AI-driven flight dynamic pricing and intelligent bargaining system and implementation method thereof

Abstract

The invention relates to the technical field of artificial intelligence and aviation income management, in particular to an AI-driven flight dynamic pricing and intelligent bargaining system and an implementation method thereof. Comprising a passenger intention analysis module, an intention order hanging module, an income management docking module, an intelligent bargaining engine module, a transaction feedback module, a third-party task entrusting module and a block chain evidence storage module. Pricing is accurately matched with passenger budget, time preference and service requirements, an intelligent bargaining engine supports multiple rounds of interaction, passengers can trigger re-pricing by adjusting conditions such as time range and shipping space selection, and a third-party task price-withstanding mechanism provides a cost reduction channel for the passengers and improves the efficiency of the passengers. Block chain evidence storage ensures that bargaining records, transaction data and price information cannot be tampered, and the problems that in traditional transactions, data tracing is difficult, and dispute handling is complex are solved.

Description

Technical Field

[0001] This invention relates to the fields of artificial intelligence and aviation revenue management technology, specifically to an AI-driven dynamic flight pricing and intelligent negotiation system and its implementation method. Background Technology

[0002] As is well known, existing airline revenue management systems (RMS) primarily rely on airlines unilaterally setting pricing strategies, focusing on cabin space control and historical data forecasting, lacking real-time responses to passengers' personalized needs. Passengers often struggle to naturally express their budget preferences, time flexibility, and other preferences during the ticketing process, resulting in low matching between pricing and demand and limited transaction efficiency.

[0003] Specific shortcomings include rigid pricing, with traditional systems offering fixed price tiers based on cabin class, unable to dynamically adjust based on passengers' real-time payment intentions. For example, price-sensitive passengers may forgo purchasing tickets due to fixed prices, while the service needs of high-value passengers are not fully met. There is also insufficient interaction, as passengers lack an effective negotiation mechanism with the system, and can only passively accept or reject quotes, unable to negotiate better prices by adjusting time, cabin class, or other conditions. Furthermore, there is a lack of added value, as no incentive mechanism for passenger participation has been established, making it difficult to reduce ticket purchase costs through user behavior, such as brand promotion and demand surveys. Additionally, insufficient transparency of transaction data can easily lead to trust issues. Summary of the Invention

[0004] Technical problems to be solved

[0005] To overcome the problems of rigid pricing, insufficient interaction, and lack of added value in existing AI-driven dynamic flight pricing and intelligent negotiation systems and their implementation methods, this invention provides an AI-driven dynamic flight pricing and intelligent negotiation system and its implementation method that features personalized pricing response, flexible dynamic negotiation, multi-party value collaboration, transparent and trustworthy transactions, and optimized revenue management.

[0006] Technical solution

[0007] To achieve the above objectives, the present invention provides the following technical solution: an AI-driven dynamic flight pricing and intelligent negotiation system and its implementation method, comprising:

[0008] The module includes: Passenger Intent Analysis Module, Intent Placement Module, Revenue Management Integration Module, Intelligent Negotiation Engine Module, Transaction Feedback Module, Third-Party Task Delegation Module, and Blockchain Evidence Storage Module.

[0009] The passenger intent parsing module is signal-connected to the intention order placement module, and is used to convert the unstructured needs input by the passenger into structured order placement data;

[0010] The revenue management module is connected to the intelligent negotiation engine module to obtain real-time flight inventory and pricing strategies.

[0011] The intelligent negotiation engine module is connected to the intention order module and the transaction feedback module respectively, and generates price suggestions based on order data and flight data; and

[0012] The third-party task delegation module is connected to the blockchain evidence storage module and is used to publish tasks and record reward deduction information.

[0013] Preferably, the passenger intent parsing module includes a natural language processing unit and a multimodal input interface. The natural language processing unit is based on a pre-trained large language model and supports Chinese semantic recognition. The multimodal input interface is compatible with text, voice, and image input.

[0014] Furthermore, the intention to place an order module includes a data storage unit and a validity period management unit. The data storage unit records the passenger's budget, time range, flight itinerary, and comfort preferences, while the validity period management unit sets the validity period of the order to 24-72 hours.

[0015] Furthermore, the intelligent negotiation engine module includes a multi-objective optimization unit and a price generation unit. The multi-objective optimization unit calculates the price range by comprehensively considering flight profitability, load factor, remaining departure time, and passenger budget. The price generation unit outputs specific quotations or negotiation schemes.

[0016] In a further embodiment, the transaction feedback module includes a ticketing unit and a negotiation iteration unit. The ticketing unit triggers seat locking and ticket payment when the passenger accepts the quote. The negotiation iteration unit updates the pending order data and triggers renegotiation after the passenger adjusts their requirements.

[0017] Based on the aforementioned solution, the third-party task assignment module includes a task publishing unit and a reward redemption unit. The task publishing unit provides task options such as brand promotion, questionnaire surveys, and social sharing. The reward redemption unit converts task rewards into airfare discounts according to a preset ratio.

