KR20230089127A - Paid mounted volume-weight prediction system using artificial Intellectual and method thereof - Google Patents

Abstract

The present invention relates to a payload load prediction system and method, and more specifically, flight condition, operating season, allowable legacy payload by aircraft (operating type), boarding customer information, paid payload information, passenger load factor information by aircraft, air cargo After collecting air flight information including reservation rate information, analyzing the payload factor algorithm for each air operation factor and condition, and generating an artificial intelligence model based on learning of the payload factor algorithm for each flight element and condition according to the time point, It relates to a system and method for predicting paid payload using artificial intelligence that predicts the paid payload of the aircraft in advance by inputting aircraft information about the aircraft to be predicted into the artificial intelligence model.

Description

Paid mounted volume-weight prediction system using artificial Intellectual and method thereof}

The present invention relates to a payload load prediction system and method, and more particularly, to a payload load prediction system and method, and more specifically, flight condition, operating season, allowable payload load by aircraft (operating type), boarding customer information, paid payload information, passenger load factor information by aircraft, air cargo Collect air operation factor information such as air operation information including reservation rate information, analyze flight operation factors and pay load factor algorithms for each condition, and artificial intelligence based on learning of the pay load factor algorithm for each air factor and condition according to time After creating a model, it relates to a payload prediction system and method using artificial intelligence that predicts the paid payload of the aircraft in advance by entering aircraft information about the aircraft to be predicted into the artificial intelligence model.

As globalization accelerates rapidly, the aviation market is gradually expanding. As of 2019, the number of air passengers reached an all-time high, and the record is being broken every year. An airline's revenue is determined by air fares and freight rates, and is mainly determined by air fares.

In general, air fares (charges) refer to a comprehensive amount charged by airlines to users of air traffic, that is, to passengers (customers) on board the aircraft, in exchange for transporting passengers and checked baggage that is less than a predefined payload.

In general, the price of jet fuel, which is the fuel of an aircraft, accounts for a high portion of airfare, and an important factor in determining the amount of jet fuel used is the amount of air cargo loaded on the aircraft. In the following description, air cargo is used as a term including checked baggage that passengers can take with them (or referred to as "air baggage") and general cargo and postal items that are sold to forwarders and loaded onto aircraft.

Therefore, payload of air cargo is very important, and the airfare includes the cost of checked baggage with the allowable paid payload (or referred to as “permissible paid payload volumetric weight”) defined in the ticket. The payload that can be loaded on an aircraft is the weight per area, and its volume is very important. For this reason, in the case of checked baggage, the volume is indirectly limited by standardized suitcases, and the volume of general cargo is also limited.

If you want to load more checked baggage with paid payload in excess of the payload payload allowed on the ticket, the passenger must pay an additional charge according to the condition.

However, depending on the passenger, some passengers take only checked baggage with a paid payload that is far less than the paid payload allowed by the ticket, some take checked baggage that exceeds the paid payload, and some take checked baggage in accordance with the paid payload allowed. also go

In this way, although the profits of airlines are determined by air fares, they lose due to the increase in air fare costs and the decrease in profits due to bloody competition caused by changes in the aviation market environment and deterioration in profitability caused by COVID-19 and consequently intensifying competition among airlines. proportion is increasing. In addition, an increase in the boarding factor does not increase airline profits, and profitability is gradually deteriorating due to oversupply compared to demand.

In addition, as the international situation becomes unstable, the exchange rate fluctuates, and airlines with a high proportion of foreign currency payments are experiencing increased financial burdens due to foreign currency exchange losses and air fuel costs, as well as aircraft leases or borrowings.

In order to improve the profitability of these airlines, it is important to accurately manage the payload and secure maximum air cargo by accurately identifying the total payload of passengers' checked baggage and general cargo in advance.

Existing air tickets include not only the passenger's fare but also the charge for a certain amount of payload, that is, the charge for checked baggage of the allowable amount of payload. It is calculated as the total paid payload. Among them, the number of passengers * paid load capacity of passengers' checked baggage + general cargo accounts for most of the total paid load capacity.

However, as described above, some passengers board with only checked baggage with a paid payload that is less than the paid payload of checked baggage purchased by the ticket, and boarded with checked baggage that is more or less paid than the payload paid for checked baggage allowed by the ticket. It is difficult to efficiently manage the entire paid load of checked baggage to be loaded on the aircraft. This is also acting as a factor that makes it difficult to operate general cargoes.

As described above, the aviation market is growing in terms of external size, but profitability is declining due to various negative factors such as bloody competition among airlines, political instability due to global unexpected situations, trade disputes, exchange rate fluctuations, and pandemic situations such as COVID-19. Accordingly, it is urgent to introduce a sales strategy based on a more flexible and active market approach and a new marketing method.

Therefore, there is a need to develop an air service system that can improve the profitability of air transport, which fluctuates according to the rapid change in the profit structure of the air market due to external environmental factors.

