Hotel multi-scene reservation order merging processing system based on AI priority scheduling

Abstract

This invention discloses an AI-priority-based hotel multi-scenario reservation order merging and processing system, belonging to the field of order processing technology. The dynamic scheduling module automatically allocates orders to matching service processes based on priority ranking, reserving a portion of service processes as idle and schedulable. It triggers order merging logic and invokes resource scheduling, dynamically migrating idle or adjustable resources to corresponding service processes. The system performs real-time matching and verification of requested resources through order identification and validation, sending the results to the execution module. The execution module controls service execution according to priority, sequentially allocating the allocated resources to the corresponding service execution channels. This processing system, through the collaborative operation of the AI-priority-based hotel multi-scenario reservation order merging and processing system, effectively solves the problems of delayed response, low resource utilization, and fluctuating user experience in traditional hotel order processing.

Description

Technical Field

[0001] This invention relates to the field of order processing technology, specifically to a hotel multi-scenario reservation order merging and processing system based on AI priority scheduling. Background Technology

[0002] The intelligent processing and resource scheduling of hotel multi-scenario reservation orders (accommodation, catering, etc.) directly impacts customer experience and revenue. However, traditional manual or rule-based methods struggle to perceive global resources in real time, accurately merge orders, and handle unexpected conflicts, leading to delays in high-value orders, resource waste, and fluctuations in satisfaction, which are particularly pronounced during peak periods or group bookings. Therefore, building an intelligent system with multi-dimensional perception, global sorting, collaborative scheduling, intelligent merging, accurate verification, and adaptive optimization has become a key challenge. This solution proposes an AI-prioritized order merging processing system that addresses the core issues of traditional models through multi-dimensional data integration, dynamic decision-making, intelligent scheduling, and closed-loop optimization, supporting intelligent management of hotel services across multiple scenarios. Summary of the Invention

[0003] The purpose of this invention is to provide a hotel multi-scenario reservation order merging and processing system with AI priority scheduling to solve the problems in the background technology.

[0004] To achieve the above objectives, the present invention provides the following technical solution: an AI-priority scheduling system for merging and processing hotel multi-scenario reservation orders, comprising: Order sorting module: Collects hotel reservation order data from different sources in real time through multi-channel interfaces, performs multi-dimensional feature analysis and tagging on each order, maps the order to different services within the hotel, and performs global priority scoring and sorting on all orders based on multi-dimensional tag information. The sorting results are sent to the dynamic scheduling module and the execution module. Dynamic scheduling module: Based on the priority ranking results, the order is automatically assigned to the matching service processing, and a portion of the service processing is reserved as idle and schedulable. The order merging logic is triggered and the resource scheduling is invoked to dynamically migrate idle or adjustable resources to the corresponding service processing. The requested resources are matched and verified in real time through order identification and verification. The real-time matching and verification results of resources are sent to the execution module. Execution module: After the order completes resource matching and verification in the corresponding service processing and is confirmed to be executed, the control service execution will allocate the allocated resources to the corresponding service execution channels in order of priority.

[0005] Preferably, after an order enters the service processing, the dynamic scheduling module performs real-time matching and verification of the requested resources through order identification and verification to determine whether the resources have been occupied, whether they meet the user's needs, and whether there are any hidden conflicts, and then feeds back the verification results.

[0006] Preferably, the dynamic scheduling module provides feedback on the verification result. If the verification passes, it returns a resource matching success signal, and the order enters the subsequent service execution and settlement process. If the verification fails, it returns a resource matching failure signal and marks the order as pending rescheduling.

[0007] Preferably, the dynamic scheduling module returns a resource matching failure signal. For orders that fail to match, it automatically releases the temporary resources they occupy and triggers a rescheduling mechanism. Based on the current global resource status and order priority, it recalculates the priority score of the order and its associated orders and performs secondary order merging, cross-scenario resource adjustment, or alternative solution recommendation.

[0008] Preferably, the dynamic scheduling module automatically triggers order merging logic for orders with similar time windows, compatible required resource types, and similar user attributes. This includes merging multiple member orders from the same team into a combined accommodation arrangement, or bundling accommodation and conference services into a package. It calls resource scheduling to dynamically migrate resources that were originally idle or adjustable to the corresponding service processing, putting them into a service-ready state, ready to allocate resources to the orders.

[0009] Preferably, after the dynamic scheduling module completes the order priority sorting, it enters the dynamic scheduling and order merging execution stage of multiple service processing. According to the priority sorting result, high-priority orders are automatically assigned to the matching service processing. Each service processing corresponds to a type of resource or service and processes multiple order requests at the same time. A portion of the service processing is reserved as idle and schedulable for dynamically accepting newly arrived orders or intelligently merging with existing orders.