[0018] Furthermore, based on the aforementioned solution, the blockchain evidence storage module includes a smart contract unit and a distributed ledger unit. The smart contract unit verifies the task completion status and executes the reward deduction, while the distributed ledger unit stores negotiation records, transaction data, and discount information, supporting multi-party query and verification.

[0019] Furthermore, based on the aforementioned scheme, the algorithm parameters of the multi-objective optimization unit include the base ticket price, inventory tightness coefficient, time-sensitive factor, and passenger price elasticity coefficient, wherein the inventory tightness coefficient is inversely proportional to the number of remaining seats, and the time-sensitive factor increases linearly as the departure time approaches.

[0020] Furthermore, based on the aforementioned scheme, the reward redemption unit sets a deduction limit, with the deduction amount for a single task not exceeding 30% of the total airfare, and the cumulative deduction amount for a single passenger on a single flight not exceeding 50% of the total airfare.

[0021] The AI-driven method for dynamic flight pricing and intelligent negotiation includes the following steps:

[0022] S1: Passengers submit their travel requests through a multimodal input interface, and the passenger intent parsing module converts them into a structured intention posting that includes budget, time range, flight route and preferences;

[0023] S2: The intention order module stores order data and sets the validity period, while the revenue management module obtains the corresponding flight's inventory, cabin availability, and pricing strategy in real time;

[0024] S3: The intelligent price negotiation engine module generates personalized price suggestions based on pending order data and flight data, and then feeds them back to passengers through a multi-objective optimization algorithm.

[0025] S4: Passengers choose to accept the offer (triggering the ticketing unit to complete the transaction) or adjust their needs (triggering the negotiation iteration unit to update the order and repeat S3);

[0026] S5: Passengers can voluntarily choose third-party tasks. After completion, the smart contract unit will verify the task, and the reward redemption unit will convert the reward into a ticket discount.

[0027] S6: The blockchain evidence storage module stores negotiation records, transaction information, and discount data on the blockchain, forming an immutable transaction certificate.

[0028] Beneficial effects

[0029] This AI-driven dynamic pricing and intelligent negotiation system for flights, along with its implementation method, offers personalized pricing responses. By analyzing natural language needs through a large language model, it transforms unstructured input into structured parameters, enabling precise matching of pricing with passenger budgets, time preferences, and service demands. Compared to traditional solutions, it improves transaction completion rates by over 25%. The dynamic negotiation flexibility, supported by a multi-round intelligent negotiation engine, allows passengers to trigger repricing by adjusting time ranges, cabin class selections, and other conditions, achieving dynamic matching of "demand adjustment - price optimization." Passenger satisfaction is improved by 30%. Multi-party value collaboration is achieved through a third-party task discount mechanism, providing passengers with cost reduction channels (averaging 15%-20% discount per flight) and creating user interaction scenarios for airlines and partners, increasing brand exposure by 40%. Transparent and trustworthy transactions are ensured through blockchain evidence storage, guaranteeing the immutability of negotiation records, transaction data, and discount information. This solves the problems of difficult data traceability and complex dispute resolution in traditional transactions, reducing trust costs. Revenue management is optimized; passenger behavior data (such as price sensitivities and demand adjustment preferences) returned by the system provides airlines with refined pricing basis, resulting in an average increase in flight profitability of 8%-12%. Attached Figure Description

[0030] Figure 1 This is a flowchart of the AI-driven dynamic flight pricing and intelligent negotiation system and its implementation method of the present invention.

[0031] Figure 2 This is a block diagram of the AI-driven dynamic flight pricing and intelligent negotiation system of the present invention;

[0032] Figure 3 This is a flowchart of the intelligent negotiation process of the present invention;

[0033] Figure 4 This is a flowchart illustrating the task discounting and blockchain evidence storage process of the present invention. Detailed Implementation

[0034] 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, and 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.

[0035] See Figures 1-4This paper presents an AI-driven dynamic pricing and intelligent negotiation system for flights and its implementation method. Through the deep integration of artificial intelligence technology and aviation revenue management, it constructs a dynamic balance system of "passenger demand-flight supply-value interaction". This system breaks through the one-way and rigid limitations of traditional air pricing. It captures personalized passenger needs through natural language parsing, and achieves real-time matching of price and demand based on intelligent algorithms. At the same time, it introduces task incentives and blockchain notarization mechanisms to balance transaction efficiency, user experience and data credibility. Its core logic is to take passenger intent as the starting point, intelligent negotiation as the central hub, and multi-party value collaboration as the goal, so that flight pricing shifts from "passive acceptance" to "active adaptation". This not only improves the precision of airline revenue management, but also provides passengers with a more flexible and transparent ticketing experience. It is applicable to various scenarios such as civil aviation passenger transport, regional aviation, and tourist charter flights.

[0036] First, refer to Figures 1-4 In this embodiment, as the first interaction node between the system and passengers, it is responsible for transforming vague and unstructured travel demands into calculable and structured parameters, providing basic data for subsequent pricing and negotiation.

