Method and apparatus for optimizing ship operations

Abstract

The present disclosure relates to a method and apparatus for optimizing ship operation. A method according to an embodiment of the present disclosure can generate recommended voyage information for a ship's navigation route based on navigation plan information for the ship's departure and arrival points, predict the amount of BOG generated by the ship and a tank pressure value of the ship based on the recommended voyage information, and obtain optimal voyage information for controlling the operation of the ship based on the amount of BOG generated and the tank pressure value.

Description

[Technical Field]

[0001] The present disclosure relates to a method and an apparatus for optimizing ship's navigation. [Background technology]

[0002] Generally, when a ship transports liquefied gas, a gas that naturally evaporates in the liquefied gas tank is generated, and this gas is called BOG (Boil-Off Gas).

[0003] When BOG is generated, the pressure inside the tank increases, and as the pressure inside the tank increases, the safety of the liquefied gas tank decreases, and there is a risk of the tank exploding. To prevent this danger, BOG is released, but this process can cause environmental pollution and waste liquefied gas.

[0004] Meanwhile, to maintain pressure inside the tank, BOG is used as fuel for the ship's propulsion engine or power generation engine, and if there is any excess fuel despite fuel use, it can be liquefied through a re-liquefaction unit and stored in a cargo tank, or removed through a gas combustion unit (GCU).

[0005] Therefore, ship operating costs vary greatly depending on factors such as the amount of BOG generated during liquefied gas transportation, the ship's speed, power consumption, BOG disposal method, the location of the sea area depending on the navigation route, and climate change, making it difficult to predict when planning economical navigation. In particular, each ship has a different cargo handling system (CHS), which makes analyzing an economical navigation method complex. Furthermore, because the amount of BOG generated varies depending on the navigation route, ship operators have difficulty in optimally operating a ship that minimizes liquefied gas consumption. This raises the need for a method for efficiently controlling liquefied gas carriers in an operation method that minimizes liquefied gas consumption.

[0006] The above-mentioned background art is technical information that the inventor possessed for the purpose of deriving the present invention or that he acquired in the process of deriving the present invention, and is not necessarily publicly known art that was made public to the general public prior to the filing of the present invention. Summary of the Invention [Problem to be solved by the invention]

[0007] The present invention provides a method and apparatus for optimizing ship operation, and a computer-readable recording medium having a program recorded thereon for executing the method on a computer.

[0008] The problems to be solved by the present invention are not limited to those described above, and other problems and advantages of the present invention not mentioned above can be understood from the following description and will be more clearly understood in the embodiments of the present invention. Furthermore, it will be understood that the problems and advantages to be solved by the present invention can be realized by the means and combinations thereof set forth in the claims. [Means for solving the problem]

[0009] As a technical means for achieving the above-mentioned technical problem, a first aspect of the present disclosure can provide a method for optimizing ship operation, the method including the steps of generating recommended navigation information regarding the ship's navigation route based on operation plan information regarding the ship's departure and arrival points, predicting the ship's BOG generation amount and the ship's tank pressure value based on the recommended navigation information, and obtaining optimal navigation information regarding the ship's operation control based on the BOG generation amount and the tank pressure value.

[0010] A second aspect of the present disclosure can provide a computing device including at least one memory and at least one processor, wherein the processor generates recommended voyage information regarding a ship's navigation route based on navigation plan information regarding the ship's departure and arrival points, predicts the ship's BOG generation amount and the ship's tank pressure value based on the recommended voyage information, and obtains optimal voyage information regarding ship operation control based on the BOG generation amount and the tank pressure value.

[0011] A third aspect of the present disclosure can provide a computer-readable recording medium having recorded thereon a program for executing the method according to the first aspect on a computer.

[0012] In addition, other methods and systems for implementing the present invention, and computer-readable recording media storing computer programs for executing the methods may also be provided.

[0013] Other aspects, features, and advantages, in addition to those described above, will become apparent from the following drawings, claims, and detailed description of the invention. [Effects of the Invention]

[0014] According to the problem-solving means of the present disclosure described above, the present disclosure generates recommended navigation information regarding the ship's navigation route based on operation plan information regarding the ship's departure and arrival points, predicts the ship's BOG generation amount and ship's tank pressure value based on the recommended navigation information, and obtains optimal navigation information regarding the ship's operation control based on the BOG generation amount and tank pressure value, thereby enabling economical operation that minimizes the ship's liquefied gas consumption.

[0015] In addition, according to the problem-solving means of the present disclosure, the amount of BOG generated and tank pressure values ​​are predicted based on environmental information including one or more of weather and climate information, tidal current information, sea condition information, and ocean current information for each location on the ship's navigation route, and optimal navigation information is obtained based on the predicted amount of BOG generated and tank pressure values ​​to calculate the entire voyage operation, thereby making it possible to present the direction of cargo management from the ship's departure point to its arrival point and enable safe cargo handling based on the predicted tank pressure values.

[0016] In addition, according to the problem-solving means of the present disclosure, by updating the ship speed and liquefied gas consumption amount regarding the navigation route and obtaining optimal navigation information, it is possible to provide the operator with a ship navigation guide to assist in navigation, and by reducing the amount of GCU incineration and the amount of engine fuel gas, it is possible to provide a carbon tax reduction effect.

[0017] The effects of the present invention are not limited to those mentioned above, and other effects not mentioned will be clearly understood by those skilled in the art from the following description. [Brief explanation of the drawings]

[0018] [Figure 1] FIG. 1 is a diagram illustrating an example of a system for optimizing ship operation according to an embodiment. [Figure 2] FIG. 10 is a diagram illustrating an example of a configuration of a user terminal according to an embodiment. [Figure 3] FIG. 2 is a configuration diagram illustrating an example of a server according to an embodiment. [Figure 4] 1 is a flowchart illustrating an example of a method for optimizing ship operation according to an embodiment. [Figure 5] 10 is a flowchart illustrating an example in which a processor according to an embodiment generates recommended voyage information based on operation plan information. [Figure 6] 10 is a flowchart illustrating an example in which a processor according to an embodiment predicts the amount of BOG generated by a ship and the tank pressure value of the ship based on recommended voyage information. [Figure 7]10 is a flowchart illustrating an example in which a processor according to an embodiment acquires optimal voyage information based on a BOG generation amount and a tank pressure value. [Figure 8] 10 is a flowchart illustrating another example in which a processor according to an embodiment acquires optimal voyage information based on the amount of BOG generated and the tank pressure value. [Figure 9] 10 is a flowchart illustrating another example of a method for optimizing ship operation according to an embodiment. [Figure 10] FIG. 1 is a diagram for explaining an example of a method for predicting the amount of BOG generated according to an embodiment. [Figure 11] 1 is a flowchart illustrating an example of a method for predicting the amount of BOG generated by a ship according to an embodiment. [Figure 12] FIG. 10 is a diagram for explaining an example of calculating gas consumption of a ship according to an embodiment. [Figure 13A] FIG. 1 is a diagram illustrating an example of a stacking model according to an embodiment. [Figure 13B] FIG. 1 is a configuration diagram illustrating an example of a BOG generation amount prediction model for a ship according to an embodiment. [Figure 14] FIG. 10 is a diagram illustrating an example of calculating correct answer data for learning according to an embodiment. [Figure 15A] 10 is a diagram illustrating the removal of lost data from data stored in a memory 120 by a processor 110 according to an embodiment. FIG. [Figure 15B] FIG. 10 is a diagram illustrating an example of separating a portion of pre-stored operation data into learning data for an input data selection model and another portion of the data into verification data for a ship BOG generation prediction model according to an embodiment. [Figure 16A] FIG. 10 is a diagram for explaining an example of acquiring current operation data of a ship from outside the ship according to an embodiment. [Figure 16B] FIG. 10 is a diagram illustrating an example of acquiring current operation data of a ship from inside the ship according to an embodiment. [Figure 17]10 is a diagram illustrating an example of calculating a final predicted value of the amount of BOG generated by applying a weight to an initial predicted value of the amount of BOG generated according to an embodiment. FIG. [Figure 18] FIG. 10 is a diagram for explaining an example of a method for predicting a tank pressure of a ship according to an embodiment. [Figure 19] 1 is a flowchart illustrating an example of a method for predicting a tank pressure of a ship according to an embodiment. [Figure 20A] 10A and 10B are diagrams illustrating an example of separating a part of pre-stored actual operation data into learning data and another part into verification data according to an embodiment. [Figure 20B] FIG. 10 is a diagram illustrating an example of a process for processing pre-stored actual operation data according to an embodiment. [Figure 21] FIG. 10 is a diagram illustrating an example of a process for learning a deep learning model for predicting tank pressure of a ship according to an embodiment. [Figure 22] A diagram to explain an example of determining the navigation mode depending on whether the water level of liquefied gas in a tank in one embodiment is above a predetermined height. [Figure 23] A diagram for explaining an example of a deep learning model corresponding to each of multiple navigation modes according to the water level of liquefied gas in a tank in one embodiment. [Figure 24] FIG. 10 is a diagram illustrating an example of applying weights to a plurality of intermediate values ​​to obtain a predicted tank pressure value according to an embodiment. [Figure 25] FIG. 1 is a diagram for explaining an example of a method for optimizing ship operation according to an embodiment. [Figure 26] 1 is a flowchart illustrating an example of a method for optimizing ship operation according to an embodiment. [Figure 27] 10A and 10B are diagrams illustrating an example of acquiring optimal navigation information in different ways depending on whether or not a speed for each navigation section included in recommended navigation information is used in the operation of a ship according to an embodiment. [Figure 28A]10 is a flowchart illustrating an example of obtaining optimal navigation information when using a speed for each navigation section included in recommended navigation information in operation of a ship according to an embodiment. [Figure 28B] 10 is a flowchart illustrating an example of obtaining optimal navigation information when not using the speeds for each navigation section included in the recommended navigation information in the operation of a ship according to an embodiment. [Figure 29] FIG. 1 is a diagram illustrating an example of a system for controlling operation of a ship according to an embodiment. [Figure 30] FIG. 2 is a configuration diagram illustrating an example of a server according to an embodiment. [Figure 31] 1 is a flowchart illustrating an example of a method for controlling operation of a marine vessel according to an embodiment. [Figure 32] FIG. 10 is a diagram illustrating an example of a screen on which preset operation information of a ship according to an embodiment is displayed. [Figure 33] FIG. 10 is a diagram illustrating an example of a screen on which either a manually set route for a ship or a route automatically set through a route optimization function can be acquired according to an embodiment. [Figure 34] FIG. 10 is a diagram for explaining an example of a method for calculating a predicted pressure value of a liquefied gas cargo tank according to an embodiment. [Figure 35A] 10A-10C illustrate an example of predicted liquefied gas cargo tank internal pressure compared to measured values ​​and average velocity compared to measured values ​​according to an embodiment. [Figure 35B] FIG. 10 is a diagram illustrating an example of the results of comparing a predicted value of the amount of liquefied gas evaporated from a liquefied gas cargo tank with a measured value according to one embodiment. [Figure 35C] FIG. 10 is a diagram illustrating an example of a result of comparing an economical operation index of a ship with a measured value according to an embodiment. [Figure 36] FIG. 10 is a diagram illustrating an example of a screen displayed by a processor according to an embodiment. BEST MODE FOR CARRYING OUT THE INVENTION

[0019] The present disclosure relates to a method and apparatus for optimizing ship operation. In one embodiment, the method generates recommended navigation information for a ship's navigation route based on navigation plan information for the ship's departure and arrival points, predicts the ship's BOG generation rate and tank pressure value based on the recommended navigation information, and obtains optimal navigation information for ship operation control based on the BOG generation rate and tank pressure value. DETAILED DESCRIPTION OF THE INVENTION

[0020] Various embodiments of the present disclosure will be described below with reference to the accompanying drawings. Since various embodiments of the present disclosure can be modified in various ways and can have multiple embodiments, specific embodiments are illustrated in the drawings and related detailed descriptions are provided. However, this is not intended to limit the various embodiments of the present disclosure to the specific embodiments, and it should be understood that the various embodiments of the present disclosure include all modifications and / or equivalents or alternatives falling within the spirit and technical scope of the various embodiments of the present disclosure. In describing the drawings, similar reference numerals are used for similar components.

[0021] The terms "comprise" or "may comprise" and the like, which may be used in various embodiments of the present disclosure, refer to the presence of the disclosed feature, operation, component, etc., and do not limit the presence of one or more additional features, operations, components, etc. Furthermore, in various embodiments of the present disclosure, the terms "comprise" or "have" and the like specify the presence of features, numbers, steps, operations, components, parts, or combinations thereof described herein, and should be understood not to preclude the possibility of the presence or addition of one or more other features, numbers, steps, operations, components, parts, or combinations thereof.

[0022] In various embodiments of the present disclosure, phrases such as "or" include any and all combinations of the words listed together. For example, "A or B" can include A, can include B, or can include both A and B.

[0023] Terms such as "first," "second," "first," or "second" used in various embodiments of the present disclosure may modify various components of various embodiments, but do not limit those components. For example, such terms do not limit the order and / or importance of the corresponding components. Such terms may be used to distinguish one component from another. For example, a first user device and a second user device are both user devices and represent different user devices. For example, a first component may be designated a second component, and similarly, a second component may be designated a first component, without departing from the scope of the various embodiments of the present disclosure.

[0024] In the embodiments of the present disclosure, the terms "module," "unit," "part," etc. are used to refer to a component that performs at least one function or operation, and such a component may be embodied in hardware or software, or a combination of hardware and software. Furthermore, multiple "modules," "units," "parts," etc. may be integrated into at least one module or chip and embodied in at least one processor, unless each needs to be embodied in specific hardware.

[0025] The terms used in the various embodiments of the present disclosure are merely used to describe particular embodiments and are not intended to limit the various embodiments of the present disclosure. The singular expressions include the plural expressions unless otherwise clearly indicated in the context.

[0026] Unless otherwise defined, all terms used herein, including technical or scientific terms, have the same meaning as commonly understood by one of ordinary skill in the art to which various embodiments of the present disclosure belong.

[0027] Terms as defined in commonly used dictionaries should be interpreted to have a meaning consistent with the contextual meaning of the relevant art, and should not be interpreted as idealized or overly formal unless expressly defined in various embodiments of the present disclosure.

[0028] In the present invention, the term "liquefied gas" may refer to any gas fuel that is generally stored in a liquid state, such as liquefied natural gas (LNG), liquefied petroleum gas (LPG), ethylene, ammonia, hydrogen, etc. For convenience, the term "liquefied gas" may also be used to refer to gas that is not in a liquid state due to heating or pressurization. This also applies to evaporated gas.

[0029] Here, for convenience, LNG can be used to encompass not only natural gas (NG) in a liquid state but also natural gas (NG) in a supercritical state, and evaporated gas can be used to encompass not only evaporated gas in a gaseous state but also liquefied evaporated gas.

[0030] Furthermore, the term "gas consumption" can be used to encompass both the consumption of liquefied gas and the consumption of BOG, or can be used to refer to either the consumption of liquefied gas or the consumption of BOG.

[0031] Here, BOG (Boil-Off Gas) refers to natural gas that naturally evaporates and vaporizes in a ship's tank. Generally, liquefied gas is transported in an ultra-low temperature liquid state, and ship tanks that store liquefied gas are constructed with an insulated structure to maintain the ultra-low temperature. However, since the temperature outside the tank is about 40°C and the temperature difference between the inside and outside of the ship's tank is more than 200°C, it is not possible to completely block the inflow of heat from the outside. As a result, BOG generated by the inflow of heat into the tank can lead to economic problems.

[0032] Hereinafter, various embodiments of the present invention will be described in detail with reference to the accompanying drawings.

[0033] FIG. 1 is a diagram illustrating an example of a system for optimizing ship operation according to an embodiment.

[0034] 1, the system 1 includes a user terminal 10 and a server 20. For example, the user terminal 10 and the server 20 are connected via wired or wireless communication, and can transmit and receive data (e.g., operation plan information, recommended voyage information, BOG generation amount, tank pressure value, optimal voyage information, etc.) between them.

[0035] 1 illustrates the system 1 as including a user terminal 10 and a server 20. For convenience of explanation, the system 1 is not limited to this. For example, the system 1 may include other external devices (not shown), and the operations of the user terminal 10 and the server 20 described below may be implemented by a single device (e.g., the user terminal 10 or the server 20) or multiple devices.

[0036] The user terminal 10 may be a computing device including a display device and a device for receiving user input (e.g., a keyboard, a mouse, etc.), and including a memory and a processor. For example, the display device may be implemented as a touch screen and may receive user input. Examples of the user terminal 10 include, but are not limited to, a notebook computer, a desktop PC, a laptop, a tablet computer, a smartphone, etc.

[0037] The server 20 may be a device that includes the user terminal 10 and communicates with external devices (not shown). As an example, the server 20 may be a device that stores various data including operation plan information, recommended voyage information, BOG generation amount, tank pressure value, optimal voyage information, etc.

[0038] Alternatively, the server 20 may be a computing device including a memory and a processor and having its own computing power. As an example, the server 20 may perform at least some of the operations of the user terminal 10 described below with reference to Figures 1 to 36. For example, the server 20 may be, but is not limited to, a cloud server.

[0039] The user terminal 10 can acquire optimal navigation information related to the operation control of the ship. For example, the user terminal 10 can generate recommended navigation information related to the ship's navigation route based on navigation plan information related to the ship's departure and arrival points. The user terminal 10 can then predict the amount of BOG generated by the ship and the tank pressure value of the ship based on the recommended navigation information. The user terminal 10 can then acquire optimal navigation information related to the operation control of the ship based on the amount of BOG generated and the tank pressure value.

[0040] Meanwhile, the user terminal 10 can control the ship in a preset navigation method using the optimal navigation information. For example, the user terminal 10 can control the ship in a navigation method that minimizes the ship's liquefied gas consumption using the learned optimal navigation information.

