Predicting tank sloshing response using statistical models trained by machine learning

Abstract

The present invention relates to an estimation of a sloshing response of a sealed, insulated tank for transporting liquefied gas. A statistical model is trained by supervised machine learning methods based on a set of test data, including marine test data, and the statistical model is capable of estimating the sloshing response of the tank as a function of the tank's filling level and current sea states, which may optionally also be a function of at least one of the vessel's draft, the vessel's speed, and the vessel's heading. The statistical model so trained is used to estimate the sloshing response of a sealed, insulated tank for transporting liquefied gas (2). In one aspect, the statistical model estimates the sloshing response of the tank from the tank's filling level and current sea states, and may optionally also estimate the sloshing response of the tank from at least one of the vessel's draft, the vessel's speed, and the vessel's heading.

Description

[Technical Field]

[0001] The present invention relates to the estimation of the sloshing response of a sealed, insulated tank for transporting liquefied gas. More particularly, the present invention relates to a method for obtaining a statistical model capable of estimating the sloshing response of such a tank, a method for obtaining a database that can be used to estimate the sloshing response of such a tank, and a ship including at least one such tank. In this specification, the term "ship" refers to a means used to transport liquefied gas between two points on the globe and / or for its own propulsion, or one or more floating processing and / or storage units for liquefied gas. Traditionally, ships fueled by liquefied gas include LNG-fueled ships, container ships, cruise ships, and bulk carriers. [Background technology]

[0002] Sealed, insulated tanks are routinely used for the storage and / or transport of liquefied gases at low temperatures, such as for example the transport of liquefied petroleum gas (also known as LPG) at temperatures above -50°C and below 0°C, or the transport of liquefied natural gas (LNG) at approximately -162°C at atmospheric pressure. These tanks may be intended to transport liquefied gases and / or to receive liquefied gases that serve as fuel for the propulsion of floating structures. Many liquefied gases may be envisaged, such as methane, ethane, propane, butane, ammonia gas, dihydrogen or ethylene, among others.

[0003] Ship tanks may be single or double sealed membrane tanks that allow transport at atmospheric pressure. The sealing membranes are generally made of thin stainless steel or Invar sheets. One membrane is generally in direct contact with the liquefied gas.

[0004] The liquid contained in the tanks is subjected to various movements during its transportation. In particular, the movement of the ship at sea, due to the influence of climatic conditions such as sea state or wind, leads to agitation of the liquid in the tanks. The agitation of the liquid, commonly called sloshing, generates stresses on the tank walls that can compromise the integrity of the tank.

[0005] These sloshing phenomena occur on natural gas (hereinafter referred to as "LNG") transport and / or user vessels (often referred to as "LNG-fueled vessels") or LNG tankers. These sloshing phenomena also occur on moored storage vessels known as floating production storage and offloading (FPSO) vessels, such as oil production platforms and natural gas liquefaction units, commonly referred to as FLNG (Floating Liquefied Natural Gas) units or Floating Storage and Regasification Units (FSRUs) (more commonly referred to as floating structures for production, storage, and export). Sloshing phenomena occur in both rough and near-calm sea conditions if the liquefied gas cargo resonates with excitations also caused by low swells experienced by the vessel. In these resonance cases, sloshing can become extremely severe, especially when waves break on vertical walls or corners, thus posing a risk of deterioration of the liquefied gas containment system or the insulation system immediately behind the containment system.

[0006] Here, tank integrity is particularly important for liquefied gas tanks, such as LNG tanks, due to the flammable or explosive nature of the liquids transported and also due to the risk of cold spots in the steel hull of the floating unit in the event of a leak.

[0007] U.S. Patent No. 8,643,509 discloses a method for reducing the risks associated with sloshing of liquefied gas cargoes. In that document, the resonant frequencies of the liquid in the tanks are estimated as a function of the tanks and their fill levels. During transit, the frequencies of the vessel's motions are evaluated as a function of climatic and ocean conditions and the vessel's speed. The predicted motion frequencies are also evaluated for the course the vessel should follow. If any of the motion frequencies are too close to the resonant frequencies of the liquid in the tanks, an alarm is issued to change course and / or the vessel's speed to avoid a dangerous situation.

[0008] Despite the provision of sloshing reduction measures, sloshing of liquid in the tank, particularly due to resonance phenomena, can often result in deformation of the primary membrane of the primary seal, damage to the underlying structures present in the primary and / or secondary spaces on which the primary seal membrane rests, falling objects, particularly from stationary equipment, which can cause damage to the primary seal membrane in the short or medium term, or more generally the risk of deformation of the primary seal membrane beyond its structural tolerance limits.

[0009] Therefore, there remains a need for a method for estimating the sloshing response of a tank when a ship is underway, and, if necessary, for applying the necessary measures to prevent the occurrence of excessive sloshing that could cause damage to the tank's primary sealing membrane. Summary of the Invention

[0010] One idea behind the present invention is to use supervised machine learning methods to train a statistical model capable of estimating the sloshing response of a tank as a function of the tank's fill level and current sea state, which may optionally also be a function of at least one of the vessel's draft, vessel's speed, and vessel's heading. The statistical model is trained based on a set of test data obtained from the results of multiple tests, each of which consists of subjecting a test tank having a given fill level to motion and measuring the pressure at at least one point on one wall of the test tank and / or the number of impacts on at least one wall of the test tank. The statistical model can then be used to estimate the tank's sloshing response in the framework of a ship's management system, for example by building a database that can be used for real-time reference.

[0011] According to an embodiment according to a first aspect, the present invention provides a method for obtaining a statistical model capable of estimating the sloshing response in at least one sealed, insulated tank for the transportation of a liquefied gas, comprising: The method includes a step of training the statistical model by a supervised machine learning method based on a set of test data; the statistical model is capable of estimating a sloshing response of the tank as a function of the tank's fill level and current sea state, the function optionally further being a function of at least one of the vessel's draft, the vessel's speed, and the vessel's heading; The set of test data is obtained from the results of a plurality of tests, each of the plurality of tests comprising subjecting a test tank having a given fill level to motion and measuring the pressure at at least one point on one wall of the test tank and / or the number of impacts on at least one wall of the test tank.

[0012] According to an embodiment according to a second aspect, the present invention provides a method for obtaining a statistical model capable of estimating the sloshing response in at least one sealed, insulated tank for the transportation of a liquefied gas, comprising: The method includes a step of training the statistical model by a supervised machine learning method based on a set of test data; the statistical model is capable of estimating a sloshing response of the tank as a function of the fill level of the tank and a current state of vessel motion, the function optionally further being a function of at least one of the vessel draft, the vessel speed, and the vessel heading; The set of test data is obtained from the results of a plurality of tests, each of the plurality of tests comprising subjecting a test tank having a given fill level to motion and measuring the pressure at at least one point on one wall of the test tank and / or the number of impacts on at least one wall of the test tank.