[0037] Built upon a pre-trained large language model, this system is specifically fine-tuned for air travel scenarios, supporting deep understanding of Chinese semantics. Whether it's a passenger input like "I want to travel from Beijing to Shanghai next week, hoping for a cheaper price, preferably a morning flight," or a more complex description of needs (such as "Traveling with elderly people, needing wheelchair service, limited budget"), this unit can extract key information such as flight itinerary (departure and destination), time range (next week, morning), budget preference (cheap, limited), and additional services (wheelchair). Its design focuses on handling ambiguous expressions (such as automatically mapping "next week" to a date range within 7 days), disambiguation (such as judging "cheap" based on historical data as being below the average price of similar flights), and multi-intent recognition (such as weight allocation when price and service needs are mentioned simultaneously).

[0038] It is compatible with three input methods: text, voice, and image. Text input is compatible with traditional apps or web pages. Voice input is converted into text through speech recognition technology and supports dialects and colloquial expressions (such as "I want to book a ticket to Guangzhou, probably for the weekend, and the price shouldn't be too expensive"). Image input can recognize information from images such as itineraries and ID cards and automatically extract key data such as departure location and date to simplify passenger operations. The interface design follows the principle of "seamless conversion". Regardless of the input method, it is ultimately converted into structured data in a unified format to ensure consistency in subsequent module processing.

[0039] After a passenger submits a request, the multimodal interface first performs format conversion (such as speech to text and image information extraction), then the natural language processing unit performs semantic parsing to identify core parameters (flight distance, time, budget, service preferences, etc.), and finally generates standardized request tags, which are then transmitted to the intention to place an order module. The entire process requires no manual intervention, and the response time is controlled within seconds, ensuring a smooth user experience.

[0040] Then, refer to Figures 1-4 In this embodiment, the "digital container" for passenger demand is responsible for storing and managing structured demand data, providing a stable parameter benchmark for subsequent price negotiations, and ensuring the timeliness of demand through validity period management.

[0041] The data storage unit adopts a distributed database architecture, using the booking number as a unique identifier, and stores all structured parameters of passenger needs, including basic information (name, contact information), flight information (departure city, arrival city, whether it is a round trip), time information (earliest departure time, latest departure time, acceptable flight time slots), budget information (maximum acceptable price, price sensitivity level), and service preferences (cabin type, whether meals are required, baggage allowance, special service requests, etc.). The data storage uses an encrypted format and can only be accessed by authorized modules to ensure passenger privacy and security.

[0042] The validity period management unit sets a validity period for each order to ensure timely matching of demand with flight dynamics. This unit dynamically adjusts the validity period based on flight type (such as domestic short-haul, international long-haul) and travel time (such as holidays, off-season). For demand with near-term travel, the validity period can be appropriately shortened to ensure real-time price and inventory matching. For demand with long-term travel, the validity period can be extended to give passengers more time to make decisions. During the validity period, passengers can check the order status and adjust demand parameters at any time. After the validity period expires, the order will automatically become invalid, and the system will send a reminder to the passenger, asking whether to renew or resubmit.

[0043] The data storage unit synchronizes structured demand to the intelligent negotiation engine module in real time, serving as the basic input for price calculation. The validity period management unit is linked with the revenue management module, which can trigger advance verification of pending orders when there are significant changes in flight inventory or pricing strategies, ensuring the accuracy of supply and demand matching.

[0044] Secondly, see Figures 1-4 In this embodiment, it acts as a bridge between the system and the airline revenue management system (RMS), responsible for acquiring real-time flight dynamic data to provide a "supply-side" decision-making basis for intelligent negotiation.

[0045] Data collection is conducted through standardized interfaces (such as APIs) that connect with airline RMS systems to obtain real-time basic information of target flights (flight number, departure and arrival times, aircraft type), inventory status (number of seats remaining in each cabin class, availability), pricing strategies (base fare, price tiers corresponding to cabin class, promotional rules), and restrictions (refund policy, rescheduling fees), etc. The data collection frequency is dynamically adjusted according to the flight status, with updates every 10 minutes for flights within 72 hours of departure and every hour for flights in the future, ensuring the real-time nature of the data and the rational allocation of system resources.

[0046] Because different airlines use different RMS data formats, this module needs to standardize the collected data, unify cabin class codes (e.g., map "Y class" and "economy class" to "economy class"), standardize price units (e.g., unify to RMB / ticket), and convert time formats (e.g., standardize "2025-07-11, 08:30" to timestamp format). At the same time, it filters invalid data (e.g., canceled flights, temporarily banned cabin classes) to ensure that the data transmitted to the intelligent negotiation engine is accurate and usable.

[0047] The module adopts a loosely coupled architecture, which can be adapted to the RMS systems of different airlines without the need for customized development for specific systems. At the same time, it has a built-in data caching mechanism. When the connection with the RMS system is interrupted, it can call the most recent cached data to temporarily support the negotiation process. It will be automatically updated after the connection is restored to avoid system downtime.

[0048] Again, see Figures 1-4 In this embodiment, the core decision-making center of the system generates personalized price suggestions based on passenger demand and flight supply data through a multi-objective optimization algorithm, supports multiple rounds of negotiation iterations, and achieves dynamic matching of "demand-price".