In addition, in order to improve the profitability of airlines, it is required to develop an aviation service system capable of predicting the total payload of air cargo to be loaded on an aircraft for a fee, that is, the payload of air checked baggage in advance.

Republic of Korea Patent Registration No. 10-1945125 (2019.04.17. Notice)

Therefore, an object of the present invention is to provide flight operation information including flight conditions, operating season, allowable heritage load capacity for each aircraft (operating aircraft type), boarding customer information, paid load capacity information, passenger load factor information for each aircraft, air cargo reservation rate information, etc. After collecting, analyzing the load factor algorithm for each air operation factor and condition, and generating an artificial intelligence model based on the learning of the pay load factor algorithm for each air factor and condition according to the time point, the aircraft for the aircraft to be predicted by the artificial intelligence model It is to provide a paid payload prediction system and method using artificial intelligence that predicts payload payload of the aircraft in advance by inputting information.

The paid payload prediction system using artificial intelligence according to the present invention for achieving the above object: sells air tickets for each of a plurality of aircraft and air payload according to the allowable paid payload that can be loaded on the aircraft, and the airline's An airline system that provides flight operation information and receives and manages the permissible paid payload and loaded payload information for each aircraft; And collects and stores flight operation information, permissible paid payload for each aircraft, and payload payload information from the airline system, analyzes air operation factors based on the air flight information and payload ratio algorithm for each condition, and analyzes the analyzed aviation elements. And after creating an artificial intelligence model based on the learning of the payload factor algorithm for each condition, receiving aircraft information about an arbitrary aircraft and predicting the payload of the aircraft, the airline system and payload manager of the airline corresponding to the aircraft It is characterized in that it includes a paid payload estimation unit provided by any one or more of the terminals.

The paid payload estimation unit may include a server communication unit enabling data communication with the airline system through a wired/wireless data communication network; Air flight information DB that stores air flight information including flight condition, flight route, flight season, reservation rate, load factor, paid load factor, passenger classification and information, ticket type, forwarder and paid load information according to past operations by aircraft; Payload registration DB that stores registered payload information registered according to the registration and transaction of paid payload of passengers by ticket type according to ticket sales and paid payload purchase by aircraft, payload load rate algorithm for paid payload by learned air operation elements and conditions A server storage unit including an artificial intelligence model DB for storing an artificial intelligence model according to target variable settings and a prediction DB for storing predicted payload information for each aircraft; And collecting air navigation information from the airline system through the server communication unit and storing it in the air flight information DB, acquiring the registered payload information from the airline system and storing it in the paid payload registration DB, and the air flight information and Analyzing the paid load factor algorithm for each air operation factor and condition based on any one or more of the registered payload load information, and generating an artificial intelligence model according to the target variable setting among the analyzed aviation factor and pay load factor algorithm for each condition, After storing in the model DB, aircraft information on a certain aircraft is received, and the paid payload of the aircraft is predicted, and through the server communication unit, at least one of the airline system and the paid payload manager terminal of the airline corresponding to the aircraft Characterized in that it comprises a server control unit provided as.

The server control unit collects flight operation information, permissible paid payload for each aircraft, and loaded payload information from the airline system through the server communication unit, stores them in the air flight information DB, and stores the registered paid payload information from the airline system. an information obtaining unit for acquiring and storing the information in the paid payload registration DB; Analyzing the payload payload ratio algorithm for each air navigation element and condition based on any one or more of the flight operation information and the registered payload load information, and artificial intelligence according to the target variable setting among the analyzed payload ratio algorithms for each flight element and condition After creating a model and storing it in the artificial intelligence model DB, an artificial intelligence prediction unit for receiving aircraft information about an arbitrary aircraft, predicting the paid payload of the aircraft, and storing the predicted payload payload in the prediction DB; and a prediction information providing unit for providing the predicted payload to at least one of the airline system and the payload manager terminal through the server communication unit.

The information acquisition unit may include: an air operation information acquisition unit that collects flight operation information, permissible paid payload information for each aircraft, and payload payload information from an airline system through the server communication unit and stores them in the flight flight information DB; Registered payload information obtained from the airline system, including input registration paid payload information by registration of boarding passengers and payload information registered according to transaction registration paid payload information, and stored in paid payload registration DB acquisition unit; It is characterized in that it includes an input information acquisition unit for receiving and outputting aircraft information to predict paid payload from a manager.

The artificial intelligence prediction unit analyzes a payload payload ratio algorithm for each air navigation element and condition based on any one or more of the flight operation information and the registered payload load information, and a target among the analyzed aviation elements and payload ratio algorithm for each condition A paid payload learning unit generating an artificial intelligence model according to variable settings and storing it in the artificial intelligence model DB; And when acquiring aircraft information for a certain aircraft through the input information acquisition unit, the aircraft information is input into the artificial intelligence model to predict the paid payload of the aircraft, and the predicted paid payload is stored in the prediction DB. It is characterized in that it includes a payload prediction unit.