[0013] "Supervised machine learning method" (in English, supervised learning) means a machine learning (also known in English as machine learning, in French as artificial learning or statistical learning) method that consists in training a prediction function on the basis of annotated examples. In other words, supervised machine learning methods make it possible to build a predictive model on the basis of multiple examples where the response to be predicted is known. Supervised machine learning methods are typically performed by a computer, i.e. the steps involving training the statistical model are typically performed by a computer.

[0014] The statistical model trained in the context of the present invention is capable of estimating a sloshing response involving one or more quantitative variables as a function of at least the tank filling level and the current state of the current sea state or vessel motion, and thus is capable of addressing a regression problem.

[0015] There is also a step of training a statistical model on the set of test data by supervised machine learning methods, whereby the statistical model is capable of computationally calculating the sloshing response of the tank as a function of at least the tank's fill level and the current state of the sea state or the vessel's motion, including the tank's fill level and the current state of the sea state and the vessel's motion for which no tests have yet been performed, and thus the statistical model can be used to estimate the tank's sloshing response under actual conditions of use on a vessel.

[0016] The test tank may be smaller than the tank for which the statistical model is to estimate the sloshing response. The test tank may have a geometric shape representative of the tank for which the statistical model is to estimate the sloshing response. Furthermore, it is clear that in the phrase or feature "training the statistical model by a supervised machine learning method based on a test data set," the "test data set" may include or consist of data from so-called "live" operations (i.e., data acquired or measured on a ship in service for liquefied gas transportation and / or use).

[0017] In embodiments, the method may include one or more of the following features.

[0018] Sloshing response means any parameter and set of qualitative and / or quantitative parameters that can describe the mechanical loads to which a tank is subjected during cargo sloshing.

[0019] According to one embodiment, the sloshing response includes at least one of the number of collisions of the fluid against the walls of the tank, the maximum pressure against the walls of the tank, and the likelihood of damage to the tank.

[0020] The statistical model is therefore capable of estimating the probability of tank damage and / or parameters that make it possible to estimate the probability of this type of damage.

[0021] According to one embodiment, the supervised machine learning method is a Gaussian process regression method.

[0022] Gaussian process regression methods are well suited for training statistical models because they allow for the creation of statistical models that can address regression problems for any input data set by training them on a relatively limited amount of data. Nevertheless, other supervised machine learning methods can be used without departing from the scope of the present invention.

[0023] According to one embodiment, at least one constraint is applied to the statistical model during training of said statistical model by said supervised machine learning method.

[0024] Thus, the training of the statistical model can be guided by fundamental physical considerations, such as the absence of sloshing in situations where the tank fill level is zero, and / or by the fact that, for example, larger movements or larger dimensions of the tank can potentially result in a larger sloshing response, thereby improving the accuracy of the estimation of the sloshing response by the statistical model.

[0025] According to one embodiment, the method further comprises, prior to the step of training the statistical model, a step comprising excluding test results characterized by a sloshing response below a threshold from said set of test data.

[0026] Therefore, the statistical model is trained only on test data that exhibits severe sloshing, more specifically with respect to the number of impacts. In fact, according to one aspect of the present invention, the number of occurrences encountered is a more important factor for statistical convergence than the severity of the impact. This results in a further improvement in the accuracy of the sloshing response estimation by the statistical model.

[0027] According to one embodiment, a statistical model considers a plurality of tanks, said statistical model being capable of estimating the sloshing response of each of said plurality of tanks as a function of the position of said tank on said vessel.

[0028] According to an embodiment according to said first aspect, the present invention provides a system for obtaining a statistical model capable of estimating a sloshing response in at least one sealed, insulated tank for the transportation of a liquefied gas, comprising: The system includes processing means for training the statistical model by a supervised machine learning method based on a set of test data; the statistical model is capable of estimating a sloshing response of the tank as a function of the tank's fill level and current sea state, the function optionally further being a function of at least one of the vessel's draft, the vessel's speed, and the vessel's heading; The system also provides a method for measuring the set of test data obtained from the results of a plurality of tests, each of which includes subjecting a test tank having a given fill level to motion and measuring the pressure at at least one point on one wall of the test tank and / or the number of impacts on at least one wall of the test tank.

[0029] According to an embodiment according to said second aspect, the present invention provides a system for obtaining a statistical model capable of estimating a sloshing response in at least one sealed, insulated tank for the transportation of a liquefied gas, comprising: The system includes processing means for training the statistical model by a supervised machine learning method based on a set of test data; the statistical model is capable of estimating a sloshing response of the tank as a function of the fill level of the tank and a current state of vessel motion, the function optionally further being a function of at least one of vessel draft, vessel speed, and vessel heading; The system also provides a method for measuring the set of test data obtained from the results of a plurality of tests, each of which includes subjecting a test tank having a given fill level to motion and measuring the pressure at at least one point on one wall of the test tank and / or the number of impacts on at least one wall of the test tank.

[0030] Such a system has the same advantages as the above method.

[0031] According to an embodiment according to said first aspect, the present invention provides a method for obtaining a database that can be used to estimate a sloshing response in at least one sealed insulated tank for the transportation of a liquefied gas, said method comprising the steps of: generating a plurality of input data vectors, each of said plurality of input data vectors including the fill level of said tank and the current sea state; obtaining an estimated sloshing response in the tank using the statistical model obtained by the method according to the first aspect for each of the generated plurality of input data vectors; storing the estimated sloshing response in the tank associated with the plurality of input data vectors in a database; Also provided is a method comprising the steps of:

[0032] According to an embodiment according to said second aspect, the present invention provides a method for obtaining a database that can be used to estimate a sloshing response in at least one sealed insulated tank for the transportation of a liquefied gas, said method comprising the steps of: generating a plurality of input data vectors, each of the plurality of input data vectors including the fill level of the tank and a current state of vessel motion; obtaining an estimated sloshing response in the tank using the statistical model obtained by the method according to the second aspect for each of the generated plurality of input data vectors; storing the estimated sloshing responses in the tank associated with the plurality of input data vectors in a database; Also provided is a method comprising the steps of:

[0033] While statistical models can be estimated by calculating the tank's sloshing response for tank fill level values ​​and current sea states or ship motion conditions for which testing has not yet been performed, the calculations required to do so may be too lengthy and / or the required computational resources may be too great to be available on board the ship. In such cases, it is important to obtain an estimate of the sloshing response as quickly as possible and using the lowest possible cost onboard systems. Therefore, one idea behind these methods is to perform most of these calculations in advance based on multiple input data vectors, which may be appropriately selected to cover the entire range of ship operations or functions, and to store the estimated tank sloshing response for each of these input data vectors in a database in association with the input data vector. The estimated tank sloshing response can be obtained by simply reading the database if the input data vector is present in the database, or by interpolation based on the database if not. This requires much less computation time and resources than estimation based on the statistical model itself. As a result, the statistical model itself may not even be necessary to perform statistical model estimation on board the ship; only the database is sufficient. In this case, the database-based estimation can be performed by a system on board the vessel, or even by a land station communicating with the vessel, for example via radio or satellite.