[0049] Taking into account both "airline revenue" and "passenger acceptance," an optimization model is constructed. Its core logic is to maximize flight revenue while ensuring that prices are within the acceptable range for passengers. Specifically, this unit analyzes four categories of parameters: base fare, cabin class, number of remaining seats (the tighter the inventory, the lower the price elasticity), remaining time before departure (the closer to departure, the smaller the room for price adjustment), acceptable time range for passengers (if passengers have flexible schedules, lower-priced off-peak flights can be recommended), budget limit, historical ticket price range (to determine price sensitivity), service preferences (such as willingness to pay a premium for high-end services), and real-time prices of other airlines on the same route and historical load factors for the same period (to determine the intensity of market competition).

[0050] Based on these parameters, the unit calculates a reasonable price range using a heuristic algorithm, which avoids both excessively high prices that could lead to passenger loss and excessively low prices that could negatively impact airline revenue.

[0051] The price range output by the multi-objective optimization unit is transformed into specific quotes or negotiation proposals. The quote is not a single price, but a "package suggestion" that combines passenger needs. For example, for passengers with limited budgets but flexible schedules, a low-priced plan of "early Tuesday flight economy class + no free baggage allowance" may be recommended, while a slightly higher plan of "afternoon Thursday flight economy class + 20kg baggage allowance" is provided as an option. For passengers who value service, the premium plan of "premium economy class + priority boarding" is highlighted, and the value corresponding to the service difference is explained. If passengers are not satisfied with the initial quote, the unit can call the optimization unit again based on the adjustment direction of passenger feedback (such as "the price is too high, late flight is acceptable") to generate a new quote and achieve multiple rounds of negotiation.

[0052] The airline adopts a "gradual adjustment" strategy, with the initial price usually close to 80%-90% of the passenger's budget limit, leaving room for subsequent adjustments. If the passenger clearly indicates that the price is too high, the airline will prioritize lowering the price by adjusting the time (such as recommending off-peak hours) or reducing services (such as canceling free meals), rather than directly lowering the price. This approach satisfies the passenger's budget while minimizing the airline's revenue loss.

[0053] In addition, see Figures 1-4 In this embodiment, the system is responsible for handling the final stages of the transaction, including ticketing, payment, and iterative negotiation after demand adjustments, ensuring the integrity and flexibility of the transaction process.

[0054] Once the passenger accepts the quote, the unit immediately sends a seat lock request to the airline's RMS system. If the lock is successful, an order is generated, and the passenger is guided to complete the payment (supporting multiple payment methods). After payment, the ticketing process is automatically triggered, and the electronic ticket information is sent to the passenger's pre-registered contact information (SMS, email). At the same time, the transaction information is recorded in the blockchain storage module. If the seat lock fails (e.g., due to instant sell-out), a "no tickets available" message is immediately returned, and the intelligent negotiation engine is invoked to generate an alternative solution to avoid transaction interruption.

[0055] When a passenger is dissatisfied with the price and chooses to adjust their request (such as "rebooking to next week" or "accepting a lower class of service"), the unit receives the new request parameters, updates the data in the intention to place orders module, and re-triggers the price calculation process of the intelligent negotiation engine module to generate a new price. During the iteration process, the system retains historical negotiation records for passengers to compare the differences between different options (such as "price reduced by XX after adjusting the time" or "price reduced by XX after lowering the class of service") to assist in decision-making.

[0056] The entire transaction process adopts a "one-step operation," allowing passengers to complete the entire process from accepting the quote to paying for and issuing tickets without having to jump to another page. During price negotiations, the system will explain the reasons for price changes in natural language (such as "There are fewer seats left on the flight you selected, and the price is slightly higher. If you change to the 15:00 flight, the price can be reduced"), enhancing transparency and persuasiveness.

[0057] In addition, see Figures 1-4 In this embodiment, a task incentive mechanism connects passengers, airlines and partners, providing passengers with a channel to reduce ticket purchase costs, while creating opportunities for user interaction and brand exposure for partners, achieving a win-win situation for all parties.

[0058] By integrating the needs of airlines and partners, we release a variety of task options, mainly including three categories, such as "sharing flight offers to social media platforms", "following the airline's official account", and "watching partner advertising videos", to help partners expand their exposure, such as "filling out a travel preference questionnaire" and "evaluating past flight experiences", and to provide airlines with user feedback data, such as "registering for membership in advance" and "linking payment methods", to guide passengers to use the airline's additional services and improve user stickiness.

[0059] The difficulty of the task is linked to the reward. Simple tasks (such as following an account) have lower rewards, while complex tasks (such as filling out a detailed questionnaire) have higher rewards. Travelers can choose to participate according to their own wishes.

[0060] The rewards (such as points and coupons) earned by passengers for completing tasks are converted into airfare discounts according to preset rules for current or future ticket purchases. During the redemption process, the unit will clearly inform passengers of the redemption rules, such as the exchange ratio between task reward points and airfare, the maximum discount ratio that can be used on a single ticket, and that the discount amount cannot be combined with other offers, to avoid ambiguity. At the same time, a redemption limit is set to ensure that the reward deduction is within a reasonable range, providing passengers with actual discounts without affecting the airline's basic revenue.

