction and an intelligent room selection function. The hotel self-service check-in terminal aims at solving the problems that existing self-service equipment is isolated in function and cannot be collaboratively optimized with a hotel operation system, and the group check-in efficiency is low. The control mainboard of the hotel self-service check-in terminal is provided with a dynamic resource collaborative allocation engine. The engine comprises an operation state interface unit which is used for obtaining a hotel room cleaning state, a task queue and future passenger flow distribution data in real time; the individual customer resource optimization unit is used for calculating and recommending an optimal room by quantifying user preferences, cleaning ready time and future congestion cost; and the group session collaboration unit is used for realizing parallel verification of members through session identification codes and intelligently distributing houses based on a graph division algorithm. Therefore, efficient collaborative allocation and intelligent management of hotel guest room resources are realized.

Description

A hotel self-check-in terminal based on multilingual interaction and intelligent room selection functions Technical Field

[0001] This invention relates to the field of smart hotel technology, specifically to a hotel self-check-in terminal based on multilingual interaction and intelligent room selection functions. Background Technology

[0002] With the deepening of digital transformation in the hotel industry, self-service check-in terminals have become key equipment for improving front desk efficiency and optimizing guest experience. Existing self-service check-in terminals generally possess basic functions such as identity verification (e.g., document reading, facial recognition), online payment, room key generation, and multilingual interfaces. Some more advanced terminals can also interface with hotel property management systems (PMS) or online travel agencies (OTAs) to automatically obtain guest booking information.

[0003] However, existing self-service check-in terminals still have significant shortcomings in achieving true "intelligence" and "collaboration," mainly in the following two aspects:

[0004] Firstly, regarding serving individual guests, the existing room selection function is essentially a static filtering process that only focuses on individual preferences. For example, when a guest selects preferences such as "high floor" or "non-smoking room," the system only filters from the pool of currently "empty and cleaned" rooms to provide a list of rooms that meet these hard criteria. This model has two major flaws: First, it completely ignores the impact of the hotel's real-time operational status. For example, the system might recommend a room located at the furthest point from the cleaning staff, even if the room meets the guest's preference, as this allocation would lead to inefficient cleaning routes or longer waiting times for the guest. Second, it lacks forward-looking considerations.

[0005] Secondly, existing solutions are weak in serving group guests. A common practice is for the group leader to handle check-in for all members at the front desk or on a single terminal, or for members to sequentially complete the process on the same terminal. This process is time-consuming and cumbersome, and room allocation heavily relies on the front desk staff's experience or simple sequential assignment, failing to efficiently and reasonably meet the complex needs of group members regarding room proximity (such as a family wanting adjacent or connected rooms). While some solutions propose sharing booking information via QR codes, this remains a one-way transmission of information and independent processing by members, failing to achieve true collaboration and intelligent room allocation. In other words, existing terminals, when handling group bookings, simply replicate the check-in process for individual guests, without designing a collaborative mechanism for parallel processing and optimized allocation specifically for group scenarios.

[0006] In summary, existing hotel self-check-in terminals are still in the "automation" stage in terms of functionality. Their decision-making process is isolated and static, failing to be deeply integrated with the hotel's real-time dynamic operation system (such as room status, cleaning, and guest flow forecasting), and failing to effectively solve the pain points of group check-in.

[0007] Therefore, how to enable self-service terminals to have dynamic resource collaborative allocation capabilities, so that they can make better room selection recommendations for individual guests based on global real-time data, and provide efficient and intelligent collaborative processing and room allocation services for group guests, has become a technical problem that urgently needs to be solved in this field. Summary of the Invention

[0008] The purpose of this invention is to overcome the shortcomings of existing self-service check-in terminals, such as isolated functions, lack of collaboration with hotel operation systems, and low efficiency in group check-in. This invention provides a hotel self-service check-in terminal based on multilingual interaction and intelligent room selection. By introducing a dynamic resource collaborative allocation engine, this terminal can perceive the hotel's operational status in real time, optimize room selection for individual guests based on global efficiency, and provide collaborative parallel check-in and intelligent room allocation services for group guests, thereby systematically improving hotel operational efficiency and guest experience. To achieve the above objectives, the embodiments of this invention disclose the following technical solutions:

[0009] A hotel self-check-in terminal based on multilingual interaction and intelligent room selection functions includes a touch screen display, an identity recognition module, a payment module, a printing module, a network communication module, and a control motherboard.

[0010] The identity recognition module is used to obtain and verify the user's identity information in order to generate an identity verification credential.

[0011] The control motherboard is configured to run a dynamic resource collaborative allocation engine, which includes:

[0012] The operation status interface unit is used to obtain data from the hotel management system in real time through the network communication module. The data includes the cleaning status identifier of each room, the order of the cleaning task queue, and the floor reservation distribution data of guests to be arrived within the future scheduled time.

[0013] The individual customer resource optimization unit, connected to the operation status interface unit, is used to perform the following steps when serving individual customer users:

[0014] Obtain the room preference vector of the current individual customer;

[0015] Based on the data from the operation status interface unit, the allocation cost of each available room that matches the room type of individual customers is calculated. The allocation cost is obtained by weighted summation of at least the first term and the second term. The first term represents the matching degree between the room preference vector and the room's static feature vector. The second term represents the cleaning readiness time cost calculated based on the room's cleaning status identifier and its position in the cleaning task queue.

[0016] Output the room with the lowest allocation cost as the first recommendation result to the touch screen;

[0017] The group session coordination unit is used to perform the following steps when serving group users:

[0018] Generate and display group session identification codes at the self-service check-in terminal of the first hotel;

[0019] In response to scanning of the group session identifier by one or more other hotel self-service check-in terminals, the user identity verified on the one or more other hotel self-service check-in terminals is associated with the same group check-in transaction.

[0020] Under the condition of satisfying the total room type constraint of the group, the graph partitioning calculation is performed based on the relationship data between group members and the room location information. With the goal of maximizing the association strength within the member group that is assigned to adjacent or same-floor rooms, the group room allocation scheme is output to the first hotel self-service check-in terminal or a hotel self-service check-in terminal specified in the group check-in transaction.

[0021] The dynamic resource collaborative allocation engine is also configured to trigger the corresponding service processes of the individual customer resource optimization unit or the group session collaborative unit based on the identity verification credentials generated by the identity recognition module.

[0022] Optionally, when the individual guest resource optimization unit calculates the allocation cost, it also includes a third item, which represents the future congestion cost. The future congestion cost is calculated in the following way: based on the floor reservation distribution data of guests to be arrived within the future scheduled time, predict the potential increase in traffic density on the floor where the current room is allocated due to the arrival of new guests within the target time window; the allocation cost is the weighted sum of the first, second and third items.

[0023] Optionally, the allocation cost performed by the individual passenger resource optimization unit The calculation formula is:

[0024]

[0025] in, For the user's room preference vector, This represents the static feature vector of the room. To calculate the vector norm of the first term; The estimated waiting time, calculated based on the cleaning status identifier "to be cleaned" and its index position in the cleaning task queue, is used to characterize the second item; The third item is defined as the number of new occupants expected to move into that floor within a specific future time period, calculated based on the floor reservation distribution data. These are preset non-negative weighting coefficients.