[0034] According to one embodiment, the present invention also provides a database obtained by the method for obtaining a database described above.

[0035] According to one embodiment, the present invention also provides a computer-readable storage medium having stored thereon a database obtained by the above method for obtaining a database.

[0036] According to an embodiment according to said first aspect, the present invention provides a method for estimating a sloshing response in at least one sealed insulated tank for the transportation of liquefied gas on a ship, said method comprising the steps of: determining the current fill level of said tank; Find the current sea conditions, generating an input data vector including the determined current fill level of the tank and the determined current sea state; estimating a sloshing response in the tank based on the generated input data vector and the database obtained by the method according to the first aspect; Also provided is a method comprising the steps of:

[0037] According to an embodiment according to said second aspect, the present invention provides a method for estimating sloshing response in a sealed insulated tank for the transport of liquefied gas on a ship, said method comprising: determining the current fill level of said tank; determining a current state of motion of said vessel; generating an input data vector comprising the determined current fill level of the tank and the determined current state of motion of the vessel; estimating a sloshing response in the tank based on the generated input data vector and the database obtained by the method according to the second aspect; Also provided is a method comprising the steps of:

[0038] These methods allow the sloshing response of a tank to be estimated by a statistical model pre-trained on test data using a database. As mentioned above, this estimation requires much less computational time and resources than an estimation based on the statistical model itself, and can be performed by a system onboard the ship or by a shore station in communication with the ship.

[0039] According to one embodiment, a plurality of tanks is considered and the method comprises a preliminary step of defining the respective positions of said plurality of tanks on the vessel.

[0040] According to one embodiment, the method includes a step comprising giving a warning to a user if the estimated sloshing response of the tank exceeds a warning threshold, and preferably further includes a step comprising decision support intended to reduce the sloshing, which may consist of a suggested change in the direction or course of the vessel, a change in heading, which is particularly suitable for stationary floating structures, a change in the speed of the vessel or a change in the fill level of one or more tanks (a change in fill level between tanks or, in the case of stationary floating structures, between a tank and a storage facility external to the vessel).

[0041] Furthermore, the alert may consist of a report of a problem to be corrected immediately or in the near future and, if possible, a warning designating one or more tanks from the ship's tanks that require inspection and maintenance operations in consideration of possible repairs.

[0042] Thus, users such as crew members can take necessary measures to limit sloshing in the tank, if necessary, for example slowing or stopping the vessel or changing the vessel's course, thereby reducing the risk of damage to the tank.

[0043] According to an embodiment according to said first aspect, the present invention provides a management system for a ship comprising at least one sealed insulated tank for the transportation of liquefied gas, said management system comprising: at least one fill level sensor for measuring the current fill level of the tank; a sea state assessment device capable of assessing current sea conditions; - processing means configured to generate an input data vector comprising the current filling level of the tank and the current sea state assessed by the sea state assessment device, and to estimate a sloshing response in the tank using the generated input data vector and the database obtained by the method according to the first aspect; A management system including:

[0044] According to an embodiment according to the second aspect, the present invention provides a management system for a ship comprising at least one sealed insulated tank for the transportation of liquefied gas, the management system comprising: at least one fill level sensor for measuring the current fill level of the tank; a vessel motion assessment device capable of assessing a current state of motion of said vessel; - processing means configured to generate an input data vector comprising the current filling level of the tank and the current state of the vessel's motion, and to estimate a sloshing response in the tank using the generated input data vector and the database obtained by the method according to the second aspect; A management system including:

[0045] Such a system has the same advantages as the above method.

[0046] According to one embodiment, the processing means is further configured to provide a warning to a user if the estimated sloshing response of the tank exceeds a warning threshold, and preferably to assist the user with decisions intended to reduce the sloshing.

[0047] According to another embodiment, the present invention provides a method for estimating sloshing response in at least one sealed insulated tank for the transportation of liquefied gas on a ship, said method comprising the steps of: determining the current fill level of said tank; Estimating future sea conditions based on meteorological information and the course of the vessel; generating a plurality of input data vectors, each of the plurality of input data vectors including the current fill level of the tank and the estimated future sea state; estimating a future sloshing response in the tank based on the generated input data vector and the database obtained by the method according to the first aspect; Also provided is a method comprising the steps of:

[0048] According to the method, a statistical model pre-trained on test data from a database can be used to estimate the future sloshing response of a tank from meteorological information and the course of a ship. As mentioned above, this estimation requires much less computational time and resources than an estimation based on the statistical model itself, and can be performed by a system onboard the ship or by a shore station in communication with the ship.

[0049] According to another embodiment, the present invention provides a method for estimating sloshing response in at least one sealed insulated tank for the transportation of liquefied gas on a ship, said method comprising the steps of: determining the current fill level of said tank; estimating future states of motion of the vessel based on meteorological information and the course of the vessel; generating a plurality of input data vectors, each of the plurality of input data vectors including the current fill level of the tank and the estimated future state of vessel motion; estimating a future sloshing response in the tank based on the generated input data vector and the database obtained by the method according to the second aspect; Also provided is a method comprising the steps of:

[0050] According to one embodiment, the method comprises a step comprising determining a change in the vessel's course and / or the tank's fill level that can reduce the future sloshing response of the tank. The expression "vessel course" means simply avoiding the vessel's heading, vessel speed or geographical zone. For stationary floating structures (ships, barges), i.e., those at a fixed location, a change in heading is reflected in a change in the angle between the north direction and the longitudinal axis of the floating structure, so as to steer or move the floating structure in a conventional manner, reducing the detrimental consequences of swells and waves on the floating structure.

[0051] Thus, a user such as a crew member can decide to have the vessel follow a course that will enable a reduction in the future sloshing response of the tank and reduce the risk of damage to the tank.

[0052] According to another embodiment, the present invention provides a management system for a ship comprising at least one sealed insulated tank for the transportation of liquefied gas, said management system comprising: at least one fill level sensor for measuring the current fill level of the tank; a sea state estimation device capable of estimating future sea states based on meteorological information and the course of the ship; processing means configured to generate a plurality of input data vectors and to estimate a future sloshing response in the tank using the generated input data vectors and the database obtained by the method according to the first aspect; Including, The present invention also provides a management system, wherein each of the plurality of input data vectors includes a current fill level of the tank and a future sea state estimated by the sea state estimation device.

