Intelligent scheduling method for seat arrangement information

Abstract

The invention discloses a seat arrangement information intelligent scheduling method, which comprises the steps of collecting seat arrangement scene information and seat arrangement personnel information, performing feature conversion on the seat arrangement scene information and the seat arrangement personnel information through a neural network, and obtaining a seat combination packet database and personnel information demand labels; managing mutual coordination operation of global seat arrangement of all the exclusive decision-making agents, and outputting a seat distribution scheme set; and performing optimal adjustment on the seat allocation scheme based on the improved whale optimization algorithm, and outputting an optimal seat allocation scheme. According to the method, by creatively fusing multi-agent cooperation, reinforcement learning dynamic strategies and an innovative real-time incremental combined auction mechanism, a comprehensive, efficient and excellent-experience solution is provided for the seat resource scheduling problem in the modern service industry. The calculation bottleneck of traditional combined auction is overcome, and the seat arrangement scheme is more effective, intelligent and real-time.

Description

Technical Field

[0001] This invention relates to the technical field of seat scheduling, and more particularly to an intelligent scheduling method for seat arrangement information. Background Technology

[0002] Seating arrangements are crucial for meetings, but traditional manual methods are inefficient and prone to errors. While manageable for small meetings, the workload increases dramatically with larger events (over 100 people), leading to significantly higher error rates (spelling of names, department order, confusion between main and guest names, omissions), severely impacting preparation efficiency and the meeting's image. In important meetings, seating errors have far-reaching consequences: improper seating of key guests can trigger a crisis of trust, departmental disorganization hinders communication, and the absence of decision-makers affects outcomes. Unexpected changes (guest absences) result in slow manual responses, and time-consuming adjustments lead to chaos (empty seats and overcrowding). Creating and updating nameplates is costly and wasteful, and manual seating arrangements can easily overlook details such as accessibility needs, isolation areas, or personalized preferences, damaging the participant experience and professionalism. In summary, traditional methods have significant shortcomings in efficiency, accuracy, flexibility, and cost control, making it difficult to meet the sophisticated needs of modern meetings.

[0003] Currently, Chinese invention patent application number CN202411835729.0 discloses a method and system for arranging examination seats based on an intelligent examination platform. This method collects environmental data through the intelligent examination platform and generates an environmental feature matrix using implicit sensing technology. It divides the examination room area based on a quantum dot distribution model and optimizes seat allocation through reinforcement learning, generating optimized seat allocation data. A multimodal verification module is used to complete candidate identity verification, binding the verification data with the optimized seat allocation data to unlock the examination interface and load the examination task. This invention effectively improves the fairness and intelligence of examination seat allocation, achieving priority allocation for candidates with special needs and dynamic management of the examination process, demonstrating significant reliability and convenience. However, existing technologies struggle with large-scale, personalized seat allocation in scenarios such as aviation, ticketing, and exhibitions, easily leading to seat arrangement errors or uneven distribution, as well as incorrect or unoptimized personnel allocation. Summary of the Invention

[0004] The technical problem solved by this invention is that existing technologies are difficult to implement on a large scale and with personalized seating arrangements in scenarios such as aviation, ticketing, and exhibitions, which can easily lead to problems such as incorrect or uneven seating arrangements, as well as incorrect or unoptimal allocation of personnel.

[0005] To solve the above-mentioned technical problems, the present invention provides the following technical solution: an intelligent scheduling method for seating information, comprising the following steps: Step S1: Collect seat arrangement scene information and seat arrangement personnel information, and transform the seat arrangement scene information and seat arrangement personnel information into features through a neural network to obtain a seat combination package database and personnel information demand tags; Step S2: When seating arrangement is required, assign a dedicated decision agent to each participant who needs a seat, manage the coordination of all dedicated decision agents in the global seating arrangement, and output a set of seat allocation schemes. Step S3: Based on the improved whale optimization algorithm, the seat allocation scheme is optimally adjusted, and the optimal seat allocation scheme is output.