[0026] Optionally, in the group conversation collaboration unit, the step of performing graph partitioning calculation based on the relationship data between group members and room location information specifically includes:

[0027] Using group members as vertices, assign edge weights to every two member vertices based on relationship data;

[0028] Obtain the room topology of the hotel floors and define adjacent or same-floor rooms as meeting preset location conditions;

[0029] Solve a constrained optimization problem: Given the required number of rooms of various types for the group, assign a specific room to each member such that the sum of the edge weights of all members assigned to room pairs that meet the preset location conditions is maximized.

[0030] Optionally, the dynamic resource collaborative allocation engine further includes a pressure feedback unit, which is used for:

[0031] Based on the current check-in waiting queue length and historical average processing speed obtained from the operation status interface unit, calculate the real-time estimated waiting time;

[0032] When the real-time estimated waiting time exceeds a first threshold, an efficiency guidance instruction is generated and sent to the control logic of the touch screen, so that when the touch screen displays the first recommendation result, it adds a visual marker to the room where the cleaning ready time cost is lower than a second threshold.

[0033] Optionally, the self-service check-in terminal further includes a near-field communication module, and the group session collaboration unit broadcasts a beacon signal containing the group session identifier code through the near-field communication module; the hotel self-service check-in terminal is configured to: when in an idle state, if it receives the beacon signal broadcast by another terminal through the near-field communication module, it activates and displays a prompt message on the touch screen to guide the user to scan the terminal screen to join the corresponding group session.

[0034] Optionally, the control motherboard is further configured with an adaptive weighting module, which is connected to the individual passenger resource optimization unit and is used to perform periodic operations.

[0035] Collect operational efficiency indicators corresponding to different combinations of weighting coefficients within a historical time period. The operational efficiency indicators include average customer waiting time and average cleaning staff moving distance.

[0036] Based on the historical data, the recommended weight coefficient combination for the next cycle is updated using regression analysis or optimization algorithms, and the updated weight coefficients are provided to the individual tourist resource optimization unit.

[0037] Optionally, the network communication module is connected to the hotel's door lock management system; the control motherboard is further configured to: after the user completes transaction confirmation through the payment module, regardless of whether it comes from the first recommendation result from the individual guest resource optimization unit or the group room allocation scheme from the group session collaboration unit, bind the final assigned room number with the corresponding user's identity, generate an authorization instruction, and send it to the door lock management system through the network communication module, so that the door lock management system authorizes the corresponding room door lock.

[0038] Optionally, when the individual guest resource optimization unit outputs the first recommendation result, it simultaneously generates a brief decision basis text for each recommended room;

[0039] The multilingual interaction module that controls the motherboard converts the brief decision-making basis text into the language currently set by the user, and displays it as an interactive control next to the recommended items for the corresponding room on the touch screen.

[0040] Optionally, the operation status interface unit is also used to acquire real-time operation status data of the hotel elevator;

[0041] When calculating the allocation cost or the future congestion cost, the individual passenger resource optimization unit also incorporates the distance between the target room's floor and the elevator lobby, as well as the current elevator operating load status, as additional factors into the calculation model.

[0042] The beneficial effects of this invention are as follows:

[0043] (1) By using a dynamic resource collaborative allocation engine, the decision-making of the self-service terminal is deeply linked with the real-time operation status of the hotel, thereby improving the overall operational efficiency and resource utilization of the hotel.

[0044] (2) By using the “group session identifier code” mechanism and combining it with the intelligent room allocation algorithm based on graph partitioning, not only is the overall processing time for groups shortened, but the satisfaction and convenience of group guests are also significantly improved.

[0045] (3) Adjust the interaction strategy dynamically according to real-time pressure and continuously optimize the algorithm parameters through historical data feedback so that the system can maintain the best decision performance in the long term and adapt to the ever-changing operating environment.

[0046] (4) By generating concise, data-driven decision-making information for each room recommendation and presenting it in the language chosen by the guest, guest trust is increased. At the same time, it provides hotel management with a data-driven view of resource allocation and adjustment methods. Attached Figure Description

[0047] To more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings in the following description are merely exemplary, and those skilled in the art can derive other embodiments based on the provided drawings without creative effort.

[0048] Figure 1 is a hardware system block diagram of a hotel self-check-in terminal based on multilingual interaction and intelligent room selection functions provided in an embodiment of the present invention;

[0049] Figure 2 is a flowchart of the optimized recommendation method for individual guest accommodation provided in an embodiment of the present invention;

[0050] Figure 3 is a flowchart of the group check-in collaborative processing and intelligent room allocation method provided in an embodiment of the present invention;

[0051] Figure 4 is a schematic diagram of the architecture of a hotel self-check-in terminal based on multilingual interaction and intelligent room selection function to interact with other hotel systems in an embodiment of the present invention. Detailed Implementation

[0052] Specific embodiments of the invention will now be described in detail. Although the invention is described in conjunction with these specific embodiments, it should be understood that it is not intended to limit the invention to these specific embodiments. Rather, these embodiments are intended to cover alternative, modified, or equivalent embodiments that may be included within the spirit and scope of the invention as defined by the claims. In the following description, numerous specific details are set forth in order to provide a thorough understanding of the invention. The invention may be practiced without some or all of these specific details.

[0053] When used in conjunction with the terms "comprising," "method comprising," or similar language in this specification and appended claims, the singular forms "a," "some," and "the" include plural references unless the context clearly indicates otherwise. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0054] The following description, with reference to the accompanying drawings, illustrates an embodiment of the hotel self-check-in terminal based on multilingual interaction and intelligent room selection functions. Addressing the problems mentioned in the background art regarding the isolated functions, lack of integration with the hotel's operating system, and low efficiency in group check-in, the present invention provides a hotel self-check-in terminal based on multilingual interaction and intelligent room selection functions. Through an integrated dynamic resource collaborative allocation engine, it intelligently fuses and globally optimizes discrete customer requests with real-time, multi-dimensional operational status data from the hotel's backend, thereby improving the experience of individual customers while systematically enhancing the overall operational efficiency of the hotel.

[0055] Specifically, a dynamic resource collaborative allocation engine is introduced as a crucial part of the decision-making process. By constructing a unified cost function model, it quantifies and incorporates multi-source heterogeneous data, such as customer static preferences, real-time cleaning scheduling costs, and future floor congestion predictions, into a unified decision framework. This transforms individual guest room selection recommendations from isolated matching actions into a collaborative optimization result that balances individual satisfaction with the overall operational load of the hotel. For group check-in scenarios, the engine reconstructs the traditional linear check-in process into a parallel and intelligent collaborative transaction through a conversational collaboration mechanism and graph partitioning optimization algorithm. This not only significantly shortens check-in time through parallel QR code verification but, more importantly, automatically solves the optimal room allocation scheme through algorithms, scientifically meeting the needs of member relationships and fundamentally addressing the pain points of low efficiency and unreasonable room allocation in group check-in. Through the aforementioned conversational collaboration mechanism, this terminal has successfully evolved from a passive, single-function business processing point into an intelligent node capable of proactively sensing, predicting, and optimizing hotel resource allocation. This ensures personalized and convenient customer service, superior to traditional self-service terminals or purely manual services that rely solely on static rules; it also achieves global optimization of hotel operational efficiency (such as cleaning paths and guest flow distribution).

[0056] Ultimately, the solution achieved a closed loop of "data-driven decision-making" at the technical level and a dual improvement in "service experience" and "operational efficiency" at the business level, providing a feasible system-level solution for the evolution of smart hotels from process automation to intelligent resource scheduling.