To achieve the above object, a paid payload prediction system using artificial intelligence according to the present invention includes: a server communication unit that enables data communication with an airline system through a wired/wireless data communication network; Air flight information DB that stores air flight information including flight condition, flight route, flight season, reservation rate, load factor, paid load factor, passenger classification and information, ticket type, forwarder and paid load information according to past operations by aircraft; A payload registration DB that stores registered payload information registered according to the registration and transaction of passengers boarding by ticket type according to the sale of tickets for each aircraft and the purchase of paid payload, and stores the payload ratio algorithm for each learned air operation element and condition, A server storage unit including an artificial intelligence model DB for storing an artificial intelligence model according to target variable settings and a prediction DB for storing predicted payload information for each aircraft; And collecting air navigation information from the airline system through the server communication unit and storing it in the air flight information DB, acquiring the registered payload information from the airline system and storing it in the paid payload registration DB, and the air flight information and Analyzing the paid load factor algorithm for each air operation factor and condition based on any one or more of the registered payload load information, and generating an artificial intelligence model according to the target variable setting among the analyzed aviation factor and pay load factor algorithm for each condition, After storing in the model DB, aircraft information on a certain aircraft is received, and the paid payload of the aircraft is predicted, and through the server communication unit, at least one of the airline system and the paid payload manager terminal of the airline corresponding to the aircraft Characterized in that it comprises a server control unit provided as.

The server control unit collects flight operation information from the airline system through the server communication unit, stores it in the air flight information DB, and provides the registered payload information, allowable payload for each aircraft, and payload payload information from the airline system. an information obtaining unit for acquiring and storing the information in the paid payload registration DB; Analyzing the paid load factor algorithm for each air navigation factor and condition based on any one or more of the flight operation information and the registered payload load information, and analyzing the AI model according to the target variable setting among the analyzed air factor and pay load factor algorithm for each condition After generating and storing in the artificial intelligence model DB, an artificial intelligence prediction unit for receiving aircraft information about an arbitrary aircraft, predicting the paid payload of the aircraft, and storing the predicted payload payload in the prediction DB; and a prediction information providing unit for providing the predicted payload to at least one of the airline system and the paid payload manager through the server communication unit.

In order to achieve the above object, a method for predicting paid payload using artificial intelligence according to another embodiment of the present invention is: The server control unit provides flight operation information from an airline system through a server communication unit, allowable paid payload for each aircraft, and payload payload An information acquisition process of collecting information and storing it in an air flight information DB; The server control unit analyzes the payload factor algorithm for each air navigation factor and condition based on the flight operation information, and generates an artificial intelligence model according to the target variable setting among the analyzed aviation factor and payload factor algorithm for each condition, thereby generating the artificial intelligence model. Learning process to save to DB; a prediction process in which the server control unit receives aircraft information about a certain aircraft, applies it to the artificial intelligence model, predicts payload of the aircraft, and outputs paid payload information; and providing paid payload information to at least one of the airline system and a paid payload manager terminal of an airline corresponding to the aircraft through the server communication unit.

The information acquisition process may include an air navigation information acquisition step of collecting air flight information from an airline system through the server communication unit and storing the air flight information DB in the air flight information DB; and obtaining the registered payload information for obtaining the registered payload information from the airline system and storing it in the paid payload registration DB, wherein the learning process is carried out by the server control unit in the flight operation information and registered paid payload information. Based on the analysis of the payload factor algorithm for each air element and condition, and generating an artificial intelligence model according to the target variable setting among the analyzed air element and condition-specific payload factor algorithms, and storing it in the artificial intelligence model DB.

The paid load factor algorithm is the paid load capacity for each operating condition, the paid load capacity for each flight route, the paid load capacity for each flight type, the paid load capacity for each operating season, the paid load capacity for each reservation rate and load factor, the paid load capacity for each passenger category and information, and the passengers on board. It is characterized in that it includes paid load capacity for each type, paid load capacity for each airline ticket, and paid load factor algorithm for each forwarder.

The present invention predicts the paid payload of the aircraft before selling the ticket for each aircraft by an artificial intelligence model learned based on the flight operation information of the aircraft, and based on the predicted paid payload, the allowable payload payload and the bundle payload in bundle units Since the price of the payload can be determined, it has the effect of determining the optimal air ticket and payload fare, which has the effect of improving the profitability of the airline while attracting more customers and air cargo.

In addition, the present invention predicts the paid payload of the aircraft by applying the ticket reservation rate, occupancy rate, registration and payload of the aircraft according to registration and transaction to the artificial intelligence model before a certain time from the departure date and time of the aircraft after the sale of the ticket. Therefore, there is an effect of predicting payload more accurately after ticket sales.

As described above, since the paid payload can be predicted for each aircraft before a certain time based on the departure date and time of the aircraft, it is possible to quickly resell the remaining paid payload out of the total paid payload of the aircraft, thereby increasing the airline's profit rate. There is an effect that can be made.