[0041] For example, the user terminal 10 can obtain optimal voyage information related to the operation control of the ship through an application installed on the user terminal 10, and control the ship in a preset navigation method using the optimal voyage information. Here, the application may be a software program installed for the purpose of the user 30's navigation planning and control activities. For example, through the application, the user 30 can perform various navigation planning and control activities, such as generating recommended voyage information, predicting BOG generation amount, predicting tank pressure value, obtaining optimal voyage information, and operating the ship.

[0042] For convenience of explanation, the description has been given throughout the specification as describing the user terminal 10 generating recommended navigation information for the ship's navigation route based on navigation plan information for the ship's departure and arrival points, predicting the ship's BOG generation amount and tank pressure value based on the recommended navigation information, and obtaining optimal navigation information for ship operation control based on the BOG generation amount and tank pressure value, but this is not limited to this. For example, at least some of the operations performed by the user terminal 10 can be performed by the server 20.

[0043] In other words, at least some of the operations of the user terminal 10 described below with reference to Figures 1 to 36 can be executed by the server 20. For example, the server 20 can generate recommended voyage information regarding the ship's navigation route based on navigation plan information regarding the ship's departure and arrival points, predict the ship's BOG generation amount and tank pressure value based on the recommended voyage information, and obtain optimal voyage information regarding ship operation control based on the BOG generation amount and tank pressure value.The server 20 can then control the ship using a preset navigation method using the optimal voyage information.

[0044] FIG. 2 is a configuration diagram illustrating an example of a user terminal according to an embodiment.

[0045] 2, the user terminal 100 includes a processor 110, a memory 120, an input / output interface 130, and a communication module 140. For convenience of explanation, only components related to the present invention are shown in FIG. 2. Therefore, in addition to the components shown in FIG. 2, other general-purpose components may also be included in the user terminal 100. Furthermore, it will be obvious to those having ordinary skill in the art to which the present invention pertains that the processor 110, the memory 120, the input / output interface 130, and the communication module 140 shown in FIG. 2 may be embodied as independent devices.

[0046] The processor 110 can process instructions of a computer program by performing basic arithmetic, logic, and input / output operations, where the instructions can be provided from the memory 120 or from an external device (e.g., the server 20, etc.). Additionally, the processor 110 can provide overall control over the operation of other components included in the user terminal 100.

[0047] First, the processor 110 generates recommended navigation information for the vessel's navigation route based on navigation plan information for the vessel's departure point and destination. For example, the processor 110 can generate recommended navigation information for the vessel's navigation route based on the departure time at the vessel's departure point, the arrival time at the vessel's destination, and location information.

[0048] The departure and arrival points may include any area or port in the sea area where the ship operates, and are not limited to the above examples.

[0049] The operation plan information may refer to information related to a ship operation plan, such as a departure time of the ship at a departure point, an arrival time of the ship at a destination point, the latitude and longitude of the departure point, the latitude and longitude of the destination point, and tank requirements of the ship at the destination point.

[0050] Meanwhile, the operation plan information can be received by user input, but can also refer to preset information. For example, if the user sets the departure point and arrival point of the ship, the arrival time of the ship, the latitude and longitude of the departure point, the latitude and longitude of the arrival point, the tank requirements of the ship at the arrival point, etc. corresponding to the set departure point and arrival point can be received from the preset information. Meanwhile, the operation plan information is not limited to the above examples.

[0051] The recommended navigation information is information about a navigation route generated based on the navigation plan information, and may include one or more of position information for each navigation section of the ship, speed information for each navigation section of the ship, and environmental information for each navigation section of the ship, where the position information includes latitude and longitude for each navigation section of the ship, the speed information includes speed for each navigation section of the ship, and the environmental information includes one or more of weather and climate information, tidal current information, sea state information, and ocean current information for each navigation section of the ship.

[0052] On the other hand, the recommended navigation information may be information about a navigation route that is generated by acquiring location-specific weather and climate information and sea condition information based on the locations of departure and arrival points and the departure and arrival times of the ship, and taking into account the propulsion resistance and BOG generation amount of the ship calculated based on the location-specific weather and climate information and sea condition information. However, the recommended navigation information is not limited to the above examples.

[0053] The navigation route may be a route that requires the lowest fuel consumption, for example, but is not limited to, a route that minimizes the ship's propulsion resistance and the amount of BOG generated.

[0054] Meanwhile, the processor 110 can obtain recommended voyage information as output data by inputting operation plan information into a recommended voyage information generation model as input data. For example, the processor 110 can obtain recommended voyage information by inputting departure time / arrival time and departure / arrival location information related to the ship's location into the recommended voyage information generation model. Here, the recommended voyage information generation model may include information such as a speed-fuel volume performance function of the ship's propulsion engine, a power-fuel volume performance function of the ship's power generation engine, a power-fuel volume performance function of the ship's compressor / pump / reliquefaction unit / GCU / subcooler, and a speed-power generation volume performance function of a shaft generator. As another example, the recommended voyage information can be obtained using a voyage information generation model, as described below, but is not limited to this.

[0055] Meanwhile, the processor 110 can acquire environmental information related to the vessel's navigation route based on the navigation plan information. The processor 110 can generate recommended navigation information based on the navigation plan information and the environmental information, using the fuel consumption and BOG generation amount related to the vessel's navigation route as a criterion. For example, the processor 110 can acquire location-specific meteorological and sea condition information based on the vessel's departure time / arrival time, the latitude and longitude of the departure point / arrival point, and the tank requirements of the destination, and predict the vessel's propulsive resistance and BOG generation amount based on the vessel's departure time / arrival time, the latitude and longitude of the departure point / arrival point, the tank requirements of the destination, and the location-specific meteorological and sea condition information, thereby generating recommended navigation information related to the vessel's navigation route that minimizes the vessel's propulsive resistance and BOG generation amount.

[0056] The processor 110 can predict the amount of BOG generated by the ship and the tank pressure value of the ship based on the recommended voyage information. Specifically, the processor 110 can predict the amount of BOG generated by the ship based on at least one of position information for each voyage section, speed information for each voyage section, and environmental information for each voyage section. The processor 110 can predict the tank pressure value of the ship based on at least one of position information for each voyage section, speed information for each voyage section, environmental information for each voyage section, and a preset liquefied gas consumption amount.

[0057] Meanwhile, the processor 110 can obtain the BOG generation amount of the ship using a prediction model. As an example, the processor 110 can obtain the BOG generation amount of the ship as output data by inputting position information for each voyage section, speed information for each voyage section, and environmental information for each voyage section into the prediction model as input data. As another example, the processor 110 can obtain a BOG generation amount prediction value using a BOG generation amount prediction model described below.

[0058] Meanwhile, the processor 110 may acquire the tank pressure value of the ship using a prediction model. As an example, the processor 110 may acquire the tank pressure value of the ship as output data by inputting position information for each voyage section, speed information for each voyage section, environmental information for each voyage section, and a preset liquefied gas consumption amount into the prediction model as input data. As another example, the processor 110 may acquire the tank pressure value using a tank pressure prediction model described below.

[0059] The liquefied gas consumption may be derived based on the gas consumption of the propulsion engine, the gas consumption of the power generation engine, the gas consumption of the gas combustion unit, and the gas consumption of the reliquefaction unit. As an example, the liquefied gas consumption may be the sum of one or more of the gas consumption of the propulsion engine, the gas consumption of the power generation engine, the gas consumption of the combustion unit (GCU), the gas consumption of the reliquefaction unit, the gas consumption of the compressor, and the gas consumption of the pump. As another example, the liquefied gas consumption may be, but is not limited to, the sum of the gas consumption of the ship itself and the gas consumption of equipment installed on the ship.

[0060] The processor 110 can then obtain optimal voyage information for controlling the operation of the vessel based on the BOG generation rate and tank pressure value. For example, the processor 110 can predict the vessel's speed for each section and the usage of equipment installed on the vessel based on the minimum liquefied gas consumption rate during the voyage. Here, the equipment may include a propulsion engine, a generator engine, a gas combustion unit, a reliquefaction unit, a shaft generator, a subcooler, etc. installed on the vessel. The processor 110 can then calculate the amount of fuel for the vessel's propulsion engine based on the predicted vessel's speed for each section and calculate the gas consumption of the equipment based on the predicted equipment usage. The processor 110 can then determine the liquefied gas consumption rate based on the calculated propulsion engine fuel amount and the gas consumption of the equipment. Finally, the processor can obtain optimal voyage information including the vessel's BOG generation rate, the vessel's tank pressure value, the vessel's speed for each section of the voyage, the vessel's liquefied gas consumption rate, and the usage of the equipment installed on the vessel.

[0061] Meanwhile, the processor 110 can generate nth intermediate voyage information for ship operation control based on the BOG generation rate and tank pressure value. Then, the processor 110 can determine the n+1th intermediate voyage information as optimal voyage information based on a comparison between the nth intermediate voyage information and the n+1th intermediate voyage information, where n may be a natural number equal to or greater than 1.

[0062] The method for generating the nth intermediate voyage information may be the same as the method for generating the optimal voyage information described above. The intermediate voyage information may include the amount of BOG generated by the ship, the tank pressure value of the ship, the speed of the ship for each section of the voyage, the liquefied gas consumption of the ship, and the usage of equipment installed on the ship. For example, the processor 110 may predict the speed of the ship for each section and the usage of equipment installed on the ship so as to minimize the liquefied gas consumption during the voyage. Here, the equipment may include a propulsion engine, a generator engine, a gas combustion unit, a reliquefaction unit, a shaft generator, a supercooler, etc. installed on the ship. The processor 110 may then calculate the amount of fuel for the ship's propulsion engine based on the predicted speed of the ship for each section, and calculate the gas consumption of the equipment based on the predicted usage of the equipment. The processor 110 may then determine the liquefied gas consumption based on the calculated fuel amount of the propulsion engine and the gas consumption of the equipment. Finally, the processor can acquire first intermediate voyage information including the amount of BOG generated by the ship, the tank pressure value of the ship, the speed of the ship by each voyage section, the liquefied gas consumption amount of the ship, and the usage amount of the equipment installed on the ship. However, the method of acquiring the intermediate voyage information is not limited to the above example.

[0063] The (n+1)th intermediate voyage information may be generated based on one or more updated values ​​of the speed information for each voyage section and the liquefied gas consumption amount included in the (n)th intermediate voyage information. As an example, the (n+1)th intermediate voyage information may be the (n+1)th intermediate voyage information related to the operation control of the vessel obtained based on the predicted BOG generation amount and tank pressure value, which are obtained by predicting the BOG generation amount and tank pressure value of the vessel based on the updated value of the speed information for each voyage section included in the (n)th intermediate voyage information. Here, the updated value of the speed information for each voyage section may be the speed for each voyage section included in the (n+1)th intermediate voyage information.

[0064] The processor 110 can predict the ship's BOG generation amount and the ship's tank pressure value using one or more of the speed and liquefied gas consumption amount for each voyage section included in the nth intermediate voyage information as update values, and obtain the n+1th intermediate voyage information regarding the ship's operation control based on the predicted BOG generation amount and tank pressure value.

[0065] Meanwhile, the processor 110 may determine the (n+1)th intermediate voyage information as the optimum voyage information in response to the difference value calculated based on the (n)th intermediate voyage information and the (n+1)th intermediate voyage information being equal to or less than a preset threshold value, and may generate the (n+2)th intermediate voyage information by updating one or more of the speed information per voyage section and the liquefied gas consumption amount included in the (n+1)th intermediate voyage information in response to the difference value exceeding a preset threshold value.

[0066] Meanwhile, the processor 110 may control the ship according to a preset navigation method using the optimal navigation information, where the navigation method may include a navigation method that minimizes the liquefied gas consumption of the ship.

[0067] Each of the generative and predictive models described above may be a trained machine learning model, for example, the predictive model may be a deep learning model that uses an artificial neural network.

[0068] A machine learning model refers to a statistical learning algorithm or a structure that executes the algorithm, which is based on machine learning technology and the structure of a biological neural network in cognitive science.

[0069] For example, a machine learning model can exhibit problem-solving capabilities by learning in a manner similar to a biological neural network, in which nodes, which are artificial neurons formed by combining synapses, repeatedly adjust synaptic weights and reduce the error between the correct output corresponding to a specific input and the inferred output. For example, machine learning models can include any probability model, neural network model, etc. used in artificial intelligence learning methods such as machine learning and deep learning.

[0070] For example, the machine learning model may be implemented as a multilayer perceptron (MLP) consisting of multiple nodes and connections between them. The machine learning model according to an embodiment of the present invention may be implemented using one of various artificial neural network model structures, including an MLP. For example, the machine learning model may include an input layer that receives input signals or data from an external device, an output layer that outputs output signals or data corresponding to the input data, and one or more hidden layers that are positioned between the input and output layers and receive signals from the input layer, extract characteristics, and transmit the extracted characteristics to the output layer. The output layer receives signals or data from the hidden layer and outputs the signals or data to the external device.

[0071] A specific example of the operation of the processor 110 according to an embodiment will be described with reference to FIGS.

[0072] Processor 110 may be embodied as an array of logic gates or as a combination of a general-purpose microprocessor and memory storing programs that can be executed by the microprocessor. For example, processor 110 may include a general-purpose processor, a central processing unit (CPU), a microprocessor, a digital signal processor (DSP), a controller, a microcontroller, a state machine, etc. In some environments, processor 110 may also include an application-specific semiconductor (ASIC), a programmable logic device (PLD), a field-programmable gate array (FPGA), etc. For example, processor 110 may refer to a combination of processing devices, such as a combination of a digital signal processor (DSP) and a microprocessor, a combination of multiple microprocessors, a combination of one or more microprocessors coupled with a digital signal processor (DSP) core, or any other such configuration.

[0073] The memory 120 may include any non-transitory computer-readable storage medium. For example, the memory 120 may include a non-volatile mass storage device such as a random access memory (RAM), a read only memory (ROM), a disk drive, a solid state drive (SSD), or a flash memory. For another example, the non-volatile mass storage device such as a ROM, an SSD, a flash memory, or a disk drive may be a permanent storage device separate from the memory. The memory 120 may also store an operating system (OS) and at least one program code (e.g., code for the processor 110 to execute the operations described below with reference to FIGS. 4 to 36).

[0074] These software components may be loaded from a computer-readable recording medium separate from memory 120. Such a separate computer-readable recording medium may be a recording medium directly connectable to user terminal 100, and may include, for example, a computer-readable recording medium such as a floppy drive, a disk, a tape, a DVD / CD-ROM drive, or a memory card. Alternatively, the software components may be loaded into memory 120 via communication module 140 rather than a computer-readable recording medium. For example, at least one program may be loaded into memory 120 based on a computer program (e.g., a computer program for processor 110 to execute the operations described below with reference to FIGS. 4 to 9) installed by a file provided via communication module 140 by a developer or a file distribution system that distributes application installation files.

[0075] The input / output interface 130 may be a means for interfacing with an input or output device (e.g., a keyboard, a mouse, etc.) that may be connected to the user terminal 100 or included in the user terminal 100. The input / output interface 130 may be configured separately from the processor 110, but is not limited thereto, and the input / output interface 130 may also be configured to be included in the processor 110.

[0076] The communication module 140 may provide a configuration or function for the server 20 and the user terminal 100 to communicate with each other via a network. The communication module 140 may also provide a configuration or function for the user terminal 100 to communicate with other external devices. For example, control signals, instructions, data, etc. provided under the control of the processor 110 may be transmitted to the server 20 and / or external devices via the communication module 140 and the network.

[0077] Meanwhile, although not shown in FIG. 2, the user terminal 100 may further include a display device. For example, the display device may be implemented as a touch screen. Alternatively, the user terminal 100 may be connected to an independent display device via wired or wireless communication to transmit and receive data therebetween. For example, recommended navigation information or optimal navigation information may be provided via the display device.

[0078] FIG. 3 is a configuration diagram illustrating an example of a server according to an embodiment.

[0079] 3, server 200 includes a processor 210, a memory 220, and a communication module 230. For convenience of explanation, only components related to the present invention are shown in FIG. 3. Therefore, in addition to the components shown in FIG. 3, other general-purpose components may also be included in server 200. It will be obvious to those skilled in the art to which the present invention pertains that processor 210, memory 220, and communication module 230 shown in FIG. 3 may be embodied as independent devices.

[0080] The processor 210 can generate recommended navigation information regarding the navigation route of the ship based on navigation plan information regarding the ship's departure and arrival points. The processor 210 can also predict the amount of BOG generated by the ship and the tank pressure value of the ship based on the recommended navigation information. The processor 210 can also obtain optimal navigation information regarding operation control of the ship based on the amount of BOG generated and the tank pressure value. The processor 210 can also control the ship using a preset navigation method using the optimal navigation information.

[0081] 2 may be executed by the processor 210. In this case, the user terminal 100 may output the information transmitted from the server 200 via a display device.

[0082] Meanwhile, an implementation example of the processor 210 is the same as the implementation example of the processor 110 described above with reference to FIG. 2, and therefore a detailed description thereof will be omitted.

[0083] The memory 220 may store various data, such as data necessary for the operation of the processor and data generated in response to the operation of the processor 210. Furthermore, the memory 220 may store an operating system (OS) and at least one program (e.g., a program necessary for the operation of the processor 210).

[0084] Meanwhile, the implementation of the memory 220 is the same as the implementation of the memory 120 described above with reference to FIG. 2, and therefore a detailed description thereof will be omitted.

[0085] The communication module 230 may provide a configuration or function for the server 200 and the user terminal 100 to communicate with each other via a network. The communication module 140 may also provide a configuration or function for the server 200 to communicate with other external devices. For example, control signals, instructions, data, etc. provided under the control of the processor 210 may be transmitted to the user terminal 100 and / or external devices via the communication module 230 and the network.