[0057] The preferred embodiments of the present invention will be described in detail below with reference to Figures 1-4, so that those skilled in the art can implement the present invention accordingly.

[0058] This hotel self-check-in terminal, based on multilingual interaction and intelligent room selection, includes a touch screen, an identity recognition module, a payment module, a printing module, a network communication module, and a control motherboard. The control motherboard is configured to run a dynamic resource collaborative allocation engine, which includes an operation status interface unit, a personal guest resource optimization unit, and a group conversation collaboration unit.

[0059] The following will describe in detail the various parts of a hotel self-check-in terminal based on multilingual interaction and intelligent room selection functions according to an embodiment of this application.

[0060] The identity verification module is used to obtain and verify the user's identity information to generate identity verification credentials. The dynamic resource collaborative allocation engine is configured to trigger the corresponding service processes of the individual customer resource optimization unit or the group session collaboration unit based on the identity verification credentials generated by the identity verification module.

[0061] Specifically, the identity recognition module includes an ID card reader, a passport scanner, and a high-definition camera. It reads document information and captures the user's facial image, compares the two to complete verification, and generates an identity verification credential containing an encrypted identity identifier. After the user is verified by the identity recognition module, the system generates an identity verification credential containing an encrypted identity identifier. Upon receiving this credential, the dynamic resource collaborative allocation engine first sends it to the hotel's backend system to check if the user has valid online booking information. If the user is an individual traveler, the corresponding service process of the individual traveler resource optimization unit is triggered; if the user selects group mode, the corresponding service process of the group session collaboration unit is triggered.

[0062] The operation status interface unit is used to obtain data from the hotel management system in real time through the network communication module. The data includes the cleaning status indicators of each room, the order of cleaning task queues, and the floor reservation distribution data of guests to be arrived within the future scheduled time.

[0063] Specifically, the Operational Status Interface Unit serves as the hub for data interaction between the dynamic resource collaborative allocation engine and the hotel's back-end operational environment. This unit collaborates with the network communication module through standardized data interface protocols (such as HTTP RESTful API or WebSocket), periodically or event-triggeredly initiating data requests to the hotel management system, or subscribing to its data update pushes. The "data from the hotel management system" it acquires is a structured dataset, specifically including:

[0064] For each physical room in the hotel, its current status is retrieved. The "cleaning status" is typically an enumeration value, such as: "VC" (Vacant Clean), "VD" (Vacant Dirty), "OC" (Occupied Clean), "OD" (Occupied Dirty), and "OOO" (Out of Order, room under maintenance). For self-check-in processes, the focus is on rooms with a "VC" status, but the "VD" status and its position in the cleaning queue are equally important.

[0065] Obtain the "Cleaning Task Queue Sequence." This is not just a list of room numbers, but a sequence of tasks with accompanying metadata. Each task entry includes at least: room number, task type (e.g., "Cleaning," "Room Check"), task creation time, current executor (or group), estimated time, and current status ("Pending Assignment," "In Progress," "Completed"). This queue sequence directly determines the expected timeline for a room to transition from "VD" to "VC." By analyzing rooms in the "VD" state in the queue and their preceding and following tasks, combined with historical average cleaning times, the "Cleaning Ready Time Cost" for a specific room can be estimated.

[0066] Obtain "Floor Reservation Distribution Data for Guests Expected to Arrive Within a Future Scheduled Time". This is typically achieved by querying all "confirmed" and "not yet checked in" reservation records in the hotel management system for a specific future time period (e.g., 2 hours or 4 hours), and aggregating the data by the floor where the reserved rooms are located. This data is represented as a mapping table or vector, for example: {"5th floor": 8, "12th floor": 3, "20th floor": 12}, indicating the expected number of new guests checking in on each floor within the target future time period. This data forms the basis for predicting the future congestion level of each floor.

[0067] When the individual passenger resource optimization unit or the pressure feedback unit is triggered, the operation status interface unit will simultaneously initiate the retrieval of the above three types of data. To ensure data real-time performance, this unit is configured to actively query the rapidly changing "cleaning status identifier" and "cleaning task queue order" at a high frequency (e.g., every 30 seconds or every minute); while for the relatively slowly changing "future booking distribution data," it will be updated at a lower frequency (e.g., every 5 minutes). All the raw data obtained will undergo preliminary cleaning and format standardization (e.g., unifying status codes from different sources into internal enumeration values), and then be cached in memory or a local database for rapid reading by other units within the dynamic resource collaborative allocation engine.

[0068] For example, when a walk-in guest checks in at 2 p.m., the instantaneous data snapshot obtained by this unit from the hotel management system is as follows: room 501 is in the status of "VC", room 502 is in the status of "VD" and is the second in the cleaning queue (the first in the queue is cleaning room 503); at the same time, the reservation distribution data for the next 2 hours (14:00-16:00) shows that 5 new reservations will arrive on the 5th floor and 10 new reservations will arrive on the 20th floor.

[0069] Understandably, this real-time, dynamic data forms the basis for all subsequent intelligent decisions—such as calculating the cleaning wait time for room 502 and assessing the future congestion caused by assigning guests to the 20th floor. Through this unit, self-service terminals can break free from their reliance on static room status data and truly "sense" the pulse of hotel operations, thereby achieving coordinated and optimized allocation of resources.

[0070] The individual customer resource optimization unit, connected to the operation status interface unit, is used to perform the following steps when serving individual customer users:

[0071] Obtain the room preference vector of the current individual customer;

[0072] Based on the data from the operation status interface unit, the allocation cost of each available room that meets the room type of individual customers is calculated. The allocation cost is obtained by weighted summation of the first and second terms. The first term represents the matching degree between the room preference vector and the room's static feature vector, and the second term represents the cleaning readiness time cost calculated based on the room's cleaning status identifier and its position in the cleaning task queue.

[0073] Output the room with the lowest allocation cost as the first recommendation to the touch screen.

[0074] In a further implementation, when the individual guest resource optimization unit calculates the allocation cost, it also includes a third term, which represents the future congestion cost. The future congestion cost is calculated in the following way: based on the floor reservation distribution data of guests to be arrived within a future scheduled time, it predicts the potential increase in traffic density on the floor where the current room is allocated due to the arrival of new guests within the target time window; the allocation cost is the weighted sum of the first, second and third terms.

[0075] Among them, the allocation cost of the individual passenger resource optimization unit is... The calculation formula is:

[0076]

[0077] in, For the user's room preference vector, This represents the static feature vector of the room. To calculate the vector norm of the first term; The estimated waiting time, calculated based on the cleaning status identifier "to be cleaned" and its index position in the cleaning task queue, is used to characterize the second item; The third item is the estimated number of new occupants on that floor within a specific future time period, calculated based on floor reservation distribution data. These are preset non-negative weighting coefficients.

[0078] Specifically, the Individual Guest Resource Optimization Unit is the main decision-making module in the dynamic resource collaborative allocation engine that handles single-guest check-in requests. Its connection to the Operational Status Interface Unit ensures that all its decisions are based on the hotel's latest real-time operational status, rather than a static room type database. The function of this unit is to transform the multi-objective optimization problem (satisfying user preferences, minimizing waiting time, and avoiding future congestion) into a computable mathematical model, and based on this, output the overall optimal room recommendation.