[0006] Preferably, step S1 includes: The seating arrangement scenario information includes a physical seating floor plan and a seating arrangement scenario theme. The seating arrangement personnel information includes the number of people that can be accommodated, basic personnel entry information, and personnel constraints. The basic personnel entry information includes personnel ID, reserved seat type, and seat preference. The personnel constraints include custom entry conditions, special entry channels for off-site staff, and staff members in their assigned positions. The physical seating plan is converted into a set of seat numbers and corresponding tradable attribute tags using image conversion software. Adjacent or identical tradable attributes are dynamically generated into seat combination packages. The tradable attributes include window seats, emergency exit seats, regular seats, adjacent seats (n), accessible seats, main seats, and VIP areas. Each seat combination package corresponds to a subset of different tradable attribute tags, where n is a natural number greater than 0.

[0007] Preferably, the seat arrangement scenario information and seat arrangement personnel information are extracted through the embedding layer to obtain feature values ​​of the seat arrangement scenario information and the seat arrangement personnel information, respectively, and then mapped to the seat combination package database and the personnel information requirement tag database.

[0008] Preferably, step S2 includes: When seating arrangements are needed, a dedicated decision agent is configured for each participant who needs a seat. The dedicated decision agent coordinates with the participants based on the data information corresponding to the participants in the personnel information demand tag database and the current global seating arrangement information to perform global seating arrangements. The current global seating arrangement information includes the remaining time, the competition intensity of a specific area, and the current highest match. The dedicated decision agent decides to compete for one or more seat packages based on the matching degree between the personnel who need to match seats and the seats. The cost is the abstract feature value of the personnel's seat demand corresponding to the current dedicated decision agent. The abstract feature value of seat demand includes the abstract feature expression of the matching percentage between the personnel information demand tag database and the seat package corresponding to the currently competing seats through one-hot encoding, or to remain on the sidelines. Set target rewards for dedicated decision-making agents. When a dedicated decision-making agent successfully obtains the seat with the highest matching degree, it receives the first positive reward. When the dedicated decision-making agent obtains a candidate seat, it receives a second positive reward; When the dedicated decision-making agent fails to acquire a seat or the cost exceeds a preset maximum threshold, it receives a negative reward.

[0009] Preferably, managing the coordinated operation of the global seating arrangement of all dedicated decision-making agents includes: The physical seating plan is divided into zones according to a preset seating competition objective, such as by the number of tables, by price, or by function. Each zone is assigned a coordinated intelligent management system, which manages the seating resources and real-time status of the corresponding seating zone. The seating resources include the seat numbers that have not been filled by competition, and the real-time status includes the vacancy, provisional allocation, and final lock status of each seat. Each seating zone is then used to construct a tree based on the seat combinations included in the current seating zone, using a disjoint-set data structure, to obtain a seating allocation lookup tree. When each person's dedicated decision agent makes a competitive request, the seat allocation search tree is first searched for the tag information of the target seat combination package. The number of times the dedicated decision agent finds the tag information of the seat combination package corresponding to the seat demand of all the persons is the highest cost that the dedicated decision agent can give. If the dedicated decision agent fails to find the target, a preset cost for failure is added to the sum of costs for successful searches to obtain the highest cost that the dedicated decision agent can provide.

[0010] Preferably, the competition mechanism includes: When the seating arrangement is initiated, each person calculates the cost based on their respective dedicated decision-making agent, and obtains the highest cost and the target seat. The target seat is the seat number included in the seat combination package corresponding to the person's information demand tag. After obtaining the highest cost and the target seat, each dedicated decision-making agent competes to bid, and the competitive bid does not exceed the highest cost. The coordination strategies for intelligent management include: Step S201: The intelligent coordinator receives a new competitive bid and updates the provisional winner list with the new competitive bid and the personnel information feature value and seat number corresponding to the competitive bid. Step S202: Coordinate the intelligent management to perform incremental updates, handle conflict sets, and quickly update the provisional winner list; Step S203: The coordinated intelligent management team will provide real-time feedback of the latest provisional winner status to all participating dedicated decision-making agents; Step S204: The exclusive decision-making agent who makes the last and highest bid obtains the seat; The coordination strategy is repeated until no dedicated decision-making agent makes a bid; The final set of seat allocation schemes is obtained.