1 is a diagram showing the configuration of a communication system to which a paid payload prediction system using artificial intelligence according to the present invention is applied.
2 is a diagram showing the configuration of a paid payload predicting unit of a paid payload predicting system using artificial intelligence according to the present invention.
3 is a flowchart showing a method for predicting payload using artificial intelligence according to the present invention.

Hereinafter, with reference to the accompanying drawings, the configuration and operation of the payload prediction system using artificial intelligence according to the present invention will be described in detail, and the payload prediction method in the system will be described.

1 is a diagram showing the structure of a communication system including a payload prediction system using artificial intelligence according to the present invention.

Referring to FIG. 1, the paid payload prediction system 600 may include an airline system 300 and a paid payload predictor 400, or may be composed of only the paid payload predictor 400. In the latter case, the airline system 300 will be configured externally.

The paid payload prediction system 600 is connected to the customer terminal units 100 through wired or wireless data communication networks 1 through either wired or wireless communication to perform mutual data communication.

The airline system 300 and the paid payload estimation unit 400 may be configured to perform data communication by either wired or wireless through the wired/wireless data communication network 1, and may be configured to perform internal Wi-Fi (WiFi) network. It may be configured to be connected through a local area network (LAN) 2 and perform data communication by either wired or wireless.

The wired/wireless data communication network 1 is an Internet network including a WiFi network, a mobile communication network such as 3G (3G), 4G, and 5G, a WiBro network, a dedicated network, and a power line communication network. It could be a data communication network.

The customer terminal 100 is any one of a smart terminal 110 such as a smart phone or smart pad that wirelessly accesses the wired/wireless data communication network 1 and a computer terminal 120 such as a desktop computer, personal computer, or laptop computer that accesses the wired connection. can become one

A customer to purchase an airline ticket and payload for a certain aircraft may access the airline system 300 through the customer terminal unit 100 . Of course, the customer may purchase an airline ticket through an administrator terminal (not shown) of an airline ticket sales manager at a corresponding airline sales booth at an airport.

The airline system 300 includes an air ticket sales unit 310 and an air cargo management server 320 .

The airline sales department 310 sells tickets for each aircraft of the airline and payload for the ticket, flight conditions for each aircraft, aircraft operating route, operating season, reservation and boarding customer information for each aircraft (operating type), and baggage payload information. , Air flight information including boarding rate information for each aircraft, reservation rate information for each aircraft, etc. is stored and managed, and flight flight information is provided upon request from the paid payload estimation unit 400. The flight operation information is determined whenever the route and flight of the aircraft are determined. It may be updated at each departure of the aircraft.

The airline sales unit 310 may be configured to automatically provide the flight operation information to the payload load estimation unit 400 whenever the flight operation information is updated.

The airline sales unit 310 applied to the present invention provides air baggage of a preset allowable payload, that is, a general ticket including a fare for checked baggage and checked baggage with the allowable payload amount of 0 (paid payload free ticket). It can be sold separately as payload-free air tickets that do not include the fare for .

The airline sales unit 310 may be configured to generate and sell paid payload bundles for regular tickets and paid payload free tickets, or may be configured to allow transactions between passengers for the paid payload bundle.

The airline sales unit 310 registers paid payloads purchased from passengers, generates input registration paid payload information, registers paid payload according to bundle paid payload transactions, generates transaction registration paid payload information, and registers the input. Payload payload and transaction registered paid payload information are integrated and managed as registered paid payload information.

The airline sales unit 310 may provide the registered payload capacity information to the payload capacity estimation unit 400 upon request from the payload capacity estimation unit 400, or may provide it to the payload capacity estimation unit 400 whenever the information is updated. It could be.

The sale and transaction of the bundle payload may be performed by configuring a separate transaction system (not shown), and the registration of the paid payload may be performed by configuring a separate registration system (not shown).

In addition, the airline sales department 310 sells to shippers and forwarders as a paid payload that can load general cargo with respect to the remaining paid payload minus the total paid payload actually loaded for checked baggage from the total payload payload of the aircraft for each aircraft , it stores and manages shipper information, forwarder information, and general cargo information.

The air cargo management server 600 stores and manages the payload payload information purchased by each passenger, shipper, and forwarder for each aircraft, and receives, stores, and manages the payload payload information that each passenger, shipper, and forwarder actually loads on the aircraft. And, the payload load information of cargo actually loaded on the aircraft is provided to one or more of the ticket sales unit 310 and the payload load prediction unit 400.

The airline system 300 may be configured individually for each airline, or may be configured integrally for a plurality of airlines to provide ticket and payload sales for each airline, various air flight information, and registration and transaction payload information. It may be configured to create and manage.

The paid payload estimation unit 400 collects and stores air flight information from the airline system 300, and collects and stores registered paid payload information according to time points.

The payload estimation unit 400 analyzes the payload factor algorithm for each air navigation factor and condition based on the flight operation information, and generates an artificial intelligence model based on learning of the analyzed payload factor algorithm for each flight factor and condition.