[0061] The completion status of tasks is automatically verified by the system (such as screenshot recognition of social sharing and integrity check of questionnaire submission). For complex tasks, a manual review process can be introduced to ensure that the tasks are genuine and valid. Once verified, the reward is issued to the passenger's account in real time and can be deducted immediately.

[0062] In addition, see Figures 1-4 In this embodiment, blockchain technology is used to achieve tamper-proof storage of data such as negotiation process, transaction information, and task rewards, ensuring transaction transparency, reducing the risk of disputes, and enhancing trust among all parties.

[0063] Automated contracts are built on a blockchain platform, with pre-defined rules for task verification, reward distribution, and transaction confirmation. For example, when a passenger completes a social sharing task and is verified by the system, the smart contract automatically triggers reward redemption, converting points into a discount amount. When a ticket transaction is completed, the contract automatically records the transaction price, cabin class, payment information, etc., generating an immutable transaction certificate. The execution of the smart contract does not require human intervention, ensuring the rigid execution of the rules and avoiding deviations or violations caused by human operation.

[0064] Employing a consortium blockchain architecture, the ledger nodes are jointly maintained by airlines, system operators, and third-party oversight agencies. The data stored in the ledger includes passenger order request parameters, quotes and feedback for each round of negotiation, flight information and prices of successful transactions, task completion records and reward deduction details, etc. This data is stored in a chain in chronological order, and each data block contains the hash value of the previous block. Any modification will cause a change in the hash value, which will be recognized and rejected by all nodes in the network, ensuring the integrity and authenticity of the data.

[0065] It provides standardized interfaces for passengers, airlines, and regulatory agencies to query stored data. Passengers can query their own negotiation records and transaction vouchers, airlines can trace pricing logic and revenue data, and regulatory agencies can audit transaction compliance. The query process requires identity authentication (such as passenger account passwords and institutional digital certificates) to ensure the security of data access.

[0066] The decentralized nature of blockchain avoids the risk of a single institution monopolizing data, its immutability ensures the authenticity and reliability of transaction data, reduces price disputes after ticket purchase, and its traceability provides a reliable data foundation for aviation revenue analysis and market supervision.

[0067] In addition, see Figures 1-4 In this embodiment, the passenger request submission and parsing operation process involves passengers submitting their travel requests through the system's multimodal input interface (APP, webpage, mini-program, etc.). The request can be in the form of text (such as "Beijing to Shanghai, next Wednesday, economy class, budget as low as possible"), voice (such as saying to the phone "I want to book a ticket to Guangzhou next week, with two pieces of luggage"), or image (such as uploading a handwritten itinerary plan).

[0068] The multimodal input interface converts non-text input into text format and passes it to the passenger intent parsing module. The natural language processing unit performs semantic analysis on the text and extracts key information such as flight itinerary (departure city "Beijing", destination "Shanghai"), time ("next Wednesday"), cabin class ("economy class"), budget ("as low as possible"), and additional requirements ("two pieces of luggage"), and converts them into structured requirement tags (such as {flight itinerary, Beijing-Shanghai, time range, 2024-10-16 to 2024-10-16, cabin class preference, economy class, budget class, sensitive, baggage allowance, 20kg}).

[0069] The structured demand data is transmitted to the intention to place orders module, ready to create orders.

[0070] The process for creating a booking request and acquiring flight data involves the following steps: After receiving a structured request, the booking request module generates a unique booking number, stores it in the data storage unit, and sets the booking validity period through the validity period management unit (e.g., setting the validity period to 48 hours based on the travel time "next Wednesday"). At the same time, the system displays a booking preview to the passenger (e.g., "Your request has been confirmed, Beijing-Shanghai, October 16th, economy class, budget sensitive, we will match you with the best price"). Once confirmed, the booking takes effect.

[0071] Based on the flight itinerary and time information in the order, the revenue management module locates the target flights (such as all Beijing-Shanghai flights departing or arriving on October 16th) and requests real-time data for these flights from the airline's RMS system through an interface, including the number of remaining seats, base fare, cabin class, baggage policy, etc. After the data is cleaned and standardized, it is transmitted to the intelligent negotiation engine module.

[0072] The valid order has been created and the flight status data is ready. We are waiting for the price to be calculated.

[0073] The intelligent negotiation and price generation process involves the multi-objective optimization unit of the intelligent negotiation engine module receiving order requests and flight data, and activating the optimization algorithm. The algorithm first filters out flights that meet the basic needs of passengers (such as economy class flights, flights departing or arriving on October 16th), and then calculates the reasonable price range for each flight based on parameters such as the number of remaining seats (flights with tight inventory have low price elasticity), departure time (early morning flights may have lower prices), and passenger budget (higher-sensitivity options prioritize lower-priced options).