[0011] Preferably, the seat allocation scheme is optimally adjusted based on the improved whale optimization algorithm, and the adjustment includes: Each seat allocation scheme is encoded as a "whale" position. The quality of each "whale" is judged by a fitness function, which aims to maximize the total amount of all accepted bids while penalizing any scheme that violates personnel constraints.

[0012] Preferably, the logic for determining the provisional winner includes: A provisional winner is determined based on a preset fixed time interval or the appearance of a new bid, and the provisional winner list is updated. When a new bid enters the competition, it is determined whether the new bid conflicts with the current provisional winner set. If a conflict exists, no update is made; otherwise, an update is made. Resource conflicts include: If there is no competition for the seat package involved in the new bid by the current exclusive decision agent, i.e. there is no conflict, the exclusive decision agent will be directly updated to the new provisional winner; If the current exclusive decision agent's new bid conflicts with the bids of one or more provisional winners, a conflict set consisting only of conflicting bids is constructed. A small-scale optimization calculation is performed using an improved whale optimization algorithm to determine the new provisional winner only within the conflict set. Preferably, when the optimal adjustment and optimization of the seat allocation scheme is completed, the current global provisional winner set is used as the initial solution, and the improved whale optimization algorithm is used for global optimization to obtain the optimal seat allocation scheme.

[0013] Preferably, after obtaining the optimal seat allocation scheme, the final result is announced, seats are allocated to the successful bidders, and settlement is carried out. All data from the competition process is used for the retraining and iteration of the intelligent scheduling method for seat allocation information.

[0014] The beneficial effects of this invention are: it transforms physical seating resources into auctionable seat packages, allowing participants (such as passengers and spectators) to express their complex preferences (such as multiple seats together, specific areas) through bidding. This method introduces a real-time incremental auction mechanism, providing users with real-time feedback on their bidding status during the auction process. The core is the use of a multi-agent system (MAS) for distributed collaboration, with reinforcement learning-enabled user agents achieving intelligent bidding and system strategy self-optimization. An improved whale optimization algorithm serves as the core engine, efficiently solving complex winner-determination problems. Ultimately, it constructs an intelligent scheduling system capable of real-time response, automatic decision-making, and maximizing overall revenue and user satisfaction. This method, by creatively integrating multi-agent collaboration, reinforcement learning dynamic strategies, and an innovative real-time incremental combinatorial auction mechanism, provides a comprehensive, efficient, and superior solution to the seating resource scheduling problem in modern service industries. It not only overcomes the computational bottleneck of traditional combinatorial auctions but also transforms resource allocation from a static background task into a dynamic, transparent, and interactive intelligent service process through intelligence and real-time processing. By employing multi-agent partitioning management and incremental computing, the computational complexity of large-scale problems is effectively decomposed, enabling support for large-scale scenarios. The introduction of reinforcement learning allows the system to automatically learn user behavior and market dynamics, autonomously optimizing bidding and auction strategies to adapt to demand fluctuations. The real-time incremental auction mechanism provides instant feedback, transforming traditional "blind auctions" into transparent, interactive, and dynamic games, significantly enhancing user engagement. The combined auction format satisfies users' personalized and complex preferences, while the intelligent optimization engine ensures that resource value is maximized while meeting these preferences. Attached Figure Description

[0015] Figure 1 This is a basic flowchart illustrating an intelligent scheduling method for seating information provided in one embodiment of the present invention. Detailed Implementation

[0016] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0017] Reference Figure 1 As an embodiment of the present invention, a method for intelligent scheduling of seating information is provided, comprising the following steps: Step S1: Collect seat arrangement scene information and seat arrangement personnel information, and transform the seat arrangement scene information and seat arrangement personnel information into features through a neural network to obtain a seat combination package database and personnel information demand tags; Step S2: When seating arrangement is required, assign a dedicated decision agent to each participant who needs a seat, manage the coordination of all dedicated decision agents in the global seating arrangement, and output a set of seat allocation schemes. Step S3: Based on the improved whale optimization algorithm, the seat allocation scheme is optimally adjusted, and the optimal seat allocation scheme is output.