[0010] Preferably, when the execution module completes the processing task of the current order in a certain service, it automatically switches the status from "in service" to "idle and schedulable", and the resource scheduling department reclaims the unused or releaseable resources back to the global resource pool for use in subsequent new orders or merged orders.

[0011] Preferably, during order execution and resource scheduling, the execution module continuously triggers inventory checks and status monitoring, performs statistics, calibration and updates on the real-time usage of various service resources, and dynamically updates the inventory status, order matching parameters and user demand information in the global resource database based on real-time data.

[0012] Preferably, after collecting the original order data, the order sorting module performs multi-dimensional feature analysis and tagging on each order, mapping the order to different service processes within the hotel, including accommodation services, catering services, conference activities, and leisure and entertainment, each corresponding to one or more resource types. A tagging system containing multiple dimensions of information is constructed for each order, including user value, expected order revenue, resource usage type, user urgency, channel cooperation priority, and potential merging correlation of other orders.

[0013] Preferably, the order sorting module collects hotel reservation order data from different sources in real time through multi-channel interfaces. The order data sources include official websites, OTA platforms, corporate contract customers, membership channels, and telephone reservation centers. Each order data includes check-in time, check-out time, room type or service type, and number of guests. It also includes user-related extended attributes, including user membership level, historical consumption behavior, channel affiliation, booking channel cost, and potential additional service needs.

[0014] The technical effects and advantages provided by the present invention in the above technical solution are as follows: This invention intelligently allocates orders to matching service processing (such as standard rooms, food stalls, etc.) based on priority ranking results, while reserving a portion as an idle schedulable resource pool. By triggering order merging logic (such as merging similar orders from adjacent time periods into consecutive rooms or packages) and invoking resource scheduling to dynamically migrate vacant rooms and unoccupied time periods to demand, it significantly improves resource reuse rate and order merging success rate. Combined with order identification and verification, it performs real-time matching and verification of resources (such as target room number, service time period) to ensure the accuracy of allocation results, avoid resource waste caused by overbooking or conflicts, and quickly feeds back the verification results to execution, forming a closed-loop scheduling chain of priority-driven - dynamic merging - accurate allocation.

[0015] This invention collects heterogeneous order data in real time through multiple channels and, based on multi-dimensional feature analysis and tagging, accurately maps orders scattered across various sources such as official websites, OTA platforms, and enterprise agreements to services such as accommodation, catering, and conferences. It also calculates global priority by combining dynamic weights such as user value, order revenue, and resource scarcity, ensuring that high-value (e.g., high-net-worth customers), high-revenue potential (e.g., orders with high likelihood of additional consumption), and high-urgency (e.g., orders close to check-in time) orders are prioritized for entry into the scheduling process, thereby improving the accuracy and strategic adaptability of order processing from the source.

[0016] After an order completes resource verification and is confirmed for execution, this invention strictly allocates the assigned resources (such as specific room numbers, catering seats, and meeting time slots) to the corresponding service execution channels (such as check-in queues and catering reception lists) according to priority order, ensuring that high-priority orders receive service responses first, shortening user waiting time, and improving overall service efficiency. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0018] Figure 1 This is a framework diagram of the identification system of the present invention. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] Example: This example provides an AI-priority scheduling system for merging hotel reservation orders across multiple scenarios. Please refer to [link / reference]. Figure 1 As shown, it includes: Order sorting module: Collects hotel reservation order data in real time from various sources through multi-channel interfaces, including but not limited to official websites, OTA platforms, corporate clients, membership channels, and telephone reservation centers. Each order record includes not only basic structured information such as check-in time, check-out time, room type or service type, and number of guests, but also extended user attributes such as user membership level, historical consumption behavior, channel affiliation, booking channel cost, and potential additional service requests (such as breakfast, airport transfer, spa, meeting rooms, etc.).

[0021] After collecting the raw order data, each order undergoes multi-dimensional feature analysis and tagging, mapping it to different service processes within the hotel, such as accommodation, catering, conferences and events, and leisure and entertainment. Each service corresponds to one or more mergeable resource types. Simultaneously, a tagging system containing multiple dimensions of information is built for each order, including user value (e.g., whether the customer is a high-net-worth individual or a high-potential repeat customer), expected order revenue (e.g., order amount, likelihood of additional spending), resource usage type (e.g., room type, time period, service resources), user urgency, channel cooperation priority, and potential merging correlation with other orders.