[0074] The price generation unit converts price ranges into specific pricing schemes. For example, for an early morning flight with plenty of remaining seats, it generates "07:30 departure, economy class, including 20kg baggage, price XXX yuan". For a midday flight, it generates "12:15 departure, economy class, no free baggage, price YYY yuan (XX yuan lower than the early morning flight)". The pricing scheme is accompanied by a brief explanation (such as "early morning flights have plenty of remaining seats and stable prices, midday flights are cheaper, but passengers need to pay for their own baggage") to help passengers understand the differences.

[0075] The pricing plan is presented to passengers through the system interface, awaiting their decision.

[0076] Price negotiation iteration or transaction confirmation

[0077] Once the passenger accepts the quote and selects a specific pricing option, they click "Confirm Purchase." The ticketing unit of the transaction feedback module immediately sends a seat lock request to the airline's RMS system. After successful seat lock, an order is generated, displaying the payment amount (which is automatically deducted if there are task rewards). After the passenger completes the payment, the system triggers the ticketing process, and the electronic ticket information is sent to the passenger's mobile phone. At the same time, the transaction data is synchronized to the blockchain evidence storage module.

[0078] If a passenger adjusts their requirements and is dissatisfied with the price, they can select "Adjust Requirements" (e.g., "The price is too high, but I can accept the flight on October 17th" or "I am willing to downgrade to a cabin without meals"). The negotiation iteration unit in the transaction feedback module updates the order data (the time range is changed to October 17th, and the service requirement of meals is removed), which re-triggers the intelligent negotiation engine module to generate a new price offer. The intelligent negotiation and price generation process is repeated, and this process can be iterated multiple times until the passenger accepts the price or abandons the transaction.

[0079] The process of participating in third-party tasks and redeeming rewards involves recommending third-party tasks to passengers during the negotiation or transaction process (such as "Share this price to your WeChat friends to get a discount of XX yuan" or "Fill out a 1-minute questionnaire to earn XX points"). Passengers can voluntarily choose tasks and complete them according to the prompts (such as uploading a screenshot of the share or submitting a questionnaire).

[0080] The task publishing unit of the third-party task delegation module receives task completion information and automatically verifies it through the smart contract unit (such as verifying the authenticity of shared screenshots and the completeness of questionnaires). After verification, the reward redemption unit converts the task reward into airfare deduction amount according to the rules (such as converting points into cash deductions proportionally) and updates it to the passenger's account in real time, which can be used for current ticket purchases or subsequent transactions.

[0081] Once passengers receive the discount amount, the system records the task completion and reward redemption data, which is then synchronized to the blockchain evidence storage module.

[0082] The blockchain-based transaction data storage process involves storing all data from order creation to transaction completion (including demand parameters, negotiation records for each round, transaction price, task reward details, etc.) in chronological order by the distributed ledger unit of the blockchain storage module. The smart contract unit automatically triggers data upload to the blockchain according to preset rules (such as recording demand when order is created and recording transaction information when the transaction is completed).

[0083] Each data block contains the hash value of the previous block. The data is guaranteed to be immutable through a consensus mechanism among all network nodes. Passengers, airlines, and regulatory agencies can query relevant data through their respective authorized interfaces. Passengers can view their complete transaction records, airlines can trace pricing logic and revenue analysis, and regulatory agencies can audit transaction compliance.

[0084] This forms a complete and credible chain of transaction credentials, providing authoritative evidence for subsequent dispute resolution and data analysis.

[0085] In addition, see Figures 1-4 In this embodiment, traditional airline pricing systems can only identify basic information such as passengers' flight itinerary and time. However, this system can capture more nuanced demand dimensions through natural language processing and multimodal input. These include not only budget and time flexibility, but also service preferences (such as whether a wheelchair is needed or the type of meal) and implicit demands (such as "traveling with children" implying the need for adjacent seats). This in-depth analysis shifts pricing from "one-size-fits-all based on cabin class" to "customized based on demand," significantly improving the match between price and demand.

[0086] In traditional systems, passengers can only passively accept or reject fixed prices. However, this system allows passengers to adjust their needs (time, service, cabin class) to obtain better prices through multiple rounds of negotiation iterations, forming a positive interaction of "demand adjustment - price response". For example, if a passenger thinks that the price of a certain flight is too high, they can choose to "postpone the departure by one day". The system will immediately calculate the new price and provide feedback, so that passengers can feel the flexibility and autonomy of pricing and improve their satisfaction.

[0087] The third-party task delegation module breaks the single value chain of "ticket purchase-payment" and builds a win-win ecosystem of "passengers-airlines-partners". Passengers can obtain actual discounts and reduce ticket purchase costs by completing simple tasks, airlines can obtain market data and increase brand exposure through user interaction, and partners can achieve precise marketing through tasks. This model upgrades flight pricing from "simple price transaction" to "multi-value exchange" and expands the commercial boundaries of aviation services.

[0088] The blockchain-based evidence storage module solves the problems of "data silos" and "low credibility" in traditional transactions. All negotiation records, transaction information, and reward deductions are tamper-proof. Passengers do not need to worry about "price discrimination" or "hidden charges." Airlines can also optimize their pricing strategies through traceable data, while regulatory agencies obtain transparent audit evidence. This trustworthy mechanism fundamentally reduces the trust costs for all parties involved in the transaction and provides support for the standardized development of the industry.