In addition, the paid payload estimation unit 400 reflects the registered payload information according to the time point, analyzes the payload factor algorithm for each air operation factor and condition, and artificially analyzes the payload factor algorithm for each analyzed aviation factor and condition. Generate an intelligence model.

The paid payload prediction unit 400 receives aircraft information about an arbitrary aircraft after generating the artificial intelligence model, applies it to the artificial intelligence model, predicts the paid payload of the aircraft, and the airline of the airline corresponding to the aircraft. Provided by the system 300. The configuration of the paid-load load estimation unit 400 will be described in detail with reference to FIG. 2 below.

Figure 2 is a diagram showing the detailed configuration of the paid payload predicting unit of the paid payload predicting system using artificial intelligence according to the present invention.

Referring to FIG. 2 , the paid payload estimation unit 400 includes a server communication unit 410 and a server control module 500 .

The server communication unit 410 is connected to the airline system 300 through the wired/wireless data communication network 1 and the internal local network 2 to perform data communication with the airline system 300.

The server control module 500 includes the server storage unit 420 and the server control unit 430 to control overall operations of the payload capacity estimation unit 400 according to the present invention.

Specifically, the server storage unit 420 includes an air flight information DB 421, a paid payload registration DB 422, an input DB 423, an artificial intelligence model DB 424, and a prediction DB 425. .

The flight operation information DB 421 includes flight conditions, flight routes, flight seasons, reservation rate, boarding rate, payload rate, passenger classification and information, ticket type, forwarder and payload information according to past operations by airline and aircraft. Saves flight information.

The paid payload registration DB 422 stores paid payload information registered and loaded according to the registration and transaction of boarding passengers for each type of ticket according to the sale of the ticket for each aircraft. As described above, the registered payload information is input registration for payload information registered and loaded by an aircraft passenger, that is, a customer, and transaction registration for volume information registered and loaded according to payload information and paid payload transaction. Includes paid payload information.

The input DB 423 stores aircraft information input from a manager about an aircraft whose payload is to be predicted. The aircraft information may include information such as flight operation conditions, flight route operation season, reservation rate, boarding rate, payload rate, passenger classification and information, ticket type, forwarder and payload information.

The artificial intelligence model DB 424 stores learned flight operation factors and paid load factor algorithms for each condition, and stores artificial intelligence models according to target variable settings. The target variable may be any one or more of the air navigation elements and conditions of the paid load factor algorithm.

The payload factor algorithm for each air operation factor and condition may be an artificial intelligence model, which is an algorithm that has learned about the payload factor for each air operation factor and condition according to past operations. The above algorithms include a payload factor algorithm for each flight condition, a payload factor algorithm for each flight route, a payload factor algorithm for each flight type, a payload factor algorithm for each flight season, a payload factor algorithm for each reservation rate, a payload factor algorithm for each passenger category, and a payload factor for each passenger. Algorithm, payload ratio algorithm by airline ticket, payload ratio algorithm by shipper, payload ratio algorithm by forwarder, payload ratio algorithm by item, payload ratio algorithm by registered payload ratio, payload ratio algorithm by payload transaction, etc. may be included.

The target variable may vary according to the prediction time point of payload. For example, if the forecasting time of paid payload is before tickets are sold, the target variable may be any one or more of flight conditions, flight route, flight type, flight season, boarding rate, passenger classification and information, payload payload, etc. If the estimated time of payload is after ticket sales start, flight conditions, flight route, flight type, flight season, reservation rate, passenger classification and information, ticket type, shipper, forwarder, item, and scheduled loading It may be any one or more of items, registered payloads, and the like.

The target variable is the amount of payload analyzed by artificial intelligence in the conditions (past DB, reservation rate, season, transaction, etc.) after the start of ticket sales of the current corresponding route based on the DB of the transported payload according to the condition of the corresponding route in the past. That is, the linear regression of the following equation 1 Based on the formula, artificial intelligence analyzes the paid payload of the route expected.

Here, w ₀ , w ₁ represent parameters identified by artificial intelligence, and x is an input value input according to each flight operation condition (operation condition, operation type, operation season, reservation rate, passenger classification and information, flight ticket Type, by shipper, by forwarder, by item to be loaded, registered paid payload, trade paid payload). y learns the weight of a linear equation that represents the relationship between a feature and a target according to the input value input according to each condition, and with this equation, the target value of a new sample, not the training data set, is predictable. is the result value.

As shown in <Mathematical Formula 2> above, W ₀ X ₀ + W ₁ X ₁ + ... is the input value entered according to each flight operation condition (operation condition, flight type, flight season, reservation rate, passenger classification and information) , flight ticket type, shipper, forwarder, scheduled load item, registered payload, transaction paid payload) are multidimensionally added as the sum of weights, and the parameters identified by artificial intelligence are shown.

y is the result of air payload, which is represented by learning the weight of a linear equation representing the relationship between a feature and a target according to each input value, and using this equation to predict the target value of a new sample, not a training dataset.