[0053] According to another embodiment, the present invention provides a management system for a ship comprising at least one sealed insulated tank for the transportation of liquefied gas, said management system comprising: at least one fill level sensor for measuring the current fill level of the tank; a motion state estimation device capable of estimating a future motion state of the vessel based on meteorological information and the course of the vessel; processing means configured to generate a plurality of input data vectors and to estimate a future sloshing response in the tank using the generated input data vectors and the database obtained by the method according to the second aspect; Including, The present invention also provides a management system, wherein each of the plurality of input data vectors includes a current fill level of the tank and the future motion state of the ship estimated by the motion state estimation device.

[0054] According to one embodiment, the processing means is further configured to determine a course for the vessel that will reduce the future sloshing response of the tank.

[0055] The present invention will be better understood with reference to the accompanying drawings, and other objects, details, features and advantages of the present invention will become more clearly apparent in the following description of some particular embodiments of the present invention, given by way of illustration only and not of limitation. [Brief explanation of the drawings]

[0056] [Figure 1] Figure 1 is a schematic diagram of a liquefied gas carrier. [Figure 2] FIG. 2 shows the management system integrated into the vessel of FIG. [Figure 3] FIG. 3 illustrates a management system according to another embodiment. [Figure 4] Figure 4 is a schematic diagram of the test tank sloshing response test device. [Figure 5]FIG. 5 is a flow chart illustrating a method for obtaining a database that can be used to estimate the sloshing response of a tank. [Figure 6] FIG. 6 is a flow chart illustrating a method for estimating the sloshing response of a tank. [Figure 7] FIG. 7 is a flow chart illustrating another method for estimating the sloshing response of a tank. [Figure 8] FIG. 8 is a flow chart illustrating yet another method for estimating the sloshing response of a tank. DETAILED DESCRIPTION OF THE INVENTION

[0057] The following embodiments are described in relation to a ship including a double hull forming a support structure in which a number of sealed, insulated tanks are arranged, in which the tanks have, for example, a polyhedral, e.g., prismatic, shape.

[0058] Such sealed, insulated tanks are provided, for example, for the transportation of liquefied gases, which are transported in such tanks at low temperatures and for which insulated tank walls are required to maintain the liquefied gas at such low temperatures. Therefore, on the one hand, it is particularly important to keep the integrity of the tank walls, including the insulating space located below the sealing membrane, intact, in order to maintain the tank's seal, avoid leakage of the liquefied gas from the tank, and not to deteriorate the insulating properties of the tank in order to maintain the gas in a liquid state.

[0059] Such a sealed insulated tank is anchored to the double hull of the vessel and also includes an insulating barrier with at least one sealed membrane. By way of example, such a tank may be manufactured according to technology commercially available under the Applicant's trademarks Mark III® or NO96®, or others.

[0060] FIG. 1 shows a vessel 1 including four sealed, insulated tanks 2. The four tanks 2 may have the same or different fill conditions. When the vessel 1 is at sea, it undergoes numerous motions associated with navigational conditions. These motions of the vessel 1 are transmitted to the liquid contained in the tanks 3, 4, 5, and 6, which are consequently subjected to internal motions. These motions of the liquid within the tanks 3, 4, 5, and 6 generate impacts on the walls of the tanks 3, 4, 5, and 6 that, if too severe, can quickly damage the tanks. Furthermore, repeated impacts on the walls of the tanks 3, 4, 5, and 6 at high, but not destructive, levels can lead to wall degradation through fatigue wear. Maintaining the integrity of the walls of the tanks 3, 4, 5, and 6 is important to maintain the sealing and insulating properties of the tanks 3, 4, 5, and 6.

[0061] It is known to avoid critical sailing conditions by preventing liquid movements that risk immediately damaging the tanks. However, there is still a need for a method that makes it possible to estimate the sloshing response of a tank when the ship is underway and, if necessary, to take the necessary measures to prevent excessive sloshing that would damage the primary sealing membrane of the tank.

[0062] 2 shows an example of a management system 100 installed on a vessel 1. The management system 100 includes a CPU (Central Processor Unit) 110 connected to a number of on-board sensors 120 that allow various parameter measurements to be obtained. The sensors 120 thus include, for example, but are not limited to, at least one fill level sensor 121 for each tank, various sensors of vessel motion 122, and sea state sensors 123. The management system 100 also includes a communication interface 130 that allows the CPU 110 to communicate with remote devices, for example, to obtain weather data, vessel position data, or other data.

[0063] The vessel motion sensor 122 determines the measured motion of the vessel, for example, by measuring accelerations experienced by the vessel on three orthogonal axes in translation and rotation. To assess the vessel motion, an inertial measurement unit (IMU) is preferably used, which may consist of one or more accelerometers and / or one or more gyroscopes, e.g., mechanical gyroscopes, and / or one or more magnetometers. If multiple measurement units (of the same type or two different types) are used, these units are preferably distributed throughout the vessel to generate accurate measurements of the vessel motion. It should be noted that IMUs are sometimes commonly referred to as motion reference units (MRUs).

[0064] Sea state sensor 123 obtains current sea conditions in the vessel's environment, such as wave height and frequency in the vessel's environment. For example, in one embodiment, wave height and / or frequency are obtained from visual observations by the crew.

[0065] The management system 100 further comprises a human-machine interface 140, which comprises a display means 41, which allows the operator to obtain the management information calculated by the system, or the measurements obtained by the sensors 120, or even the current sloshing state, which can be estimated as explained in more detail below.

[0066] The human-machine interface 140 further includes acquisition means 42 that enable the operator to manually provide numerical values ​​to the CPU 110, typically providing data that cannot be obtained by sensors because the vessel is not equipped with the necessary sensors or the vessel is damaged. For example, in one embodiment the acquisition means allows the operator to input information regarding wave height and / or frequency based on visual observation, and / or to manually input the vessel's heading and / or speed.

[0067] The management system 100 further includes a database 150, which can be used to estimate the sloshing response of the tank, as will be explained in more detail below.

[0068] 3 shows an example of a management system 200 located on land and communicating with a vessel 1. The vessel includes a CPU 110, a sensor 120, and a communication interface 130. The management system 200 includes a CPU 210, a communication interface 230, a human-machine interface 240, and a database 250. The functionality of the management system 200 is similar to that of the management system 100, differing only in that information measured by the sensor 120 on the vessel 1 is transmitted to the management system 200 located on land via the communication interfaces 130 and 230. For example, the communication interface may use terrestrial or satellite radio transmission of data.

[0069] Next, how the database 150 is obtained will be explained with reference to FIGS.

[0070] 4 shows a schematic example of a test device 1000 that allows a test to be carried out on a test tank 1010. The test consists of subjecting the test tank 1010, which is filled with a fluid 1011 at a given filling level, to a movement and measuring the pressure at at least one point on the wall 1010a of the test tank 1010 using a pressure sensor 1012 and / or measuring the number of impacts on at least one wall of the test tank 1010.