[0018] Step S1 includes: The seating arrangement scenario information includes a physical seating floor plan and a seating arrangement scenario theme. The seating arrangement personnel information includes the number of people that can be accommodated, basic personnel entry information, and personnel constraints. The basic personnel entry information includes personnel ID, reserved seat type, and seat preference. The personnel constraints include custom entry conditions, special entry channels for off-site staff, and staff members in their assigned positions. The physical seating plan is converted into a set of seat numbers and corresponding tradable attribute tags using image conversion software. Adjacent or identical tradable attributes are dynamically generated into seat combination packages. The tradable attributes include window seats, emergency exit seats, regular seats, adjacent seats (n), accessible seats, main seats, and VIP areas. Each seat combination package corresponds to a subset of different tradable attribute tags, where n is a natural number.

[0019] Seating packages, such as “three window seats together,” include the tradable attributes of three consecutive seats and a window seat, allowing users’ complex needs to be met by a single auction item.

[0020] The feature values ​​of the seating arrangement scenario information and the seating arrangement personnel information are extracted through the embedding layer. The feature values ​​of the seating arrangement scenario information and the seating arrangement personnel information are then mapped to the seating combination package database and the personnel information requirement tag database, respectively.

[0021] Step S2 includes: To reduce the complexity and burden of manually "bidding" for seats, each participant who needs a seat is equipped with an intelligent agent based on a reinforcement learning deep Q-network. When seating arrangements are needed, a dedicated decision agent is configured for each participant who needs a seat. The dedicated decision agent coordinates with the participants based on the data information corresponding to the participants in the personnel information demand tag database and the current global seating arrangement information to perform global seating arrangements. The current global seating arrangement information includes the remaining time, the competition intensity of a specific area, and the current highest match. The dedicated decision agent decides to compete for one or more seat packages based on the matching degree between the personnel who need to match seats and the seats. The cost is the abstract feature value of the personnel corresponding to the current dedicated decision agent for seat demand. The abstract feature value of seat demand includes the abstract feature expression of the matching percentage between the personnel information demand tag database and the seat package corresponding to the currently competing seats using one-hot encoding, or it may remain on the sidelines to select other seats. Set target rewards for dedicated decision-making agents. When a dedicated decision-making agent successfully obtains the seat with the highest matching degree, it receives the first positive reward. When the dedicated decision-making agent obtains a candidate seat, it receives a second positive reward; The dedicated decision-making agent receives a negative reward when it fails to acquire a seat or when the cost exceeds a preset maximum threshold. Through continuous learning, the agent can increasingly accurately simulate the user's true intentions, achieving automated and personalized bidding.

[0022] Managing the coordinated operation of the global seating arrangement of all dedicated decision-making agents includes: A central intelligent coordinating mechanism is established to receive competition requests from all dedicated decision-making agents, manage the entire competition lifecycle, and maintain the overall competition status.

[0023] The physical seating plan is divided into zones according to a preset seating competition objective, such as by the number of tables, by price, or by function. Each zone is assigned a coordinated intelligent management system, which manages the seating resources and real-time status of the corresponding seating zone. The seating resources include the seat numbers that have not been filled by competition, and the real-time status includes the vacancy, provisional allocation, and final locking status of each seat. Each seating zone is then used to construct a tree based on the seat combinations included in the current seating zone, using a disjoint-set data structure, to obtain a seating allocation lookup tree. When each person's dedicated decision agent makes a competitive request, the seat allocation search tree is first searched for the tag information of the target seat combination package. The number of times the dedicated decision agent finds the tag information of the seat combination package corresponding to the seat demand of all the persons is the highest cost that the dedicated decision agent can give. If the dedicated decision agent fails to find the target, a preset cost for failure is added to the sum of costs for successful searches to obtain the highest cost that the dedicated decision agent can provide.