[0079] The specific steps performed by this unit are as follows:

[0080] The first step is to obtain the room preference vector of the current individual guest. This step aims to transform the user's unstructured, natural language preference choices (such as "I want a high-floor, quiet, non-smoking room") into quantifiable input that can be processed by computer algorithms. The system provides the user with a preference selection interface via a touch screen, with options typically including floor (high / medium / low), orientation (south / north, etc.), whether it is non-smoking, and whether it is close to the elevator. After the user selects these options, the system encodes them into a multi-dimensional room preference vector. For example, if the system defines six dimensions: [high floor, middle floor, low floor, no smoking, south-facing, near elevator], and the user selects "high floor" and "no smoking," then the generated preference vector will... In one alternative embodiment, the vector is [1, 0, 0, 1, 0, 0]. This vector serves as the benchmark for subsequent calculations of user satisfaction.

[0081] Step 2: Calculate the allocation cost for each available room that matches the room type. For each room that matches the guest's booking basic room type (such as a king room) and is in a "available" status (including "VC" and some "VD" rooms with quick cleaning), perform the following calculations:

[0082] Constructing room static feature vectors Each room has a corresponding static feature vector. Its dimensions and preference vector Consistent, the numerical value represents the room's attributes in that dimension (can be represented using binary or continuous values). For example, room A is located on the 20th floor, is non-smoking, faces north, and is far from the elevator; its vector... [1, 0, 0, 1, 0, 0].

[0083] The first term, preference matching cost, is calculated by calculating a vector. and The difference between them is used to quantify the "vector norm". This is a mathematical method for measuring this difference. A common implementation is to calculate the Manhattan distance (the sum of the absolute differences across all dimensions) or Euclidean distance. The larger the difference, the less the room matches the user's preferences, and the higher the cost value. For example, if a user wants a "high floor" but the room is on a low floor, the difference in this dimension is 1, contributing to the cost.

[0084] Calculate the second item, cleaning-ready time cost. Its calculations heavily rely on real-time data provided by the operational status interface unit.

[0085] If the room cleanliness status is marked as "VC", then .

[0086] If the status is "VD", the calculation needs to be based on its position in the cleaning task queue. For example, if the room is in the queue... The location is estimated based on historical data, and the average cleaning time per room is also estimated. minutes, then its estimated waiting time A more refined implementation would also consider the time elapsed while rooms are being cleaned in the queue. This cost directly quantifies the experience metric of "how long it takes to check into a room."

[0087] Calculate the third term (optimization term), future congestion costs. This demonstrates the forward-looking nature of the decision-making process. This unit uses the operations status interface unit to obtain floor reservation distribution data for guests expected to arrive within the next scheduled timeframe. This value is directly taken as the floor where the target room is located, and the estimated number of new guests expected to check in during a specific future time period (such as the next 2 hours). The principle is that if a floor is about to receive a large number of new guests, assigning a guest to that floor now will exacerbate congestion in the elevators, corridors, and other public areas of that floor in the future, affecting guest experience and logistical efficiency. Therefore, The higher the value, the higher the cost.

[0088] The total allocation cost is obtained by weighted summation. : Apply the above three costs according to the formula Perform synthesis. Weighting coefficients. It is a preset, non-negative harmonic parameter, the value of which is set by hotel managers according to operational strategies. For example, when emphasizing customer satisfaction, a larger value can be set. When there is a peak occupancy period and you want to speed up room turnover, you can set a larger [configuration value]. When it is desired to distribute passenger flow evenly and avoid overcrowding on specific floors, a larger [floor height] can be set. These coefficients can also be dynamically optimized by the adaptive weighting module based on historical data.

[0089] As an optional implementation, when the individual guest resource optimization unit outputs the first recommendation result, it simultaneously generates a brief decision basis text for each recommended room; the multilingual interaction module that controls the mainboard converts the brief decision basis text into the language currently set by the user, and displays it in the form of an interactive control next to the recommendation item of the corresponding room on the touch screen.

[0090] Specifically, after completing the cost calculation and determining the first recommendation result, this functional module did not terminate its processing flow. Instead, it further executed a step to improve system transparency and user interaction experience. This step transforms the "black box" process of machine decision-making into a "white box" explanation that users can perceive and understand, thereby enhancing users' trust and acceptance of the recommendation results.

[0091] This involves simultaneously generating a brief decision-making basis text for each recommended room. The generation rules for this text typically combine conditional statements with template filling. For example:

[0092] Regarding preference matching degree: if If the value is extremely small (close to 0), it generates "Completely meets your [X, Y] requirements"; if there are some differences, it generates "Basically meets your requirements (meets your [X] requirements, [Y] is slightly different)". Placeholders such as [X] and [Y] are replaced with specific preference dimensions, such as "high floor" or "no smoke".

[0093] Regarding the time cost of cleaning readiness: If If so, it will generate "Cleaned and ready for immediate move-in"; if According to The value is converted into a specific time description, such as "cleaning and check-in are expected to be completed in about 30 minutes".

[0094] Regarding future congestion costs (if implemented): based on Based on the value's range (e.g., low, medium, high), different prompts are generated, such as "The passenger flow on this floor is stable" or "This floor is slightly busier in the evening."

[0095] Finally, combine the above sentences logically into a complete sentence or two. For example: "Reasons for recommendation: It perfectly meets your requirements for a non-smoking environment and a high floor, it has been cleaned and is ready for immediate check-in, and the floor has stable evening occupancy."

[0096] Furthermore, the generated initial (e.g., Chinese) brief decision-making basis text is not directly displayed, but is passed to the multilingual interaction module controlling the motherboard for processing. The multilingual interaction module is responsible for the language localization of the entire terminal user interface. Based on the language mode set by the user in the first step of the process (e.g., "English," "Japanese"), it calls upon the built-in multilingual dictionary and translation engine (or pre-translated templates) to accurately convert (translate) the original text into the target language. This ensures that guests of any nationality can understand the recommendation reasons in their familiar language. The converted text does not occupy screen space flatly, but is integrated into the user interface in an elegant, non-intrusive way. On the touchscreen display, the system dynamically generates an interactive control next to each recommended room entry (usually containing room number, image, price, etc.). This control is typically designed as an information icon "ⓘ", a "Why?" button, or an expandable arrow. The text itself is associated and hidden behind this control. When a user has a question or is interested in a recommended room, they can trigger a view of the detailed reasons by clicking or touching the interactive control next to it. At this point, the system will display the corresponding, language-translated, concise decision-making basis text in a nearby location (such as below, to the side, or in a pop-up bubble) with a smooth animation effect. After the user has finished reading, they can click to close or automatically collapse the interface, restoring its clean and simple appearance.

[0097] Understandably, this design significantly enhances the system's credibility and users' sense of control. Users are no longer passive recipients but can understand and accept the rationality of their choices. Secondly, it serves as a gentle educational and guiding tool. For example, when the system recommends a room that doesn't perfectly match a user's original preferences due to operational efficiency reasons (such as a slightly lower floor but faster cleaning), explanatory text can mitigate potential user dissatisfaction and help them understand the value of "fast check-in." Finally, the unified multilingual interaction module ensures a consistent experience for international travelers, eliminating language barriers and representing a significant user experience optimization in the international hotel setting.

[0098] As an optional implementation, the operation status interface unit is also used to obtain real-time operation status data of the hotel elevators; when calculating allocation costs or future congestion costs, the individual guest resource optimization unit also incorporates the distance between the floor where the target room is located and the elevator lobby, as well as the current elevator operation load status, as additional factors into the calculation model.