[0022] Based on the aforementioned multi-dimensional tag information, all orders are given a global priority score and sorted. This priority assessment is not based on a single indicator, but rather adopts a multi-factor dynamic weighting mechanism, comprehensively considering overall revenue, current resource scarcity, order timeliness, and cross-scenario synergy (for example, the combination of accommodation and catering or conferences can significantly improve overall consumption and customer loyalty).

[0023] Dynamic Scheduling Module: After order priority ranking, the module enters the dynamic scheduling and order merging execution phase involving multiple service processing. Based on the priority ranking results, high-priority orders are automatically assigned to matching service processing (e.g., standard rooms, suites, catering, conference services, etc.). Each service processing corresponds to a core resource or service type and can handle multiple order requests simultaneously. To improve overall resource utilization and order processing efficiency, a portion of service processing is reserved as idle, schedulable services for dynamically accepting newly arriving orders or intelligently merging them with existing orders.

[0024] For orders with similar time windows, compatible required resource types (such as the same room type, adjacent time slots, and similar additional services), and similar user attributes, order merging logic will be automatically triggered. This will aggregate these orders into a unified execution unit, thereby achieving centralized resource allocation and packaged service provision. For example, multiple orders from the same team members can be merged into a connecting room arrangement, or accommodation and meeting services can be bundled into a package. At the same time, resource scheduling will be invoked to dynamically migrate previously idle or adjustable resources (such as vacant rooms, unused service slots, and reusable meeting venues) to the corresponding service processing, putting them into a service-ready state, ready to allocate specific resources to orders at any time.

[0025] Once an order enters the specific service processing stage, the requested resources are matched and verified in real time through order identification and validation. This determines whether the resource is already occupied, whether it meets the user's needs, and whether there are any hidden conflicts (such as overlapping times or mismatched restrictions). The validation results are then fed back. If the validation passes, a resource matching success signal is returned, and the order proceeds to the subsequent service execution and settlement process. If the validation fails (e.g., the target room is already booked or the required service time slot is unavailable), a resource matching failure signal is returned, and the order is marked as pending rescheduling.

[0026] For orders that fail to match, instead of simply discarding or delaying processing, the temporary resources they occupy are automatically released, and a rescheduling mechanism is triggered. Based on the latest global resource status and order priority, the priority score of the order and its associated orders is recalculated, and attempts are made to merge the orders again, adjust resources across scenarios, or recommend alternative solutions (such as upgrading room type, adjusting service time, or providing additional compensation options).

[0027] Execution Module: Once an order has completed resource matching and verification in the corresponding service processing and is confirmed for execution, the module will control the service execution according to priority, and allocate the allocated resources (such as specific room numbers, service time periods, package contents, etc.) to the corresponding service execution channels (such as check-in queues, catering reception lists, meeting execution plans, etc.) in sequence. This ensures that high-priority orders receive service responses first, thereby improving overall order processing efficiency, reducing user waiting time, and optimizing customer experience.

[0028] Meanwhile, once a service completes its processing of the current order, it will automatically switch its status from "service" to "idle and schedulable." The resource scheduler will then reclaim the unused or releaseable resources (such as vacant rooms or unused service capacity) back into the global resource pool for use in scheduling new or merged orders. This mechanism ensures efficient resource flow and full utilization, avoiding the waste of idle service capacity.

[0029] Throughout the entire order execution and resource scheduling process, inventory checks and status monitoring are continuously triggered, and the real-time usage of various service resources (including rooms, catering seats, meeting venues, SPA time slots, etc.) is statistically analyzed, calibrated, and updated. Based on this real-time data, the inventory status, order matching parameters, and user demand information in the global resource database are dynamically updated, thereby providing accurate and real-time data support for the next round of order scheduling decisions.

[0030] Furthermore, this application not only makes scheduling decisions based on the current order and resource status, but also analyzes cross-dimensional information such as historical booking patterns, user behavior trends, and channel traffic changes to predict resource demand fluctuations in different time periods, service scenarios, and user groups in the future. This allows for advance adjustments to resource allocation strategies, optimization of order merging possibilities, and prediction of potential resource bottlenecks and service conflicts, thus achieving an intelligent leap from passive response scheduling to proactive scheduling.

[0031] The specific implementation methods for each module of this application are described in detail below: Direct sales channels (official website / APP / mini-program) carry in-depth behavioral data such as user registration ID, device fingerprint, and click flow path; OTA platforms (Ctrip / Booking / Expedia, etc.) include channel-specific fields such as platform commission rate, traffic source keywords, and promotional activity tags; enterprise agreements include B-end attributes such as customer-associated enterprise credit limit, contract discount rules, and group agreement number; the membership system integrates RFM model data (recent consumption time, consumption frequency, consumption amount), membership level (such as silver card / gold card / black diamond), and historical loyalty points; the telephone reservation center records urgent requests noted by human agents (such as customers requesting a window room must be booked) and service priority prompts derived from voice sentiment analysis (such as complaint risk warnings).