[0086] FIG. 4 is a flowchart illustrating an example of a method for optimizing ship operation according to an embodiment.

[0087] Referring to Figure 4, the method for optimizing ship operation is composed of steps that are processed in time series by the user terminal 10, 100 or processor 110 shown in Figures 1 and 2. Therefore, even if the content is omitted below, the content described above regarding the user terminal 10, 100 or processor 110 shown in Figures 1 and 2 can also be applied to the method for optimizing ship operation in Figure 4.

[0088] Also, as described above with reference to FIGS. 1 and 2, at least one of the steps of the method for optimizing vessel operations of FIG. 4 may be processed by the server 20, 200 or the processor 210.

[0089] In step 310, processor 110 generates recommended voyage information for the vessel's route based on the vessel's planned departure and arrival points.

[0090] For example, the processor 110 can generate recommended voyage information as output data by inputting the ship's departure time at the departure point, the ship's arrival time at the destination point, the latitude and longitude of the departure point, and the latitude and longitude of the destination point into the recommended voyage information generation model. The recommended voyage information generation model, which includes information such as the ship's propulsion engine speed-fuel volume performance function, the ship's power generation engine power-fuel volume performance function, the ship's compressor / pump / reliquefaction unit / GCU / subcooler power-fuel volume performance function, and the shaft generator speed-power generation volume performance function, can derive the most fuel-efficient route by analyzing the ship's resistance performance according to the ship's propulsion resistance. Here, fuel efficiency may refer to the amount of fuel consumed by the ship per unit mileage or per unit time.

[0091] As another example, the processor 110 can obtain location-specific weather, climate, and sea condition information based on the departure time of the vessel at the departure point, the arrival time of the vessel at the destination point, the latitude and longitude of the departure point, the latitude and longitude of the destination point, and the tank requirements at the destination point, and can predict the vessel's propulsion resistance and BOG generation amount based on the departure time / arrival time of the vessel, the latitude and longitude of the departure point / arrival point, the tank requirements at the destination point, the location-specific weather, climate, and sea condition information, and generate recommended navigation information regarding the navigation route that will minimize the vessel's propulsion resistance and BOG generation amount.

[0092] Hereinafter, an example in which the processor generates recommended voyage information based on the operation plan information will be described with reference to FIG.

[0093] FIG. 5 is a flowchart illustrating an example of how a processor according to an embodiment generates recommended voyage information based on operation plan information.

[0094] 5, the processor 110 acquires environmental information related to the vessel's navigation route based on the navigation plan information in step 410. For example, the processor 110 may acquire environmental information related to the vessel's navigation route based on one or more of the departure time of the vessel at the departure point, the arrival time of the vessel at the destination point, the latitude and longitude of the departure point, the latitude and longitude of the destination point, and the vessel's tank requirements at the destination point.

[0095] Here, the environmental information may include one or more of meteorological and climate information, tidal current information, sea state information, and ocean current information for each location included in the navigation route of the ship. Meanwhile, the environmental information may refer to information about the environment in the navigation route of the ship and is not limited to the above examples.

[0096] In step 420, the processor 110 generates recommended voyage information based on the ship's navigation plan information and environmental information, using the fuel consumption and BOG generation amount for the ship's navigation route as a criterion. For example, the processor 110 can generate recommended voyage information for a navigation route that minimizes the ship's propulsive resistance and BOG generation amount by predicting the ship's propulsive resistance and BOG generation amount based on the ship's departure time / arrival time, the latitude and longitude of the departure point / arrival point, the tank requirements of the destination point, location-specific weather and climate information, and sea condition information. Here, the processor 110 can predict the BOG generation amount for the ship by taking into account changes in outside air temperature.

[0097] 4, in step 320, the processor 110 predicts the amount of BOG generated by the vessel and the tank pressure value of the vessel based on the recommended voyage information. The obtained amount of BOG generated and tank pressure value may be, but are not limited to, time-series data for each voyage section.

[0098] As an example, the processor 110 can predict the amount of BOG generated by the ship based on the position information for each voyage section and the speed information for each voyage section included in the recommended voyage information, and can predict the tank pressure value of the ship based on a preset liquefied gas consumption amount.

[0099] As another example, the processor 110 can use a predictive model to obtain the BOG generation rate of the vessel and the tank pressure value of the vessel.

[0100] However, the method of acquiring the amount of BOG generated by the ship and the tank pressure value of the ship is not limited to the above-mentioned example.

[0101] Hereinafter, with reference to FIG. 6, an example will be described in which the processor predicts the amount of BOG generated by the ship and the tank pressure value of the ship based on the recommended voyage information.

[0102] FIG. 6 is a flowchart illustrating an example in which a processor according to an embodiment predicts the amount of BOG generated by a ship and the tank pressure value of the ship based on recommended voyage information.

[0103] Referring to FIG. 6, in step 510, the processor 110 predicts the amount of BOG generated by the ship based on the position information for each voyage section, the speed information for each voyage section, and the environmental information for each voyage section.

[0104] As one example, the processor 110 can obtain the BOG generation amount of the ship as output data by inputting position information for each voyage section, speed information for each voyage section, and environmental information for each voyage section into a prediction model as input data. As another example, the processor 110 can obtain the BOG generation amount using a BOG generation amount prediction model described below. Here, the BOG generation amount prediction model includes multiple deep learning models and predicts the BOG generation amount over time using ship operation data. In addition, the BOG generation amount prediction model can include multiple deep learning models corresponding to multiple navigation modes based on the liquefied gas water level in the tank. However, the method of obtaining the BOG generation amount is not limited to the above example.

[0105] In step 520, the processor 110 predicts the tank pressure value of the vessel based on the position information for each voyage segment, the speed information for each voyage segment, the environmental information for each voyage segment, and the preset liquefied gas consumption amount.

[0106] As one example, the processor 110 can obtain the tank pressure value of the vessel as output data by inputting position information for each voyage section, speed information for each voyage section, environmental information for each voyage section, and a preset liquefied gas consumption amount as input data into the prediction model. As another example, the processor 110 can obtain the tank pressure value using a tank pressure prediction model described below. Here, the tank pressure prediction model outputs a predicted value of the vessel's tank pressure over time. The tank pressure prediction model can also include multiple deep learning models corresponding to multiple navigation modes based on the liquefied gas water level in the tank. However, the method of obtaining the tank pressure value is not limited to the above example.

[0107] Referring again to FIG. 4, in step 330, the processor 110 obtains optimal voyage information for controlling the operation of the vessel based on the BOG generation rate and the tank pressure value.

[0108] For example, the processor 110 may predict the vessel's speed for each section and the usage of equipment installed on the vessel based on the minimum liquefied gas consumption during the voyage. The processor 110 may then calculate the amount of fuel for the vessel's propulsion engine based on the predicted vessel's speed for each section, and calculate the gas consumption of the equipment based on the predicted equipment usage. The processor 110 may then determine the liquefied gas consumption based on the calculated amount of fuel for the propulsion engine and the gas consumption of the equipment. Finally, the processor may obtain optimal voyage information including the vessel's BOG generation rate, the vessel's tank pressure value, the vessel's speed for each section of the voyage, the vessel's liquefied gas consumption, and the usage of the equipment installed on the vessel.

[0109] As another example, the processor 110 can generate optimal voyage information that takes constraints into account using a navigation optimization model, which will be described later.

[0110] However, the method for generating the optimum voyage information is not limited to the above example.

[0111] Hereinafter, with reference to FIG. 7, an example will be described in which the processor acquires optimal voyage information based on the amount of BOG generated and the tank pressure value.

[0112] FIG. 7 is a flowchart illustrating an example of a process in which a processor according to an embodiment acquires optimal voyage information based on the amount of BOG generated and the tank pressure value.

[0113] Referring to FIG. 7, in step 610, the processor 110 generates n-th intermediate voyage information related to the operation control of the vessel based on the BOG generation amount and the tank pressure value.

[0114] For example, the processor 110 may predict the vessel's speed for each section and the usage of equipment installed on the vessel to minimize liquefied gas consumption during the voyage. Here, the equipment may include a propulsion engine, a generator engine, a gas combustion unit, a reliquefaction unit, a shaft generator, a subcooler, etc., installed on the vessel. The processor 110 may calculate the amount of fuel for the vessel's propulsion engine based on the predicted vessel's speed for each section, and calculate the gas consumption of the equipment based on the predicted equipment usage. The processor 110 may then determine the liquefied gas consumption based on the calculated propulsion engine fuel amount and the gas consumption of the equipment. Finally, the processor may obtain nth intermediate voyage information including the vessel's BOG generation rate, the vessel's tank pressure value, the vessel's speed for each section, the vessel's liquefied gas consumption, and the usage of the equipment installed on the vessel.

[0115] As another example, the processor 110 can generate nth intermediate voyage information that takes constraints into account using a navigation optimization model described below.

[0116] However, the method for generating the nth intermediate voyage information is not limited to the above example.

[0117] In step 620, the processor 110 determines the n+1th intermediate voyage information as the optimum voyage information based on a comparison of the nth intermediate voyage information and the n+1th intermediate voyage information using a preset threshold as a criterion.

[0118] As an example, the processor 110 may determine the n+1 intermediate voyage information as the optimal voyage information based on a comparison of the section-specific speed included in the nth intermediate voyage information and the section-specific speed included in the n+1th intermediate voyage information, using a preset threshold as a criterion.

[0119] As another example, the processor 110 may determine the n+1 intermediate voyage information as the optimal voyage information based on a comparison of the liquefied gas consumption amount included in the nth intermediate voyage information and the liquefied gas consumption amount included in the n+1th intermediate voyage information, using a preset threshold as a criterion.

[0120] However, the method for determining the optimum navigation information is not limited to the above example.

[0121] Hereinafter, with reference to FIG. 8, an example in which the processor acquires optimal voyage information based on the amount of BOG generated and the tank pressure value will be described.

[0122] FIG. 8 is a flowchart illustrating another example in which a processor according to an embodiment acquires optimal voyage information based on the amount of BOG generated and the tank pressure value.

[0123] Referring to Figure 8, in step 710, the processor 110 determines the n+1th intermediate voyage information as the optimal voyage information in response to the difference value calculated based on the nth intermediate voyage information and the n+1th intermediate voyage information being less than or equal to a predetermined threshold.

[0124] For example, the processor 110 may calculate a difference value based on the section speed included in the nth intermediate voyage information and the section speed included in the (n+1)th intermediate voyage information, and may determine the (n+1)th intermediate voyage information as the optimal voyage information in response to the calculated difference value being 0.01 or less.

[0125] As another example, the processor 110 may calculate a difference value based on the liquefied gas consumption amount included in the nth intermediate voyage information and the liquefied gas consumption amount included in the n+1th intermediate voyage information, and may determine the n+1th intermediate voyage information as the optimal voyage information in response to the calculated difference value being 0.01 or less.

[0126] However, the method for determining the optimum navigation information is not limited to the above example.

[0127] In step 720, in response to the difference value exceeding a predetermined threshold, the processor 110 generates the (n+2)th intermediate voyage information by updating at least one of the speed information for each voyage section and the liquefied gas consumption amount included in the (n+1)th intermediate voyage information.

[0128] For example, if the difference between the section speed included in the nth intermediate voyage information and the section speed included in the n+1th intermediate voyage information exceeds 0.01, the processor 110 may generate the n+2th intermediate voyage information by updating the section speed included in the n+1th intermediate voyage information. The processor 110 may predict the amount of BOG generated by the ship and the tank pressure value of the ship based on the section speed included in the n+1th intermediate voyage information, as described above, and obtain the n+2th intermediate voyage information related to operation control of the ship based on the predicted BOG generation amount and tank pressure value. Here, as the speed for each section of the voyage is updated, the gas consumption of the propulsion engine and the gas consumption of the power generation engine according to the speed of the ship are updated, and therefore the liquefied gas consumption is also updated. In other words, the n+2th intermediate voyage information may be information generated based on the updated section speed and liquefied gas consumption.

[0129] As another example, if the difference between the liquefied gas consumption amount included in the nth intermediate voyage information and the liquefied gas consumption amount included in the n+1th intermediate voyage information exceeds 0.01, the processor 110 may generate the n+2nd intermediate voyage information by updating the voyage liquefied gas consumption amount included in the n+1th intermediate voyage information. The processor 110 may predict the tank pressure value of the ship as described above based on the liquefied gas consumption amount included in the n+1th intermediate voyage information, and obtain the n+2nd intermediate voyage information related to operation control of the ship based on the predicted BOG generation amount and tank pressure value. In other words, the n+2nd intermediate voyage information may be information generated based on the updated value of the liquefied gas consumption amount.

[0130] The above-mentioned preset value may be a value set by a user or developer, or may be a value obtained by the processor 110 by repeatedly generating the nth intermediate voyage information. As an example, the processor 110 may obtain optimal voyage information, set a correct value corresponding to the optimal voyage information, and obtain a preset value by associating the optimal voyage information with the correct value. As another example, the processor 110 may obtain optimal voyage information, calculate an error by associating the optimal voyage information with the correct value by using a loss function, and obtain a preset value based on the calculated error.

[0131] Another example of a method for optimizing ship operation will now be described with reference to FIG.

[0132] FIG. 9 is a flowchart illustrating another example of a method for optimizing ship operation according to an embodiment.

[0133] Steps 810 to 830 in Fig. 9 correspond to steps 310 to 330 in Fig. 4. Therefore, overlapping content regarding steps 810 to 830 will be omitted below.

[0134] In step 840, the processor 110 controls the ship in a preset operating manner using the optimal voyage information, where the operating manner may include an operating manner that minimizes the ship's liquefied gas consumption.

[0135] When acquiring the optimal voyage information, the processor 110 may predict the vessel's speed for each section and the usage of equipment installed on the vessel based on the minimum liquefied gas consumption during the voyage, calculate the amount of fuel for the vessel's propulsion engine based on the predicted vessel's speed for each section, and calculate the gas consumption of the equipment based on the predicted equipment usage. The processor 110 may then determine the liquefied gas consumption based on the calculated amount of fuel for the propulsion engine and the gas consumption of the equipment. Therefore, the vessel's BOG generation rate, vessel's tank pressure value, vessel's speed for each section, vessel's liquefied gas consumption, and vessel's equipment usage included in the optimal voyage information may be values ​​generated based on the minimum liquefied gas consumption during the voyage. In other words, when the processor 110 controls the vessel using the optimal voyage information acquired by the above-mentioned method, it may control the vessel in a navigation method that minimizes the vessel's liquefied gas consumption.

[0136] FIG. 10 is a diagram for explaining an example of a method for predicting the amount of BOG generated according to an embodiment.

[0137] 10, there is shown an example of a ship 1010 including at least one tank 1020. During operation of the ship 1010, BOG may be generated 1030 in the tank 1020.

[0138] There are several ways to treat BOG, including using it as fuel for the main engine or generator engine, reliquefaction, or incineration in a gas combustion unit (GCU). In order to treat BOG efficiently using the above methods, it is necessary to predict the amount of BOG generated.

[0139] Previously, there were limitations to accurately predicting the amount of BOG generated because the amount changed constantly depending on weather conditions, temperature, sloshing, etc. along the shipping route.

[0140] According to one embodiment of a method for predicting the amount of BOG generated by a ship, input data for a BOG generation amount prediction model is selected from operation data of the ship 1010. Then, the BOG generation amount prediction model is trained using the input data, and the BOG generation amount is predicted using the trained prediction model.

[0141] Therefore, it is possible to operate a liquefied gas carrier in accordance with the BOG generation amount guaranteed in the cargo transportation contract. In addition, since weather conditions change when a ship is operating, it is possible to derive a correlation between the BOG generation amount from a ship and weather conditions.

[0142] Hereinafter, a method and apparatus for predicting the amount of BOG generated from a ship according to an embodiment of the present disclosure will be described in detail with reference to FIGS. 2 and 11 to 17.

[0143] 2, the communication module 140 may receive information about existing flight data, information about current flight data, weather information, etc. from an external server or device. In addition, the memory 120 may store various data, such as existing flight data, current flight data, weather information, and data generated in response to the operation of the processor 110.

[0144] The processor 110 can control the operation of the device 100 by executing a program stored in the memory 120. As an example, the processor 110 can execute at least a part of a method for predicting the amount of BOG generated by a ship, which will be described with reference to Figures 11 to 17.

[0145] That is, the processor 110 selects input data for the BOG generation amount prediction model from existing operation data using an input data selection model, learns a BOG generation amount prediction model including multiple deep learning models using the input data, and can predict the BOG generation amount from the BOG generation amount prediction model learned using the ship's current operation data.

[0146] The processor 110 can use the input data selection model to select input data for the BOG generation prediction model from existing operation data.

[0147] For example, the processor 110 can calculate a correlation coefficient between some pre-stored operational data and the amount of BOG generated, learn an input data selection model using the calculated correlation coefficient, and use the learned input data selection model to select data whose correlation coefficient is greater than or equal to a predetermined value as input data.

[0148] In addition, at least one of the plurality of deep learning models may include a stacking model, and the stacking model may include a plurality of sub-deep learning models.

[0149] Processor 110 can train a BOG generation prediction model including multiple deep learning models using input data. For example, processor 110 can calculate correct answer data for training, train using the input data and the calculated correct answer data, and verify the BOG generation prediction model using other pre-stored operational data.

[0150] Here, the correct data can be calculated using at least one of the gas consumption amount, the gas temperature change value, and the heat insulating material temperature change value.

[0151] The processor 110 can predict the amount of BOG generation from a BOG generation prediction model trained using current operation data of the ship. For example, the processor 110 can output initial BOG generation prediction values ​​of each of a plurality of deep learning models and calculate a final BOG generation prediction value by applying different weights to each of the initial BOG generation prediction values.