[0071] The test tank 1010 may be smaller than the tank for which the sloshing response is to be estimated and / or may have a geometry representative of the tank for which the sloshing response is to be estimated.

[0072] Of course, the fluid 1011 is preferably of the same nature, ideally having the same temperature, density and viscosity as that transported by the tank for which the sloshing response estimation is performed, and may in particular be, for example, liquefied petroleum gas (LPG) at a temperature between -50°C and 0°C or liquefied natural gas (LNG) at about -162°C at atmospheric pressure. Many liquefied gases may be envisaged, such as methane, ethane, propane, butane, ammonia gas, dihydrogen or ethylene, among others.

[0073] Furthermore, it is possible to measure pressure at multiple points on the wall 1010a (or some or all of the walls) of the test tank 1010, with the number and placement of pressure sensors 1012 adjusted accordingly. If the number of impacts on at least one wall of the test tank 1010 is measured, the measurement is made using multiple pressure sensors 1012 appropriately positioned on that wall. It is possible to measure the number of impacts on multiple walls of the test tank 1010 or on all of the walls of the test tank 1010.

[0074] As mentioned above, the test tank 1010 undergoes movement during testing. Thus, in the illustrated example, the apparatus 1000 comprises a platform 1013 to which the test tank 1010 is fixed. The platform 1013 is driven in movement by the action of six hydraulic cylinders 1015, one of whose ends is connected to the platform at three fixed points 1014 and the other end of which is connected to the frame or ground 1001. This allows the test tank 1010 to be driven in movement with six degrees of freedom in translation and rotation. Of course, the test tank 1010 may be driven in movement by different means.

[0075] The apparatus 1000 further comprises a test control unit 1020. The test control unit 1020 is configured to control the hydraulic cylinder 1015 to impart predetermined movements within a test program to the test tank 1010. In one embodiment, these movements are movements representative of given movements of the vessel, preferably taking into account the tank's position on the vessel and / or its geometric shape. In another embodiment, these movements are movements representative of given sea states that are converted into corresponding movements of the vessel, preferably taking into account the tank's position on the vessel and / or its geometric shape. Evaluating the corresponding movements of the vessel based on given sea states is a routine task in assessing the seaworthiness of a vessel. Furthermore, the test control unit 1020 stores values ​​measured during the test by the at least one pressure sensor 1012.

[0076] The test controller 1020 communicates with a test data processor 1030. The test data processor 1030 comprises a communication interface 1031 that allows it to receive from the test controller 1020 values ​​measured during the test by the at least one pressure sensor 1012 and movements imparted to the test tank 1010 during the test. The test data processor 1030 further comprises a memory 1033 and a CPU 1032.

[0077] The test data processor 1030 is configured to train a statistical model in a CPU 1032 in communication with a memory 1033 by machine learning methods. The statistical model can estimate a sloshing response of a tank as a function of the tank's fill level and current sea state, which may optionally also be a function of at least one of the vessel's draft, the vessel's speed, and the vessel's heading. The sloshing response includes at least one of the number of fluid collisions against the tank wall, the maximum pressure against the tank wall, and the likelihood of damage to the tank. In one aspect, the statistical model considers multiple tanks, and the statistical model can estimate the sloshing response of each of the multiple tanks as a function of the tank's position on the vessel.

[0078] More specifically, the statistical model is trained by a supervised machine learning method. For example, the supervised machine learning method may be a Gaussian process regression method. Gaussian process regression is well known per se and is well suited for training statistical models because it can generate statistical models that can address regression problems for any input data set by training them based on a relatively limited amount of data. Nevertheless, other supervised machine learning methods can be employed.

[0079] The statistical model is trained based on test results generated using the test tank 1010. More specifically, in a preferred example, the statistical model is trained based on the sloshing response of the test tank 1010 during each test, which sloshing response is pre-calculated based on values ​​measured during the test by at least one pressure sensor 1012. The sloshing response of the test tank 1010 may include at least one of the number of fluid impacts on one or more walls 1010a of the test tank 1010 and the maximum pressure on the walls 1010a of the test tank 1010 over a given period of time. In one aspect, the statistical model may be trained based on both the results of tests conducted on the test tank 1010 and test data acquired or measured in one or more tanks of vessels in service as liquefied gas transport and / or service vessels, serving as the test tank 1010. In other embodiments, the statistical model may be trained solely based on test data acquired or measured in one or more tanks of vessels in service as liquefied gas transport and / or service vessels, those vessels serving as test tanks 1010.

[0080] A method 300 for obtaining the database 150 will be described with reference to Figure 5. Steps 301 to 305 may be performed by a CPU 1032 in communication with a memory 1033.

[0081] The method 300 may optionally further include step 301 consisting of excluding test results that indicate a sloshing response of the test tank 1010 below a threshold value from the set of test data for training the statistical model. Thus, the statistical model is trained based only on test data that indicated significant sloshing in the test tank 1010, thereby improving the accuracy of the estimation of the sloshing response using the statistical model.

[0082] After optional step 301, method 300 includes step 302 of training the statistical model described above.

[0083] Optionally, at least one constraint may be applied to the statistical model during training of the statistical model by the supervised machine learning method in step 302. The constraints may be defined based on fundamental physical considerations, e.g., based on the absence of sloshing in situations where the tank filling level is zero, and / or based on practical experience, e.g., the fact that larger movements or larger dimensions of the tank can potentially result in a larger sloshing response, thereby improving the accuracy of the estimation of the sloshing response by the statistical model.

[0084] After completion of step 302, a statistical model is obtained that can estimate the sloshing response of the tank as a function of the tank's fill level and the current sea state, which may optionally also be a function of at least one of the vessel's draft, vessel's speed, and vessel's heading, which may be any value for those quantities, including those for which tests have not yet been performed in the test tank 1010. However, the calculations required to do this may be too lengthy and / or the required computing resources may be too great to be available on board the vessel, in which case it is important to obtain an estimate of the sloshing response as quickly as possible and using the lowest cost onboard systems possible. For this purpose, step 302 is followed by step 303, in which a plurality of input data vectors are generated, each of the plurality of input data vectors comprising the tank filling level and the current sea state, and then step 304 is performed, which consists of obtaining, for each input data vector generated in step 303, an estimated sloshing response of the tank with the help of the statistical model in step 302, and storing the estimated sloshing response of the tank together with the associated input data vector in a database.

[0085] In step 305, the database obtained in step 304 may optionally be transmitted to management system 100 or stored on a computer readable storage medium. Database 150 is also obtained, the use of which is described below.