[0032] Based on the cleaned base data, composite tags are generated, specifically divided into the following levels: The system directly extracts membership levels (e.g., Black Diamond members have higher weight than Silver members), historical repurchase intervals (number of orders in the last 30 / 90 / 365 days), and cumulative spending tiers (e.g., high-net-worth clients are defined as those with annual spending exceeding 50,000 yuan). It outputs potential repurchase probabilities (based on user behavior sequence similarity and competitor channel price comparisons) and price sensitivity coefficients (calculated based on the response to discounts during historical bookings). If the order is associated with a company agreement number, the system also incorporates the company's credit score (e.g., the company's past fulfillment rate and outstanding payment records).

[0033] The calculation includes: base room rate × number of nights + fees for explicitly selected additional services (such as airport transfers selected in advance); additional consumption probability calculated using a regression model trained on historical data (e.g., 62% probability of high-level members choosing SPA, and 28% probability for regular users), multiplied by the average price per customer for the corresponding service (e.g., 300 yuan average for SPA); channel cost rate (e.g., 15% commission rate for a certain OTA platform); and special discounts and subsidies (e.g., 50 yuan discount for new customers' first order).

[0034] Room type code (e.g., Deluxe King), specific time period (check-in 14:00 to check-out 12:00), non-changeable service resources (e.g., Meeting Room A is already booked from 10:00 to 12:00); time period proximity (e.g., room type resources may be released if orders are not checked in after 18:00 on the same day); room type alternatives (e.g., Superior Queen can be recommended and marked with a downgrade risk label if Deluxe King is in short supply); if the order booked a meeting room and lunch package at the same time, mark it with a meeting package bundle label, and associate the meeting room capacity (e.g., can accommodate 20 people) and the catering time (e.g., 12:00-13:30).

[0035] Unconfirmed orders that have exceeded the time limit (e.g., orders not paid within 30 minutes of a phone reservation will trigger a reminder), urgent orders close to the check-in date (e.g., unassigned room orders placed before 18:00 the day before check-in), urgent requests actively marked by users (e.g., needing priority check-in due to flight delays), and SLA time limits in the company agreement (e.g., VIP customers requiring a response within 2 hours).

[0036] Orders from strategic partners (such as OTA platforms invested in by the company's major shareholders) are given higher priority than those from ordinary channels by default; orders participating in limited-time promotions (such as Double 11 pre-sales) need to ensure resource supply to avoid breach of contract compensation; channels with low commissions but high user value (such as orders converted from private domain traffic) may receive extra points.

[0037] Combination bookings of accommodation, catering, and meetings are marked with a high stickiness potential tag (historical data shows that such users have a 40% higher repurchase rate in the following 30 days); the same user booking the same room type for two consecutive nights (which can reduce room change and cleaning costs) or multiple people booking a shared meeting room (which reduces the per capita service cost).

[0038] Based on the above tagging system, a multi-factor dynamic weighted scoring model is used to calculate the comprehensive priority score (Score∈[0,100]) for each order, as follows: The weighting of core resource constraints (such as orders during periods of room shortage) accounts for 40%-50% (e.g., luxury room orders in tourist cities during peak season, where resource scarcity directly impacts overall occupancy rates); high-yield expectation orders (such as high-member orders with a high probability of hidden consumption) account for 20%-30% (e.g., orders with an additional consumption probability >50%, whose potential revenue may exceed the base room rate); user value and channel priority orders account for 15%-25% (e.g., even if Black Diamond member orders have average revenue, service experience must be guaranteed to maintain loyalty); and timeliness and synergy order orders account for 10%-15% (e.g., unallocated orders close to the check-in date need to be processed urgently to avoid customer churn).

[0039] Automatically adjust weight parameters based on real-time operational status: When the remaining inventory of a certain room type is less than 10% of the total number of rooms, the weight of resource scarcity increases from 40% to 60%; when the quarterly contract signing target completion rate of corporate clients is less than 70%, the priority weight of order channels under the agreement with affiliated companies increases by 10%; during peak periods such as holidays, the weight of timeliness (such as orders placed 1 day before check-in) increases by an additional 5%.