[0152] Here, the processor 110 may apply the highest weight to the first predicted value of the BOG generation amount output from the stacking model among the first predicted values ​​of the BOG generation amount.

[0153] Hereinafter, a method for predicting the amount of BOG generated by a ship according to an embodiment of the present disclosure will be described in detail with reference to FIGS.

[0154] FIG. 11 is a flowchart illustrating an example of a method for predicting the amount of BOG generated in a ship according to an embodiment.

[0155] 11, the method for predicting the amount of BOG generated from a ship may include steps 1110 to 1130. However, the method is not limited thereto, and other general steps other than those shown in FIG. 11 may also be included in the method for predicting the amount of BOG generated from a ship. Furthermore, as described above with reference to FIGS. 2 and 10, at least one of the steps of the flowchart shown in FIG. 11 may be processed by the processor 110.

[0156] In step 1110, the processor 110 can use an input data selection model to select input data for the BOG generation prediction model from existing operational data.

[0157] For example, the processor 110 can calculate the correlation coefficient between existing operational data and the amount of BOG generated, learn an input data selection model using the calculated correlation coefficient, and use the learned input data selection model to select as input data data whose correlation coefficient is greater than or equal to a predetermined value.

[0158] First, the processor 110 calculates a correlation coefficient. For example, the processor 110 can calculate a linear correlation coefficient between existing operational data and the amount of BOG generated. Then, the processor 110 can use data from the existing operational data with a linear correlation coefficient of 0.1 or greater as learning data for the input data selection model. Here, the linear correlation coefficient may refer to the Pearson correlation coefficient. The Pearson correlation coefficient refers to the value obtained by dividing the covariance of two variables by the product of their standard deviations. However, examples of the linear correlation coefficient calculated by the processor 110 are not limited to the Pearson correlation coefficient described above.

[0159] The processor 110 learns the input data selection model. For example, the processor 110 can select input data from existing operational data using the learned input data selection model.

[0160] Here, the term "deep learning model" refers to a set of machine learning algorithms that use a layered algorithm structure based on a deep neural network in machine learning technology and cognitive science.

[0161] For example, a deep learning model may be composed of an input layer that receives input signals or data from the outside, an output layer that outputs output signals or data corresponding to the input data, and at least one hidden layer that is located between the input and output layers and receives signals from the input layer, extracts characteristics, and transmits them to the output layer. The output layer receives signals or data from the hidden layer and outputs them to the outside.

[0162] Therefore, the input data selection model can receive some pre-stored operational data and learn to extract data (such as gas consumption) with a linear correlation coefficient of 0.1 or greater.

[0163] The processor 110 also selects input data. For example, the processor 110 may select input data including gas consumption, cargo load (e.g., liquefied gas level in a tank), etc. Here, the gas consumption may be calculated using gas consumption in a main engine, a generator engine, a gas combustion unit, and a reliquefaction unit. The processor 110 may also select input data including gas consumption, cargo load (e.g., liquefied gas level in a tank), water temperature, air temperature, air pressure, ship speed, wave height, swell height, etc.

[0164] An example of how the processor 110 calculates the gas consumption amount will be described below with reference to FIG.

[0165] FIG. 12 is a diagram for explaining an example of calculating the gas consumption amount of a ship according to an embodiment.

[0166] 12, the processor 110 calculates the gas consumption 1210 using the gas consumption of the propulsion engine 1220, the gas consumption of the power generation engine 1230, the gas consumption of the gas combustion device 1240, and the gas consumption of the reliquefaction device 1250. As an example, the difference between the combined value of the gas consumption of the propulsion engine 1220, the gas consumption of the power generation engine 1230, and the gas consumption of the gas combustion device 1240 and the gas consumption of the reliquefaction device 1250 can be calculated.

[0167] When the conventional gas vaporization amount was used as input data, there was no information on whether the vaporized gas was used, whether it was discharged, whether it was re-liquefied, etc., making it impossible to accurately predict the amount of BOG generated.

[0168] Therefore, by calculating the gas consumption amount 1210 instead of the gas vaporization amount and using this as input data, it is possible to predict the amount of BOG generated on an hourly basis, enabling efficient ship operation. It is also possible to check in real time whether the BOG generation standard stipulated in the cargo transportation contract is being met.

[0169] Referring again to FIG. 11, in step 1120, the processor 110 may use the input data to train a BOG generation prediction model that includes multiple deep learning models.

[0170] For example, at least one of the plurality of deep learning models may include a stacking model, and the stacking model may further include a sub-deep learning model.

[0171] Here, a stacking model refers to an algorithm that uses the output data of multiple sub-deep learning models as training data for one deep learning model.

[0172] An example of a BOG generation amount prediction model will be described below with reference to FIGS. 13A to 13C.

[0173] FIG. 13A is a configuration diagram illustrating an example of a stacking model according to an embodiment.

[0174] 13A , the stacking model 1310 may include a first sub-deep learning model 1311 to an n-th sub-deep learning model 1312 and an n+1 sub-deep learning model 1313, where n is a natural number equal to or greater than 3. As an example, the first sub-deep learning model 1311 to the n+1 sub-deep learning model 1313 may be the same deep learning model. Alternatively, the first sub-deep learning model 1311 to the n+1 sub-deep learning model 1313 may be different deep learning models. As another example, at least one of the first sub-deep learning model 1311 to the n-th sub-deep learning model 1312 may include multiple deep learning models.

[0175] For example, the processor 110 uses the same learning data to train a first sub-deep learning model 1311 to an n-th sub-deep learning model 1312. The trained sub-deep learning models 1311, 1312 output a first predicted value 1314 to an n-th predicted value 1315. The output first predicted value 1314 to n-th predicted value 1315 are used as learning data for an n+1-th sub-deep learning model 1313. The output data of the n+1-th sub-deep learning model 1313 is used as an m-th initial output value 1316 of a BOG generation amount prediction model for the ship 1010.

[0176] FIG. 13B is a configuration diagram showing an example of a BOG generation amount prediction model for a ship according to one embodiment.

[0177] 13B , the BOG generation amount prediction model 1320 of the ship 1010 may include a first deep learning model 1321 to an m-th deep learning model 1322, where m is a natural number equal to or greater than 4. For example, at least one of the first deep learning model 1321 to the m-th deep learning model 1322 may be the stacking model 1310.

[0178] The ship BOG generation prediction model includes multiple deep learning models, which can improve the prediction accuracy of ship BOG generation.

[0179] For example, diversity can be ensured by using the prediction results of multiple deep learning models. Therefore, by using various prediction results, the generalization performance of a ship BOG generation prediction model that includes multiple deep learning models can be improved.

[0180] Here, generalization refers to the ability of a deep learning model to make accurate predictions on new data other than the training data. In other words, it means maintaining accuracy on new data without overfitting or underfitting. Here, overfitting means that a deep learning model shows high accuracy only on the training data and low accuracy on data other than the training data, while underfitting means that a deep learning model cannot learn the training data sufficiently and shows low accuracy on both the training data and data other than the training data.

[0181] Furthermore, since the prediction results are independent of each other, even if the accuracy of one predicted value decreases, the accuracy can be improved by using a plurality of other highly accurate predicted values.

[0182] For example, the processor 110 can calculate correct answer data for learning, learn a BOG generation prediction model using the input data and the calculated correct answer data, and verify the BOG generation prediction model using some other pre-stored operational data.

[0183] For example, the processor 110 can calculate the correct data using at least one of the gas consumption amount, the gas temperature change value, and the insulation temperature change value.

[0184] First, the processor 110 calculates correct answer data and trains a BOG generation amount prediction model using the calculated correct answer data and input data. For example, the processor 110 can calculate correct answer data for supervised learning. The accuracy of the BOG generation amount prediction model can be improved by the processor 110 performing supervised learning on the BOG generation amount prediction model using the correct answer data.

[0185] Here, supervised learning is a machine learning method that uses training data to learn and output values ​​that are close to the correct data. Also, the performance of the supervised learning model is evaluated using test data.

[0186] An example of how processor 110 calculates correct answer data will be described below with reference to FIG.

[0187] FIG. 14 is a diagram illustrating an example of calculating correct answer data for learning according to an embodiment.

[0188] 14, the processor 110 can calculate correct data using a gas consumption amount 1420, a gas temperature change value 1430, and an insulation temperature change value 1440. As an example, the processor 110 can calculate the difference between the total value of the gas consumption amount 1420 and the gas temperature change value 1430 and the insulation temperature change value 1440 as correct data 1410. As another example, the processor 110 can calculate correct data using the gas consumption amount 1420, the gas temperature change value 1430, the insulation temperature change value 1440, and the gas emission amount (license gas cosumption).

[0189] Here, since the gas temperature inside the tank is affected by the temperature outside the tank, the gas temperature change value 1430 may be a gas temperature change value 1430 that takes into account both the gas temperature inside the tank and the temperature outside the tank.

[0190] The processor 110 can then verify the BOG generation prediction model using another portion of the pre-stored operation data. For example, the pre-stored operation data can be separated into learning data for the input data selection model and verification data for the ship's BOG generation prediction model. Furthermore, by verifying the ship's BOG generation prediction model, the performance of the model can be evaluated and further learning can be performed.

[0191] Hereinafter, with reference to FIGS. 15A and 15B, an example in which processor 110 separates pre-stored operation data into learning data and verification data will be described.

[0192] FIG. 15A is a diagram for explaining that the processor 110 according to one embodiment removes lost data from the data stored in the memory 120. As shown in FIG.

[0193] 15A, the processor 110 removes lost data 1512 from the operation data 1511 pre-stored in the memory 120. The lost data 1512 refers to a portion of the pre-stored operation data 1511 for which data has not been stored due to a measurement error. The lost data 1512 may affect the accuracy of the model, so the processor 110 can remove the lost data 1512.

[0194] Figure 15B is a diagram illustrating an example of separating a portion of pre-stored operation data into learning data for an input data selection model and another portion of the data into verification data for a ship BOG generation prediction model in one embodiment.

[0195] Then, the processor 110 separates the pre-stored operation data 1520 from which the lost data 1512 has been removed into learning data 1521 for the input data selection model and verification data 1522 for the ship BOG generation amount prediction model.

[0196] Here, the purpose of separating the pre-stored operation data 1520 into training data 1521 and validation data 1522 is to prevent the model from being insufficiently trained or from over-fitting. Here, over-fitting means that if the model learns too much from the training data, the accuracy is very high when the training data is input, but the accuracy of the model drops significantly when other data is input.

[0197] Referring again to FIG. 11, in step 1130, the processor 110 can predict the amount of BOG generation from a BOG generation prediction model trained using the current operating data of the ship.

[0198] First, the processor 110 acquires current operational data of the ship. For example, the current operational data of the ship may include gas consumption in the propulsion engine, gas consumption in the power generation engine, gas consumption in the gas combustion device, gas consumption in the reliquefaction device, temperature in the tank, pressure in the tank, cargo load (e.g., liquefied gas level in the tank), water temperature, air temperature, air pressure, wave height, and gas consumption of the ship. However, this is merely an example, and the current operational data of the ship is not limited to this. In addition, the communication module 140 may acquire data related to air temperature, air pressure, wave height, etc. from a weather information center.

[0199] An example of how the processor 110 acquires the current operation data of the ship will be described below with reference to FIGS. 16A and 16B.

[0200] FIG. 16A is a diagram illustrating an example of acquiring current operation data of a ship from outside the ship according to an embodiment.

[0201] 16A, the processor 110 may acquire current operating data of the vessel from the exterior of the vessel 1610. As an example, the processor 110 may acquire data related to water temperature, wave height, etc. from the exterior of the vessel 1610.

[0202] FIG. 16B is a diagram illustrating an example of acquiring current operation data of a ship from inside the ship according to an embodiment.

[0203] 16B, the processor 110 can acquire current operational data of the ship from inside the ship 1620. Specifically, data such as the temperature in the tank, the pressure in the tank, and the cargo load can be acquired from a tank 1621 inside the ship 1620. In addition, data such as the gas consumption in the propulsion engine from a propulsion engine 1622, the gas consumption in the power generation engine from a power generation engine 1623, the gas consumption from a gas combustion device 1624, and the gas consumption from a reliquefaction device 1625 can be acquired. However, this is merely an example, and the location from which the current operational data of the ship described above is acquired is not limited to this.

[0204] For example, the processor 110 can output the initial predicted value of the BOG generation amount for each of the multiple deep learning models, and calculate the final predicted value of the BOG generation amount by applying different weights to each of the initial predicted values ​​of the BOG generation amount.

[0205] For example, the processor 110 may apply the highest weight to the first predicted value of BOG generation output from the stacking model among the first predicted values ​​of BOG generation.

[0206] First, the processor 110 can output the initial BOG generation amount prediction value of each of the multiple deep learning models. For example, the processor 110 can output the first initial output value to the mth initial output value as the output values ​​of the first deep learning model to the mth deep learning model included in the BOG generation amount prediction model for the ship. In addition, the second deep learning model may be a stacking model.

[0207] The processor 110 may apply different weights to the initial predicted values ​​of the BOG generation amount to calculate the final predicted value of the BOG generation amount. The processor 110 may also apply the highest weight to the initial predicted value of the BOG generation amount output from the stacking model.

[0208] Here, weights represent the importance of each piece of data, and applying weights to each piece of data can improve the learning accuracy of the model. Therefore, the more important the data, the higher the weight that can be applied.

[0209] As one example, the processor 110 may calculate a final predicted value of BOG generation amount by applying the same weight to each of the initial predicted values ​​of BOG generation amount. As another example, the processor 110 may calculate a final predicted value of BOG generation amount by applying different weights to each of the initial predicted values ​​of BOG generation amount. As yet another example, the processor 110 may apply the highest weight to the initial predicted value of BOG generation amount output from the stacking model among the initial predicted values ​​of BOG generation amount. Because the stacking model includes a sub-deep learning model and may therefore be highly important, a high weight may be applied to the initial predicted value of BOG generation amount output from the stacking model.

[0210] Hereinafter, an example of predicting the amount of BOG generation from a ship BOG generation amount prediction model will be described with reference to FIG.

[0211] FIG. 17 is a diagram illustrating an example of calculating a final predicted value of the amount of BOG generation by applying a weight to an initial predicted value of the amount of BOG generation according to an embodiment.

[0212] 17, the processor 110 may output initial output values ​​of the first deep learning model 1720 to the m-th deep learning model 1730 included in the ship BOG generation amount prediction model 1710. Here, the m-th deep learning model 1730 may be a stacking model. Also, the m-th initial output value 1750 may be an initial output value output from the stacking model.

[0213] Of the initial output values ​​of the first deep learning model 1720 to the m-th deep learning model 1730 included in the ship BOG generation amount prediction model 1710, the stacking model, that is, the m-th deep learning model 1730, may have the highest accuracy.

[0214] Weights W11770 to W1750 are assigned to the first initial output value 1740 to the mth initial output value 1750. TIFF2025538486000002.tif331780 can be applied. TIFF2025538486000003.tif541770~Weight W m 1780 may have the same value. m 1770~Weight W m The weights W11770 to W1780 may be different values. m Among 1780, the weight W of the mth initial output value output from the stacking model m 1780 may be the largest value.

[0215] Therefore, the processor 110 can obtain a final output value 1790, i.e., a predicted BOG generation amount value, that is close to the actual BOG generation amount from the BOG generation amount prediction model of the ship.

[0216] FIG. 18 is a diagram for explaining an example of a method for predicting a tank pressure of a ship according to an embodiment.

[0217] 18, there is shown an example of a vessel 1810 including at least one tank 1820. During the operation of the vessel 1810, BOG may be generated in the tank 1820, which may cause a pressure increase 1830 inside the tank 1820.

[0218] In this way, BOG is generated, increasing the volume of the liquefied gas stored in the tank 1820, ultimately increasing the pressure inside the tank 1820. The increase in pressure inside the tank 1820 reduces the safety of the tank 1820 and may lead to the tank 1820 exploding. To prevent this danger, BOG is released, but this process may cause environmental pollution.

[0219] Previously, there were limitations to accurately predicting the pressure in Tank 1820 because the pressure in Tank 1820 changes constantly depending on weather conditions, temperature, sloshing, etc. along the route.

[0220] According to one embodiment of the method for predicting tank pressure of a ship, multiple deep learning models are trained using pre-stored actual operation data, one of multiple navigation modes is selected depending on the water level of liquefied gas in the tank, and the pressure in a tank 1820 of a ship 1810 is predicted using the deep learning model corresponding to the selected navigation mode.

[0221] Therefore, it is possible to deal with future situations by predicting the real-time pressure inside the tank 1820. In addition, according to the means for solving the problems of the present invention, it is possible to efficiently control the usage of equipment required for operating the ship 1810.

[0222] Hereinafter, a method and apparatus for predicting tank pressure of a ship according to an embodiment of the present disclosure will be described in detail with reference to FIGS. 2 and 19 to 24.

[0223] 2, the communication module 140 can receive information about existing navigation data, information about current navigation data, recommended navigation information, weather information, etc. from an external server or external device. In addition, the memory 120 can store various data such as existing navigation data, current navigation data, recommended navigation information, weather information, and data generated in response to the operation of the processor 110.

[0224] The processor 110 can control the operation of the device 100 by executing a program stored in the memory 120. As an example, the processor 110 can execute at least a part of a method for predicting tank pressure of a ship, which will be described with reference to Figures 19 to 24.

[0225] In other words, the processor 110 can learn a deep learning model corresponding to each of a plurality of navigation modes for the ship's voyage using pre-stored actual operation data, select one of the plurality of navigation modes according to the water level of the liquefied gas in the tank, and predict the tank pressure from the recommended navigation information using the deep learning model corresponding to the selected navigation mode.

