Method for classifying an event associated with a container

Abstract

The invention relates to a method for detecting an event associated with a container on board a vehicle, the method comprising: • choosing (SLC1) a first predictive model of the power consumption of the container, the model being trained using a first machine learning operation; • implementing (MO1) the first predictive power consumption model in order to determine at least one item of predicted power consumption data for the container characterising an expected operating state of the container; • detecting (DTC1) at least one event associated with the container by means of a characterisation function that identifies and automatically characterises at least one difference between the expected operating state of the container and an actual operating state of the container by means of the predicted power consumption value and the item of operating data.

Description

[0001] METHOD FOR CLASSIFYING AN EVENT ASSOCIATED WITH A CONTAINER

[0002] Field of invention

[0003] The invention relates to the field of methods for detecting events taken from classes of events.

[0004] More particularly, the invention relates to the field of methods for detecting an event from a function identifying a difference between a predicted value and an actual value.

[0005] More particularly, the invention relates to the field of methods for detecting an event associated with a controlled atmosphere container, also commonly called a refrigerated container.

[0006] State of the art

[0007] Container trucks and ships are known in the prior art, which transport containers with a controlled atmosphere on their trailer, on their deck and in their holds, which contain perishable goods (meat, vegetables, fruit, medicines to cite a few examples), and which incorporate a refrigeration unit so as to ensure the preservation of these goods during transport.

[0008] Also known in the prior art are refrigerated rooms, warehouses or rooms which contain perishable goods (meat, vegetables, fruit, medicines, to cite a few examples), and which integrate one or more refrigeration systems (also commonly called cold units) so as to ensure the preservation of these goods.

[0009] During transport or storage, maintaining a specific set temperature, humidity level and sometimes CO2 level in containers, cold rooms, cold rooms or cold warehouses is essential to avoid deterioration of the cargo or stored contents. For example, when transporting bananas, which are one of the most transported foodstuffs, a variation of 1°C from the set temperature (13.3°C on average) can lead to premature aging of the entire batch. In the case of transporting medicines or vaccines, a variation of a few degrees in the packaging temperature can lead to total loss of batches.To ensure that these transport or storage condition parameters (set temperature, humidity level and CO2 level) are maintained, refrigeration units must be sized to take into account several factors, including the lowering of air temperature during loading or storage (and sometimes the temperature of the cargo or stock), heat exchanges with the outside by radiation and conduction through the walls of containers or cold rooms, the heat released by the cargo or stock to be cooled (such as the respiration heat of the fruit), the energy required to treat the outside air supplied in the event of air renewal for the purpose of cooling and / or adjusting humidity and CO2 levels to compensate for the production of gases from the fruit (ethylene, CO2), etc.

[0010] This sizing in air conditioning (cooling, heating, adaptation of the humidity level, etc.) results in an energy requirement necessary to operate the refrigeration systems (compressor, evaporator, fan, but also management and communication electronics), or even heating. The energy used to power these mobile refrigeration systems is electricity, itself coming from thermal generators (running on fuel, diesel, gasoline, LNG, etc.), fuel cells (hydrogen, methanol, biomethanol, etc.), on-board turbines (running on kerosene) or even electric storage batteries.Also, whatever the primary energy used to operate the refrigeration systems, the measurement of the energy consumption of the refrigeration system will be made in electrical energy (kW for the total power drawn by the systems, and kWh for the energy consumed by these systems over a period of time).

[0011] Thus, the transport or refrigerated storage of food results in significant energy consumption, partly due to the power required to operate the refrigeration systems.

[0012] However, the energy consumption of a container or a cold room, linked to the operation of refrigeration systems, is likely to vary significantly when specific events occur. To cite just a few examples, these events may include a variation in the interior temperature, a poorly closed door, a leak in a refrigeration circuit, a fault in a component of a refrigeration unit such as a compressor, a fault in the power supply, a sealing defect, are all events that modify the efficiency of refrigeration systems and result in increased energy consumption to produce a given quantity of cold. Thus, these events result in higher energy consumption of the refrigerated system to maintain its set temperature.

[0013] Therefore, there is a technical need, both preventive and curative, to deal with these events which lead to excess energy consumption.

[0014] The invention aims to propose a solution to this technical need, so as to, at a minimum, limit the aforementioned drawbacks.

[0015] Summary of the invention

[0016] According to a first aspect, the invention relates to a method for detecting an event associated with a container on board a vehicle, the method comprising the following steps:

[0017] • First reception, by a first computer, of first input data comprising at least one temperature data item relating to an internal temperature of the container;

[0018] • Second reception, by said first computer or by another computer, of second input data comprising at least one operating data relating to the operation of at least one item of equipment of the container;

[0019] • Selection of a first predictive model of electrical consumption of the container, said model being trained from a first machine learning;

[0020] • Implementation of the first predictive electricity consumption model to determine at least one piece of predicted electricity consumption data for the container, called predicted consumption data, and characterizing an expected operating state of the container;

[0021] • Detection of at least one event associated with the container from a set of event classes, by means of a characterization function automatically identifying and characterizing at least one deviation between the expected operating state of the container and an actual operating state of the container, by means of the predicted consumption data and the operating data.

[0022] According to one embodiment, the operating data comprises actual electrical consumption data of the container, called actual consumption data, and the characterization function identifies and characterizes at least one difference between a value associated with said actual consumption data and a value associated with the predicted consumption data. The values ​​associated with the actual and predicted consumption data are, for example, values ​​of consumed electrical power.

[0023] One advantage is to identify and characterize the event based on a deviation between an expected electricity consumption value and an actual electricity consumption value.