[0099] Specifically, the operation status interface unit, based on the existing data interface, further establishes a data connection with the hotel's elevator monitoring system or building automation system. This unit acquires real-time elevator operation status data through a network communication module, using either polling or event subscription. This data is structured and dynamically changing, and typically includes, but is not limited to:

[0100] Real-time status of each elevator: such as "going up", "going down", "stopped", "idle", "under maintenance"; car position information: the floor the elevator is currently on; call signal queue: requests for elevators on each floor and their waiting time;

[0101] Car load information: Real-time load or passenger estimate obtained through weight sensors (e.g., "70% full load"); Operating mode: Whether it is in peak mode, energy-saving mode, or fire emergency landing mode, etc.

[0102] After obtaining the raw data, the individual passenger resource optimization unit adds a quantitative assessment of the following two elevator-related factors to its cost calculation model: the distance cost between the floor where the target room is located and the elevator lobby, and the cost of the current elevator operating load status.

[0103] The cost of the distance between the target room's floor and the elevator lobby focuses on planar movement. The system either has built-in floor plan topology information or retrieves it from the hotel database. For each room, the walking distance to the nearest elevator lobby can be pre-calculated or calculated in real-time (or expressed as a score representing "remoteness," such as "room at the end of the corridor" scoring high, and "room near the elevator" scoring low). This distance... Transformed into a cost-added value The greater the distance, the higher the cost, reflecting the inconvenience of passengers carrying luggage on foot. The current elevator operating load cost reflects the real-time pressure on vertical transportation. The acquired elevator status data is aggregated and analyzed to calculate a comprehensive index reflecting the current elevator operating load status. The calculation of this indicator takes into account average waiting time, elevator car crowding, and service pressure on specific floors.

[0104] Furthermore, these elevator-related cost factors are flexibly integrated into the existing cost allocation calculation framework. The formula is expanded to:

[0105]

[0106] in, For the user's room preference vector, This represents the static feature vector of the room. To calculate the vector norm of the first term; The estimated waiting time, calculated based on the cleaning status identifier "to be cleaned" and its index position in the cleaning task queue, is used to characterize the second item; The third item is the estimated number of new occupants on that floor within a specific future time period, calculated based on floor reservation distribution data. The distance between the floor where the target room is located and the elevator lobby; For elevator load; These are preset non-negative weighting coefficients.

[0107] Understandably, this alternative implementation extends resource optimization from the "room" itself to the "movement path" of guests after check-in, achieving a more comprehensive experience optimization. It effectively reduces guests' anxiety while waiting for elevators and fatigue from long walks during peak hours, especially beneficial for guests with heavy luggage or elderly guests. From an operational perspective, it also helps guide the distribution of guest rooms across floors and areas, indirectly balancing elevator usage and improving the overall efficiency of hotel facilities and guest satisfaction. This marks a deeper evolution of the system from "static resource allocation" to "dynamic circulation optimization."

[0108] The group session coordination unit is used to perform the following steps when serving group users:

[0109] Generate and display group session identification codes at the self-service check-in terminal of the first hotel;

[0110] In response to scanning of the group session identifier by one or more other hotel self-service check-in terminals, the user identity verified on one or more other hotel self-service check-in terminals is associated with the same group check-in transaction.

[0111] Under the condition of satisfying the total room type constraint of the group, the graph partitioning calculation is performed based on the relationship data between group members and the room location information. With the goal of maximizing the association strength within the group of members who are assigned to adjacent or same-floor rooms, the group room allocation scheme is output to the first hotel self-service check-in terminal or a hotel self-service check-in terminal specified in the group check-in transaction.

[0112] Within the group conversation collaboration unit, the graph division calculation is performed based on the relationship data between group members and room location information, specifically including:

[0113] Using group members as vertices, assign edge weights to every two member vertices based on relationship data;

[0114] Obtain the room topology of the hotel floors and define adjacent or same-floor rooms as meeting preset location conditions;

[0115] Solve a constrained optimization problem: Given the required number of rooms of various types for the group, assign a specific room to each member such that the sum of the edge weights of all members assigned to room pairs that meet the preset location conditions is maximized.

[0116] Specifically, the Group Session Collaboration Unit is an intelligent collaboration module within the dynamic resource collaborative allocation engine, dedicated to handling group (multi-person) check-in transactions. It logically aggregates physically dispersed self-service terminals through a "digital session" mechanism and utilizes graph theory optimization algorithms to address the efficiency and fairness challenges of traditional manual room allocation, aiming to achieve a disruptive restructuring of the group check-in process.

[0117] The specific steps include: Once the group leader or first member selects the "Group Check-in" mode and passes authentication at any self-service terminal (defined in this transaction as the First Hotel Self-Service Check-in Terminal), this unit is activated. The unit generates a group session identifier (usually encoded as a QR code) containing encrypted session information. This identifier contains key information such as the group booking number, a unique session ID, a timestamp, and a token used for security verification. Upon generation, the code is immediately displayed on the First Hotel Self-Service Check-in Terminal's touchscreen.

[0118] Other group members can disperse to any available self-service check-in terminal in the hotel lobby and use its camera to scan the identification code displayed on the screen of the first self-service check-in terminal. The scanning action is a "join request." After scanning, the terminal immediately sends a request to the server via the internal network to verify the validity of the identification code. Upon successful verification, the terminal switches from an independent state to a state of "joined the XXX group transaction managed by the group session collaboration unit." Its screen interface updates accordingly, directly entering the context of the group check-in transaction, typically skipping repetitive steps such as language selection and booking inquiries, and directly starting the identity verification process. After a member completes verification such as document reading and facial recognition on the terminal, their successfully verified user identity information (such as name and ID number) is not processed independently but is automatically tagged with the group session and associated with the same group check-in transaction. The server maintains a "session-member" mapping table, which is updated in real time. This means that the verification operations of different members on different terminals are highly parallel, and their data is aggregated in the background through a shared session ID. The traditional model requires N verifications to be completed linearly, but under this model, the theoretical time required is close to that of a single verification, greatly reducing the overall time spent at the front desk of the group.

[0119] Once the system detects that all booking members have completed verification (or the tour leader manually triggers it), it enters the intelligent room allocation phase. This unit transforms the complex social room allocation requirements into a computable graph partitioning optimization problem, specifically including the following sub-steps: Each group member is abstracted as a graph node (vertex). Based on relationship data, an edge is established between every two member vertices, and a weight is assigned to this edge. The relationship data can come from the travel companion relationship field in the booking information; or be confirmed by the tour leader or members through a simplified interface during registration (e.g., selecting "family" or "colleague"). Preset weight values ​​are used, for example: "immediate family member" weight is 10, "colleague / friend" weight is 5, and "no special relationship" weight is 1 or lower. The weight values ​​quantify the strength of members' desire for adjacent rooms. Further, room location information, i.e., room topology, is obtained from the hotel database. Preset location conditions are used to define what constitutes "adjacent or on the same floor."

[0120] Finally, among all room allocation schemes, a scheme is sought that maximizes the sum of edge weights for all members assigned to room pairs that meet preset location conditions. The scheme must satisfy the group's total room type constraints, meaning the total number of allocated rooms and the number of each room type (e.g., king / twin) must be exactly the same as the group booking. Constraints such as floor preference and non-smoking room requirements are also included. The final scheme is sent to the first hotel self-service check-in terminal (leader's terminal) or a hotel self-service check-in terminal designated by the leader / system (such as a public display terminal already in the session) for final presentation and confirmation. After the leader confirms, the system can complete payment authorization for all rooms with one click and instruct members to print their individual room cards at their respective verified terminals.