[0086] Up to this point, we have described a situation in which the statistical model is capable of estimating the tank's sloshing response as a function of at least the tank's fill level and the current sea state. However, in one aspect, the statistical model is capable of estimating the tank's sloshing response as a function of the tank's fill level and the current state of the vessel's motion, which may optionally also be a function of at least one of the vessel's draft, vessel's speed, and vessel's heading. Steps 302, 303, and 304 are therefore modified accordingly.

[0087] Next, a method 400 for estimating the sloshing response of a tank using the database 150 will be described with reference to Figure 6. The method 400 may consider multiple tanks rather than just one tank. In this case, a preliminary step of defining the respective positions of the multiple tanks on the ship may be performed before the method 400 is performed.

[0088] According to a first embodiment, the flowchart of Figure 6 is executed in its entirety in CPU 110, which forms a single processing means. According to a second embodiment, the flowchart of Figure 6 is executed in part in a land-based management system 200 in communication with the vessel. According to this second embodiment, the vessel 1 transmits all information from the sensors 120 to a land-based station, and CPU 110 and CPU 210 form a shared processing means.

[0089] The method 400 comprises a first step 401 consisting of determining the current fill level of the tank and the current sea state. The current fill level of the tank is typically determined based on fill information provided by the tank fill level sensor 121. The current sea state may also be determined from information provided by the sea state sensor 123 and / or from information provided by terrestrial or satellite radio communication with a network of weather stations.

[0090] In step 401, optionally, the vessel draft and / or vessel heading may also be determined based on information typically provided by the vessel's on-board systems. The vessel draft is typically provided to the vessel's on-board systems by one or more float and / or hydrostatic sensors. The vessel heading is typically provided to the vessel's on-board systems by one or more navigation compasses.

[0091] The method 400 further includes a second step 402 that includes generating an input data vector that includes the data determined in step 401 .

[0092] The method 400 further includes a third step 403 that includes estimating the sloshing response of the tank based on the input data vector generated in step 402 and the database 150. More specifically, if the input data vector is indicated to exist in the database 150, the sloshing response is obtained by simply reading the database 150. Nevertheless, the database 150 will typically not contain the input data vector, but rather will contain input data that is close to that contained in the input data vector. In this case, the sloshing response is obtained by interpolation from sloshing responses associated with two or more adjacent input data vectors present in the database 150.

[0093] After step 403, the obtained sloshing response can be compared with a warning threshold and, if the sloshing response exceeds the warning threshold, a warning can be displayed to the user, for example on the display means 41. The display of this warning is preferably followed by a decision support step intended to reduce sloshing. This decision support step may consist of a proposed change in the direction or course of the vessel, a change in heading, which is particularly suitable for stationary floating structures, a change in the vessel's speed or a change in the fill level of one or more tanks (between tanks or, in the case of stationary floating structures, between the tanks and storage facilities external to the vessel). Furthermore, the warning may consist of reporting a problem to be corrected immediately or in the near future and specifying, if possible, one or more tanks from the vessel's tanks that require inspection and maintenance operations, with a view to possible repairs.

[0094] Another method 500 for estimating the sloshing response of a tank using the database 150 is described with reference to Figure 7. In this embodiment, the database 150 is derived from a statistical model capable of estimating the sloshing response of a tank as a function of the tank's fill level and the current state of the vessel's motion, as described above with reference to Figure 4, and optionally also as a function of at least one of the vessel's draft, vessel's speed, and vessel's heading. The method 500 may also consider multiple tanks rather than just one tank. In this case, there may be a preliminary step of defining the respective positions of the multiple tanks on the vessel before performing the method 500.

[0095] The method 500 comprises a first step 501 consisting of determining the current fill level of the tank and the current state of vessel motion. The current fill level of the tank is typically determined based on fill information provided by the tank fill level sensor 121. The current state of vessel motion may also be determined based on information provided by the vessel motion sensor 122.

[0096] In step 501, optionally, the vessel draft and / or vessel heading may also be determined based on information typically provided by the vessel's on-board systems. The vessel draft is typically provided to the vessel's on-board systems by one or more float and / or hydrostatic sensors. The vessel heading is typically provided to the vessel's on-board systems by one or more navigation compasses.

[0097] The vessel motion sensor 122 typically has an acquisition frequency much higher than the duration of a typical evolution of tank sloshing, and the information provided by the vessel motion sensor 122 may be averaged over an acquisition period, and then the other data determined in step 501 is averaged over the same acquisition period.

[0098] The method 500 further includes a second step 502 , similar to step 402 , which consists of generating an input data vector containing the data determined in step 501 .

[0099] The method 500 further includes a third step 503 consisting of estimating the sloshing response of the tank based on the input data vector generated in step 502 and the database 150. Step 503 is similar to step 403 and therefore will not be described in detail again.

[0100] After step 503, the obtained sloshing response can be compared with a warning threshold and, if the sloshing response exceeds the warning threshold, a warning can be displayed to the user, for example on the display means 41. The display of this warning is preferably followed by a decision support step intended to reduce sloshing. This decision support step may consist of a suggested change in the direction or course of the vessel, a change in heading, which is particularly suitable for stationary floating structures, a change in the vessel's speed or a change in the fill level of one or more tanks (between tanks or, in the case of stationary floating structures, between the tanks and storage facilities external to the vessel). Furthermore, the warning may consist of reporting a problem to be corrected immediately or in the near future and specifying one or more tanks from the vessel's tanks that require inspection and maintenance operations, if possible, to allow for possible repairs.

[0101] Another method 600 for estimating the sloshing response of a tank using a database 150 is described with reference to Figure 8. In this embodiment, the database 150 is derived from a statistical model capable of estimating the sloshing response of a tank as a function of the tank's fill level and current sea state, which may optionally also be a function of at least one of the vessel's draft, the vessel's speed, and the vessel's heading.

[0102] The method 600 comprises a first step 601 consisting of determining the current filling level of the tank and estimating future sea states. The current filling level of the tank is typically determined based on filling information provided by the tank's filling level sensor 121. The future sea states are estimated based on meteorological information and the vessel's course. The vessel's course is typically obtained from information provided by the vessel's on-board systems, such as the vessel's speed and vessel heading. The meteorological information may be provided by the sea state sensor 123 and / or by terrestrial or satellite radio communication with a network of weather stations.

[0103] The method 600 further includes a second step 602 that comprises generating a plurality of input data vectors, each of the plurality of input data vectors including the current fill level of a tank and an estimated future sea state.

[0104] In step 601, optionally, vessel draft, vessel heading, and vessel speed may also be determined based on information typically provided by onboard systems of the vessel. The vessel draft is typically provided to the vessel's onboard systems by one or more float-type and / or hydrostatic pressure sensors. The vessel heading is typically provided to the vessel's onboard systems by one or more navigation compasses. The vessel speed is typically provided to the vessel's onboard systems by an IMU and / or a GPS-based satellite navigation receiver.