[0024] According to one embodiment, the operating data comprises at least one piece of data relating to an operating duration and / or an operating regime of at least one piece of equipment of the container. This is, for example, the operating regime of a compression system of the container.

[0025] An advantage is to identify and characterize the event from a real parameter allowing to deduce a consumption of the container. Typically, an operating regime or an operating duration of a compression system of a container allows to deduce an energy consumption of this container.

[0026] According to one embodiment, the characterization function comprises an inference engine implemented from a knowledge base and from at least one predefined rule.

[0027] One advantage is to automatically identify and characterize the event from a set of predefined rules and a dataset. The inference engine is, for example, implemented by an expert system.

[0028] According to one embodiment, the step of selecting the first predictive consumption model is implemented automatically by means of a first function whose parameters depend on data taken from the first input data and / or data taken from the second input data. An advantage is to allow an optimized selection of the model according to the available input data. For example, in the case where little data is available, a generic model may be chosen, whereas a more precise model may be chosen when more data relating to the container is available.

[0029] According to one embodiment, the first function is a learning function trained using a second machine learning.

[0030] One advantage is that it allows automatic selection of the predictive model based on available data from a function trained on a large amount of data, which therefore allows selection of the most optimized model based on the specific case.

[0031] According to one embodiment, the temperature data comprises:

[0032] • An interior temperature of the container and / or;

[0033] • A container set temperature and / or;

[0034] • A differential between an interior temperature of the container and a temperature outside the container and / or;

[0035] • A differential between a container set temperature and an internal container temperature.

[0036] According to one embodiment, the first input data comprises at least one external environmental data measured by at least one sensor positioned on the container or on the vehicle.

[0037] One advantage is to refine the choice of model based on parameters external to the container.

[0038] According to one embodiment, at least one external environment data item comprises at least one data item taken from:

[0039] • An outside ambient temperature;

[0040] • An external temperature of the container walls;

[0041] • External humidity;

[0042] • Rainfall data;

[0043] • Data relating to wind speed.

[0044] According to one embodiment, at least one external environment data item is estimated from a time-stamped position data item of the vehicle received from a location device. One advantage is to obtain data making it possible to refine the choice of the predictive model without having to directly measure this data.

[0045] According to one embodiment, the first input data comprises at least one meteorological data item received by means of a communication interface and from at least one remote server.

[0046] One advantage is to refine the choice of the predictive consumption model based on data from a remote database.

[0047] According to one embodiment, the meteorological data comprises at least one piece of data taken from:

[0048] • Data on the temperature outside the container;

[0049] • A temperature data relative to a dew point;

[0050] • Cloudiness data;

[0051] • Azimuth data;

[0052] • Radiative power data;

[0053] • Humidity data;

[0054] • Data relating to wind speed.

[0055] According to one embodiment, the second input data comprise at least a first characteristic data item of the container, called first characteristic data item, and relating to an operating mode of the container characterized by a limited interval of internal set temperatures of the container.

[0056] One advantage is being able to refine the choice of the predictive consumption model.

[0057] According to one embodiment, the second input data comprises at least one second characteristic data item of the container, called second characteristic data item, and relating to a model and / or a brand of the container characterizing technical characteristics specific to the container.

[0058] One advantage is being able to refine the choice of the predictive consumption model.

[0059] According to one embodiment, the second characteristic data comprises data produced by a second function trained using a third machine learning, called the second function, taking as input the actual consumption data. An advantage is that it makes it possible to refine the choice of the predictive model by finding a model or a brand of the container from its energy consumption profile.

[0060] According to one embodiment, the first machine learning receives as input data taken from at least one history of first input data and second input data of at least one container.

[0061] According to one embodiment, the first machine learning training the first predictive model of electricity consumption receives as input data comprising histories of the same input data relating to different container models or different histories of input data relating to the same container model.

[0062] One advantage is to train the predictive model from varied data to optimize its performance in choosing the most suitable model.

[0063] According to one embodiment, the second machine learning training the first function receives as input data taken from at least one history of first input data and second input data of at least one container.

[0064] According to one embodiment, the third machine learning driving the second function receives as input data characteristic of the electrical consumption of a plurality of containers each having specific technical characteristics.

[0065] According to one embodiment, the event comprises at least one event taken from:

[0066] • A predicted or estimated deviation from an internal container temperature;

[0067] • A fault in an element of a container cold group;

[0068] • A container door opening;

[0069] • A fault in the container’s power supply;

[0070] • A leak of a refrigerant fluid from a refrigeration circuit of the container;

[0071] • A container leak.

[0072] According to one embodiment, the event is detected when at least one deviation between the expected operating state of the container and an actual operating state of the container is greater than a predefined threshold value or variable according to predefined parameters.

[0073] One advantage is being able to quantify a significant deviation from the occurrence of an event.

[0074] According to one embodiment, the threshold value is calculated from an arithmetic mean carried out on values ​​taken from the first input data and / or taken from the second input data and from a standard deviation associated with said values.

[0075] According to another aspect, the invention relates to a method for detecting an event associated with a container, comprising the implementation of a predictive model to determine at least one predicted parameter characterizing an operating state of the container. This is for example a predictive model of an operating mode, an operating regime, an operating time, or any other parameter associated with the container or with equipment of the container (such as its refrigeration system). Thus, the invention generally relates to a method comprising the implementation of a predictive model to predict a parameter likely to have a link on the energy consumption of the container.

[0076] According to another aspect, the invention relates to a system for detecting an event associated with a container configured to implement the steps of the method.