[0226] First, the processor 110 can use pre-stored actual operational data to train deep learning models corresponding to each of a plurality of navigation modes related to the navigation of the ship.

[0227] For example, the processor 110 can acquire correct answer data for learning, learn a deep learning model corresponding to each of a plurality of navigation modes using a portion of the pre-stored actual operation data and the correct answer data, and verify the learned deep learning model using another portion of the pre-stored actual operation data.

[0228] Here, learning can be performed using an error and back propagation algorithm between the ground truth data and the output data of the deep learning model.

[0229] Additionally, the processor 110 can select one of a number of sailing modes depending on the level of liquefied gas in the tank.

[0230] For example, the processor 110 can measure the level of liquefied gas in the tank and select a first mode from among a plurality of navigation modes if the water level is above a predetermined height, and select a second mode from among a plurality of navigation modes if the water level is below the predetermined height.

[0231] The processor 110 can predict tank pressure from the recommended voyage information using a deep learning model corresponding to the selected voyage mode.

[0232] For example, the processor 110 may obtain recommended voyage information for a vessel, obtain a plurality of intermediate values ​​using the recommended voyage information as input data, and apply weights to the plurality of intermediate values ​​to obtain a predicted pressure value for the tank.

[0233] FIG. 19 is a flowchart illustrating an example of a method for predicting a tank pressure of a ship according to an embodiment.

[0234] 19, the method for predicting tank pressure of a ship may include steps 1910 to 1930. However, the method is not limited thereto, and other general steps other than the steps shown in FIG. 19 may also be included in the method for predicting tank pressure of a ship. Furthermore, as described above with reference to FIGS. 2 and 18, at least one of the steps of the flowchart shown in FIG. 19 may be processed by the processor 110.

[0235] In step 1910, the processor 110 can use pre-stored actual operational data to train deep learning models corresponding to each of a plurality of navigation modes related to the navigation of the vessel.

[0236] For example, the processor 110 can acquire correct answer data for learning, learn a deep learning model corresponding to each of a plurality of navigation modes using a portion of the pre-stored actual operation data and the correct answer data, and verify the learned deep learning model using another portion of the pre-stored actual operation data.

[0237] Here, the ground truth data refers to the target that the deep learning model is trying to predict, as data necessary for supervised learning of the deep learning model. The ground truth data is used to measure the performance of the deep learning model during the learning process and to evaluate the results predicted by the deep learning model.

[0238] In addition, supervised learning is one of the machine learning methods that aims to learn the relationship between input data and correct answer data, and when new input data is input, to predict the result for the new input data.

[0239] First, processor 110 acquires correct answer data for learning.

[0240] For example, the processor 110 may acquire, as the correct answer data, actual tank pressure values ​​under sailing conditions similar to those of the current voyage from pre-stored actual operation data. The sailing conditions may include weather information, sailing speed by section, total average speed, and the amount of liquefied gas in the tank 1820 at the time of arrival / departure.

[0241] For example, the pre-stored actual operational data may include the vessel 1810's latitude, longitude, liquefied gas discharge amount, speed, wave height, wave period, wave direction, swell height, swell period, swell direction, wind speed, air temperature, air pressure and water temperature.

[0242] For example, the actual tank pressure value may be a tank pressure value for each section of a voyage, or may be an average tank pressure value during operation.

[0243] For example, the processor 110 uses a portion of pre-stored actual operation data and correct answer data to train a deep learning model corresponding to each of a plurality of navigation modes.

[0244] Thus, a deep learning model corresponding to each of the multiple voyage modes can be trained to predict the pressure in tank 1820 from the recommended voyage information.

[0245] For example, the processor 110 can separate pre-stored actual operational data and use it as training data and validation data.

[0246] Hereinafter, with reference to FIG. 20A, an example will be described in which processor 110 separates a part of pre-stored actual operation data into learning data and the other part into verification data.

[0247] FIG. 20A is a diagram illustrating an example of separating a part of pre-stored actual operation data into learning data and another part into verification data according to an embodiment.

[0248] 20A , processor 110 separates pre-stored actual operation data 2011 into a portion 2012 of pre-stored actual operation data and another portion 2013 of pre-stored actual operation data. The separated portion 2012 of pre-stored actual operation data is used as training data for the deep learning model, and the other portion 2013 is used as validation data for the deep learning model.

[0249] Hereinafter, with reference to FIG. 20B, an example of a process in which the processor 110 processes the pre-stored actual operation data 2011 will be described in detail.

[0250] FIG. 20B is a diagram illustrating an example of a process of processing pre-stored actual operation data according to an embodiment.

[0251] 20B, the processor 110 uses a noise filter to filter noise from the already stored actual flight data 2021. Here, the noise filter may refer to a Savitzky-Golay filter. However, examples of the noise filter used by the processor 110 are not limited to the Savitzky-Golay filter.

[0252] The processor 110 normalizes the noise-filtered data 2023. Here, data normalization 2023 is a preprocessing process that adjusts the scale of input data when training a deep learning model to increase the training speed and improve the performance of the deep learning model.

[0253] As an example, data normalization 2023 may use a min-max normalization method. The min-max normalization method is a method of converting input data values ​​into a range between 0 and 1. As another example, data normalization 2023 may use a standard normalization method. The standard normalization method is a method of converting input data values ​​into a distribution with a mean of 0 and a variance of 1. However, examples of normalization methods used by processor 110 are not limited to the min-max normalization method or standard normalization method described above.

[0254] The processor 110 divides the normalized 2023 data according to the voyage order, where voyage order means the order of the voyage.

[0255] In addition, the processor 110 converts the separated per-stage data 2024 into a tensor 2025 format. Here, the tensor 2025 is a mathematical concept representing a multi-dimensional array, where a vector represents a one-dimensional array, a matrix represents a two-dimensional array, and the tensor 2025 represents a three-dimensional or higher array. Therefore, the deep learning model can represent input data and parameters of the deep learning model using the tensor 2025.

[0256] The processor 110 can configure one batch 2026 with 10 tensors 2025. Here, the batch 2026 refers to a collection of input data that can be processed at one time as input data for a deep learning model. Generally, one batch 2026 includes multiple pieces of input data, all of which have the same size. Using the batch 2026 can improve the generalization performance of the deep learning model and enable efficient calculations. Therefore, by configuring the batch 2026, the processor 110 can improve the learning efficiency when repeatedly learning the deep learning model.

[0257] By using the processed actual operation data 2027, the processor 110 can improve the learning efficiency of the deep learning model and increase the prediction accuracy of the deep learning model.

[0258] The processor 110 can also train the deep learning model using an error and back propagation algorithm between the ground truth data and the output data of the deep learning model.

[0259] Here, the backpropagation algorithm is an algorithm for training deep learning models used in the field of supervised learning. The backpropagation algorithm calculates the error between output data and ground truth data, and calculates weights and biases using the calculated error. It then updates the calculated weights and biases to minimize the error between the output data and ground truth data.

[0260] An example in which the processor 110 learns a deep learning model using a backpropagation algorithm will now be described with reference to FIG.

[0261] FIG. 21 is a diagram illustrating an example of a process of learning a deep learning model that predicts the tank pressure of a ship according to one embodiment.

[0262] 21, the processor 110 obtains output data 2120 and correct answer data 2130 of a deep learning model 2110. Then, the processor 110 calculates an error 2140 between the output data 2120 and the correct answer data 2130 using a loss function. The processor 110 calculates the weight and bias of the error 2140 and learns the deep learning model 2110 by repeating the process of correcting the weight using a backpropagation algorithm 2150.

[0263] The processor 110 initializes the weights using the Xavier initialization method to prevent the weights from being lost during the process of using the backpropagation algorithm 2150. Here, the Xavier initialization is one method for initializing the weights, and initializes the weights using a normal distribution. However, examples of the method for initializing the weights used by the processor 110 are not limited to the Xavier initialization method described above.

[0264] Furthermore, the processor 110 uses a root mean square propagation (RMSprop) technique to find optimal weight values ​​when using the backpropagation algorithm 2150. Here, RMSprop is an algorithm that supports fast and stable learning of a deep learning model, and can provide optimal weights for each parameter to perform efficient learning. However, an example of a method for finding optimal weight values ​​used by the processor 110 is not limited to the above-mentioned RMSprop technique.

[0265] The processor 110 can verify the deep learning model trained using another portion 2013 of the pre-stored actual operational data 2011.

[0266] Here, validation is the process of evaluating the performance of a deep learning model and checking whether the deep learning model generalizes. In other words, it is a process to prevent overfitting or underfitting.

[0267] The steps of the processor 110 learning the deep learning model described above can be equally applied to deep learning models corresponding to each of multiple navigation modes for the voyage of the vessel 1810.

[0268] Referring again to FIG. 19, in step 1920, processor 110 may select one of a plurality of sailing modes depending on the level of liquefied gas in the tank.

[0269] For example, the processor 110 can measure the level of liquefied gas in the tank and select a first mode from among a plurality of navigation modes if the water level is equal to or greater than a predetermined height, and select a second mode from among the plurality of navigation modes if the water level is less than the predetermined height.

[0270] Hereinafter, with reference to FIGS. 22 and 23, an example will be described in which the processor 110 selects the first mode or the second mode from among a plurality of navigation modes according to the water level of the liquefied gas.

[0271] FIG. 22 is a diagram for explaining an example of determining the navigation mode depending on whether the water level of the liquefied gas in the tank is equal to or higher than a predetermined level according to one embodiment.

[0272] 22, the processor 110 measures the level of the liquefied gas in the tank 2140 and determines whether the level of the liquefied gas is equal to or greater than a predetermined height 2220, or whether the level of the liquefied gas is less than the predetermined height 2120. For example, the predetermined height may be half the height of the tank 2140. Therefore, the processor 110 can determine that the liquefied gas level is equal to or greater than half the height of the tank 2140 2220 as a laden voyage, and can determine that the liquefied gas level is less than half the height of the tank 2140 2120 as a ballast voyage.

[0273] FIG. 23 is a diagram illustrating an example of a deep learning model corresponding to each of a plurality of navigation modes according to the water level of liquefied gas in a tank according to one embodiment.

[0274] 23 , when the level of the liquefied gas is equal to or greater than half the height of the tank, the processor 110 selects the first mode 2330 in 2310, and when the level of the liquefied gas is less than half the height of the tank, the processor 110 selects the second mode 2340 in 2320. For example, the first mode 2330 may be a full-ship sailing mode, and the second mode 2340 may be an empty-ship sailing mode. Furthermore, the deep learning model 2350 of the first mode 2330 may use the LSTM algorithm, and the deep learning model 2360 of the second mode 2340 may use the GRU algorithm. However, the algorithms used by the deep learning models 2350 and 2360 are not limited to the above-mentioned LSTM algorithm and GRU algorithm.

[0275] Referring again to FIG. 19, in step 1930, the processor 110 may predict tank pressure from the recommended voyage information using a deep learning model corresponding to the selected voyage mode.

[0276] For example, the processor 110 may obtain recommended voyage information for a vessel, obtain a plurality of intermediate values ​​using the recommended voyage information as input data, and apply weights to the plurality of intermediate values ​​to obtain a predicted pressure value for the tank.

[0277] First, the processor 110 acquires the recommended voyage information for the ship. The processor 110 can acquire the recommended voyage information by the method described above with reference to FIGS.

[0278] The processor 110 may use the recommended voyage information as input data to obtain a plurality of intermediate values ​​and apply weights to the plurality of intermediate values ​​to obtain a predicted tank pressure value.

[0279] An example in which the processor 110 applies weights to multiple intermediate values ​​to obtain a predicted tank pressure will now be described with reference to FIG.

[0280] FIG. 24 is a diagram illustrating an example of obtaining a predicted tank pressure value by applying weights to a plurality of intermediate values ​​according to an embodiment.

[0281] 24 , the processor 110 may use the recommended voyage information 2410 as input data for the deep learning model 2420 to obtain the intermediate value 2430. As an example, in the first mode, i.e., the full-ship voyage mode, the deep learning model 2420 may use the LSTM algorithm. As another example, in the second mode, i.e., the empty-ship voyage mode, the deep learning model 2420 may use the GRU algorithm. However, the algorithm used by the deep learning model 2420 is not limited to the LSTM algorithm and the GRU algorithm described above.

[0282] Furthermore, the processor 110 can apply 2440 weights to the intermediate values ​​2430 to obtain the predicted tank pressure value 2450. For example, the processor 110 can apply 2440 weights to the intermediate values ​​2430 using a linear layer to obtain the predicted tank pressure value 2450. However, the method by which the processor 110 applies 2440 weights to the intermediate values ​​2430 is not limited to the method using a linear layer.

[0283] Therefore, the processor 110 can predict the tank pressure value of the ship 1810 using different deep learning models depending on the water level of the liquefied gas in the tank 1820 of the ship 1810. In other words, the processor 110 can improve the accuracy of predicting the tank pressure of the ship by predicting the tank pressure of the ship using different deep learning models for each sailing mode.

[0284] FIG. 25 is a diagram for explaining an example of a method for optimizing ship operation according to an embodiment.

[0285] 25, a ship operation optimization model 2520 can acquire optimal voyage information 2530 that takes into account ship constraints 2540 using a BOG generation amount and a tank pressure value 2510. Here, the ship operation optimization model 2520 is a model used to implement a method for optimizing ship operation according to an embodiment, and may be a model that implements the method for optimizing ship operation described above with reference to FIGS. 1 to 9. Therefore, the following description will focus on differences from the model that implements the method for optimizing ship operation described above with reference to FIGS. 1 to 9.

[0286] According to one embodiment of a method for optimizing ship operation, the amount of BOG generated by the ship and the tank pressure of the ship are predicted using the ship's recommended voyage information. Then, the predicted amount of BOG generated and the tank pressure values ​​are used to obtain optimal voyage information for each voyage section that satisfies the ship's constraints, and the operation of the ship is controlled using the obtained optimal voyage information for each voyage section. Specific details of the constraints will be described later.

[0287] Therefore, according to one embodiment of the method for optimizing ship operation, it is possible to perform economical operation by minimizing the gas consumption of the ship using optimal voyage information for each voyage section.

[0288] Hereinafter, a method and apparatus for optimizing ship operation according to an embodiment of the present disclosure will be described in detail with reference to Figures 2 and 26 to 28B. Here, the apparatus may refer to a user terminal 100.

[0289] 2, the communication module 140 can receive information about existing navigation data, information about current navigation data, recommended navigation information, navigation plans, weather information, etc. from an external server or external device. In addition, the memory 120 can store various data such as existing navigation data, current navigation data, recommended navigation information, navigation plans, weather information, and data generated in response to the operation of the processor 110.

[0290] The processor 110 can control the operation of the device 100 by executing a program stored in the memory 120. As an example, the processor 110 can execute at least a part of a method for optimizing ship operation, which will be described with reference to Figures 26 to 28B.

[0291] That is, the processor 110 can obtain optimal navigation information for each voyage section using the ship's BOG generation amount and the ship's tank pressure value, update the optimal navigation information taking into account the ship's constraints, and control the operation of the ship using the updated optimal navigation information.

[0292] The processor 110 can obtain optimal voyage information for each voyage section using the BOG generation rate of the ship and the tank pressure value of the ship. Furthermore, the processor 110 can update the optimal voyage information in consideration of constraints of the ship. Here, the constraints can be set based on at least one of mass, energy, power, maximum / minimum liquefied gas consumption of equipment, equipment efficiency, efficiency of a propulsion engine (main engine), efficiency of a generator engine, efficiency of a shaft generator, tank pressure, the relationship between the maximum / minimum speed of the ship and the average speed of the ship.

[0293] For example, the processor 110 may acquire optimal navigation information in different ways depending on whether or not a speed for each navigation section included in the recommended navigation information is used during navigation of the ship.

[0294] For example, when using the speed for each voyage section included in the recommended voyage information when operating a ship, the processor 110 can obtain optimal voyage information using the BOG generation amount, tank pressure value, and recommended voyage information, and the recommended voyage information can include preset weather information.

[0295] In addition, the processor 110 may calculate the difference between the gas consumption amount included in the (n+1)th intermediate voyage information and the gas consumption amount included in the (n)th intermediate voyage information, determine whether the difference is within a preset range, and determine the (n+1)th intermediate voyage information as the optimal voyage information based on the determination result, where n is a natural number equal to or greater than 1, and the (n+1)th intermediate voyage information may be generated based on at least one parameter update value included in the (n)th intermediate voyage information.

[0296] As another example, if the speed for each voyage section included in the recommended voyage information is not used when operating a ship, the processor 110 can obtain optimal voyage information using the BOG generation amount, tank pressure value, and recommended voyage information, and the recommended voyage information can include updated weather information according to the operation of the ship.

[0297] In addition, the processor 110 may calculate the difference between the gas consumption amount included in the (n+1)th intermediate voyage information and the gas consumption amount included in the (n)th intermediate voyage information, determine whether the difference is within a preset range, and determine the (n+1)th intermediate voyage information as the optimal voyage information based on the determination result, where n is a natural number equal to or greater than 1, and the (n+1)th intermediate voyage information may be generated based on at least one parameter update value included in the (n)th intermediate voyage information.

[0298] Furthermore, the processor 110 may control the operation of the vessel using the updated optimal voyage information. For example, the processor 110 may control the vessel in a set navigational maneuver using the optimal voyage information, and the navigational maneuver may include a navigational maneuver that minimizes gas consumption of the vessel.

[0299] FIG. 26 is a flowchart illustrating an example of a method for optimizing ship operation according to an embodiment.

[0300] 26, the method for optimizing ship operation may include steps 2610 to 2630. However, the method is not limited to these steps, and other general steps other than those shown in FIG. 26 may also be included in the method for optimizing ship operation. Furthermore, as described above with reference to FIGS. 2 and 25, at least one of the steps in the flowchart shown in FIG. 26 may be processed by processor 110.