[0024] Competition mechanisms include: When the seating arrangement is initiated, each person calculates the cost based on their respective dedicated decision-making agent, and obtains the highest cost and the target seat. The target seat is the seat number included in the seat combination package corresponding to the person's information demand tag. After obtaining the highest cost and the target seat, each dedicated decision-making agent competes to bid, and the competitive bid does not exceed the highest cost. The coordination strategies for intelligent management include: Step S201: The intelligent coordinator receives a new competitive bid and updates the provisional winner list with the new competitive bid and the personnel information feature value and seat number corresponding to the competitive bid. Step S202: Coordinate the intelligent management to perform incremental updates, handle conflict sets, and quickly update the provisional winner list; Step S203: The coordinated intelligent management team will provide real-time feedback of the latest provisional winner status to all participating dedicated decision-making agents; Step S204: The exclusive decision-making agent who makes the last and highest bid obtains the seat; The coordination strategy is repeated until no dedicated decision-making agent makes a bid; The final set of seat allocation schemes is obtained.

[0025] The seat allocation scheme is optimally adjusted based on the improved whale optimization algorithm. The adjustments include: Each seat allocation scheme is encoded as a "whale" position. The quality of each "whale" is judged by a fitness function. The fitness function aims to maximize the total amount of all accepted bids while penalizing any scheme that violates personnel constraints, thus ensuring the validity of the final result.

[0026] Traditional whale optimization algorithms rely on fixed internal parameters, which can cause them to get trapped in local optima prematurely in complex problems. Our coordinated intelligent management system monitors the search process of the whale optimization algorithm in real time. When the coordinated intelligent management system detects that the algorithm may be stagnating, such as when the optimal solution has not been updated for several generations, it dynamically adjusts the exploration and development parameters of the whale optimization algorithm through its reinforcement learning model. This guides the "whale swarm" out of local traps, thereby significantly improving the probability and efficiency of finding the global optimum.

[0027] The provisional winner selection logic includes: A provisional winner is determined based on a preset fixed time interval or the appearance of a new bid, and the provisional winner list is updated. When a new bid enters the competition, the system does not need to perform global optimization on all bids again. It determines whether there is a resource conflict between the new bid and the current provisional winner set. If there is a conflict, it does not update; if there is no conflict, it updates. Resource conflicts include: If there is no competition for the seat package involved in the new bid by the current exclusive decision agent, i.e. there is no conflict, the exclusive decision agent will be directly updated to the new provisional winner; If the current exclusive decision agent's new bid conflicts with the bids of one or more provisional winners, a conflict set consisting only of conflicting bids is constructed. A small-scale optimization calculation is performed using an improved whale optimization algorithm to determine the new provisional winner only within the conflict set. Once the optimal adjustment and optimization of the seat allocation scheme is completed, the current global provisional winner set is used as a high-quality initial solution. A complete and thorough global optimization is then performed using the improved whale optimization algorithm to ensure the global optimality of the final allocation scheme and obtain the optimal seat allocation scheme.

[0028] After obtaining the optimal seating allocation scheme, the final results are announced, seats are allocated to the successful bidders, and settlements are made. All data from the competition process is used for the retraining and iteration of the intelligent scheduling method for seating information.

[0029] In one embodiment, tickets are sold for popular concerts: 1. Users (or their agents) bid on combination packages such as "two adjacent seats in the first three rows of the inner field" and "family four-person package tickets in section A of the stands".

[0030] 2. After the sale begins, users can see in real time whether their bid is in the "provisional win" state.

[0031] 3. When another user bids higher, the original leader will immediately receive a notification that "your bid has been surpassed" and can choose whether to raise the bid.

[0032] 4. The entire ticketing process becomes a dynamic and transparent game, with the final allocation results announced by the system at the end of the ticketing period. This approach not only maximizes box office revenue through price discovery mechanisms but is also highly favored by users for its fairness and interactivity.