[0077] According to one embodiment, the system comprises:

[0078] • At least one calculator for: o Receiving the first input data comprising at least one temperature data item related to an interior temperature of the container; o Receiving the second input data comprising at least one operating data item relating to an operation of at least one piece of equipment of the container; o Selecting a first predictive model of electrical consumption of the container from a set of predictive models of electrical consumption; o Implementing the first predictive model of electrical consumption to determine at least one predicted electrical consumption data item of the container, called predicted consumption data item, and characterizing an expected operating state of the container;o Detect at least one event associated with the container from a set of event classes, by means of a characterization function automatically identifying and characterizing at least one deviation between the expected operating state of the container and an actual operating state of the container, and at least by means of the predicted power consumption value and the operating data.;

[0079] According to one embodiment, the system comprises:

[0080] • The container;

[0081] • The vehicle;

[0082] • A terminal connected to the container.

[0083] According to another aspect, the invention relates to a system for classifying an event associated with a container on board a vehicle, the system comprising:

[0084] • a plurality of containers each generating at least one temperature data item relating to their internal temperature and at least one operating data item relating to the operation of at least one piece of equipment in the container;

[0085] • an administration console for the operation of the plurality of containers;

[0086] • at least one calculator configured to: o implement a first predictive model of electrical consumption to determine at least one predicted electrical consumption data of the container characterizing an expected operating state of the container; o implement at least one function for classifying an event characterized by at least one difference between an expected operating state of the container and an actual operating state of the container, by means of the predicted electrical consumption value and the operating data.

[0087] According to another aspect, the invention relates to a computer program product comprising instructions which, when the program is executed by a computer, cause the latter to implement the steps of the method of the invention.

[0088] According to another aspect, the invention relates to a method for detecting an event associated with a cold room comprising at least one cold unit, the method comprising the following steps:

[0089] • First reception, by a first computer, of first input data comprising at least one temperature data item relating to an interior temperature of the cold room;

[0090] • Second reception, by said first computer or by another computer, of second input data comprising at least one operating data item relating to the operation of at least one piece of equipment in the cold room;

[0091] • Selection of a first predictive model of electrical consumption of the refrigeration unit, said model being trained from a first machine learning;

[0092] • Implementation of the first predictive electricity consumption model to determine at least one predicted electricity consumption data of at least one refrigeration unit, called predicted consumption data, and characterizing an expected operating state of the refrigeration unit;

[0093] • Detection of at least one event associated with the cold room among a set of event classes, by means of a characterization function automatically identifying and characterizing at least one deviation between the expected operating state of the cold unit and an actual operating state of the cold unit, by means of the predicted consumption data and the operating data.

[0094] According to one embodiment, the operating data comprises actual electrical consumption data of at least one refrigeration unit, called actual consumption data, and the characterization function identifies and characterizes at least one difference between a value associated with said actual consumption data and a value associated with the predicted consumption data. The values ​​associated with the actual and predicted consumption data are, for example, values ​​of electrical power consumed. According to one embodiment, the operating data comprises at least one piece of data relating to an operating duration and / or an operating regime of at least one piece of equipment in the cold room. This is, for example, the operating regime of a compression system in the cold room.

[0095] An advantage is to identify and characterize the event from a real parameter allowing to deduce a consumption of the cold room. Typically, an operating regime or an operating duration of a compression system allows to deduce an energy consumption of the cold room.

[0096] According to another aspect, the invention relates to a system for detecting an event associated with a cold room configured to implement the steps of the method.

[0097] According to one embodiment, the system comprises:

[0098] • At least one calculator for: o Receiving the first input data comprising at least one temperature data item related to an interior temperature of the cold room; o Receiving the second input data comprising at least one operating data item relating to an operation of at least one piece of equipment in the cold room; o Selecting a first predictive model of electrical consumption of the cold unit from a set of predictive models of electrical consumption; o Implementing the first predictive model of electrical consumption to determine at least one predicted electrical consumption data item, called predicted consumption data item, and characterizing an expected operating state of the cold unit;o Detect at least one event associated with the cold room from a set of event classes, by means of a characterization function automatically identifying and characterizing at least one deviation between the expected operating state of the cold unit and an actual operating state of the cold unit, and at least by means of the predicted electrical consumption value and the operating data.;

[0099] According to one embodiment, the system comprises:

[0100] • The cold room including at least one cold unit;

[0101] • A terminal connected to the cold room;

[0102] According to another aspect, the invention relates to a system for classifying an event associated with a cold room, the system comprising:

[0103] • a plurality of cold groups each generating at least one temperature datum relating to their set temperature and at least one operating datum relating to the operation of at least one piece of equipment in the cold room;

[0104] • an administration console for the operation of the plurality of cold groups;

[0105] • at least one calculator configured to: o implement a first predictive model of electrical consumption to determine at least one predicted electrical consumption data of the cold unit characterizing an expected operating state of the cold room; o implement at least one classification function of an event characterized by at least one difference between an expected operating state of the cold room and an actual operating state of the cold room, by means of the predicted electrical consumption value and the operating data.

[0106] Brief description of the figures

[0107] Other characteristics and advantages of the invention will emerge on reading the detailed description which follows, with reference to the appended figures which illustrate:

[0108] Fig. 1: A schematic view of a vehicle transporting a plurality of containers.

[0109] Fig.2: A flowchart of several steps in the process of detecting an event associated with a container on board a vehicle or a cold room. Fig.3: A flowchart of several steps in the process of detecting an event

[0110] Fig.4: A graphical representation of predicted container power consumption values ​​and actual container power consumption values ​​in curve form and as a function of time.