The prediction DB 425 stores predicted paid payload information for each aircraft.

The server control unit 430 controls the overall operation of the server control module 500, including the information acquisition unit 440, the artificial intelligence prediction unit 450, and the prediction information providing unit 460.

The information acquisition unit 440 includes an air flight information acquisition unit 441, a registered payload load information acquisition unit 442, and an input information acquisition unit 445.

The air flight information acquisition unit 441 collects air flight information from the airline system 300 through the server communication unit 410 and stores the air flight information DB 421 .

The registered paid payload information acquisition unit 442 generates the input registered paid payload information by registering the boarding passenger from the airline system 300 and stores the input paid payload information acquisition unit 443 in the paid payload registration DB 422 and a transaction payload acquisition unit 444 for acquiring transaction registered paid payload information registered according to payload payload transactions and storing it in the payload payload registration DB 422 .

The input information acquisition unit 445 receives aircraft information to predict payload payload from the manager, obtains it from the airline system 300, stores it in the input DB 423, or outputs it to the artificial intelligence prediction unit 450.

The artificial intelligence prediction unit 450 analyzes a payload factor algorithm for each air navigation element and condition based on any one or more of the flight operation information and the registered payload load information, and selects a target among the analyzed aviation factor and payload factor algorithm for each condition. The paid payload learning unit 451 generates an artificial intelligence model according to variable settings and stores it in the artificial intelligence model DB 424, and receives aircraft information about an arbitrary aircraft and applies the artificial intelligence model to the artificial intelligence model. It includes a paid payload prediction unit 452 that predicts payload and stores the predicted payload payload in the prediction DB 425 .

The paid payload learning unit 451 can predict the paid payload for each aircraft at various times, and performs learning by changing the target variable, that is, the applied paid payload ratio algorithm, according to the predicted time point, and the artificial intelligence model according to the learning. will create Each artificial intelligence model for each prediction time point may be stored and managed in the artificial intelligence model DB 424, or the artificial intelligence model may be continuously updated.

It is preferable that the input data of the aircraft information applied to the payload load prediction unit 452 also differs according to the target variable at that time.

When the predicted paid payload information is generated by predicting the paid payload amount, the prediction information providing unit 460 transmits the generated predicted paid payload information to the airline system 300 through the server communication unit 410 or the manager terminal of the paid payload manager of the airline ( Not shown) or may be provided by e-mail of the paid payload manager.

3 is a diagram showing a flow chart of a payload prediction method using artificial intelligence according to the present invention.

Referring to FIG. 3 , the server control unit 430 checks whether a learning event occurs from the airline system 300 (S111). The learning event may occur when registering a certain flight, when ticket sales for the corresponding flight start, when a payload manager requests, and at a certain period.

When a learning event occurs, the server control unit 430 collects air flight information from the airline system 300, and further collects registration and transaction payload information according to the time point and stores it in the server storage unit 420 (S113).

When information including air flight information among air flight information and registered paid payload information is collected, the server controller 430 generates a paid load factor algorithm corresponding to the collected information and stores it in the artificial intelligence model DB 424 (S115). .

When the payload ratio algorithm is generated, the server controller 430 checks whether a payload ratio prediction event occurs (S117). The payload estimation event may be generated when the payload ratio algorithm is analyzed, may be generated at regular intervals, or may be generated at the request of a payload manager.

When a payload prediction event occurs, the server control unit 430 sequentially determines whether to start selling air tickets for at least one or more aircraft to be analyzed (S119).

If tickets for the corresponding aircraft are not on sale, the server control unit 430 generates an artificial intelligence model by applying a pre-ticket load factor algorithm, that is, a target variable before ticket sales (S120), and applies the predicted event to the generated artificial intelligence model. The aircraft information obtained when the occurrence of , is applied to the artificial intelligence model to predict the predicted paid payload before ticket sales, and the predicted paid payload information is generated and stored in the prediction DB 425 (S121). At this time, based on the predicted paid payload information, determine the total price of the payload-free ticket payload excluding the above-mentioned allowable payload payload for the payload-free ticket from the airfare of the ticket for the payload payload and the total payload payload, It is possible to determine a base price for a standard unit of the payload-free airline ticket payload once, generate a plurality of bundle payloads based on the price for the reference unit, and determine the price of the generated bundle payload.

The determined payload payload information and price information will be provided to any one or more of the airline system 300 and payload manager terminal unit (S123).

After storing the predicted payload amount information, the server controller 430 determines whether a learning event occurs again (S124) or whether a paid payload prediction event occurs (S117).

On the other hand, when ticket sales for the aircraft, that is, flight, start to be sold, the server control unit 430 creates an artificial intelligence model based on a pre-set paid load factor algorithm after ticket sales (S125).