[0040] For each order, generate the final score by following these steps: Resource scarcity sub-scores are calculated based on room type availability (e.g., the ratio of currently available DeluxeKing rooms to total rooms). The lower the ratio, the higher the sub-score, linearly mapped to 0-30 points. Time slot conflict penalties are added (e.g., 5-10 points are deducted if the time slot overlaps with a confirmed high-priority order). Explicit revenue is scored directly based on amount (e.g., 5 points for 1000-2000 yuan, 10 points for 2000-5000 yuan). Implicit revenue is calculated by multiplying the model's output additional consumption probability by the average order value to generate expected incremental revenue (e.g., probability 60% × 300 yuan = 180 yuan → 3 points). Net revenue after deducting channel costs and subsidies is calculated as (total revenue points - cost points). Membership levels are mapped to base points (15 points for Black Diamond members, 10 points for Gold members, and 5 points for Silver members). The points are calculated as follows: Repurchase probability score (>80% adds 5 points, 60%-80% adds 3 points, <60% adds no points); 2 extra points are awarded for high-potential new customers (no prior purchases but similar behavior to high-net-worth individuals); Remaining days before check-in are mapped to emergency points (≤1 day adds 20 points, 2-3 days adds 15 points, 4-7 days adds 10 points, >7 days adds 5 points); 10-15 extra points are awarded for overdue unconfirmed orders or urgent corporate SLA requirements; Strategic partner orders have a base score of 10 points, ordinary OTA platforms add 5 points, telephone booking centers (with human follow-up for key needs) add 3 points, and orders associated with promotional activities add 5 points; Cross-scenario combination orders (e.g., accommodation + catering + conference) add 8 points, resource reuse orders (e.g., two consecutive nights in the same room type) add 5 points, and single accommodation orders do not receive points. The final overall score is calculated as follows: (S1 × Resource Weight) + (S2 × Revenue Weight) + (S3 × User Weight) + (S4 × Timeliness Weight) + (S5 × Channel Weight) + (S6 × Collaboration Weight). Each weight is dynamically adjusted based on real-time operational strategies (e.g., during peak season, the resource weight increases to 60%, while the revenue weight decreases to 20%).

[0041] All orders are sorted in descending order of their final scores, and a priority queue is generated. Top 10% high-priority orders: trigger automatic resource reservation (e.g., locking room type until payment confirmation) and push to human customer service for priority processing; Middle 60% regular orders: enter the regular scheduling process, and allocate remaining resources according to their scores; Bottom 30% low-priority orders: marked as deferred processing (e.g., regular member orders during off-peak seasons), and allocated only when resources are redundant. After completing the multi-dimensional priority ranking of orders, the dynamic scheduling module first automatically routes high-priority orders (usually the top 30%-40% of orders, the specific percentage can be dynamically adjusted according to the hotel's peak and off-peak seasons or operational strategies) to service processing units that match their core needs. Each service processing unit corresponds to a type of core resource or service within the hotel, for example: The accommodation service processing unit is responsible for the allocation of room types such as standard rooms, suites, and apartments, and needs to link to real-time room status data in PMS (Property Management System) (including statuses such as sold rooms, dirty rooms, rooms under maintenance, and rooms available for sale). The catering service processing unit handles catering needs such as breakfast, lunch, dinner, and afternoon tea, and takes into account constraints such as table capacity, chef scheduling, and food inventory in the catering industry. The meeting service processing unit manages venue resources such as meeting rooms and banquet halls, and needs to verify venue dimensions, projection equipment configuration, and conflicts with other meeting bookings at the same time. The leisure and entertainment service processing unit covers services such as spas, gyms, and swimming pools, and associates service personnel scheduling (such as the availability of massage therapists) and equipment usage status (such as the time for pool cleaning and maintenance).

[0042] Each service processing unit adopts a multi-order concurrent processing mechanism, meaning that a single unit can receive and process multiple order requests simultaneously (for example, a catering service unit can allocate multiple group orders to different tables at the same time), but must strictly adhere to the physical capacity limits of resources (for example, if a meeting room can accommodate a maximum of 50 people, then the total number of people allocated in the same time period must not exceed this threshold).

[0043] To achieve intensive resource utilization and improved service efficiency, the dynamic scheduling module scans the set of orders to be assigned in real time, automatically identifies order groups that meet the following conditions, and triggers the merging logic: Time window proximity: The overlap rate of check-in / use time periods of orders exceeds the threshold (e.g., the overlap between check-in and check-out time periods of accommodation orders and other orders is ≥6 hours, or the overlap between the use time period of conference orders and the service time period of other orders is ≥50%). Resource type compatibility: The required core resource types are consistent and can be shared (such as all being DeluxeKing room types, rooms on adjacent floors / facing the same direction, or similar additional services such as breakfast + airport transfer packages). User attribute similarity: The user groups associated with the orders have a basis for merging (e.g., multiple employee orders from customers of the same enterprise, orders from members of the same tour group, or high-value customer groups with similar historical repurchase behavior).