[0111] Fig.5: a flowchart of a step of the method for detecting an event, in an embodiment in which the method comprises a step of implementing the first machine learning function from two data characteristics of the container or the cold room.

[0112] Fig.6: a flowchart of a step of a method for detecting an event, in an embodiment in which the method comprises a step of implementing the second function.

[0113] Fig.7: A graphical representation of the training of the consumption predictive model, the first function and the second function using the first, second and third machine learning.

[0114] Fig.8: a schematic representation of a cold room including a cold unit.

[0115] Description of the invention

[0116] General information

[0117] According to a first aspect, the invention relates to a method for detecting an Evm event associated with a CTN container on board a Vhc vehicle.

[0118] According to another aspect, the invention relates to a method for detecting an Evm event associated with a cold room SFD comprising at least one cold unit GF. In the present application, when technical characteristics are described for the cold room SFD, it will be understood that these technical characteristics may for example relate to a cold unit GF of the cold room SFD. For example, data relating to technical or refrigeration characteristics of the cold room SFD may comprise data relating to technical or refrigeration characteristics of at least one cold unit GF of the cold room SFD.

[0119] The term "container" is used in this description for reasons of simplification. A "CTN container" means any type of container with a controlled atmosphere. For example, this may be a refrigerated container, also known as a refrigerated container.

[0120] In the reference, such a container will be referred to interchangeably by the terms “container”, “refrigerated container” or “refrigerated container” to refer to any type of container with a controlled atmosphere.

[0121] A Vhc vehicle is any means of transport capable of carrying one or more containers. The invention is not limited to ships.

[0122] The Vhc vehicle is, for example, an airplane or any other means of air transport, a train or any other means of rail transport, a ship or any other means of maritime transport, a car, a truck, or any other means of road transport.

[0123] An example of a Vhc vehicle carrying a plurality of CTN containers is illustrated in Figure 1. This is an example in which the Vhc vehicle is a ship.

[0124] The term "cold room" is used in this description for reasons of simplification. "Cold room" means any type of storage location, cold room, warehouse, or room with a controlled atmosphere equipped with one or more cold units or refrigeration units.

[0125] For example, it could be a cold room, also called a cold room, or a refrigerated cabinet.

[0126] In the reference, such a storage location will be referred to interchangeably by the terms “cold room”, “refrigerated room” or “refrigerated room” to refer to any type of controlled atmosphere storage location equipped with one or more refrigeration units.

[0127] Figure 8 illustrates an example of an SFD cold room including a GF cold unit.

[0128] The method comprises several steps for detecting DTCi an Evm event associated with the CTN container or the SFD cold room.

[0129] An "Evm event" is a cause of a deviation between an expected operating state of the CTN container or the SFD cold room and an actual operating state of the CTN container or the SFD cold room. Such a deviation corresponds, for example, to a deviation between a predicted energy consumption value and an actual energy consumption value over time, as illustrated in Figure 4. According to another example, it is a deviation between an expected operating time and / or operating regime of the CTN container or the SFD cold room, and an actual operating time and / or operating regime of the CTN container or the SFD cold room, or of equipment thereof, for example one of their compression systems.

[0130] The expected operation of the CTN container or the SFD cold room is characterized by a predicted electrical consumption data of the CTN container or the SFD cold room, such as an electrical consumption data of the GF refrigeration unit. The actual operation of the SFD container or cold room is characterized by an operating data Dtfct of the CTN container or the SFD cold room, which includes for example an actual electrical consumption data of the CTN container or the SFD cold room, or even a data relating to an operating time and / or an operating regime of an item of equipment of the CTN container or the SFD cold room.

[0131] Examples of Evm events associated with the CTN container include a poorly closed door of the container, a loss of tightness of the container walls, a fault in the electrical supply of the container, a fault in the refrigeration system of the container such as a refrigerant leak or a fault in an element of a cooling unit of the CTN container, for example a fault in a compressor or a variation in an interior temperature of the CTN container, predicted or estimated from a set of data such as historical data. All these events cause a drift in energy consumption of the CTN container.

[0132] Examples of Evm events associated with the cold room include a poorly closed door, a loss of wall tightness, a fault in the electrical supply of the GF cold unit, a fault in the refrigeration system such as a refrigerant leak, fouling of an element or a fault in an element of a GF cold unit, for example a fault in a compressor or a variation in an interior temperature, predicted or estimated from a set of data such as historical data. All these events cause a drift in energy consumption of the SFD cold room. In the present description, the invention is described through different embodiments illustrated by examples.

[0133] Features described in one embodiment may be directly applicable to another embodiment.

[0134] The features described in an example may be independent of one another. Thus, if an embodiment is illustrated through an example comprising a plurality of steps, the implementation of this embodiment is not limited to the joint implementation of all of the steps. For example, if an embodiment describes the training of a learning function through an example comprising an input data preprocessing step and an input data labeling step, the invention is not limited to the implementation of these two steps for this embodiment, and the training could not include an input data preprocessing step. In addition, formulations such as “according to an example” followed by “according to another example” for the same embodiment do not limit the invention to one or the other of these examples when the latter are not incompatible.

[0135] First Dtini input data: general information

[0136] With reference to figures 2 and 3, the method comprises a step of first reception RECi of first input data Dtini.

[0137] The first input data Dtini are received by a first calculator CLCi . They are for example received by the first calculator CLCi from a first memory MEMi , which can be a local memory or a remote memory.

[0138] In the case of a remote memory, the first CLCi calculator comprises, for example, means for acquiring said first input data Dtim.