[0301] In step 2610, the processor 110 can obtain optimal voyage information for each voyage section using the BOG generation amount of the ship and the tank pressure value of the ship.

[0302] First, the processor 110 can predict the amount of BOG generated by the ship. As one example, the processor 110 can predict the amount of BOG generated by the ship using the method described above with reference to Fig. 6. As another example, the processor 110 can predict the amount of BOG generated by the ship using the method and device for predicting the amount of BOG generated by the ship described above with reference to Figs. 2 and 10 to 17. However, the method for predicting the amount of BOG generated by the ship is not limited to the method described above.

[0303] The processor 110 can also predict the tank pressure of the vessel. As one example, the tank pressure of the vessel can be predicted using the method described above with reference to Fig. 6. As another example, the tank pressure of the vessel can be predicted using the method and device for predicting the tank pressure of the vessel described above with reference to Figs. 2 and 18 to 24. However, the method for predicting the tank pressure of the vessel is not limited to the above-described method.

[0304] The processor 110 may acquire optimal navigation information for each navigation section using the amount of BOG generated by the ship and the tank pressure value of the ship. For example, the processor 110 may acquire optimal navigation information in different ways depending on whether or not a speed for each navigation section included in the recommended navigation information is used during operation of the ship.

[0305] Here, the recommended voyage information may be recommended voyage information for a navigator to efficiently navigate a ship from a departure point to a destination point, and the processor 110 may generate the recommended voyage information in the manner described above with reference to Figures 4 and 5. The recommended voyage information may further include the speed for each voyage section, the overall average speed of the ship, the latitude / longitude of the ship over time, and weather information over time, in addition to the recommended voyage information described above with reference to Figures 4 and 5. However, the method for generating the recommended voyage information is not limited to the manner described above, and the recommended voyage information is not limited to the recommended voyage information described above.

[0306] Hereinafter, with reference to FIG. 27, an example will be described in which the processor 110 classifies the ship into different modes depending on whether or not the speed for each voyage section included in the recommended voyage information is used during the operation of the ship.

[0307] Figure 27 is a diagram illustrating an example of obtaining optimal navigation information in different ways depending on whether or not the speeds for each navigation section included in the recommended navigation information are used in the operation of a ship according to one embodiment.

[0308] Referring to FIG. 27, the processor 110 can acquire optimal navigation information in different ways depending on whether or not a speed for each navigation section included in the recommended navigation information is used 2710 during navigation of the ship.

[0309] For example, when the processor 110 uses the speed for each voyage section included in the recommended voyage information in operating the ship (2720), the processor 110 navigates at the speed for each voyage section included in the recommended voyage information. Here, the speed for each voyage section included in the recommended voyage information may be the same for all voyage sections. Alternatively, the speed for each voyage section included in the recommended voyage information may be different for all voyage sections. Alternatively, the speed for each voyage section included in the recommended voyage information may be the same for only some voyage sections.

[0310] As another example, when the processor 110 does not use the speed for each voyage segment included in the recommended voyage information in operating the ship 2730, it navigates at the speed for each voyage segment included in the optimal voyage information. Here, the speed for each voyage segment included in the optimal voyage information may be the same for all voyage segments. Alternatively, the speed for each voyage segment included in the optimal voyage information may be different for all voyage segments. Alternatively, the speed for each voyage segment included in the optimal voyage information may be the same for only some voyage segments. However, even when the processor 110 does not use the speed for each voyage segment included in the recommended voyage information 2730, it may use the total average speed included in the recommended voyage information.

[0311] Hereinafter, the process by which the processor 110 obtains the optimum voyage information will be described in detail for cases where the speeds for each voyage section included in the recommended voyage information are used and not used in the operation of the ship.

[0312] For example, when using the speed for each voyage section included in the recommended voyage information for operating a ship, the processor 110 can obtain optimal voyage information using the BOG generation amount, tank pressure value, and recommended voyage information, and the recommended voyage information can include preset weather information.

[0313] In addition, the processor 110 may calculate the difference between the gas consumption amount included in the (n+1)th intermediate voyage information and the gas consumption amount included in the (n+1)th intermediate voyage information, determine whether the difference is within a preset range, and determine the (n+1)th intermediate voyage information as the optimal voyage information based on the determination result. Here, the (n+2)th intermediate voyage information may be generated based on at least one parameter update value included in the (n+1)th intermediate voyage information, where n is a natural number equal to or greater than 1.

[0314] Hereinafter, with reference to FIG. 28A, an example will be described in which the processor 110 acquires optimal navigation information when using the speeds for each navigation section included in the recommended navigation information in navigation of a ship.

[0315] FIG. 28A is a flowchart illustrating an example of acquiring optimal navigation information when using a speed for each navigation section included in recommended navigation information in operation of a ship according to one embodiment.

[0316] Referring to FIG. 28A, when using the speed for each voyage section included in the recommended voyage information in the operation of a ship 2810, the processor 110 can obtain optimal voyage information using the recommended voyage information, the predicted BOG generation amount of the ship, and the predicted tank pressure of the ship.

[0317] First, in step 2811, the processor 110 obtains recommended voyage information.

[0318] Here, the recommended voyage information may include the above-described recommended voyage information as is. In step 2812, the processor 110 may predict the amount of BOG generated by the ship and the tank pressure of the ship using the above-described method. Here, the amount of BOG generated may vary depending on weather conditions such as wind speed, wave height, and swell height. Therefore, the predicted value of the BOG generation amount and the predicted value of the tank pressure may also vary depending on weather conditions.

[0319] In step 2813, the processor 110 acquires intermediate voyage information using the recommended voyage information, the predicted BOG generation amount of the vessel, and the predicted tank pressure of the vessel. Here, the intermediate voyage information may include the optimal speed of the vessel for each voyage section, the gas consumption of the vessel, and the usage of equipment installed on the vessel. Here, the equipment may include a subcooler, shaft generator, main engine, generator engine, gas combustion unit, reliquefaction unit, etc. installed on the vessel. In addition, the gas consumption may be calculated using the gas consumption of the propulsion engine, the gas consumption of the generator engine, the gas consumption of the gas combustion unit, and the gas consumption of the reliquefaction unit.

[0320] In step 2814, the processor 110 calculates the difference between the gas consumption amount included in the (n+1)th intermediate voyage information and the gas consumption amount included in the nth intermediate voyage information. The processor 110 may also determine whether the difference between the (n+1)th gas consumption amount and the nth gas consumption amount is within a preset range. Here, the preset range may be from 0 to less than 0.01.

[0321] If the difference between the (n+1)th gas consumption amount and the nth gas consumption amount is not within the preset range, the processor 110 executes the process again from step 2812.

[0322] In step 2815, the weather information included in the recommended voyage information is not changed because the processor 110 does not use the optimal speed of the vessel included in the (n+1)th intermediate voyage information. Therefore, the processor 110 uses the preset weather information included in the recommended voyage information when obtaining the (n+2)th intermediate voyage information.

[0323] When processor 110 re-executes step 2812, the (n+2)th interim voyage information can be generated based on the updated value of at least one parameter included in the (n+1)th interim voyage information. For example, the at least one parameter may be gas consumption.

[0324] For example, when the processor 110 re-executes step 2812, the predicted BOG generation amount of the vessel may be changed by using the (n+1)th gas consumption amount instead of the nth gas consumption amount. Also, the intermediate voyage information may be changed by changing the predicted BOG generation amount of the vessel.

[0325] In step 2816, if the difference between the (n+1)th gas consumption and the (n)th gas consumption is within a preset range, the processor 110 may determine the (n+1)th intermediate voyage information as the optimal voyage information. Here, the optimal voyage information may include the predicted BOG generation amount of the ship, the predicted tank pressure of the ship, the gas consumption amount of the ship included in the (n+1)th intermediate voyage information, the optimal speed of the ship, and the usage amount of the equipment installed on the ship.

[0326] For example, if the speed for each section of a voyage included in the recommended voyage information is not used when operating a ship, the processor 110 can obtain optimal voyage information using the amount of BOG generated, the tank pressure value, and the recommended voyage information, and the recommended voyage information can include weather information updated according to the operation of the ship.

[0327] In addition, the processor 110 may calculate the difference between the gas consumption amount included in the (n+1)th intermediate voyage information and the gas consumption amount included in the (n+1)th intermediate voyage information, determine whether the difference is within a preset range, and determine the (n+1)th intermediate voyage information as the optimal voyage information based on the determination result. Here, the (n+2)th intermediate voyage information may be generated based on at least one parameter update value included in the (n+1)th intermediate voyage information, where n is a natural number equal to or greater than 1.

[0328] Hereinafter, with reference to FIG. 28B, an example will be described in which the processor 110 acquires optimal navigation information when the speeds for each navigation section included in the recommended navigation information are not used in navigation of a ship.

[0329] FIG. 28B is a flowchart illustrating an example of acquiring optimal navigation information when not using the speeds for each navigation section included in the recommended navigation information in the operation of a ship according to an embodiment.

[0330] Referring to FIG. 28B, if the speeds for each voyage section included in the recommended voyage information are not used in the operation of the ship 2820, the processor 110 can obtain optimal voyage information using the recommended voyage information, the predicted BOG generation amount of the ship, and the predicted tank pressure of the ship.

[0331] First, in step 2821, the processor 110 obtains recommended voyage information.

[0332] Here, the recommended voyage information may include the above-described recommended voyage information as is, and may further include weather information updated according to the operation of the ship. In step 2822, the processor 110 may predict the BOG generation amount and tank pressure of the ship using the above-described method. Here, the BOG generation amount may change depending on weather conditions such as wind speed, wave height, and swell height. Therefore, the predicted BOG generation amount and tank pressure may also change depending on weather conditions.

[0333] In step 2823, the processor 110 acquires intermediate voyage information using the recommended voyage information, the predicted BOG generation amount of the vessel, and the predicted tank pressure of the vessel. Here, the intermediate voyage information may include the optimal speed of the vessel for each voyage section, the gas consumption amount of the vessel, and the usage amount of equipment installed on the vessel. Here, the equipment may include a supercooling device, a shaft generator, a power generation engine, a propulsion engine, a gas combustion device, a reliquefaction device, etc. installed on the vessel.

[0334] In step 2824, the processor 110 calculates the difference between the gas consumption amount included in the (n+1)th intermediate voyage information and the gas consumption amount included in the nth intermediate voyage information. The processor 110 may also determine whether the difference between the (n+1)th gas consumption amount and the nth gas consumption amount is within a preset range. Here, the preset range may be from 0 to less than 0.01.

[0335] If the difference between the (n+1)th gas consumption amount and the nth gas consumption amount is not within the preset range, the processor 110 executes the process again from step 3021.

[0336] When processor 110 re-executes step 2821, the (n+2)th interim voyage information may be generated based on the updated value of at least one parameter included in the (n+1)th interim voyage information. As an example, the at least one parameter may be gas consumption. As another example, the at least one parameter may be an optimal speed for each voyage segment.

[0337] In step 2825, processor 110 uses the vessel's optimum speed for each voyage segment included in the (n+1)th intermediate voyage information. Therefore, the vessel's optimum speed included in the (n+1)th intermediate voyage information may differ from the vessel's speed for each voyage segment included in the recommended voyage information. If the vessel's speed changes, the vessel's latitude and longitude may change over time. Therefore, the weather information may also differ from the weather information included in the recommended voyage information. Furthermore, if the weather information changes, the vessel's BOG generation rate and the vessel's tank pressure may change.

[0338] For example, in step 2825, processor 110 updates the weather information using the optimal speed of the vessel included in the (n+1)th intermediate voyage information, and processor 110 uses the updated weather information when re-executing step 2822.

[0339] For example, when processor 110 re-executes step 2822, the predicted BOG generation amount of the vessel may be changed by using the (n+1)th gas consumption amount instead of the nth gas consumption amount. Therefore, the intermediate voyage information may be changed by changing the predicted BOG generation amount of the vessel. Also, when processor 110 re-executes step 2822, the tank pressure value of the vessel may be changed by using updated weather information instead of the weather information included in the recommended voyage information, and the intermediate voyage information may be changed by changing the tank pressure value of the vessel.

[0340] In step 2826, if the difference between the (n+1)th gas consumption and the (n)th gas consumption is within a preset range, the processor 110 may determine the (n+1)th intermediate voyage information as the optimal voyage information. Here, the optimal voyage information may include the predicted BOG generation amount of the ship, the predicted tank pressure of the ship, the gas consumption amount of the ship included in the (n+1)th intermediate voyage information, the optimal speed of the ship, and the usage amount of the equipment installed on the ship.

[0341] 26 again, in step 2620, the processor 110 can update the optimal voyage information taking into account constraints of the vessel, where the constraints can be set based on at least one of mass, energy, power, maximum / minimum liquefied gas consumption of equipment, efficiency of equipment, efficiency of the propulsion engine (main engine), efficiency of the generator engine, efficiency of the shaft generator, tank pressure, the relationship between the maximum / minimum speed of the vessel and the average speed of the vessel.

[0342] In other words, the constraints may include mass conditions, energy conditions, power consumption conditions, maximum / minimum liquefied gas consumption conditions for equipment, equipment efficiency conditions, propulsion engine (main engine) efficiency conditions, generator engine efficiency conditions, shaft generator efficiency conditions, tank pressure conditions, maximum / minimum ship speed conditions, and ship average speed conditions, and each constraint will be described in detail below.

[0343] For example, the mass condition can be set based on the mass of BOG consumed by equipment installed on the ship. Here, the mass of BOG consumed can be calculated in consideration of the relationship between the mass of BOG used in the main engine and generator engine, the mass of BOG discharged from the gas combustion unit, and the mass of BOG reliquefied in the reliquefaction unit.

[0344] For example, the energy condition can be set based on the difference in tank pressure change amount for each section of the ship's voyage. Here, the difference in tank pressure change amount for each section of the ship's voyage can be calculated taking into consideration the relationship between the amount of BOG generated for each section of the voyage, the amount of BOG returned to the tanks, the amount of liquefied gas returned to the tanks, and the amount of BOG consumed by equipment installed on the ship.

[0345] For example, the power consumption condition may be set based on the power consumption of the equipment installed on the ship, which may be calculated in consideration of the relationship with the usage of each of the compressor, gas combustion device, reliquefaction device, subcooler, and fuel gas pump.

[0346] For example, the maximum / minimum liquefied gas consumption requirements for equipment can be set based on a turndown ratio. If the required liquefied gas consumption for each equipment is greater than the equipment's minimum liquefied gas consumption requirement, the equipment will operate at the minimum liquefied gas consumption.

[0347] For example, the efficiency conditions of the equipment can be set based on the amount of power change of the equipment depending on the amount of liquefied gas used in each equipment. Here, the efficiency of the equipment means the performance of the equipment relative to the amount of liquefied gas used in the equipment. As an example, the performance of the compressor can mean the target compression pressure. As another example, the performance of the reliquefaction device can mean the target reliquefaction temperature.

[0348] For example, the efficiency condition of the propulsion engine may be set based on the ship's speed and the required amount of liquefied gas fuel. Here, the ship's speed may be the optimal speed for each voyage section included in the optimal voyage information. The required liquefied gas fuel may be LNG gas or BOG.

[0349] For example, the efficiency condition of the power generating engine may be set based on the power amount of the equipment installed on the ship and the basic power consumption of the ship. Here, the power amount of the equipment installed on the ship may be the sum of the power amounts of all the equipment. Furthermore, the basic power consumption of the ship may refer to the power basically used on the ship.

[0350] For example, the efficiency condition of the shaft generator can be set based on the amount of electrical power generated by the shaft generator according to the rotational speed of the propulsion engine, and the sum of the amount of electrical power generated by the shaft generator and the amount of electrical power generated by the power generation engine can be the total amount of electrical power generated by the ship.

[0351] For example, the tank pressure condition can be set based on the maximum and minimum tank pressure values, where the tank pressures for a full voyage and a ballast voyage can be obtained using different methods.

[0352] For example, maximum and minimum speed conditions for a vessel can be set based on engine performance, where if the vessel is sailing below the minimum speed, the speed will be zero and the vessel will come to a halt.

[0353] For example, the average speed condition of the vessel may be set based on the speed of each section of the vessel. As an example, when a section speed included in the recommended voyage information is used in vessel operation, the average of the section speeds may be the same as the total average speed calculated from the entire navigation time and the entire navigation distance. As another example, when a section speed included in the optimal voyage information is used in vessel operation, the average of the section speeds may be the same as the total average speed calculated from the entire navigation time and the entire navigation distance.

[0354] Therefore, the processor 110 can take into account each of the constraints described above and update the optimal voyage information that satisfies each of the constraints.

[0355] Referring again to FIG. 26, in step 2630, the processor 110 may control the operation of the vessel using the updated optimal voyage information.

[0356] For example, the processor 110 may control the vessel in a navigation method set using the optimal voyage information, where the navigation method may include a navigation method that minimizes the gas consumption of the vessel.

[0357] For example, when the processor 110 controls the operation of the vessel using the optimal voyage information, the gas consumption of the vessel can be minimized. In other words, the processor 110 can control the vessel in a navigation manner that minimizes the gas consumption of the vessel.

[0358] Thus, the processor 110 can control the vessel in a navigation strategy that minimizes the vessel's gas consumption using optimal voyage information that is updated taking into account constraints.

[0359] FIG. 29 is a diagram illustrating an example of a system for controlling the operation of a ship according to an embodiment.

[0360] The ship 3220 may include a user terminal 3210 and a server 3240. For example, the user terminal 3210 and the server 3240 are connected via wired or wireless communication and can transmit and receive data (e.g., navigation information) between them.