[0044] Orders that meet certain criteria are aggregated into a unified execution unit (called a merged order group) using a clustering algorithm. The steps are as follows: For each order to be evaluated, extract key feature vectors, including core resource type (e.g., room type code), time period (check-in time - check-out time), additional service list (e.g., breakfast / SPA), user attribute tags (e.g., enterprise agreement number / membership level), and priority score; calculate the similarity between order feature vectors based on Euclidean distance or cosine similarity (e.g., 1 point for the same room type, 0.8 points for adjacent time periods, 1 point for the same enterprise agreement number, and 0 points for different ones), and generate a similarity matrix between orders; use hierarchical clustering or DBSCAN density clustering algorithms to group orders with similarity higher than a threshold (e.g., overall similarity ≥ 0.7) into the same cluster to form merged order groups; summarize the resource requirements of the merged order groups (e.g., total number of rooms = sum of the number of rooms in each order within the group, total number of diners = sum of the number of diners in each order), and check whether the integrated demand exceeds the physical capacity of a single service processing unit (e.g., if the total number of rooms after merging exceeds the current available DeluxeKing rooms, then split some orders to other room types).

[0045] The merged execution unit will enjoy package service optimizations: for example, bookings that bundle accommodation and meetings can automatically receive meeting room discounts (such as reducing the original price of RMB 2,000 / hour to RMB 1,800 / hour), and cleaning fees can be waived for connecting room arrangements (such as waiving the additional service fee of RMB 20 / night for each connecting room). At the same time, the service processing unit can reduce operating costs through centralized allocation (such as booking 10 adjacent rooms for a team at once is 50% more efficient than processing 10 orders separately).

[0046] Once an order (including independent orders or combined order groups) enters a specific service processing unit, an order identification and verification process is triggered. This process performs multi-dimensional real-time matching and verification of the core resources requested. Specific verification dimensions include: Check if the resources have been occupied by other confirmed orders (e.g., whether the target Deluxe King room has been allocated to other guests during the check-in period required by the order); verify if there are any service time conflicts (e.g., whether the 10:00-12:00 time required by the meeting order conflicts with the booked banquet). Verify whether the user's special needs can be met (e.g., if the order requires a non-smoking room, verify the smoking status of the target room type; if a sea view room is required, verify whether the room faces the sea); confirm the availability of additional services (e.g., whether there are available vehicles for the airport transfer service selected in the order during the target time period). Overlapping time conflicts (e.g., the check-out cleaning time of an accommodation order conflicts with the check-in time of the next order, resulting in the inability to deliver on time); mismatched resource constraints (e.g., SPA services require guests to arrive 30 minutes in advance for preparation, but the order requires a start time earlier than that constraint); The verification process obtains the latest data by calling real-time resource status APIs (such as the room status interface of PMS, the table interface of catering, and the venue interface of conferences), and verifies each item based on the rule engine. If all verification dimensions pass, a resource matching success signal is returned, and the order enters the service execution process (such as generating room key pre-authorization, reserving catering tables, and locking conference room time slots); if any dimension fails (e.g., the target room has been booked), a resource matching failure signal is returned, and the order is marked as pending rescheduling, while recording the reason for failure (e.g., room type conflict and time slot unavailable).

[0047] For orders that fail to match, they will not be discarded or delayed indefinitely. Instead, an intelligent rescheduling mechanism will be activated, as follows: Automatically release any temporary resources previously occupied by this order (e.g., the lock-up period for reserved meeting room slots), returning them to the idle resource pool to ensure resources can be reused by other orders. Based on the latest global resource status (e.g., 3 vacant rooms of a certain room type were just released due to order cancellation) and the order priority score (reassessing the timeliness weight; for example, an order originally scheduled 3 days before check-in is now scheduled for 1 day before check-in due to delayed processing, increasing the timeliness weight), recalculate the overall priority score of this order and its related orders (e.g., other sub-orders in a merged order group). Attempt to merge this order a second time (e.g., after the original merge failed, newly arrived orders from the same company can be re-aggregated with it), or adjust resources across scenarios (e.g., when the target room type for an accommodation order is unavailable, check the availability of other room types on the same floor; when meeting orders have conflicting time slots, recommend vacant meeting rooms in adjacent time slots).