[0139] In another case, the first input data Dtim are acquired in a preliminary step, and recorded in a local memory to then be transmitted to the first CLCi computer which receives them.

[0140] The first Dtini input data may be transmitted in real time to a remote server by means of a telecommunications device, such as a radio or satellite device. According to another example, the first Dtini data are retrieved a posteriori from a local memory, for example for a posteriori processing of these data.

[0141] The first input data Dtini includes for example at least one data which is either:

[0142] - measured locally using one or more sensors;

[0143] - calculated locally using measured data and / or received data, for example using the first CLCi calculator;

[0144] - calculated remotely and received locally, for example by means of a communication interface.

[0145] According to one embodiment, the first input data Dtini comprises a mixture of measured and / or calculated data, locally and / or remotely.

[0146] According to one embodiment, the first input data Dtini comprises a mixture of measured and / or calculated data, locally and / or remotely, and recorded locally in a local memory and / or remotely on a remote server.

[0147] According to one embodiment, the first input data Dtini comprises real-time measured data.

[0148] According to one embodiment, the first input data Dtini comprises calculated and / or estimated data corresponding to a physical parameter associated with a current position of the vehicle.

[0149] According to one embodiment, the first input data Dtini comprises predicted data. This is, for example, data corresponding to predicted values ​​associated with physical parameters. For example, it may be predicted temperature data based on a predicted trajectory of the vehicle or predicted temperature data based on meteorological data of the site. These predicted values ​​are, for example, estimated from other estimated values, or from measured physical parameters, or from a combination of predicted values ​​and measured values.

[0150] First Dtini input data: temperature data

[0151] Dtempl The first input data Dtini includes at least one temperature data related to an internal temperature of the container Dtempi , called temperature data Dtempi .

[0152] The temperature data Dtempi includes, for example, at least one temperature value, measured, calculated or estimated. This is, for example, a value expressed in degrees Celsius (°C) or in Kelvin (K), or any other suitable unit of measurement.

[0153] The temperature data Dtempi includes, for example, at least one temperature value measured by a sensor. This is, for example, a temperature value measured inside the CTN container or the SFD cold room by means of a temperature probe.

[0154] According to one embodiment, the temperature data Dtempi comprises at least one temperature value calculated, estimated or deduced from other parameters.

[0155] These parameters include, for example, other values ​​of measured, calculated or estimated physical parameters, such as other temperature values ​​(e.g. an outside ambient temperature value).

[0156] According to one example, the temperature data Dtempi comprises a calculated value obtained from a differential between an interior temperature value of the CTN container or the SFD cold room and at least one other temperature value, such as a temperature value outside the CTN container or the SFD cold room. Such a value is for example calculated by means of the first calculator CLCi. It can also be calculated by means of another calculator, locally or remotely. If the value is calculated from a remote entity, it is for example transmitted by this remote entity and received by the first calculator CLCi by means of a communication interface.

[0157] According to various examples, the temperature data Dtempi comprises at least one data item taken from the following data and / or calculated and / or estimated from the following data:

[0158] - an internal set temperature of the CTN container or the SFD cold room;

[0159] - an actual interior temperature of the CTN container or SFD cold room; - a temperature outside the CTN container or SFD cold room, such as an ambient temperature.

[0160] First Dtini input data: Dtext external environment data

[0161] According to one embodiment, the first input data Dtini comprises at least one external environment data Dtext.

[0162] The external environment data Dtext characterizes a physical parameter external to the container or the SFD cold room.

[0163] For example, it includes data measured using one or more sensors. These include, for example, sensors located on the CTN container or on the building housing the SFD cold room, or near them.

[0164] According to another example, the external environmental data Dtext comprises data calculated or estimated from one or more physical parameters, external or not to the CTN container or the SFD cold room, themselves measured, calculated or estimated.

[0165] According to one embodiment, the external environment data Dtext comprises at least one piece of data received from a remote device. This is, for example, a device connected to a server or a database comprising data relating to physical parameters surrounding the vehicle Vhc or to a site on which the cold room SFD is located, such as a meteorological database.

[0166] According to one embodiment, the external environment data Dtext comprises at least one data item calculated from at least one data item from a remote server or a remote database, such as a weather database.

[0167] According to various examples, the external environment data Dtext comprises at least one data taken from:

[0168] - temperature data, for example a temperature value outside the CTN container or the SFD cold room, such as an ambient air temperature value, a temperature value of an external face of the container walls, a temperature of an external wall of the cold room, or a dew point temperature; pressure data, for example an atmospheric pressure value; humidity data, for example a relative humidity level

[0169] - wind data, for example a wind speed value;

[0170] - cloudiness data, for example a value relating to a proportion of clouds covering the sky.

[0171] First input data Dtini: position data Dt pos

[0172] According to one embodiment, the first input data Dtini comprises position data Dtpos of the vehicle Vhc or of a site on which the cold room SFD is located.

[0173] The position data Dtpos includes, for example, a position of the vehicle Vhc, a plurality of positions of the vehicle Vhc corresponding, for example, to a trajectory of the vehicle on a given route, or even an address of the site or GPS coordinates of the site.

[0174] The position data Dtpos is for example received by the first CLCi computer. It can also be received by another computer.

[0175] The position data Dtpos is for example received by a computer following its transmission by a third-party system such as a satellite geolocation system, also designated by the acronym GNSS. It is for example recorded in a memory, for example the first memory MEMi, or in a different memory.

[0176] According to one embodiment, the position data Dtpos comprises at least one temporal data item associated with it, such as a timestamp of one or more positions of the vehicle Vhc.