[0033] This method transforms physical seating resources into auctionable seat packages, allowing participants (such as passengers and spectators) to express complex preferences (e.g., multiple seats together, specific areas) through bidding. It introduces a real-time incremental auction mechanism, providing users with real-time feedback on their bidding status during the auction process. The core of this method utilizes a multi-agent system (MAS) for distributed collaboration, leveraging reinforcement learning-enabled user agents to achieve intelligent bidding and system strategy self-optimization. An improved whale optimization algorithm serves as the core engine, efficiently solving complex winner-determination problems. Ultimately, it constructs an intelligent scheduling system capable of real-time response, automatic decision-making, and maximizing overall revenue and user satisfaction. This method creatively integrates multi-agent collaboration, reinforcement learning dynamic strategies, and an innovative real-time incremental combinatorial auction mechanism, providing a comprehensive, efficient, and superior solution to the seating resource scheduling problem in modern service industries. It not only overcomes the computational bottlenecks of traditional combinatorial auctions but also transforms resource allocation from a static background task into a dynamic, transparent, and interactive intelligent service process through intelligence and real-time processing. By employing multi-agent partitioning management and incremental computing, the computational complexity of large-scale problems is effectively decomposed, enabling support for large-scale scenarios. The introduction of reinforcement learning allows the system to automatically learn user behavior and market dynamics, autonomously optimizing bidding and auction strategies to adapt to demand fluctuations. The real-time incremental auction mechanism provides instant feedback, transforming traditional "blind auctions" into transparent, interactive, and dynamic games, significantly enhancing user engagement. The combined auction format satisfies users' personalized and complex preferences, while the intelligent optimization engine ensures that resource value is maximized while meeting these preferences.

[0034] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media containing computer-usable program code. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0035] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for intelligent scheduling of seating information, characterized in that, Includes the following steps: Step S1: Collect seat arrangement scene information and seat arrangement personnel information, and transform the seat arrangement scene information and seat arrangement personnel information into features through a neural network to obtain a seat combination package database and personnel information demand tags; Step S2: When seating arrangement is required, assign a dedicated decision agent to each participant who needs a seat, manage the coordination of all dedicated decision agents in the global seating arrangement, and output a set of seat allocation schemes. Step S3: Based on the improved whale optimization algorithm, the seat allocation scheme is optimally adjusted, and the optimal seat allocation scheme is output.

2. The intelligent scheduling method for seating information as described in claim 1, characterized in that, Step S1 includes: The seating arrangement scenario information includes a physical seating floor plan and a seating arrangement scenario theme. The seating arrangement personnel information includes the number of people that can be accommodated, basic personnel entry information, and personnel constraints. The basic personnel entry information includes personnel ID, reserved seat type, and seat preference. The personnel constraints include custom entry conditions, special entry channels for off-site staff, and staff members in their assigned positions. The physical seating plan is converted into a set of seat numbers and corresponding tradable attribute tags using image conversion software. Adjacent or identical tradable attributes are dynamically generated into seat combination packages. The tradable attributes include window seats, emergency exit seats, regular seats, adjacent seats (n), accessible seats, main seats, and VIP areas. Each seat combination package corresponds to a subset of different tradable attribute tags, where n is a natural number greater than 0.

3. The intelligent scheduling method for seating information as described in claim 2, characterized in that, The seating arrangement scenario information and seating arrangement personnel information are processed through an embedding layer to extract feature values ​​for the seating arrangement scenario information and seating arrangement personnel information, respectively. These feature values ​​are then mapped to the seating combination package database and the personnel information requirement tag database.

4. The intelligent scheduling method for seating information as described in claim 3, characterized in that, Step S2 includes: When seating arrangements are needed, a dedicated decision agent is configured for each participant who needs a seat. The dedicated decision agent coordinates with the participants based on the data information corresponding to the participants in the personnel information demand tag database and the current global seating arrangement information to perform global seating arrangements. The current global seating arrangement information includes the remaining time, the competition intensity of a specific area, and the current highest match. The dedicated decision agent decides to compete for one or more seat packages based on the matching degree between the personnel who need to match seats and the seats. The cost is the abstract feature value of the personnel's seat demand corresponding to the current dedicated decision agent. The abstract feature value of seat demand includes the abstract feature expression of the matching percentage between the personnel information demand tag database and the seat package corresponding to the currently competing seats through one-hot encoding, or to remain on the sidelines. Set target rewards for dedicated decision-making agents. When a dedicated decision-making agent successfully obtains the seat with the highest matching degree, it receives the first positive reward. When the dedicated decision-making agent obtains a candidate seat, it receives a second positive reward; When the dedicated decision-making agent fails to acquire a seat or the cost exceeds a preset maximum threshold, it receives a negative reward.