[0121] Understandably, the group conversation collaboration unit not only achieves extreme processing efficiency (parallel verification), but also provides a fair and reasonable room allocation scheme through scientific algorithms, fundamentally solving the pain points of group check-in and improving the overall satisfaction of group guests and the professional image of the hotel.

[0122] As an optional implementation, the dynamic resource collaborative allocation engine also includes a pressure feedback unit, which is used for:

[0123] Based on the current check-in waiting queue length and historical average processing speed obtained from the operation status interface unit, calculate the real-time estimated waiting time;

[0124] When the real-time estimated waiting time exceeds the first threshold, an efficiency guidance instruction is generated and sent to the control logic of the touch screen, so that when the touch screen displays the first recommendation result, it adds a visual mark to the room whose cleaning ready time cost is lower than the second threshold.

[0125] Specifically, the pressure feedback unit receives the current check-in waiting queue length and historical average processing speed from the operation status interface unit. The real-time estimated waiting time is calculated using the formula: Real-time estimated waiting time = Current check-in waiting queue length ÷ Historical average processing speed. When the real-time estimated waiting time exceeds a first threshold, the unit determines that it is currently in a peak operational pressure period. At this point, it no longer passively displays the first recommended result calculated by the individual guest resource optimization unit (which is already the optimal solution), but actively generates an efficiency guidance instruction. This instruction contains explicit UI modification commands: when displaying the recommended room list to the user, add a specific, prominent visual marker to all rooms in the list that meet the condition of "cleaning readiness time cost is lower than the second threshold".

[0126] The first threshold is a time threshold, set by the hotel management based on its service standards (such as "promising an average waiting time of no more than 5 minutes"). The second threshold is a cost per unit of time for cleaning to be ready. The threshold (usually set to a very small value close to 0, such as...) This is used to filter out rooms that are "already cleaned" or "will be cleaned soon".

[0127] As an optional implementation, the self-service check-in terminal also includes a near-field communication (NFC) module. The group session collaboration unit broadcasts a beacon signal containing a group session identifier code through the NFC module. The hotel self-service check-in terminal is configured such that, when idle, if it receives a beacon signal broadcast by another terminal through the NFC module, it activates and displays a prompt message on the touch screen, guiding the user to scan the terminal screen to join the corresponding group session.

[0128] Specifically, when the tour leader successfully initiates a group check-in transaction at the self-service check-in terminal in the first hotel, the group session collaboration unit generates a visible group session identifier (QR code). Simultaneously, this unit drives the terminal's short-range wireless communication module into broadcast mode. The short-range wireless communication module continuously broadcasts a special beacon signal. Other terminals in the hotel that are not in service operate in scanning and listening mode by default. Once these hotel self-service check-in terminals identify a valid group session beacon and are idle, the terminal immediately interrupts its current standby display and activates a specific interactive interface on the touchscreen.

[0129] Because the terminal displaying the prompts does not directly show the QR code on the tour leader's terminal (due to the physical distance), its guiding function is to direct users to operate this terminal. After seeing the prompts, users walk to this terminal. At this point, users need to use the terminal's camera, or more commonly, their own smartphone, to scan the large, clear group session identification code (QR code) on the screen of the First Hotel self-service check-in terminal (the terminal where the tour leader is located). The scanning action is completed on this terminal, thus binding this terminal to group affairs, and then users can complete their identity verification on this terminal.

[0130] Understandably, through the above methods, the terminal is no longer an isolated information island, but an intelligent node capable of sensing the status of its surrounding companions and proactively changing its behavior (interface) to adapt to environmental needs. This eliminates the confusion among group members about which machine to use and the inefficiency of walking back and forth, making the guidance process extremely natural and smooth, significantly enhancing the sense of technology and convenience. Even if the leader's terminal (the primary terminal) screen is obstructed or is far away, other terminals can still serve as effective guidance points, expanding the effective service area of ​​the collaborative system, which is particularly suitable for large or complex hotel lobbies. Utilizing mature low-power wireless technology, a significant enhancement in experience is achieved with almost no additional energy consumption or cost.

[0131] As an optional implementation, the control motherboard is also equipped with an adaptive weighting module, which is connected to the individual passenger resource optimization unit and is used for periodic execution:

[0132] Collect operational efficiency indicators corresponding to different combinations of weight coefficients within historical time periods. These operational efficiency indicators include average guest waiting time and average cleaning staff moving distance. Based on historical data, use regression analysis or optimization algorithms to update the recommended weight coefficient combinations for the next period and provide the updated weight coefficients to the individual guest resource optimization unit.

[0133] Specifically, the adaptive weighting module automatically performs data collection tasks at the end of each running cycle. It extracts two key datasets from the database associated with the system logs and the operational status interface unit: historical weight coefficient combinations and their corresponding operational efficiency indicators. By precisely mapping and storing each set of "weight coefficient combinations" to its generated "operational efficiency indicators," the module constructs a historical experience database for analyzing the causal relationship between decision parameters and the final results.

[0134] Based on accumulated historical data, the adaptive weighting module uses a data-driven approach to derive better weighting coefficients. Its main methods include regression analysis and optimization algorithms. After calculating the better weighting coefficients, the adaptive weighting module updates the recommended weighting coefficient combination for the next cycle and provides the updated weighting coefficients to the individual customer resource optimization unit.

[0135] As an optional implementation, the network communication module is connected to the hotel's door lock management system. The control board is also configured to: after the user completes the transaction confirmation through the payment module, regardless of whether it is the first recommendation result from the individual guest resource optimization unit or the group room allocation scheme from the group session collaboration unit, bind the final assigned room number with the corresponding user's identity, generate an authorization instruction, and send it to the door lock management system through the network communication module so that the door lock management system can authorize the corresponding room door lock.

[0136] Specifically, the hotel's door lock management system is an independent backend system responsible for managing electronic access permissions for all guest room door locks. The network communication module establishes a standardized data interface connection with this system via the hotel's internal network (typically a wired Ethernet or secure Wi-Fi network). When a user completes the final step of the check-in process—payment confirmation—the control board is configured to immediately trigger the authorization process. This triggering is deterministic and applies to two check-in scenarios: individual guest scenarios, where the user confirms a room in the first recommended result output by the individual guest resource optimization unit and completes payment; and group scenarios, where the tour leader confirms the group room allocation plan output by the group session collaboration unit and completes group payment. After triggering, the control board binds the final assigned room number with the corresponding user's identity identifier. Based on this binding relationship, the control board generates an authorization command. After the command is generated, the control board sends the command to the door lock management system via the network communication module in real-time or near real-time. Upon receiving the authorization command, the door lock management system verifies and processes it.

[0137] Verify command: Check the command format, permissions, and validity.

[0138] Update door lock permission list: Add new authorization information (room number - identity identifier - time period) to the authorization list of the electronic door lock for the target room. For networked electronic locks, this update can be directly sent to the door lock via the network; for offline door locks, the card issuing system's database is updated, awaiting writing when the room card is made.

[0139] Feedback result (optional): Return a success or failure response to the self-service terminal.