[0048] If the issue cannot be resolved through merging or reorganization, alternative options will be generated for the user to choose from (or the optimal solution will be applied automatically), including: Resource upgrade / downgrade: If the originally booked Deluxe King room is unavailable, we recommend a Superior Queen room and compensate 10% of the room rate difference, or downgrade to a Standard Queen room but include free breakfast; Service time adjustment: If there is a conflict between the 10:00-12:00 time slot of the meeting order, we recommend the 14:00-16:00 time slot and will waive the rescheduling fee; Additional compensation options include providing dining vouchers (e.g., a 200 RMB dining voucher), SPA discount coupons (e.g., 20% off), and doubled points (e.g., an extra 1000 loyalty points) to compensate users for the inconvenience caused by resource adjustments.

[0049] For accommodation services, the execution module inserts the assigned room number (e.g., Deluxe King room 1001 on the 10th floor) into the check-in queue with priority, ensuring that customers with high-priority orders can proceed with identity verification and room key collection first. For catering services, the corresponding dining time (e.g., 12:30-13:30) and table information (e.g., table 2 in restaurant A area, accommodating 6 people) are added to the top of the catering service list, prioritizing chef preparation and waiter guidance. For conference services, the assigned meeting room (e.g., Crystal Hall on the 3rd floor, accommodating 30 people) and usage time (e.g., 14:00-16:00) are included in the priority scheduling of the conference execution plan, ensuring that conference equipment (e.g., projector, sound system) is prepared and the venue is set up in advance. This priority-based allocation mechanism, by strictly adhering to the order priority, avoids the problem of low-priority orders occupying key resources, causing high-priority orders to have excessively long waiting times, thus achieving optimal resource allocation.

[0050] During resource allocation, the execution module calls the real-time resource status interface of each service processing unit to ensure that the allocated resources are available immediately upon allocation. For example, before assigning a room number to the check-in queue, it checks the PMS again to confirm that the room number has not been temporarily occupied by other orders; before assigning a meeting time slot to the meeting execution plan, it verifies with the meeting management that the meeting room during that time slot has not been booked by other urgent meetings. This dual verification mechanism further ensures the accuracy and effectiveness of resource allocation and reduces the risk of order execution failure due to resource conflicts.

[0051] Once a service processing unit completes its current order processing task (e.g., guest successfully checks in, catering service ends and tables are cleared, meeting room is returned after a meeting), the execution module automatically triggers a dynamic switch in service status, updating the service status corresponding to that order from "idle and available" to "available for scheduling" in real time. This status switching operation is achieved by sending status update instructions to each service processing unit, ensuring that the order status in each business unit remains consistent with the global view of the execution module.

[0052] The resource scheduling module will immediately intervene and reclaim any unused or releaseable resources (such as vacant rooms and unused service capacity) in the service processing unit back to the global resource pool. For accommodation services, if a guest checks out early, causing the room to become available early, the execution module will provide real-time feedback on the room's available time period (e.g., if the original check-in time is 14:00 and the actual check-out time is 12:00, then the time period from 12:00 to 14:00 becomes available) to the PMS. The PMS will then mark the room status for that time period as available and resell it. For catering services, if a table still has remaining capacity after the meal (e.g., a table originally booked for 6 people is actually used for 4 people), the execution module will pass the remaining capacity information of the table (e.g., 2 seats remaining) to the catering management, and the catering management will re-include part of the table's capacity in the schedulable resources. For conference services, if a conference ends early and the conference room is returned, the execution module will mark the conference room from its current occupied state as available and update the available time period information in the conference room management.

[0053] Throughout the entire order execution and resource scheduling process, the execution module continuously triggers inventory checks and status monitoring functions to comprehensively and accurately statistically analyze, calibrate, and update the real-time usage of various service resources (including rooms, catering seats, meeting venues, SPA time slots, etc.). Specifically, the execution module obtains real-time resource status information for each service processing unit through a combination of timed polling and event-driven methods. For room resources, the PMS pushes the real-time status of rooms (e.g., sold, available, cleaning, under maintenance) to the execution module at regular intervals (e.g., every 5 minutes). The execution module summarizes and calibrates this information to ensure that the room status in the global resource database is consistent with the actual situation. For catering seats, catering management provides real-time feedback on table usage (e.g., occupied, available) after each order is completed. The execution module updates the real-time available number of catering seats based on this feedback. For meeting venues, meeting management sends notifications to the execution module at the start and end of meetings. The execution module updates the meeting room usage status and availability for future time slots accordingly.