When the artificial intelligence model is generated, the server control unit 430 applies the acquired aircraft information to the artificial intelligence model generated after the sale of the air ticket to predict the predicted payload amount after ticket sales, and generates the predicted paid payload information. (S127). When the predicted paid payload information is generated, the server controller 430 may determine a price for the remaining paid payload for the total paid payload of the corresponding aircraft according to the generated payload payload information.

When the predicted paid payload information is generated, the server control unit 430 updates the predicted paid payload by updating the artificial intelligence model according to newly generated information (registered payload information) until a certain time based on the departure date and time of the corresponding aircraft. will (S129).

When a predetermined time is reached based on the departure date and time of the corresponding aircraft, the server control unit 430 will stop updating the predicted payload for the corresponding aircraft and repeat the above-described processes from S111 onwards.

On the other hand, it is common knowledge in the art that the present invention is not limited to the above-described typical preferred embodiments, but can be implemented by various improvements, changes, substitutions, or additions within the scope of the present invention. If you have the , you can easily understand. If the implementation by such improvement, change, substitution or addition falls within the scope of the appended claims below, the technical idea should also be regarded as belonging to the present invention.

100: customer terminal 300: airline system
400: paid payload prediction unit 410: server communication unit
420: server storage unit 430: server control unit
440: information acquisition unit 441: flight information acquisition unit
442: Registered payload information acquisition unit 443: Input paid payload information acquisition unit
444: transaction paid payload information acquisition unit 445: input information acquisition unit
450: artificial intelligence prediction unit 451: paid payload learning unit
452: paid payload prediction unit 460: prediction information providing unit
500: server control module 600: paid payload prediction system

Claims (10)