[0140] Once the authorization list is updated, the corresponding physical door lock enters the "authorized" state. Afterwards, when a physical room card (generated by the printing module) or virtual room card (issued via a mobile application) created based on this authorization is swiped on the door lock or approaches it, the door lock verifies that the identity information on the card or in the mobile phone matches the authorization list and then executes the unlocking operation.

[0141] This invention, by employing the aforementioned technical solution, offers the following beneficial effects: Through a dynamic resource collaborative allocation engine, the decisions of self-service terminals are deeply integrated with the hotel's real-time operational status, thereby improving the overall operational efficiency and resource utilization of the hotel. By using a "group session identifier" mechanism combined with a graph partitioning-based intelligent room allocation algorithm, not only is the overall check-in time for groups shortened, but the satisfaction and convenience of group guests are also significantly improved. Interaction strategies are dynamically adjusted based on real-time pressure, and algorithm parameters are continuously optimized through historical data feedback, enabling the system to maintain optimal decision-making performance over the long term and adapt to constantly changing operating environments. By generating concise, data-driven decision-making criteria for each room recommendation and displaying them in the language selected by the guest, guest trust is increased. Simultaneously, it provides hotel management with a data-driven view of resource allocation and adjustment methods.

[0142] The following will illustrate a hotel self-check-in terminal based on multilingual interaction and intelligent room selection functions according to specific embodiments of the present invention.

[0143] This embodiment provides a hotel self-check-in terminal based on multilingual interaction and intelligent room selection functions, mainly comprising the following hardware components: a high-resolution multi-touch display screen, an identity recognition module integrating an ID card reader, passport scanner, and high-definition camera, a payment module supporting bank cards, mobile payments, and pre-authorization (such as a PIN pad and QR code scanner), a thermal / thermal transfer printing module for printing room cards and invoices, a network communication module responsible for wired / wireless network connections, and a control motherboard integrating a CPU, memory, and storage. All modules are connected to the control motherboard via an internal bus.

[0144] The control motherboard carries the operating system and the software system of this invention. A key component of this software system is the "dynamic resource collaborative allocation engine." Logically, this engine includes an operation status interface unit, a personal resource optimization unit, a group session collaboration unit, a pressure feedback unit, and an adaptive weighting module.

[0145] Implementation Scenario 1: Smart Room Selection for Individual Customers

[0146] When an individual customer arrives at the terminal, the check-in process is as follows:

[0147] First, the user selects their language (e.g., English) on the touchscreen. Then, they complete ID card and facial recognition verification as prompted. The system synchronizes their online booking information (e.g., a king-size bed booked through Booking.com) via the OTA interface. Next, the individual guest resource optimization unit is activated. First, it retrieves real-time data from the hotel's PMS through the operational status interface unit: the current cleaning status of all king-size rooms (e.g., room A "cleaned," room B "awaiting cleaning" and 3rd in the cleaning queue), and the number of guests who have booked rooms on each floor within the next 2 hours. A room preference selection window appears on the user interface (floor, orientation, non-smoking, etc.), and the user selects "high floor" and "non-smoking room." The system generates a preference vector P.

[0148] For each available room that meets the criteria of "double room", "high floor", and "non-smoking", the individual guest resource optimization unit calculates its allocation cost. Taking room B as an example:

[0149] Calculate the difference between the room characteristics (e.g., located on the 20th floor, facing north) and the user's preference (high floor, no orientation specified) to obtain a scalar value.

[0150] Based on the fact that room B is in the "awaiting cleaning" status and is the 3rd in the queue, and combined with the historical average cleaning time for each room (e.g., 15 minutes), its cleaning readiness time is estimated to be about 45 minutes.

[0151] According to the query results for the next two hours, there are three new arrival bookings for the 20th floor, indicating significant congestion pressure.

[0152] Substitute into the formula (Assuming) , , The adaptive weighting module has set the values ​​to 0.5, 0.3, and 0.2.

[0153] Also calculate the cost of other optional rooms (such as room A, which is "cleaned"). Due to the cleaning readiness time of room A. And future congestion pressure Unlike other implementations, in one alternative embodiment, its Value lower than .

[0154] The system displays the 2-3 lowest-cost rooms (e.g., room A and another slightly lower-cost room) as the "first recommendation" on the touchscreen. Simultaneously, if the stress feedback unit detects a queue in the lobby, it can add a "Quick Check-in" icon next to room A's display.

[0155] After the user selects room A and completes payment, the system controls the printing module to produce a room key and simultaneously sends the user information and room A authorization instructions to the hotel door lock management system.

[0156] Implementation Scenario 2: Group Collaborative Check-in and Intelligent Room Allocation

[0157] A group of seven people, consisting of a family of three (parents and child) and two pairs of colleagues (four people in total), checked in.

[0158] Step 1. The tour leader (father) selects "Group Check-in" on the first terminal, verifies his / her identity, and links the group booking number. The terminal generates a QR code (group meeting session identifier code) which is displayed on the screen.

[0159] Step 2. At the same time, the group session coordination unit broadcasts this session beacon via the terminal's Bluetooth module.

[0160] Step 3. On the second available terminal next to them, the mother and child see a screen prompt: "A group session has been detected nearby. Scan the code to join quickly." They scan the QR code on the screen, and the second terminal immediately enters the group's session context. The mother and child then complete facial verification in turn.

[0161] Step 4. The two colleagues repeat Step 3 on the third and fourth terminals respectively to complete the identity verification. Thus, the seven members completed identity verification in parallel across four terminals within a few minutes, with all data linked to the same transaction.

[0162] Step 5. On the tour leader's (father's) terminal, the system prompts "Members have been verified, room assignments will begin." The system knows that 3 twin rooms and 1 double room (family room) have been booked.

[0163] Step 6. The group conversation collaboration unit activates the intelligent room allocation algorithm. It acquires member relationship data ("Family: Father, Mother, Child"; "Colleague Group 1: Colleague A, B"; "Colleague Group 2: Colleague C, D"), and assigns high relationship weights to family members and medium weights to colleagues.

[0164] Step 7. The algorithm obtains the hotel floor plan and knows that there are consecutive twin and king rooms available on the 5th floor.

[0165] Step 8. The algorithm performs graph partitioning optimization to solve the problem, aiming to assign closely related members to adjacent rooms. The solution is as follows: the family is assigned to an adjacent double room and a twin room (connected) on the 5th floor, and the two colleagues are assigned to two other adjacent twin rooms on the 5th floor.

[0166] Step 9. The room allocation plan is presented on the tour leader's terminal in the form of a highlighted floor plan. After the tour leader confirms, the system completes the pre-authorization of payment for all rooms in one go and instructs each member to print their own room card on their respective terminal, or the tour leader can print and distribute them collectively.

[0167] The self-check-in terminal of this invention interacts with the hotel management system (PMS), online booking platform (OTA), door lock management system, and even elevator monitoring system via the hotel's internal network. The operational status interface unit is used to aggregate this data, ensuring that the dynamic resource allocation engine's decisions are based on comprehensive and real-time information.