[0177] According to one embodiment, at least one first data item Dtini is calculated and / or estimated from the position data item Dtpos.

[0178] For example, ambient temperature data can be estimated based on the position of the vehicle Vhc or the site at a given time, based on a forecast weather database.

[0179] According to one embodiment, at least one of the first data Dtini received relates to a position data Dtpos of the vehicle Vhc or of a site on which the cold room SFD is located. The term "relating to the position data Dt" means pos » data comprising a measured, calculated or predicted value for the position.

[0180] For example, if the first input data Dtini comprises a plurality of ambient temperature values ​​and the position data Dtpos comprises a set of geographic coordinates over a given period, then the first input data Dtini comprises, for example, ambient temperature values ​​measured at said coordinates and over the given period. It may also be an average temperature over said period.

[0181] According to one embodiment, at least a first input data Dtini is obtained by means of observations from meteorological and / or satellite stations.

[0182] First input data Dtini: meteorological data Dtmt

[0183] According to one embodiment, the first input data Dtini comprises at least one meteorological data Dtmt.

[0184] The meteorological data comes from a database, for example. This could be a database of a meteorological service capable of providing meteorological data from position data associated with temporal data (for example, a time-stamped geographical position).

[0185] According to several examples, Dtmt meteorological data may include:

[0186] - temperature data, for example ambient air temperature or dew point;

[0187] - solar radiation data;

[0188] - azimuth data;

[0189] - cloud cover data, for example cloudiness;

[0190] - humidity data, for example a relative humidity value;

[0191] - wind speed data,

[0192] - data relating to wind direction,

[0193] - rainfall data. In one embodiment, the method comprises implementing a function for determining at least one external environmental data Dtext and / or at least one meteorological data Dtmt, or for determining a variation of at least one of these data from at least one other external environmental data Dtext and / or from at least one meteorological data Dtmt, and / or from a variation of one of these data.

[0194] According to one embodiment, at least one first input data Dtini comprises a value calculated as a function of a predicted trajectory of the vehicle Vhc, of a predicted position data of the vehicle Vhc, or of a predicted variation of at least one external environment data Dtext or meteorological data Dtmt.

[0195] More generally, the first Dtini input data is likely to include any type of data that may have an impact on the internal temperature of the CTN container or the SFD cold room.

[0196] According to one embodiment, the first input data Dtini comprises data estimated, calculated or predicted by means of a mathematical model taking as input at least one position data Dtpos of the vehicle Vhc received, measured, estimated, calculated and / or predicted.

[0197] According to one embodiment, the first input data Dtini comprises position data Dtpos of the vehicle Vhc or of the site on which the cold room SFD is located, estimated, calculated or predicted using a mathematical model taking as input meteorological data Dtmt received, measured, calculated, estimated and / or predicted.

[0198] Second Dtm2 input data: general information

[0199] The method comprises a second step of receiving REC2 second input data Dtin2.

[0200] The second input data Dtin2 is received by a computer, which can be the first computer CLC1 or another computer.

[0201] The second input data Dtin2 are for example received by the computer from the first memory MEM1 or from another memory, which may be a local memory or a remote memory. When the data are received from a remote memory, the computer comprises for example means for acquiring said second input data Dtin2.

[0202] In another case, the second input data Dtin2 are acquired in a prior step and recorded in a local memory, to then be transmitted to the computer which receives them.

[0203] Second input data Dtm2: operating data Dtfct

[0204] The second input data Dtin2 comprises at least one operating data Dtfct relating to the operation of at least one piece of equipment in the CTN container or the SFD cold room. This may be data relating to the operation of equipment in the CTN container, or relating to the operation of the CTN container itself. It may also be at least one piece of actual energy consumption data of the SFD cold room or equipment in the GF cold group, for example a compression system in the SFD cold room.

[0205] The second input data Dtin2 comprises, for example, at least one piece of actual energy consumption data of the CTN container or of a piece of equipment of the CTN container, for example a compression system of the CTN container. This is, for example, actual electricity consumption data Cre of the CTN container, or at least one piece of actual energy consumption data of the cold room SFD or of a piece of equipment of the cold group GF, for example a compression system of the cold room SFD. The operating data Dtfct comprises, for example, at least one piece of data relating to energy consumption over time, such as power data, for example total power drawn expressed in kW, or current or voltage data expressed respectively in Amperes and Volts, for example drawn current data, or energy consumption data expressed in kWh or Joules.

[0206] According to another example, the operating data Dtfct comprises at least one piece of data measured by a piece of equipment of the CTN container or the SFD cold room, or even a piece of measured or calculated data relating to a piece of equipment of the CTN container or the SFD cold room. This is, for example, data making it possible to determine an operating time and / or an operating regime of one or more pieces of equipment of the CTN container or the SFD cold room, for example data relating to one or more compression / expansion cycles of a refrigeration system, data relating to air ventilation inside the SFD cold room, data relating to air ventilation or heating inside the CTN container, or even data characteristic of events occurring during the operation of the container or a piece of equipment thereof, such as data recorded in a memory in the form of a “log”.Such data characterizes, for example, a history of events occurring over time during the operation of at least one piece of equipment in the CTN container or the SFD cold room. This data makes it possible, for example, to deduce energy consumption over time of the CTN container or the SFD cold room.

[0207] According to one embodiment, the second input data Dtin2 comprises air temperature data from a refrigeration system of the CTN container or the SFD cold room. This is for example a setpoint temperature, a supply air temperature, also referred to as Supply Air Temperature in English literature, or an extract air temperature, also referred to as Return Air Temperature in English literature.