[0105] Method 600 further includes a third step 603 consisting of estimating the future sloshing response of the tank based on each of the input data vectors generated in step 602 and on database 150. Step 603 is similar to step 403 and therefore will not be described in detail again.

[0106] After step 603, a vessel course may be determined that allows for a reduction in the future sloshing response of the tank relative to the vessel's sloshing response that would occur if the vessel maintained its current course. The expression "vessel course" simply refers to avoiding the vessel's heading, vessel speed, or geographical zone. For stationary floating structures (ships, barges), i.e., those at a fixed location, a change in heading is reflected in a change in the angle between the north direction and the longitudinal axis of the floating structure, so as to guide or move the floating structure in a conventional manner, thereby reducing the adverse effects of swells and waves on the floating structure. Additionally or alternatively, a change in the tank's filling level may be determined to allow for a reduction in the future sloshing response of the tank.

[0107] An embodiment of the method 600 will be described with reference to Figure 8. In this embodiment, the database 150 is derived from a statistical model capable of estimating the sloshing response of the tank as a function of the tank's fill level and the current state of the vessel's motion, which may optionally also be a function of at least one of the vessel's draft, the vessel's speed, and the vessel's heading.

[0108] The method 601 comprises determining a current filling level of a tank and determining a future state of a vessel's motion. The current filling level of the tank is typically determined based on filling information provided by a tank filling level sensor 121. The future state of a vessel's motion is estimated based on meteorological information and the vessel's course. The vessel's course is typically obtained from information provided by the vessel's onboard systems, such as the vessel's speed and vessel's heading. The meteorological information may be provided by a sea state sensor 123 and / or by terrestrial or satellite radio communication with a network of weather stations. In one example, the future state of a vessel's motion is estimated by first estimating future sea states estimated from the meteorological information and the vessel's course, and then estimating the future state of the vessel's motion from the thus estimated future sea states. As already mentioned above, assessing the corresponding vessel's motion based on given sea states is a routine task in assessing the seakeeping performance of a vessel.

[0109] The method 602 comprises generating a plurality of input data vectors, each of the plurality of input data vectors including a current fill level of a tank and an estimated future state of vessel motion.

[0110] In step 601, optionally, the vessel draft, vessel heading, and vessel speed may also be determined, as described above.

[0111] Method 603 consists of estimating the future sloshing response of the tank based on each of the input data vectors generated in step 602 and database 150. Step 603 is similar to step 403 and therefore will not be described in detail again.

[0112] As described above, after step 603, a course for the vessel may be determined that will enable a reduction in the future sloshing response of the tank relative to the sloshing response of the vessel that would occur if the vessel maintained its current course.

[0113] Although the present invention has been described with reference to some particular embodiments, it is clear that the invention is in no way limited thereto, but includes all technical equivalents of the means described, as well as combinations thereof, if such combinations fall within the scope of the invention.

[0114] Furthermore, it is clear that a feature or combination of features described with reference to a method applies to the corresponding system, and vice versa.

[0115] Use of the verb "comprise" ("comporter") or "comprendre") and its conjugations does not exclude the presence of elements or steps other than those stated in a claim.

[0116] In the claims, any reference signs placed between parentheses shall not be construed as limiting the claim.

Claims

1. 1. A method for obtaining a statistical model capable of estimating a sloshing response in at least one sealed, insulated tank for the transportation of liquefied gas, comprising: The method includes a step (302) including training the statistical model by a supervised machine learning method based on a set of test data; the statistical model is capable of estimating a sloshing response of the tank as a function of the fill level of the tank and current sea state, the function optionally being further a function of at least one of the vessel draft, the vessel speed and the vessel heading; the set of test data is obtained from the results of a plurality of tests, each of the plurality of tests comprising subjecting a test tank (1010) having a given filling level to a movement and measuring the pressure at at least one point on one wall (1010a) of the test tank (1010) and / or the number of impacts on at least one wall of the test tank (1010); The method, wherein the sloshing response includes at least one of a number of fluid collisions against the wall of the tank, a maximum pressure against the wall of the tank, and a likelihood of damage to the tank.

2. 1. A method for obtaining a statistical model capable of estimating a sloshing response in at least one sealed, insulated tank for the transportation of liquefied gas, comprising: The method includes a step (302) including training the statistical model by a supervised machine learning method based on a set of test data; the statistical model is capable of estimating a sloshing response of the tank as a function of the fill level of the tank and a current state of vessel motion, the function optionally being further a function of at least one of the vessel draft, the vessel speed and the vessel heading; the set of test data is obtained from the results of a plurality of tests, each of the plurality of tests comprising subjecting a test tank (1010) having a given filling level to a movement and measuring the pressure at at least one point on one wall (1010a) of the test tank (1010) and / or the number of impacts on at least one wall of the test tank (1010); The method, wherein the sloshing response includes at least one of a number of fluid collisions against the wall of the tank, a maximum pressure against the wall of the tank, and a likelihood of damage to the tank.

3. The method according to claim 1 or 2, wherein the supervised machine learning method is Gaussian process regression.

4. 4. The method according to any one of claims 1 to 3, characterized in that at least one constraint is applied to the statistical model during training of said statistical model by said supervised machine learning method.

5. 5. The method according to any one of claims 1 to 4, further comprising a step (301) prior to said step (302) of training said statistical model, comprising excluding test results characterized by a sloshing response below a threshold from said set of test data.

6. 6. The method of any one of claims 1 to 5, wherein the statistical model considers a plurality of tanks, and wherein the statistical model is capable of estimating the sloshing response of each of the plurality of tanks as a function of the position of the tank on the ship.

7. 1. A system (1030) for obtaining a statistical model capable of estimating a sloshing response in at least one sealed, insulated tank for the transportation of liquefied gas, comprising: The system (1030) includes processing means (1032, 1033) for training the statistical model by a supervised machine learning method based on a set of test data; the statistical model is capable of estimating a sloshing response of the tank as a function of the fill level of the tank and current sea state, the function optionally being further a function of at least one of the vessel draft, the vessel speed and the vessel heading; the set of test data is obtained from the results of a plurality of tests, each of the plurality of tests comprising subjecting a test tank (1010) having a given filling level to a movement and measuring the pressure at at least one point on one wall (1010a) of the test tank (1010) and / or the number of impacts on at least one wall of the test tank (1010); The system, wherein the sloshing response includes at least one of a number of fluid collisions against the wall of the tank, a maximum pressure against the wall of the tank, and a likelihood of damage to the tank.