[0089] The passenger behavior data returned by the system (such as price sensitivity, demand adjustment preferences, and task participation) provides airlines with valuable market insights. For example, if the data shows that passengers on a certain route pay more attention to prices on weekends, targeted weekend promotions can be launched. If it is found that most passengers are willing to pay a lower price for the "no baggage allowance" option, cabin service design can be optimized. This refined management based on real-time data enables airlines to shift their revenue strategy from "experience-based judgment" to "data-driven", thereby improving overall revenue efficiency.

[0090] Finally, see Figures 1-4 In this embodiment, for business travelers who are time-sensitive and have low price elasticity, the system can prioritize recommending high-frequency and high-punctuality flights, and add service premium options such as "fast security check" and "VIP lounge" to meet their needs for efficiency and experience.

[0091] For price-sensitive travelers with flexible schedules, the system can recommend off-peak flights and offer discounted "flight + hotel" packages. It also provides additional discounts through social sharing tasks to reduce travel costs.

[0092] For special groups such as the elderly and disabled, the system identifies their special service needs (such as wheelchairs and priority boarding) through demand analysis, integrates relevant service costs into the pricing, and provides a "one-stop" quote to avoid the hassle of additional charges later.

[0093] For regional flights with unstable passenger numbers and fewer flights, the system can attract flexible passengers through dynamic pricing. For example, when there are many seats remaining on a regional flight, low-price offers can be pushed to passengers with flexible schedules to increase the load factor. As the departure time approaches, prices can be appropriately increased for passengers who need to travel urgently to balance revenue.

[0094] Low-cost airlines rely on a profit model of "basic fare + value-added services". The system can break down service options (such as baggage allowance, meals, and seat selection) into independent modules, allowing passengers to combine them freely. Through intelligent negotiation, "personalized service packages" can be generated, which not only meets passengers' pursuit of low cost, but also increases the sales rate of ancillary services.

[0095] Charter operators can publish "charter availability" information through the system and offer dynamic prices for individual passengers. When there are many available seats, they can offer low-price promotions and gradually increase prices as departure approaches, maximizing the utilization rate of charter flights.

[0096] For groups of 10 or more, the system can develop a "group negotiation" function, allowing group representatives to negotiate prices with the system (e.g., "If we add 5 more people, can we lower the price by XX yuan?"). This allows them to leverage the scale advantage of bulk orders to obtain discounts while simplifying the cumbersome process of group ticket purchases.

[0097] By integrating the system with airline membership programs, differentiated negotiation privileges can be provided based on membership level (e.g., platinum members can get more negotiation opportunities and higher task reward ratios), thereby enhancing user stickiness.

[0098] Expand the language capabilities of the natural language processing unit to support input in multiple languages ​​such as English, Japanese, and Korean, adapting to international flight scenarios and serving outbound passengers.

[0099] Based on passengers' historical data (such as preferred airlines and frequently chosen cabin classes), flights that passengers may prefer are prioritized during the negotiation process, reducing decision-making time and improving the experience.

[0100] Working principle:

[0101] The AI-driven dynamic flight pricing and intelligent negotiation system and its implementation method, when in use, first involve demand analysis and order generation:

[0102] After a passenger inputs their natural language request, the system extracts structured parameters through a large language model. For example, "I want to go from Guangzhou to Chengdu next Tuesday, preferably cheap, and I can accept red-eye flights" is parsed as: route (Guangzhou-Chengdu), time window (next Tuesday ± 1 day), budget preference (low price), special conditions (accepting red-eye flights), and a booking is generated and put into validity management.

[0103] Dynamic pricing calculation:

[0104] The revenue management module obtains real-time data: the basic economy class fare for this flight is 600 yuan, with 8 seats remaining (inventory tightness coefficient K1 = 1.1), 48 hours before departure (time sensitivity factor K2 = 1.1), and the passenger price elasticity coefficient K3 = 0.3 (moderate sensitivity). The intelligent negotiation engine calculates the suggested price as 600 × 1.1 × 1.1 × (1 - 0.3 × (budget - 600) / 600). If the passenger's budget is 500 yuan, the suggested price is 600 × 1.21 × (1 - 0.3 × (-100) / 600) = 600 × 1.21 × 1.05 ≈ 762 yuan (exceeding the budget). Therefore, it automatically recommends a lower-priced cabin or adjusts the time (such as next Wednesday, K2 = 1.0), recalculates to 600 × 1.1 × 1.0 × 1.05 ≈ 693 yuan, and if it still exceeds the budget, it triggers a negotiation range feedback.

[0105] Multiple rounds of negotiation and transaction:

[0106] If a passenger rejects the initial quote and adjusts their requirements (e.g., "accept next Thursday, budget 550 yuan"), the negotiation iteration unit updates the pending order data, the engine recalculates and provides a matching price, until the passenger accepts (triggering ticketing) or abandons the offer.