[0054] In the description of this specification, references to terms such as "an embodiment," "example," "specific example," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0055] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to any specific implementation. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A hotel multi-scenario reservation order merging processing system with AI priority scheduling, characterized in that: include: Order sorting module: Collects hotel reservation order data from different sources in real time through multi-channel interfaces, performs multi-dimensional feature analysis and tagging on each order, maps the order to different services within the hotel, and performs global priority scoring and sorting on all orders based on multi-dimensional tag information. The sorting results are sent to the dynamic scheduling module and the execution module. Dynamic scheduling module: Based on the priority ranking results, the order is automatically assigned to the matching service processing, and a portion of the service processing is reserved as idle and schedulable. The order merging logic is triggered and the resource scheduling is invoked to dynamically migrate idle or adjustable resources to the corresponding service processing. The requested resources are matched and verified in real time through order identification and verification. The real-time matching and verification results of resources are sent to the execution module. Execution module: After the order completes resource matching and verification in the corresponding service processing and is confirmed to be executed, the control service execution will allocate the allocated resources to the corresponding service execution channels in order of priority. 2.The AI priority scheduling hotel multi-scenario reservation order consolidation processing system of claim 1, wherein: After an order enters the service processing, the dynamic scheduling module performs real-time matching and verification of the requested resources through order identification and verification to determine whether the resources have been occupied, whether they meet the user's needs, and whether there are any hidden conflicts, and then feeds back the verification results.

3. The AI-priority scheduling system for merging hotel reservation orders across multiple scenarios, as described in claim 2, is characterized in that: The dynamic scheduling module provides feedback on the verification result. If the verification passes, it returns a resource matching success signal, and the order enters the subsequent service execution and settlement process. If the verification fails, it returns a resource matching failure signal and marks the order as pending rescheduling.

4. The AI-priority scheduling system for merging hotel reservation orders according to claim 3, characterized in that: The dynamic scheduling module returns a resource matching failure signal. For orders that fail to match, it automatically releases the temporary resources they occupy and triggers a rescheduling mechanism. Based on the current global resource status and order priority, it recalculates the priority score of the order and its associated orders and performs secondary order merging, cross-scenario resource adjustment, or alternative solution recommendation.

5. The AI-priority scheduling system for merging hotel reservation orders according to claim 2, characterized in that: The dynamic scheduling module automatically triggers order merging logic for orders with similar time windows, compatible required resource types, and similar user attributes. This includes merging multiple orders from the same team into a combined accommodation arrangement, or bundling accommodation and conference services into a package. It calls resource scheduling to dynamically migrate resources that were originally idle or adjustable to the corresponding service processing, putting them into a service-ready state, ready to allocate resources to the orders.

6. The AI-priority scheduling system for merging hotel reservation orders according to claim 2, characterized in that: After completing the order priority sorting, the dynamic scheduling module enters the dynamic scheduling and order merging execution stage of multi-service processing. According to the priority sorting result, high-priority orders are automatically assigned to the matching service processing. Each service processing corresponds to a type of resource or service and processes multiple order requests at the same time. A portion of service processing is reserved as idle and schedulable for dynamically accepting newly arrived orders or intelligently merging with existing orders.

7. The AI-priority scheduling system for merging hotel reservation orders according to claim 1, characterized in that: Once the execution module completes the processing task of the current order in a certain service, it automatically switches the status from "in service" to "idle and schedulable". The resource scheduler then reclaims the unused or releaseable resources back to the global resource pool for use in subsequent new orders or merged orders.

8. The AI-priority scheduling system for merging hotel reservation orders according to claim 7, characterized in that: During order execution and resource scheduling, the execution module continuously triggers inventory checks and status monitoring, and performs statistics, calibration and updates on the real-time usage of various service resources. Based on real-time data, it dynamically updates the inventory status, order matching parameters and user demand information in the global resource database.

9. The AI-priority scheduling system for merging hotel reservation orders according to claim 1, characterized in that: After collecting the original order data, the order sorting module performs multi-dimensional feature analysis and tagging on each order, mapping the order to different service processes within the hotel, including accommodation, catering, conferences and events, and leisure and entertainment. Each of these corresponds to one or more resource types. A tagging system containing multiple dimensions of information is constructed for each order, including user value, expected order revenue, resource usage type, user urgency, channel cooperation priority, and potential merging correlations with other orders.

10. The AI-priority scheduling system for merging hotel reservation orders according to claim 8, characterized in that: The order sorting module collects hotel reservation order data from different sources in real time through multi-channel interfaces. The order data sources include official websites, OTA platforms, corporate contract customers, membership channels, and telephone reservation centers. Each order data includes check-in time, check-out time, room type or service type, and number of guests. It also includes user-related extended attributes, such as user membership level, historical consumption behavior, channel affiliation, booking channel cost, and potential additional service needs.

Images