[0361] 29 illustrates system 3200 as including a user terminal 3210 and a server 3240, but is not limited to this. For example, system 3200 may include other external devices (not shown), and the operations of user terminal 3210 and server 3240 described below may be embodied by a single device (e.g., user terminal 3210 or server 3240) or more devices.

[0362] The user terminal 3210 may be a computing device including a display device and / or a device for receiving user input (e.g., a keyboard, a mouse, etc.), and including a memory and a processor. For example, the user terminal 3210 may include, but is not limited to, a notebook computer, a desktop PC, a laptop, a tablet computer, a smartphone, etc.

[0363] The server 3240 may include a communication module capable of communicating with external devices (not shown) including the user terminal 3210 and with other devices. The communication module may communicate with external devices (not shown) including the user terminal 3210 via a network. For example, the network may include a value-added network (VAL), a mobile radio communication network, a satellite communication network, and / or a combination thereof. The network is a comprehensive data communication network that enables the components shown in FIG. 29 to smoothly communicate with each other and may include a wired communication network or a wireless communication network. As an example, the server 3240 may transmit preset flight information and preset route information 3250 to the user terminal 3210. Alternatively, the server 3240 may be a computing device including a memory and a processor and having its own computing capabilities. For example, the server 3240 may store various data including the flight information and route information 3250.

[0364] Conventionally, operators of vessels 3220 have controlled the pressure in tanks 3230 and the amount of liquefied gas evaporated from tanks 3230 using an integrated automation system (IAS), an integrated control device. However, the IAS was unable to accurately predict changes in the pressure in tanks 3230 and changes in the amount of liquefied gas evaporated from tanks 3230 due to weather changes during navigation. As a result, operators of vessels 3220 have repeatedly had to determine the amount of evaporated gas to use in order to control the pressure in tanks 3230 while the vessel 3220 is operating. This has led to uneconomical operation of the vessel 3220.

[0365] The user terminal 3210 according to one embodiment may calculate at least one predicted value for operation control of the ship 3220 and an economic operation index 3260 of the ship 3220. In addition, the user terminal 3210 may acquire measurements taken in real time according to the operation of the ship 3220 and compare the predicted value and the economic operation index 3260 with the measurements.

[0366] As a result, the user terminal 3210 controls the ship 3220 using the results of comparing the measured values ​​with the predicted values ​​and the economic operation index 3260, thereby minimizing the operator's decisions. This solves the problem of uneconomical operation caused by requiring excessive decisions from the operator. In addition, the operator can check whether economical operation is being performed during operation without using IAS.

[0367] The ship 3220 may include a tank 3230. Furthermore, a tank 3290 may be disposed inside the destination 3280. Therefore, when the ship 3220 is in operation, the tank 3230 may be located at sea, and the tank 3290 may be located on land. For example, the tank 3230 and the tank 3290 may have a double-wall structure and may be composed of an inner tank and an outer tank. The inner tank is a space for storing liquefied gas. The outer tank surrounds the inner tank and plays a role in keeping the liquefied gas warm.

[0368] The ship 3220 travels to a destination 3280 via a preset route 3270. A specific method for setting the route 3270 will be described later with reference to FIG.

[0369] An example of the operation of the user terminal 3210 will be described below with reference to FIGS. 2 and 30 to 36.

[0370] Referring again to FIG. 2, the user terminal 100 in FIG. 2 may be the same device as the user terminal 3210 in FIG.

[0371] The processor 110 can acquire preset operation information and preset route information 3250 of the ship 3220 from the server 3240. For example, the processor 110 can acquire the pressure of the tank 3230 upon arrival at the destination, the arrival and departure information of the ship 3220, the total operation distance, the total operation time, the average speed, and weather information from the server 3240. The processor 110 can also acquire the latitude, longitude, and speed of the ship 3220 for at least one preset time point.

[0372] The processor 110 can calculate at least one predicted value for controlling the navigation of the ship 3220 using the navigation information and route information 3250. For example, the processor 110 can calculate at least one predicted value among a predicted value of the pressure inside the tank 3230, a predicted value of the amount of liquefied gas evaporated from the tank 3230, a predicted value of the flow rate at which the liquefied gas pressurized by the pump from the tank 3230 is supplied from the vaporizer to a device inside the ship 3220, and a predicted value of the flow rate at which the liquefied gas evaporated from the tank 3230 is supplied from the compressor to a device inside the ship 3220.

[0373] The processor 110 can then compare the predicted values ​​with measurements taken in real time as the vessel 3220 operates and control the vessel 3220.

[0374] Meanwhile, although not shown in Fig. 2, the user terminal 100 may further include a display device. Alternatively, the user terminal 100 may be connected to an independent display device via a wired or wireless communication method, and data may be transmitted and received between the display device and the independent display device. For example, a result of comparing a predicted value with a measured value may be provided to the user via the display device.

[0375] FIG. 30 is a configuration diagram illustrating an example of a server according to an embodiment.

[0376] The server 3300 includes a processor 3310, a memory 3320, and a communication module 3330. For convenience of explanation, only components related to the present invention are shown in FIG. 30. Therefore, in addition to the components shown in FIG. 30, other general-purpose components may also be included in the server 3300. Furthermore, it will be obvious to a person having ordinary knowledge in the technical field related to the present invention that the processor 3310, the memory 3320, and the communication module 3330 shown in FIG. 30 may be embodied as independent devices. Furthermore, the server 3300 in FIG. 30 may be the same device as the server 3240 in FIG. 29.

[0377] The processor 3310 can set route information using the operation information. The processor 3310 can also transmit the operation information, route information, and the like to the user terminal 3210.

[0378] The memory 3320 can store operation information and route information 3250, etc. Furthermore, the memory 3320 can store an operating system (OS) and at least one program (e.g., a program necessary for the processor 3310 to operate, etc.).

[0379] The communication module 3330 may provide a configuration or function for the server 3300 and the user terminal 3210 to communicate with each other via a network. The communication module 3330 may also provide a configuration or function for the server 3300 to communicate with other external devices. For example, control signals, instructions, data, etc. provided under the control of the processor 3310 may be transmitted to the user terminal 3210 and / or external devices via the communication module 3330 and the network.

[0380] FIG. 31 is a flowchart illustrating an example of a method for controlling operation of a marine vessel according to an embodiment.

[0381] The method for controlling the operation of a vessel 3220 comprises steps that are processed in time series by the user terminal 3210, 100 or processor 110 shown in Figures 2 and 32. Therefore, even if the content is omitted below, the content described above regarding the user terminal 100, 3210 or processor 110 shown in Figures 2 and 29 can also be applied to the method for controlling the operation of a vessel 3220 in Figure 29.

[0382] In step 3410, the processor 110 obtains preset operation information of the ship 3220 from the server 3240.

[0383] For example, the operation information may include arrival and departure information, total navigation distance, total navigation time, average speed, and weather information of the ship 3220. The operation information may also include the pressure of the tank 3230 included in the ship 3220 at the time when the ship 3220 is scheduled to arrive at the destination. The pressure of the tank 3230 at the time of scheduled arrival may correspond to either the pressure of the tank 3290 provided at the destination or a preset pressure.

[0384] In step 3420, the processor 110 obtains from the server 3240 the route information of the ship 3220 that has been set using the operation information.

[0385] For example, the route information may include at least one of the latitude, longitude, and speed of the vessel 3220 for at least one preset point in time.

[0386] Hereinafter, the operation information and route information will be specifically described with reference to FIGS.

[0387] FIG. 32 is a diagram illustrating an example of a screen on which preset operation information of a ship according to an embodiment is displayed.

[0388] 32 shows an example of a screen 3510 on which various pieces of flight information are output. Specifically, various pieces of flight information can be divided and displayed in areas 3511, 3512, 3513, and 3514 of the screen 3510. However, the layout of the screen 3510 and the output areas for the flight information are not limited to the example shown in FIG.

[0389] Area 3511 may display the name of the location from which vessel 3220 departs, the departure time of vessel 3220, and the pressure of the liquefied gas cargo tanks within the departure location. For example, the pressure at the time of departure in tank 3230 may correspond to the pressure of the liquefied gas cargo tanks within the departure location.

[0390] Area 3512 can also display the total navigation distance, total navigation time, and average speed of the ship 3220. For example, the total navigation distance can be calculated as the distance traveled from the time the ship 3220 departs from the departure point to the time it arrives at the destination 3280. The total navigation time can be calculated as the time required from the time the ship 3220 departs from the departure point to the time it arrives at the destination 3280. The average speed can also be calculated by dividing the total navigation distance by the total navigation time.

[0391] Area 3513 may also display the name of the destination of vessel 3220, the time of arrival of vessel 3220 at the destination, and the pressure of tank 3290. For example, the pressure of tank 3230 at the time of arrival may correspond to either the pressure of tank 3290 or a preset pressure.

[0392] However, the operation information does not have to include the pressure of the tank 3290. In this case, the processor 110 can calculate the predicted value by assuming the pressure of the tank 3290 to be a predetermined value (for example, 100 mbarg).

[0393] Additionally, area 3514 can display information indicating the distance traveled by ship 3220 from the time of departure to the time of measurement, obtained in real time, as a percentage of the total distance traveled.

[0394] Furthermore, although not shown in Fig. 32, the operation information may include weather information. For example, the weather information may refer to weather information about the route 3270 from the time the ship 3220 departs from the departure point to the time it arrives at the destination 3280. The weather information may also include air pressure, temperature, wave height, etc.

[0395] FIG. 33 illustrates an example of a screen on which either a manually set route for a vessel or a route automatically set through a route optimization function can be obtained according to one embodiment.

[0396] 33 shows an example of a screen 3610 on which the processor 110 can acquire route information. Specifically, various selection areas can be displayed in areas 3611, 3612, 3613, and 3614 of the screen 3610. However, the layout of the screen 3610 and the output area for the selected information are not limited to the example shown in FIG.

[0397] The processor 110 can obtain from the server 3240 either a manually set route for the vessel 3220 or an automatically set route via a route optimization function.

[0398] Here, the manually set route and the automatically set route are set using preset operation information.

[0399] For example, when the user selects area 3611, the processor 110 can respond by activating inactivated areas 3612 and 3613. When the user selects activated area 3612, the processor 110 can request the server 3240 to transmit route information for the automatically set route via the Integrated Smartship Solution (ISS). This allows the processor 110 to obtain the route information from the server 3240.

[0400] Furthermore, when the user selects the activated area 3613, the processor 110 can respond by requesting the server 3240 to transmit route information for the automatically set route via an ECDIS (Electronic Chart Display Information System). This allows the processor 110 to obtain route information from the server 3240.

[0401] Furthermore, when the user selects area 3614, the processor 110 can respond by requesting the server 3240 to transmit route information for a manually set route. This allows the processor 110 to obtain route information from the server 3240.

[0402] Route optimization is a technology that optimizes the route and speed of vessel 3220, minimizes liquefied gas fuel consumption, and optimizes navigation time. Route optimization takes into account various factors such as vessel 3220's position, speed, and weather information. ISS and ECDIS are among the programs that include route optimization.

[0403] ISS is a system that manages the operation of ships 3220 and manages the operation of equipment, and is a program that collects key data from ships 3220 in real time and provides analysis services for major equipment including engines and operation optimization functions. In addition, ECDIS is a program that provides users with navigation information and maps related to sea areas.

[0404] In addition, information regarding the manually set route of the ship 3220 and information regarding the route automatically set through the route optimization function includes the latitude, longitude, and speed of the ship 3220 for at least one preset point in time.

[0405] For example, the server 3240 may set N time points after the departure of the ship 3220. For example, the number of time points may be preset according to a predetermined cycle or may be adjusted according to a user's input. The 0th time point may correspond to the time point when the ship 3220 departs from the departure point, and the Nth time point may correspond to the time point when the ship 3220 arrives at the destination. The route information may include at least one of the latitude, longitude, and speed of the ship 3220 for at least one of the N time points. Here, N is a natural number equal to or greater than 1.

[0406] For example, if the period is two days and the total operating time is 24 days, the route information may include information on the vessel 3220 for a total of 12 time points. The route information may include information on the latitude, longitude, and speed of the vessel 3220 for every two days from the departure of the vessel 3220. Furthermore, the pressure in the tank 3230 at the 12th time point may correspond to the pressure in the tank 3290 at the destination.

[0407] Referring again to FIG. 31, in step 3430, the processor 110 calculates at least one predicted value for operation control of the vessel 3220 using the operation information and route information 3250.

[0408] For example, the predicted value may include at least one of a predicted internal pressure value of the tank 3230, a predicted amount of liquefied gas evaporated from the tank 3230, and a predicted flow rate at which the liquefied gas evaporated in the tank 3230 is supplied from the compressor to a device inside the ship 3220. For example, the device may include at least one of a compressor included in the tank 3230, a vaporizer included in the tank 3230, a gas combustion device, a reliquefaction device, a propulsion engine, and a power generation engine.

[0409] Hereinafter, a process of calculating a predicted pressure value inside the tank 3230 for controlling the operation of the ship 3220 using the operation information and route information 3250 will be described with reference to FIG.

[0410] FIG. 34 is a diagram for explaining an example of a method for calculating a predicted pressure value of a liquefied gas cargo tank according to an embodiment.

[0411] In step 3710, the processor 110 may obtain the preset pressure inside the tank 3290 at the destination from the server 3240.

[0412] If the pressure of the tank 3290 is not set, the processor 110 may calculate the predicted pressure of the tank 3230 assuming the pressure of the tank 3290 to be a predetermined value (eg, 100 mbarg).

[0413] In step 3720, the processor 110 can divide the total operation time of the ship 3220 into N time points, with a predetermined time interval as the period. The number of time points (i.e., N) is determined by dividing the total operation time by the period. For example, the number of time points may be determined in advance or arbitrarily by the user. Here, N is a natural number equal to or greater than 1.

[0414] In step 3730, the processor 110 may calculate the pressure of the tank 3230 at the Nth time point as corresponding to either the pressure of the tank 3290 at the destination or a preset pressure.

[0415] In step 3740, the processor 110 can calculate the amount of change in the pressure of the tank 3230 at each time point compared to the pressure of the tank 3230 at the previous time point. For example, when the processor 110 calculates the amount of change in the pressure of the tank 3230 at each time point compared to the pressure of the tank 3230 at the previous time point, the amount of change in the pressure of the tank 3230 is as shown in Equation 1 below.

[0416]

number

[0417] The processor 110 can set the values ​​of α, β, and γ taking into consideration the cargo characteristics of the ship 3220, the pressure change characteristics of the tank 3230, the shape characteristics of the tank 3230, etc. The evaporation energy of the liquefied gas evaporated from the tank 3230, the energy of the gas flowing into the tank 3230, and the energy of the gas discharged from the tank 3230 can be determined as constants by mass flow meters at gas consumption points of the ship 3220. For example, the gas consumption points may include a propulsion engine, a power generation engine, a gas combustion device, and a reliquefaction device.

[0418] In step 3750, the processor 110 can calculate a predicted value of the internal pressure of the tank 3230 at each time point based on the amount of change in the pressure of the tank 3230 at each time point calculated in step 540.

[0419] Referring again to FIG. 31, in step 3430, the estimates calculated by processor 110 may include an estimate of the amount of liquefied gas evaporated from tank 3230.

[0420] Conventionally, the predicted value of the amount of liquefied gas evaporated from tank 3230 has been calculated by substituting unpredictable data, such as the temperature of the liquefied gas inside tank 323, into a simple mathematical formula. This method has the problem of being unable to accurately predict the amount of evaporated liquefied gas.

[0421] To solve the problem of inaccurate prediction in the prior art, the processor 110 can use a deep learning model to calculate a predicted value of the amount of evaporated liquefied gas, where the deep learning model can be trained to calculate a predicted value of the amount of evaporated liquefied gas using training data such as operation information and route information.

[0422] In deep learning technology and cognitive science, a machine learning model refers to a statistical learning algorithm embodied based on the structure of a biological neural network or a structure that executes that algorithm.

[0423] For example, a deep learning model can exhibit problem-solving capabilities by learning in a way that, like a biological neural network, nodes, which are artificial neurons formed by combining synapses, repeatedly adjust the weights of their synapses and learn to reduce the error between the correct output corresponding to a specific input and the inferred output.

[0424] For example, a deep learning model may be implemented as a multilayer perceptron (MLP) consisting of multiple nodes and connections between them. The deep learning model according to this embodiment may be implemented using one of various artificial neural network model structures, including an MLP. For example, a deep learning model may include an input layer that receives input signals or data from the outside, an output layer that outputs output signals or data corresponding to the input data, and at least one hidden layer located between the input and output layers, receiving signals from the input layer, extracting characteristics, and transmitting them to the output layer. The output layer receives signals or data from the hidden layer and outputs them to the outside.

[0425] In addition, a deep learning model that calculates the amount of liquefied gas evaporated from the tank 3230 can be stored and operated in the user terminal 3210.

[0426] In step 3440, the processor 110 may calculate an economic operating index 3260 for the vessel 3220 using the predicted value.

[0427] Here, the economical operation index 3260 may include at least one of the amount of liquefied gas lost inside the tank 3230, the amount of liquefied gas consumed by the engine of the ship 3220, the amount of liquefied gas incinerated by the gas combustion device, the liquefied gas re-liquefaction flow rate inside the tank 3230, and the boil-off rate (BOR) per day.

[0428] For example, the processor 110 can calculate the economical operation index 3260 using operation information, route information, a predicted amount of liquefied gas evaporated from the tank 3230, and a predicted pressure in the tank 3230.

[0429] For example, the processor 110 may calculate the economic operation index 3260 using a flight optimization analytical model.

[0430] The operation optimization analysis model is a model for reducing the amount of liquefied gas loss by determining the flow rate of liquefied gas supplied to equipment such as tanks 3230, engines, and gas combustion equipment, as well as the flow rate of liquefied gas consumed by each equipment.