[0168] Through the aforementioned technical solution, this application achieves dynamic and intelligent collaborative management of hotel room resources and check-in processes. The system, through a dynamic resource collaborative allocation engine, deeply integrates discrete guest preferences, real-time cleaning task status, and future passenger flow pressure predictions to construct a multi-objective optimization decision model. This ensures that each individual guest's room selection recommendation effectively smooths the overall operational load of the hotel while meeting individual needs. In group check-in scenarios, the system, through a unique conversational collaboration mechanism and graph partitioning algorithm, reconstructs the traditional sequential check-in process into a parallel and intelligent collaborative workflow. This not only significantly reduces overall check-in time but also scientifically meets the interconnected needs of group members. Ultimately, the entire solution transforms the self-service terminal from an isolated service node into an intelligent hub driving simultaneous improvements in the efficiency of both front-of-house services and back-office operations.

[0169] The detailed description of the above specific embodiments fully illustrates the feasibility, preferred implementation, and technical effects achieved by the technical solution of the present invention. Those skilled in the art can make several modifications and substitutions based on the above description without departing from the principles and spirit of the present invention, and these modifications and substitutions should also be considered within the scope of protection of the present invention.

[0170] The present invention and its embodiments have been described above. This description is not restrictive, and the accompanying drawings are only one embodiment of the present invention. The actual method is not limited to this. In conclusion, if those skilled in the art are inspired by this description and design similar methods and embodiments without departing from the spirit of the present invention, they should all fall within the protection scope of the present invention.

Claims

1. A hotel self-check-in terminal based on multilingual interaction and intelligent room selection functions, comprising a touch screen display, an identity recognition module, a payment module, a printing module, a network communication module, and a control motherboard, characterized in that, The identity recognition module is used to obtain and verify the user's identity information in order to generate an identity verification credential. The control motherboard is configured to run a dynamic resource collaborative allocation engine. The dynamic resource collaborative allocation engine includes: an operation status interface unit, used to acquire data from the hotel management system in real time through the network communication module. This data includes the cleaning status identifier of each room, the cleaning task queue order, and the floor reservation distribution data of guests to arrive within a future scheduled time. A personal guest resource optimization unit, connected to the operation status interface unit, is used to perform the following steps when serving personal guests: acquiring the room preference vector of the current personal guest; calculating the allocation cost of each available room matching the personal guest's room type based on the data from the operation status interface unit, where the allocation cost is at least a weighted sum of a first term and a second term. The first term represents the matching degree between the room preference vector and the room's static feature vector, and the second term represents the cleaning readiness time cost calculated based on the room's cleaning status identifier and its position in the cleaning task queue order; and outputting at least one room with the lowest allocation cost. The first recommended room is displayed on the touch screen. The group session collaboration unit, when serving group users, performs the following steps: generating and displaying a group session identifier code on the first hotel self-service check-in terminal; in response to scanning the group session identifier code by one or more other hotel self-service check-in terminals, associating the user identity verified on the one or more other hotel self-service check-in terminals with the same group check-in transaction; under the condition of satisfying the total room type constraint of the group, performing graph partitioning calculation based on the relationship data between group members and room location information, with the goal of maximizing the association strength within the member group assigned to adjacent or same-floor rooms, and outputting the group room allocation scheme to the first hotel self-service check-in terminal or a hotel self-service check-in terminal specified in the group check-in transaction; wherein, the dynamic resource collaborative allocation engine is also configured to trigger the corresponding service process of the individual guest resource optimization unit or the group session collaboration unit based on the identity verification credential generated by the identity recognition module.

2. The hotel self-check-in terminal according to claim 1, characterized in that, When the individual guest resource optimization unit calculates the allocation cost, it also includes a third item, which represents the future congestion cost. The future congestion cost is calculated in the following way: based on the floor reservation distribution data of guests to be arrived within the future scheduled time, it predicts the potential increase in traffic density on the floor where the current room is allocated due to the arrival of new guests within the target time window; the allocation cost is the weighted sum of the first, second and third items.

3. The hotel self-check-in terminal according to claim 2, characterized in that, The allocation cost performed by the individual passenger resource optimization unit The calculation formula is: in, For the user's room preference vector, This represents the static feature vector of the room. To calculate the vector norm of the first term; The estimated waiting time, calculated based on the cleaning status identifier "to be cleaned" and its index position in the cleaning task queue, is used to characterize the second item; The third item is defined as the estimated number of new occupants on that floor within a specific future time period, calculated based on the floor reservation distribution data. These are preset non-negative weighting coefficients.

4. The hotel self-check-in terminal according to claim 1, characterized in that, In the group conversation collaboration unit, the step of performing graph partitioning calculation based on the relationship data and room location information between group members specifically includes: using group members as vertices, assigning edge weights to every two member vertices based on the relationship data; obtaining the room topology relationship of the hotel floors, defining adjacent or same-floor rooms as satisfying preset location conditions; and solving a constrained optimization problem: under the premise of satisfying the required number of various room types for the group, assigning a specific room to each member, such that the sum of the edge weights of all members assigned to room pairs that satisfy the preset location conditions is maximized.

5. The hotel self-check-in terminal according to claim 1, characterized in that, The dynamic resource collaborative allocation engine also includes a pressure feedback unit, which is used to: calculate the real-time estimated waiting time based on the current check-in waiting queue length and historical average processing speed obtained by the operation status interface unit; when the real-time estimated waiting time exceeds a first threshold, generate an efficiency guidance instruction to the control logic of the touch screen, so that when the touch screen displays the first recommendation result, it adds a visual mark to the room whose cleaning ready time cost is lower than a second threshold.

6. The hotel self-check-in terminal according to claim 1, characterized in that, The hotel self-check-in terminal also includes a near-field communication module. The group session collaboration unit broadcasts a beacon signal containing the group session identifier code through the near-field communication module. The hotel self-check-in terminal is configured such that when it is idle, if it receives the beacon signal broadcast by another terminal through the near-field communication module, it activates and displays a prompt message on the touch screen to guide the user to scan the terminal screen to join the corresponding group session.

7. The hotel self-check-in terminal according to claim 1, characterized in that, The control motherboard is also equipped with an adaptive weight adjustment module, which is connected to the individual customer resource optimization unit and is used to periodically perform the following: collect the corresponding operational efficiency indicators when using different combinations of weight coefficients within a historical time period, including average customer waiting time and average cleaning staff moving distance; based on the historical data, update the recommended weight coefficient combination for the next period using regression analysis or optimization algorithms, and provide the updated weight coefficients to the individual customer resource optimization unit.

8. The hotel self-check-in terminal according to claim 1, characterized in that, The network communication module is connected to the hotel's door lock management system. The control board is also configured to: after the user completes the transaction confirmation through the payment module, regardless of whether it is the first recommendation result from the individual guest resource optimization unit or the group room allocation scheme from the group session collaboration unit, bind the final assigned room number with the corresponding user's identity, generate an authorization instruction, and send it to the door lock management system through the network communication module so that the door lock management system can authorize the corresponding room door lock.

9. The hotel self-check-in terminal according to claim 1, characterized in that, When the individual guest resource optimization unit outputs the first recommendation result, it simultaneously generates a brief decision basis text for each recommended room; the multilingual interaction module that controls the mainboard converts the brief decision basis text into the language currently set by the user, and displays it in the form of an interactive control next to the recommendation entry of the corresponding room on the touch screen.

10. The hotel self-check-in terminal according to claim 1, characterized in that, The operation status interface unit is also used to obtain real-time operation status data of the hotel elevators; when calculating the allocation cost or the future congestion cost, the individual guest resource optimization unit also incorporates the distance between the floor where the target room is located and the elevator lobby, as well as the current elevator operation load status, as additional factors into the calculation model.

Images