5. The intelligent scheduling method for seating information as described in claim 4, characterized in that, Managing the coordinated operation of the global seating arrangement of all dedicated decision-making agents includes: The physical seating plan is divided into zones according to a preset seating competition objective, such as by the number of tables, by price, or by function. Each zone is assigned a coordinated intelligent management system, which manages the seating resources and real-time status of the corresponding seating zone. The seating resources include the seat numbers that have not been filled by competition, and the real-time status includes the vacancy, provisional allocation, and final lock status of each seat. Each seating zone is then used to construct a tree based on the seat combinations included in the current seating zone, using a disjoint-set data structure, to obtain a seating allocation lookup tree. When each person's dedicated decision agent makes a competitive request, the seat allocation search tree is first searched for the tag information of the target seat combination package. The number of times the dedicated decision agent finds the tag information of the seat combination package corresponding to the seat demand of all the persons is the highest cost that the dedicated decision agent can give. If the dedicated decision agent fails to find the target, a preset cost for failure is added to the sum of costs for successful searches to obtain the highest cost that the dedicated decision agent can provide.

6. The intelligent scheduling method for seating information as described in claim 5, characterized in that, Competition mechanisms include: When the seating arrangement is initiated, each person calculates the cost based on their respective dedicated decision-making agent, and obtains the highest cost and the target seat. The target seat is the seat number included in the seat combination package corresponding to the person's information demand tag. After obtaining the highest cost and the target seat, each dedicated decision-making agent competes to bid, and the competitive bid does not exceed the highest cost. The coordination strategies for intelligent management include: Step S201: The intelligent coordinator receives a new competitive bid and updates the provisional winner list with the new competitive bid and the personnel information feature value and seat number corresponding to the competitive bid. Step S202: Coordinate intelligent management to perform incremental updates, handle conflict sets, and quickly update the provisional winner list; Step S203: The coordinated intelligent management team will provide real-time feedback of the latest provisional winner status to all participating dedicated decision-making agents; Step S204: The exclusive decision-making agent who makes the last and highest bid obtains the seat; The coordination strategy is repeated until no dedicated decision-making agent makes a bid; The final set of seat allocation schemes is obtained.

7. The intelligent scheduling method for seating information as described in claim 6, characterized in that, The seat allocation scheme is optimally adjusted based on the improved whale optimization algorithm. The adjustments include: Each seat allocation scheme is encoded as a "whale" position. The quality of each "whale" is judged by a fitness function, which aims to maximize the total amount of all accepted bids while penalizing any scheme that violates personnel constraints.

8. The intelligent scheduling method for seating information as described in claim 7, characterized in that, The logic for determining the provisional winner includes: A provisional winner is determined based on a preset fixed time interval or the appearance of a new bid, and the provisional winner list is updated. When a new bid enters the competition, it is determined whether the new bid conflicts with the current provisional winner set. If a conflict exists, no update is made; otherwise, an update is made. Resource conflicts include: If there is no competition for the seat package involved in the new bid by the current exclusive decision agent, i.e. there is no conflict, the exclusive decision agent will be directly updated to the new provisional winner; If the current exclusive decision agent's new bid conflicts with the bids of one or more provisional winners, a conflict set consisting only of conflicting bids is constructed. A small-scale optimization calculation is then performed using an improved whale optimization algorithm to determine the new provisional winner only within the conflict set.

9. The intelligent scheduling method for seating information as described in claim 8, characterized in that, Once the optimal adjustment and optimization of the seat allocation scheme is completed, the current global provisional winner set is used as the initial solution, and the improved whale optimization algorithm is used for global optimization to obtain the optimal seat allocation scheme.

10. The intelligent scheduling method for seating information as described in claim 1, characterized in that, After obtaining the optimal seating allocation scheme, the final results are announced, seats are allocated to the successful bidders, and settlements are made. All data from the competition process is used for the retraining and iteration of the intelligent scheduling method for seating information.

Images