[0208] In one embodiment, the second input data Dtin2 comprise at least one piece of data characteristic of an operating mode of the container or the cold room Dd, called first characteristic data Dd. The first characteristic data Dd comprises, for example, data characteristic of an operating temperature range of the CTN container or the SFD cold room, for example, characteristic data relating to a limited range of setpoint temperatures that can be applied to the CTN container or the SFD cold room.

[0209] In one embodiment, the second input data Dtin2 comprises at least one characteristic data item of a model and / or a brand of CTN container or GF refrigeration unit, called second characteristic data item D C2. This is, for example, data relating to a brand of CTN container or GF refrigeration unit to which specific technical characteristics are associated (energy consumption, operating temperature range, etc.).

[0210] According to one embodiment, the second input data Dtin2 comprises data from a CTN container management system or the SFD cold room.

[0211] For example, this is the first characteristic data Dd, the second characteristic data D C2, data relating to technical settings of the CTN container or the SFD cold room, or data relating to the refrigeration characteristics of the CTN container or the SFD cold room, history of events occurring during the operation of the CTN container or of equipment in the container, such as its refrigeration system, or occurring during the operation of one or more items of equipment in the SFD cold room, such as during the operation of a GF refrigeration unit.

[0212] Link between first and second input data

[0213] According to one embodiment, the second input data Dtin2 comprises the first input data Dtini.

[0214] According to one embodiment, the first input data Dtini comprises the second input data Dtin2.

[0215] Selecting the model from a set of models

[0216] The method comprises a step of selecting SLCi a first predictive model of electrical consumption MOD P of the CTN container or the SFD cold room, called the first predictive model MOD P .

[0217] The first predictive model MOD P is selected from a set of predictive electricity consumption models.

[0218] The choice of the first predictive model MOD P among the set of models depends for example on the first input data Dtini and / or the second input data Dtin2. Thus, the first predictive model MOD P is for example selected based on data taken from the first input data Dtini and / or taken from the second input data Dtin2.

[0219] For example, when the second input data Dtin2 does not include the second characteristic data D C2, we can choose a generic predictive model suitable for any brand of CTN container or GF refrigeration unit. According to another example, when the second input data Dtin2 includes the second characteristic data D C 2, we can choose the first predictive model from a smaller subset of predictive models.

[0220] In one embodiment, the first predictive model MOD P is manually chosen by a user from the set of models.

[0221] Training the MOD consumption predictive model P : machine learning

[0222] Referring to Figure 7, the first predictive model MOD P is trained TRi from a first AMI machine learning model.

[0223] The first AMI machine learning model comprises, for example, at least one model taken from various models such as a neural network, for example a recurrent neural network RNN, a neural network of the CNN type, or a convolutional neural network, or a random forest, also called a decision tree forest, or a linear regression, or an ARIMA (AutoRegressive Integrated Moving Average) model.

[0224] In one embodiment, the first AMI machine learning model receives as input at least one data item taken from the first input data Dtini and / or taken from the second input data Dtin2.

[0225] For example, this is data taken from the first temperature data Dtem P i, the position data Dt pos, weather data Dtmt, outdoor environment data Dtext and / or operating data Dtfct. These are, for example, historical data.

[0226] Training data is preferably vectorized and normalized. Each input vector thus defined includes, for example, a set of training data.

[0227] According to one embodiment, the implementation of the first AMI machine learning model comprises a labeling of each input vector. The labels define, for example, characteristic states of the CTN container or the SFD cold room, such as a characteristic state of a GF cold unit, characteristic states of the Vhc vehicle, data relating to an operation carried out on the Vhc vehicle, the CTN container, the SFD cold room or the GF cold unit, or any other context data, for example a state of loading a payload inside the CTN container or the SFD cold room, a state of movement of the CTN container within the Vhc vehicle, or a state of introduction or exit of the CTN container from the Vhc vehicle.

[0228] One advantage of using a larger number of different data as input to the first AMI machine learning model is to facilitate labeling. Typically, a case of loading a container with a payload and a case of opening a container during a customs clearance are different events that can lead to similar consumption variations. In this case, the Dtpos position data can help discriminate between the two events.

[0229] According to one embodiment, output data of the first AMI machine learning model comprises classified data. The output data of the first AMI machine learning model comprises, for example, labeled data.

[0230] According to one embodiment, the implementation of the first AMI machine learning model comprises a step aimed at correcting data classified by the first AMI machine learning. This step comprises, for example, a modification of a label, an addition of a label, an enrichment of a label, etc.

[0231] According to one embodiment, the method comprises a step of preprocessing the input data of the first AMI machine learning model. The input data is for example collected, cleaned and preprocessed.

[0232] According to one embodiment, the method comprises a step of defining the learning model for the first AMI machine learning model.

[0233] For example, this is a supervised model such as a classification model or a regression model.

[0234] Another example is an unsupervised model that takes unlabeled data as input, such as a clustering model, often called a "clustering" model.

[0235] According to one embodiment, data taken from the input data of the first AMI machine learning model are divided into several subsets. These are for example a training data set and a validation data set. The model is then for example trained on the training data set.

[0236] According to one embodiment, parameters of the first AMI machine learning model are adjusted to reduce an error function (or cost function). Such a step is for example implemented by means of an optimization algorithm such as gradient descent, mean square error or binary cross entropy.

[0237] According to one embodiment, the first AMI machine learning model is evaluated using data taken from the validation data.

[0238] An advantage is to estimate the capacity of the predictive model of electricity consumption MOD P to generalize its predictions to new input data.