[0107] Task Discount and Evidence Storage:

[0108] Passengers select the "Share to get a 50 yuan discount" task. After completion, the system verifies the validity of the sharing through a smart contract, and the reward redemption unit applies the 50 yuan discount to the ticket price (actual payment of 500 yuan).

[0109] The blockchain evidence storage module records the order data, three rounds of negotiation records, the transaction amount of 550 yuan, and the 50 yuan deduction information on the chain, generating an immutable transaction certificate for all parties to query and verify.

[0110] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. An AI-driven dynamic pricing and intelligent negotiation system for flights, characterized in that, include: The module includes: Passenger Intent Analysis Module, Intent Placement Module, Revenue Management Integration Module, Intelligent Negotiation Engine Module, Transaction Feedback Module, Third-Party Task Delegation Module, and Blockchain Evidence Storage Module. The passenger intent parsing module is signal-connected to the intention order placement module, and is used to convert the unstructured needs input by the passenger into structured order placement data; The revenue management module is connected to the intelligent negotiation engine module to obtain real-time flight inventory and pricing strategies. The intelligent negotiation engine module is connected to the intention order module and the transaction feedback module respectively, and generates price suggestions based on the order data and flight data. as well as The third-party task delegation module is connected to the blockchain evidence storage module and is used to publish tasks and record reward deduction information.

2. The AI-driven dynamic flight pricing and intelligent negotiation system according to claim 1, characterized in that, The passenger intent parsing module includes a natural language processing unit and a multimodal input interface. The natural language processing unit is based on a pre-trained large language model and supports Chinese semantic recognition. The multimodal input interface is compatible with text, voice and image input.

3. The AI-driven dynamic flight pricing and intelligent negotiation system according to claim 2, characterized in that, The intention to place an order module includes a data storage unit and a validity period management unit. The data storage unit records the passenger's budget, time range, flight route and comfort preference, while the validity period management unit sets the validity period of the order to 24-72 hours.

4. The AI-driven dynamic flight pricing and intelligent negotiation system according to claim 3, characterized in that, The intelligent negotiation engine module includes a multi-objective optimization unit and a price generation unit. The multi-objective optimization unit calculates the price range by comprehensively considering flight profitability, load factor, remaining departure time, and passenger budget. The price generation unit outputs specific quotations or negotiation schemes.

5. The AI-driven dynamic flight pricing and intelligent negotiation system according to claim 4, characterized in that, The transaction feedback module includes a ticketing unit and a negotiation iteration unit. The ticketing unit triggers seat locking and ticket payment when the passenger accepts the price quote. The negotiation iteration unit updates the order data and triggers renegotiation after the passenger adjusts their needs.

6. The AI-driven dynamic flight pricing and intelligent negotiation system according to claim 5, characterized in that, The third-party task assignment module includes a task publishing unit and a reward redemption unit. The task publishing unit provides task options such as brand promotion, questionnaire surveys, and social sharing. The reward redemption unit converts task rewards into airfare discounts according to a preset ratio.

7. The AI-driven dynamic flight pricing and intelligent negotiation system according to claim 6, characterized in that, The blockchain evidence storage module includes a smart contract unit and a distributed ledger unit. The smart contract unit verifies the completion of tasks and executes reward deductions, while the distributed ledger unit stores negotiation records, transaction data, and discount information, supporting multi-party queries and verification.

8. The AI-driven dynamic flight pricing and intelligent negotiation system according to claim 7, characterized in that, The algorithm parameters of the multi-objective optimization unit include the base ticket price, inventory tightness coefficient, time sensitivity factor, and passenger price elasticity coefficient. The inventory tightness coefficient is inversely proportional to the number of remaining seats, and the time sensitivity factor increases linearly as the departure time approaches.

9. The AI-driven dynamic flight pricing and intelligent negotiation system according to claim 8, characterized in that, The reward redemption unit has a deduction limit: the deduction amount for a single task reward shall not exceed 30% of the total ticket price, and the cumulative deduction amount for a single passenger on a single flight shall not exceed 50% of the total ticket price.

10. The AI-driven dynamic flight pricing and intelligent negotiation method according to any one of claims 1 to 9, characterized in that, Includes the following steps: S1: Passengers submit their travel requests through a multimodal input interface, and the passenger intent parsing module converts them into a structured intention posting that includes budget, time range, flight route and preferences; S2: The intention order module stores order data and sets the validity period, while the revenue management module obtains the corresponding flight's inventory, cabin availability, and pricing strategy in real time; S3: The intelligent price negotiation engine module generates personalized price suggestions based on pending order data and flight data, and then feeds them back to passengers through a multi-objective optimization algorithm. S4: Passengers choose to accept the offer (triggering the ticketing unit to complete the transaction) or adjust their needs (triggering the negotiation iteration unit to update the order and repeat S3); S5: Passengers can voluntarily choose third-party tasks. After completion, the smart contract unit will verify the task, and the reward redemption unit will convert the reward into a ticket discount. S6: The blockchain evidence storage module stores negotiation records, transaction information, and discount data on the blockchain, forming an immutable transaction certificate.

Images