8. 1. A system (1030) for obtaining a statistical model capable of estimating a sloshing response in at least one sealed, insulated tank for the transportation of a liquefied gas, comprising: The system (1030) includes processing means (1032, 1033) for training the statistical model by a supervised machine learning method based on a set of test data; the statistical model is capable of estimating a sloshing response of the tank as a function of the fill level of the tank and a current state of vessel motion, the function optionally being further a function of at least one of vessel draft, vessel speed and vessel heading; the set of test data is obtained from the results of a plurality of tests, each of the plurality of tests comprising subjecting a test tank (1010) having a given filling level to a movement and measuring the pressure at at least one point on one wall (1010a) of the test tank (1010) and / or the number of impacts on at least one wall of the test tank (1010); The system, wherein the sloshing response includes at least one of a number of fluid collisions against the wall of the tank, a maximum pressure against the wall of the tank, and a likelihood of damage to the tank.

9. 1. A method (300) for obtaining a database (150) that can be used to estimate a sloshing response in at least one sealed, insulated tank for the transportation of liquefied gas, said method (300) comprising: generating (303) a plurality of input data vectors, each of said plurality of input data vectors comprising the fill level of said tank and the current sea state; obtaining (303) an estimated sloshing response in the tank for each of the generated plurality of input data vectors using the statistical model obtained by the method according to claim 1 or any one of claims 3 to 6 dependent on claim 1; storing (303) the estimated sloshing responses in the tank associated with the plurality of input data vectors in a database; A method comprising the steps of:

10. 1. A method (300) for obtaining a database (150) that can be used to estimate a sloshing response in at least one sealed, insulated tank for the transportation of liquefied gas, said method (300) comprising: generating (303) a plurality of input data vectors, each of said plurality of input data vectors comprising the fill level of said tank and a current state of vessel motion; obtaining (304) an estimated sloshing response in the tank for each of the generated plurality of input data vectors using the statistical model obtained by the method according to claim 2 or any one of claims 3 to 6 dependent on claim 2; storing (305) the estimated sloshing responses in the tank associated with the plurality of input data vectors in a database; A method comprising the steps of:

11. 1. A method (400) for estimating sloshing response in at least one sealed insulated tank for the transportation of liquefied gas on a ship, the method comprising: determining the current fill level of the tank (401); Find the current sea conditions (401), generating 402 an input data vector including the determined current fill level of the tank and the determined current sea state; estimating (403) a sloshing response in the tank based on the generated input data vector and the database (150) obtained by the method of claim 9; A method comprising the steps of:

12. 1. A method (500) for estimating sloshing response in a sealed insulated tank for the transportation of liquefied gas on a ship, the method comprising: determining the current fill level of the tank (501); determining (501) a current state of motion of said vessel; generating (502) an input data vector including the determined current fill level of the tank and the determined current state of motion of the vessel; estimating (503) a sloshing response in the tank based on the generated input data vector and the database (150) obtained by the method of claim 10; A method comprising the steps of:

13. 13. A method according to claim 11 or 12, characterised in that a plurality of tanks is considered, the method comprising a preliminary step of defining the respective positions of said plurality of tanks on the vessel.

14. 14. The method (400, 500) of any one of claims 11 to 13, further comprising the step of providing a warning to a user if the estimated sloshing response of the tank exceeds a warning threshold.

15. A management system (100) for a ship (1) equipped with at least one sealed insulated tank (2) for the transportation of liquefied gas, said management system (100) comprising: at least one filling level sensor (121) for measuring the current filling level of said tank (2); a sea state assessment device (123) capable of assessing current sea states; processing means (110) configured to generate an input data vector including the current filling level of the tank and the current sea state assessed by the sea state assessment device (123), and to estimate a sloshing response in the tank (2) by means of the generated input data vector and the database (150) obtained by the method (300) of claim 9; A management system comprising:

16. A management system (100) for a ship (1) equipped with at least one sealed insulated tank (2) for the transportation of liquefied gas, said management system (100) comprising: at least one filling level sensor (121) for measuring the current filling level of said tank (2); a vessel motion assessment device (122) capable of assessing a current state of motion of said vessel; processing means (110) configured to generate an input data vector comprising the current filling level of the tank and the current state of the vessel's motion, and to estimate a sloshing response in the tank by means of the generated input data vector and the database (150) obtained by the method of claim 10; A management system comprising:

17. A method (600) for estimating sloshing response in at least one sealed, insulated tank (2) for the transportation of liquefied gas on a ship, the method comprising: determining the current fill level of the tank (601); estimating future sea conditions based on meteorological information and the course of the vessel (601); generating (602) a plurality of input data vectors, each of the plurality of input data vectors including the current fill level of the tank and the estimated future sea state; estimating (603) a future sloshing response in the tank based on the generated input data vector and the database (150) obtained by the method of claim 9; A method comprising the steps of:

18. A method (600) for estimating sloshing response in at least one sealed, insulated tank (2) for the transportation of liquefied gas on a ship, the method comprising: determining the current fill level of the tank (601); estimating (601) future states of motion of the vessel based on weather information and the course of the vessel; generating (602) a plurality of input data vectors, each of the plurality of input data vectors including the current fill level of the tank and the estimated future state of vessel motion; estimating (603) a future sloshing response in the tank based on the generated input data vector and the database (150) obtained by the method of claim 10; A method comprising the steps of:

19. 19. A method according to claim 17 or 18, comprising the step of: determining a change in the vessel's course and / or the fill level of the tank that can reduce the future sloshing response of the tank.

20. A management system (100) for a ship (1) equipped with at least one sealed insulated tank (2) for the transportation of liquefied gas, said management system (100) comprising: at least one filling level sensor (121) for measuring the current filling level of said tank (2); a sea state estimation device (123) capable of estimating future sea states based on meteorological information and the course of the ship (1); processing means (110) configured to generate a plurality of input data vectors and to estimate a future sloshing response in the tank (2) using the generated input data vectors and the database (150) obtained by the method of claim 9; Including, 10. A management system according to claim 9, wherein each of said plurality of input data vectors includes a current fill level of said tank and a future sea state estimated by said sea state estimation device (123).

21. A management system (100) for a ship (1) equipped with at least one sealed insulated tank (2) for the transportation of liquefied gas, said management system (100) comprising: at least one filling level sensor (121) for measuring the current filling level of said tank (2); a motion state estimation device (123) capable of estimating the future motion state of the vessel (1) based on meteorological information and the course of the vessel (1); a processing means (110) configured to generate a plurality of input data vectors and to estimate a future sloshing response in the tank (2) using the generated input data vectors and the database (150) obtained by the method of claim 10; Including, 10. A management system comprising: a plurality of input data vectors each including a current fill level of the tank and a future motion state of the vessel estimated by the motion state estimation device;

22. 22. The system according to claim 20 or 21, wherein the processing means (110) is further configured to determine a course for the vessel that can reduce the future sloshing response of the tank.

Images