[0431] For example, the processor 110 can calculate the economic operation index 3260 using operation information, route information, predicted pressure in the tank 3230, and predicted amount of liquefied gas evaporated from the tank 3230 via an operation optimization analysis model.

[0432] The operation optimization analysis model can also calculate the economic operation index 3260 by taking into account several constraints. For example, the model can calculate the economic operation index 3260 by taking into account the relationship between mass, energy, amount of power, maximum / minimum liquefied gas consumption of the equipment, efficiency of the equipment, efficiency of the propulsion engine, efficiency of the power generation engine, efficiency of the shaft generator, pressure of the tank 3230, maximum / minimum speed of the ship 3220, and average speed of the ship 3220.

[0433] In addition, the operation optimization analysis model for calculating the economic operation index 3260 can be stored and operated in the user terminal 3210.

[0434] In step 3450, the processor 110 can compare measurements taken in real time as the vessel 3220 operates with predicted values ​​and economic operating indicators, and display the results of the comparison.

[0435] In step 3460, the processor 110 may use the results of the comparison to control the vessel 3220.

[0436] An example of the results of processor 110 comparing the measured values ​​with the predicted values ​​and the economical operation index will be described below with reference to FIGS. 35A to 35C.

[0437] FIG. 35A illustrates an example of a comparison of predicted liquefied gas cargo tank internal pressure with measured values ​​and average velocity with measured values, according to one embodiment.

[0438] As shown in FIG. 35A, the results of comparing the predicted values ​​with the measured values ​​can be calculated in the form of a graph 3810, but the manner in which the results of comparing the predicted values ​​with the measured values ​​are presented is not limited to graph 3810.

[0439] In the graph 3810, a first dotted line 3811 represents the predicted internal pressure value of the tank 3230, and a first solid line 3812 represents the measured internal pressure value of the tank 3230.

[0440] The processor 110 can control the ship 3220 so that the pressure in the tank 3230 falls within a pressure control range set based on the predicted internal pressure value of the tank 3230. For example, when the ship 3220 is sailing at full capacity, the pressure control range can mean a range of 60 mbarg or more and 190 mbarg or less.

[0441] When the pressure in the tank 3230 of the ship 3220 exceeds the pressure control region, the processor 110 can control the ship 3220 to reduce the pressure in the tank 3230. For example, the processor 110 can reduce the pressure in the tank 3230 by increasing the flow rate at which the tank 3230 supplies evaporated liquefied gas to the gas combustion device. The processor 110 can also reduce the pressure in the tank 3230 by changing the combination of the device that supplies the liquefied gas and the device that is supplied with the liquefied gas.

[0442] Also, if the measured value is less than the predicted value, in order to increase the pressure in the tank 3230, the processor 110 can control the tank 3230 so that the amount of liquefied gas that the tank 3230 supplies to the gas combustion device, the propulsion engine, the power generation engine, and the reliquefaction device is reduced.

[0443] In the graph 3810, a second dotted line 3813 represents the predicted velocity of the vessel 3220 and a second solid line 3814 represents the measured velocity of the vessel 3220.

[0444] The processor 110 may control the vessel 3220 so that the measured value corresponds to the predicted value. For example, if the measured value exceeds the predicted value, the processor 110 may control the vessel 3220 to decrease the speed of the vessel 3220. Alternatively, if the measured value is less than the predicted value, the processor 110 may control the vessel 3220 to increase the speed of the vessel 3220.

[0445] In graph 3810, a third dotted line 3815 represents the predicted load factor of the vessel's 3220 internal propulsion engine, and a third solid line 3816 represents the measured load factor of the vessel's 3220 internal propulsion engine.

[0446] The processor 110 can control the vessel 3220 so that the load factor of the internal propulsion engine of the vessel 3220 corresponds to the predicted value. For example, the processor 110 can change the combination of a device that supplies liquefied gas and a device that is supplied with liquefied gas. The processor 110 can also control the supply flow rate of the device that supplies liquefied gas.

[0447] For example, if the load factor of the internal propulsion engine of the vessel 3220 exceeds a predicted value, the processor 110 can control the vessel 3220 to increase the flow rate of liquefied gas supplied to the propulsion engine from the tank 3230. Also, if the load factor of the internal propulsion engine of the vessel 3220 is less than a predicted value, the processor 110 can control the vessel 3220 to increase the flow rate of liquefied gas supplied to the propulsion engine from the tank 3230.

[0448] FIG. 35B illustrates an example of a comparison of predicted and measured amounts of liquefied gas evaporated from a liquefied gas cargo tank according to one embodiment.

[0449] As shown in FIG. 35B, the results of comparing the predicted values ​​with the measured values ​​can be output in the form of a graph 3820, but the manner in which the results of comparing the predicted values ​​with the measured values ​​are displayed is not limited to graph 3820.

[0450] In graph 3820 , a first dotted line 3821 indicates the predicted amount of liquefied gas evaporated from tank 3230 , and a first solid line 3822 indicates the measured amount of liquefied gas evaporated from tank 3230 .

[0451] For example, the processor 110 may calculate a predicted amount of evaporated liquefied gas via a deep learning model.

[0452] The processor 110 can control the ship 3220 according to the predicted value of the amount of evaporated liquefied gas. For example, the processor 110 can control the type of device that supplies the evaporated liquefied gas according to the predicted value. Also, the processor 110 can control the supply flow rate of the device that supplies the evaporated liquefied gas.

[0453] For example, the processor 110 may, in response to the forecast, control the vessel 3220 to limit the supply of evaporated liquefied gas to the compressor included in the tank 3230. The processor 110 may also, in response to the forecast, control the vessel 3220 so that the compressor supplies a flow rate of evaporated liquefied gas of 2500 kg / h to the equipment of the vessel 3220. The processor 110 may also, in response to the forecast, control the vessel 3220 so that the compressor supplies a flow rate of evaporated liquefied gas of 1000 kg / h to the gas combustion equipment and a flow rate of evaporated liquefied gas of 1500 kg / h to the propulsion engine and the power generation engine.

[0454] In graph 3820, a first region 3823 represents the predicted flow rate of vaporized liquefied gas supplied to the gas combustion device by the compressor contained within tank 3230. A second region 3824 represents the predicted flow rate of vaporized liquefied gas supplied to the propulsion engine and the power generation engine by the compressor contained within tank 3230.

[0455] Also, although not shown in graph 3820, the predicted values ​​may include a predicted flow rate of evaporated liquefied gas that the compressor supplies to the reliquefaction device, a predicted flow rate of evaporated liquefied gas that the vaporizer contained within tank 3230 supplies to the propulsion engine and power generation engine, and a predicted flow rate of evaporated liquefied gas that the tank contained within tank 3230 supplies to the supercooling device.

[0456] FIG. 35C is a diagram illustrating an example of a result of comparing an economical operation index of a ship with a measured value according to one embodiment.

[0457] As shown in FIG. 35C, the results of comparing the economic operation index 3260 with the measured values ​​can be calculated in the form of a table 3830, but the manner in which the results of comparing the economic operation index 3260 with the measured values ​​are displayed is not limited to table 3830.

[0458] The first item 3831 means the amount of liquefied gas lost inside the tank 3230. For example, the processor 110 can calculate the measurement value of the first item 3831 as the sum of the cumulative value of the flow rate of liquefied gas supplied by the tank 3230 to the propulsion engine from the time when the ship 3220 departed to the time of measurement, the cumulative value of the flow rate of liquefied gas supplied by the tank 3230 to the power generation engine, and the cumulative value of the flow rate of liquefied gas supplied by the tank 3230 to the gas combustion device.

[0459] Additionally, second item 3832 refers to the amount of liquefied gas consumed by the engine of ship 3220. For example, processor 110 can calculate the measured value of first item 3831 as the sum of the cumulative value of the flow rate of liquefied gas supplied by tank 3230 to the propulsion engine from the time of departure to the time of measurement and the cumulative value of the flow rate of liquefied gas supplied by tank 3230 to the power generation engine.

[0460] Additionally, the third item 3833 means the amount of liquefied gas incinerated by the gas combustion device of the ship 3220. For example, the processor 110 can calculate the measurement value of the third item 3833 as the cumulative value of the flow rate of liquefied gas supplied from the tank 3230 to the gas combustion device from the time of departure to the time of measurement.

[0461] Additionally, the fourth item 3834 means the reliquefaction flow rate inside the tank 3230. For example, the processor 110 can calculate the measurement value of the fourth item as the cumulative value of the liquefied gas flow rate passing through the reliquefaction device from the time of departure to the time of measurement.

[0462] Furthermore, the fifth item 3835 means the evaporation rate per day. For example, the processor 110 can calculate the measurement value of the fifth item 3835 as the daily average value of the evaporation rate of the liquefied gas relative to the total volume of the liquefied gas stored in the tank 3230 from the time of departure to the time of measurement.

[0463] FIG. 36 is a diagram illustrating an example of a screen displayed by a processor according to an embodiment.

[0464] FIG. 36 shows an example of an execution screen 3900 of a program for executing the method described above with reference to FIGS. 29 to 35C.

[0465] For example, the area 3910 can display preset operation information of the ship 3220. Specifically, at least one of the pressure of the tank 3230 when the ship 3220 arrives at the destination, arrival and departure information, total operation distance, total operation time, average speed, and weather information can be displayed.

[0466] Additionally, area 3920 may display a screen that allows processor 110 to obtain route information. Processor 110 may obtain from server 3240 either a manually set route for vessel 3220 or a route that has been automatically set via a route optimization function.

[0467] Additionally, area 3930 may display the results of comparing the predicted pressure of the tank 3230 with the measured value, the results of comparing the average speed of the vessel 3220 with the measured value, and the results of comparing the predicted pressure of the tank 3230 with the measured value.

[0468] In addition, area 3940 can display the predicted flow rate of the liquefied gas that the internal compressor of tank 3230 supplies to the power generation engine and propulsion engine, and the predicted flow rate of the liquefied gas that the internal compressor of tank 3230 supplies to the gas combustion device, as a result of comparing the predicted value of the amount of liquefied gas evaporated from tank 3230 with the measured value.

[0469] Additionally, the area 3950 can display the results of comparing the economical operation index 3260 with the measured value. Specifically, the economical operation index 3260 can include at least one of the amount of liquefied gas lost inside the tank 3230, the amount of liquefied gas consumed by the engine of the ship 3220, the amount of liquefied gas burned by the gas combustion device, the liquefied gas reliquefaction flow rate inside the tank 3230, and the evaporation rate per day.

[0470] In addition, the area 3960 may display a comparison result between the flow rate supplied in real time to the device inside the ship 3220 and a preset flow rate. Furthermore, it may display whether power is being supplied to the device. Specifically, the device may include a compressor included in the tank 3230, a vaporizer included in the tank 3230, a gas combustion device, a reliquefaction device, etc.

[0471] According to the above, the present invention generates recommended navigation information for a ship's navigation route based on navigation plan information for the ship's departure and arrival points, predicts the ship's BOG generation amount and tank pressure value based on the recommended navigation information, and obtains optimal navigation information for ship operation control based on the BOG generation amount and tank pressure value.Furthermore, the present invention can control the ship using a preset navigation method using the optimal navigation information.

[0472] According to the above, the present invention predicts the amount of BOG generated and tank pressure values ​​based on environmental information including one or more of weather and climate information, tidal current information, sea condition information, and ocean current information for each location on the ship's navigation route, and obtains optimal voyage information based on the predicted amount of BOG generated and tank pressure values ​​to calculate the entire voyage, thereby making it possible to present the direction of cargo management from the ship's departure point to its arrival point and enable safe cargo handling based on the predicted tank pressure values.

[0473] Furthermore, the present invention can provide a navigation method that can maintain the final target tank pressure by increasing fuel consumption and lowering tank pressure at locations where more fuel is needed during the voyage, and by reducing fuel consumption and increasing tank pressure at locations where less fuel is needed.

[0474] Furthermore, the present invention can provide ship operators with ship navigation guidance and assist in navigation by controlling the ship using an operation method that minimizes liquefied gas consumption based on optimal navigation information, and can provide a carbon tax reduction effect by reducing the amount of GCU incineration and engine fuel gas.

[0475] Furthermore, the present invention can support ship operators in making decisions regarding operations by predicting the amount of BOG generated and tank pressure values ​​on unfamiliar ship routes.

[0476] Meanwhile, the above-described method can be created as a computer-executable program and can be implemented on a general-purpose digital computer that runs the program using a computer-readable recording medium. Furthermore, the data structure used in the above-described method can be recorded on a computer-readable recording medium by various means. The computer-readable recording medium includes magnetic storage media (e.g., ROM, RAM, USB, floppy disk, hard disk, etc.), optically readable media (e.g., CD-ROM, DVD, etc.), etc.

[0477] Meanwhile, the above-described method can be provided in a computer program product. The computer program product can be traded as a commodity between a seller and a buyer. The computer program product can be distributed in the form of a machine-readable storage medium (e.g., a compact disc read-only memory (CD-ROM)) or can be distributed online (e.g., downloaded or uploaded) via an application store (e.g., Play Store™) or directly between two user devices. In the case of online distribution, at least a portion of the computer program product can be at least temporarily stored or temporarily generated in a machine-readable storage medium, such as the memory of a manufacturer's server, an application store server, or an intermediary server.

[0478] Those skilled in the art will understand that the present invention can be embodied in various modified forms without departing from the essential characteristics of the above description. Therefore, the disclosed method should be considered from an illustrative rather than a restrictive perspective, and the scope of protection is defined in the claims rather than the above description, and should be interpreted to include all differences within the scope of equivalents.

Claims

1. 1. A method of optimizing ship operations, comprising: generating recommended voyage information for a vessel's navigation route based on navigation plan information for the vessel's departure and arrival points; predicting the amount of BOG generated by the ship and the tank pressure value of the ship based on the recommended voyage information; and obtaining optimal navigation information for operational control of the ship based on the BOG generation amount and the tank pressure value. method.

2. The flight plan information is including departure time, arrival time and location information, The generating step includes: acquiring environmental information related to the navigation route of the ship based on the navigation plan information; generating the recommended voyage information based on the operation plan information and the environmental information, and using fuel consumption and BOG generation amounts related to the operation route of the ship as a reference; The method of claim 1.

3. The recommended voyage information is The navigation route includes at least one of position information for each navigation section, speed information for each navigation section, and environmental information for each navigation section. The method of claim 1.

4. The predicting step includes: predicting the amount of BOG generated by the ship based on at least one of the position information for each navigation section, the speed information for each navigation section, and the environmental information for each navigation section; and predicting a tank pressure value of the ship based on at least one of the position information for each voyage section, the speed information for each voyage section, the environmental information for each voyage section, and a predetermined liquefied gas consumption amount. The method of claim 3.

5. The step of acquiring the optimum navigation information includes: generating n-th intermediate voyage information related to operational control of the vessel based on the BOG generation amount and the tank pressure value; and determining the (n+1)th intermediate navigation information as the optimum navigation information based on a comparison between the (n)th intermediate navigation information and the (n+1)th intermediate navigation information using a preset threshold as a reference; The n+1 intermediate voyage information is The nth intermediate voyage information is generated based on at least one updated value of the speed information for each voyage section and the liquefied gas consumption amount included in the nth intermediate voyage information, The n is a natural number of 1 or more. The method of claim 1.

6. The determining step includes: determining the (n+1)th intermediate navigation information as the optimal navigation information in response to a difference value calculated based on the (n)th intermediate navigation information and the (n+1)th intermediate navigation information being equal to or less than a preset threshold value; and generating (n+2)th intermediate voyage information by updating at least one of the speed information for each voyage section and the liquefied gas consumption amount included in the (n+1)th intermediate voyage information in response to the difference value exceeding a predetermined threshold. The method of claim 5.

7. The optimum voyage information is The information includes at least one of the BOG generation amount of the ship, the tank pressure value of the ship, the speed information of the ship by each voyage section, the liquefied gas consumption amount of the ship, and the usage amount of the equipment installed on the ship. The method of claim 1.

8. and further comprising controlling the ship in a preset navigation method using the optimal navigation information. The operation method is: including an operating procedure that minimizes the vessel's consumption of liquefied gas; The method of claim 1.

9. A computer-readable recording medium having a program recorded thereon for causing a computer to execute the method of claim 1.

10. at least one memory; at least one processor; The processor: generating recommended voyage information regarding a navigation route for the vessel based on navigation plan information regarding the vessel's departure and arrival points; predicting the amount of BOG generated by the ship and the tank pressure value of the ship based on the recommended voyage information; obtaining optimal navigation information for operation control of the ship based on the BOG generation amount and the tank pressure value; Computing equipment.

11. 1. A method for predicting BOG generation from a ship, comprising: selecting input data for a BOG generation prediction model from a portion of pre-stored operational data using an input data selection model; training a BOG generation prediction model including a plurality of deep learning models using the input data; and predicting the amount of BOG generated from the learned BOG generation prediction model using current operation data of the ship. method.

12. The selecting step includes: calculating a correlation coefficient between the portion of pre-stored operational data and the amount of BOG generated; training the input data selection model using the calculated correlation coefficients; and selecting, as the input data, data whose correlation coefficient is equal to or greater than a predetermined value using the learned input data selection model. The method of claim 11.

13. The learning step includes: calculating correct answer data for the learning; a learning step using the input data and the calculated correct answer data; and validating the BOG generation prediction model using another portion of pre-stored operational data. The method of claim 11.

14. The predicting step includes: outputting an initial predicted value of BOG generation for each of the plurality of deep learning models; and calculating a final predicted value of the amount of BOG generated by applying different weights to each of the initial predicted values ​​of the amount of BOG generated. The method of claim 11.

15. The calculating step applying a highest weight to the first predicted value of BOG generation output from the stacking model among the first predicted values ​​of BOG generation; 15. The method of claim 14.

Images