Sells air tickets for each of a number of aircraft and air payloads according to the allowable paid payloads that can be loaded on the aircraft, provides air operation information of the airline, receives and manages the allowable paid payloads for each aircraft and paid payload information airline system; and
Collects and stores flight operation information, permissible paid payload for each aircraft and payload payload information from the airline system, analyzes air operation factors based on the air flight information and payload rate algorithm for each condition, and analyzes the analyzed aviation factors and After creating an artificial intelligence model based on the learning of the payload factor algorithm for each condition, receiving aircraft information about an arbitrary aircraft and predicting the payload of the aircraft, the airline system and payload manager terminal of the airline corresponding to the aircraft Payload prediction system using artificial intelligence, characterized in that it comprises a paid payload prediction unit provided by any one or more of.
According to claim 1,
The paid payload prediction unit,
a server communication unit for performing data communication with the airline system through a wired/wireless data communication network;
Air flight information DB that stores air flight information including flight condition, flight route, flight season, reservation rate, load factor, paid load factor, passenger classification and information, ticket type, forwarder and paid load information according to past operations by aircraft; Payload registration DB that stores registered payload information registered according to the registration and transaction of paid payload of passengers by ticket type according to ticket sales and paid payload purchase by aircraft, payload load rate algorithm for paid payload by learned air operation elements and conditions A server storage unit including an artificial intelligence model DB for storing an artificial intelligence model according to target variable settings and a prediction DB for storing predicted payload information for each aircraft; and
The server communication unit collects flight operation information from the airline system and stores it in the air flight information DB, obtains the registered payload information from the airline system and stores it in the paid payload registration DB, and Analyzing the paid load factor algorithm for each air operation factor and condition based on any one or more of the registered payload load information, and generating an artificial intelligence model according to the target variable setting among the analyzed air factor and pay load factor algorithm for each condition, and the artificial intelligence model After storing in the DB, it receives aircraft information about a certain aircraft, predicts the paid payload of the aircraft, and sends it to any one or more of the airline system and the paid payload manager terminal of the airline corresponding to the aircraft through the server communication unit. Payload prediction system using artificial intelligence, characterized in that it comprises a server control unit to provide.
According to claim 2,
The server control unit,
Through the server communication unit, air operation information, permissible paid payload for each aircraft, and paid payload information are collected from the airline system and stored in the air flight information DB, and the registered paid payload information is obtained from the airline system to collect the payload payload information. Information obtaining unit to store in the registration DB;
Analyzing the payload payload ratio algorithm for each air navigation element and condition based on any one or more of the flight operation information and the registered payload load information, and artificial intelligence according to the target variable setting among the analyzed payload ratio algorithms for each flight element and condition After creating a model and storing it in the artificial intelligence model DB, an artificial intelligence prediction unit for receiving aircraft information about an arbitrary aircraft, predicting the paid payload of the aircraft, and storing the predicted payload payload in the prediction DB; and
A payload prediction system using artificial intelligence, characterized in that it comprises a prediction information providing unit for providing the predicted payload to any one or more of the airline system and the payload manager terminal through the server communication unit.
According to claim 3,
The information acquisition unit,
an air navigation information acquisition unit that collects air flight information, allowable paid payload for each aircraft, and payload payload information from an airline system through the server communication unit and stores them in the air flight information DB;
Registered payload information obtained from the airline system, including input registration paid payload information by registration of boarding passengers and payload information registered according to transaction registration paid payload information, and stored in paid payload registration DB acquisition unit;
A paid payload prediction system using artificial intelligence, characterized in that it comprises an input information acquisition unit for receiving and outputting aircraft information to predict paid payload from a manager.
According to claim 4,
The artificial intelligence prediction unit,
Analyzing the payload payload ratio algorithm for each air navigation element and condition based on any one or more of the flight operation information and the registered payload load information, and artificial intelligence according to the target variable setting among the analyzed payload ratio algorithms for each flight element and condition Payload learning unit for generating a model and storing it in the artificial intelligence model DB; and
When acquiring aircraft information on a certain aircraft through the input information acquisition unit, the aircraft information is input into the artificial intelligence model to predict the paid payload of the aircraft, and the predicted paid payload is stored in the prediction DB. Payload prediction system using artificial intelligence, characterized in that it comprises a prediction unit.
A server communication unit that enables data communication with an airline system through a wired/wireless data communication network;
Flight operation information that stores flight operation information including flight conditions, flight route, flight season, reservation rate, load factor, load factor, classification and information of passengers, type of ticket, forwarder and payload information according to past operations by aircraft DB, payload registration DB that stores registered payload information registered according to registration and transaction of passengers boarding by ticket type according to ticket sales and paid payload purchase by aircraft, and payload load factor algorithm for each learned air operation element and condition A server storage unit including an artificial intelligence model DB for storing an artificial intelligence model according to target variable settings and a prediction DB for storing predicted payload information for each aircraft; and
Through the server communication unit, air navigation information is collected from the airline system and stored in the air flight information DB, and the registered paid payload information is obtained from the airline system and stored in the paid payload registration DB. Analyzing the paid load factor algorithm for each air operation factor and condition based on any one or more of the registered payload load information, and generating an artificial intelligence model according to the target variable setting among the analyzed air flight element and pay load factor algorithm for each condition, the artificial intelligence model After storing in the DB, it receives aircraft information about a certain aircraft, predicts the paid payload of the aircraft, and transfers it to any one or more of the airline system and the paid payload manager terminal of the airline corresponding to the aircraft through the server communication unit. Payload prediction system using artificial intelligence, characterized in that it comprises a server control unit to provide.
According to claim 6,
The server control unit,
Air flight information is collected from the airline system through the server communication unit and stored in the air flight information DB, and the registered paid payload information and allowable paid payload information for each aircraft and loaded payload information are obtained from the airline system to obtain the paid payload amount. Information acquisition unit to store in the registration DB;
Analyzing the paid load factor algorithm for each air navigation factor and condition based on any one or more of the flight operation information and the registered payload load information, and analyzing the AI model according to the target variable setting among the analyzed air factor and pay load factor algorithm for each condition After generating and storing in the artificial intelligence model DB, an artificial intelligence prediction unit for receiving aircraft information about an arbitrary aircraft, predicting the paid payload of the aircraft, and storing the predicted payload payload in the prediction DB; and
A payload prediction system using artificial intelligence, characterized in that it comprises a prediction information providing unit for providing the predicted payload to any one or more of the airline system and the payload manager through the server communication unit.
Information acquisition process in which the server control unit collects air flight information, allowable paid payload for each aircraft, and payload payload information from the airline system through the server communication unit and stores them in an air flight information DB;
The server control unit analyzes the payload factor algorithm for each air navigation factor and condition based on the flight operation information, and generates an artificial intelligence model according to the target variable setting among the analyzed aviation factor and payload factor algorithm for each condition, thereby generating the artificial intelligence model. Learning process to save to DB;
a prediction process in which the server control unit receives aircraft information about a certain aircraft, applies it to the artificial intelligence model, predicts payload of the aircraft, and outputs paid payload information; and
A method for predicting paid payload using artificial intelligence, characterized in that it comprises a process of providing paid payload information provided by the server control unit to any one or more of the airline system and payload payload manager of the airline corresponding to the aircraft through the server communication unit. .
According to claim 8,
The information acquisition process,
an air navigation information acquisition step of collecting air flight information from an airline system through the server communication unit and storing the air flight information DB in the air flight information DB; and
Acquiring the registered payload information from the airline system and storing the registered paid payload information in the paid payload registration DB,
The learning process is
The server control unit analyzes the payload factor algorithm for each aviation factor and condition based on the flight operation information and registered payload load information, and generates an artificial intelligence model according to the target variable setting among the analyzed aviation factor and payload factor algorithm for each condition Payload prediction method using artificial intelligence, characterized in that stored in the artificial intelligence model DB.
According to claim 9,
The paid load factor algorithm,
Payload capacity by operation condition, Payload capacity by flight route, Payload capacity by flight type, Paid load capacity by operating season, Paid load capacity by reservation rate and occupancy rate, Paid load capacity by passenger category and information, Paid load capacity by boarding passenger, Air ticket type A method for predicting payload using artificial intelligence, characterized in that it includes payload and payload rate algorithm for each forwarder.

Images