[0239] According to one embodiment, the method comprises several training and evaluation cycles to optimize the predictive power consumption model MOD P .

[0240] According to one embodiment, training the MOD power consumption model Pincludes a parameter adjustment step. These are, for example, adjustable parameters defined before the start of training. Parameters include, for example, hyperparameters. Examples of hyperparameters include a learning rate, a number of epochs (or iterations), a size of the training data set, or an activation function. This advantageously allows for optimizing the model's performance.

[0241] In another example, a neural network of the RNN type, meaning "recurrent neural network" in Anglo-Saxon terminology, is implemented.

[0242] The output values ​​of the predictive electricity consumption model MOD P include, for example, continuous values, for example, a continuous value of electricity consumption in kilowatt-hours, or a discrete value of electricity consumption, for example, a rounded value.

[0243] According to another example, the output values ​​of the predictive model of electricity consumption MOD P include output vectors including a classification of the electrical consumption of the CTN container or the SFD cold room.

[0244] The output values ​​of the predictive electricity consumption model MOD P include, for example, a time component. For example, the output values ​​each include a CTN container or SFD cold room electricity consumption value per hour, or a CTN container or SFD cold room electricity consumption value per day.

[0245] In another example, a random forest is implemented and comprises several decision trees, which are created from subsets of training data each trained on a different sample of data, which are for example randomly selected from a training data set. Each decision tree calculates for example a predicted value as a function of the input variables, which are for example continuous variables such as values ​​of pressure, humidity, temperature, cloudiness, wind speed etc. to output a predicted value of electricity consumption.

[0246] First Fi function: general information

[0247] In one embodiment, with reference to Figure 5, the first predictive model MOD P is chosen by means of a first function Fi.

[0248] The first function Fi comprises, for example, a function dependent on the first input data Dtini and / or the second input data Dtin2. For example, it is an expert system configured to implement an inference engine based on a knowledge base and a rule base. Examples of an inference engine include a forward, backward, or mixed chaining engine.

[0249] In one embodiment, the first function Fi comprises a learning function, or machine learning function.

[0250] The first function Fi receives, for example, as input at least one piece of data taken from the first input data Dtini and / or taken from the second input data Dtin2. This is, for example, a piece of temperature data, for example the temperature data Dtem P i. According to another example, this is the second characteristic data D C2. According to another example, the first function Fi receives as input a plurality of data taken from the first input data Dtini and / or taken from the second input data Dtin2.

[0251] One advantage is to choose as precisely as possible the most suitable predictive model based on the available data.

[0252] The first function Fi, for example, implements at least one model taken from: a neural network, for example a recurrent neural network RNN, a random forest, also called a decision tree forest, or a linear regression, or an ARIMA (AutoRegressive Integrated Moving Average) model.

[0253] First Fi function: training

[0254] According to one embodiment, with reference to FIG. 7, the first function Fi is trained TR2 by means of a second machine learning AM2.

[0255] According to one embodiment, the first function F1 is trained by means of the second machine learning AM2 from data taken from the first input data Dtini and / or taken from the second input data Dtin2 and relating to the same model or the same brand of CTN container or GF refrigeration unit.

[0256] This is, for example, a history of first Dtini input data and / or second Dtin2 input data relating to the same model or brand of CTN container or GF refrigeration unit.

[0257] According to one embodiment, the first function F1 is trained from data taken from the first input data Dtini and / or taken from the second input data Dtin2 and relating to different models or different brands of CTN container or GF refrigeration units.

[0258] This is, for example, a history of first Dtini input data and / or second Dtin2 input data relating to different models or brands of CTN container or GF refrigeration units.

[0259] Training data is preferably vectorized and normalized. Each input vector thus defined includes, for example, a set of training data.

[0260] According to one embodiment, the implementation of the second machine learning model AM2 comprises a labeling of each input vector. The labels define, for example, characteristic states of the CTN container or the SFD cold room, for example of a GF cold group, characteristic states of the Vhc vehicle, data relating to an operation carried out on the vehicle or the container or any other contextual data, for example a state of loading a payload inside the CTN container or the SFD cold room, a state of movement of the CTN container within the Vhc vehicle, or a state of introduction or exit of the CTN container from the Vhc vehicle.

[0261] According to one embodiment, output data of the first AMI machine learning model comprises classified data. The output data of the first AMI machine learning model comprises, for example, labeled data.

[0262] According to one embodiment, the implementation of the first AMI machine learning model comprises a step aimed at correcting data classified by the first AMI machine learning. This step comprises, for example, a modification of a label, an addition of a label, an enrichment of a label, etc.

[0263] According to one embodiment, the method comprises a step of preprocessing the input data of the first AMI machine learning model. The input data is for example collected, cleaned and preprocessed.

[0264] According to one embodiment, the method comprises a step of defining the learning model for the first AMI machine learning model.

[0265] For example, this is a supervised model such as a classification model or a regression model.

[0266] Another example is an unsupervised model that takes unlabeled data as input, such as a clustering model, often called a "clustering" model.

[0267] According to one embodiment, data taken from the input data of the second machine learning model AM2 are divided into several subsets. This is for example a training data set and a validation data set. The model is then for example trained on the training data set. According to one embodiment, parameters of the second machine learning model AM2 are adjusted to reduce an error function (or cost function). Such a step is for example implemented by means of an optimization algorithm such as gradient descent, mean square error or binary cross entropy.

[0268] According to one embodiment, the second machine learning model AM2 is evaluated using data taken from the validation data.

[0269] One advantage is to estimate the ability of the first function Fi to generalize predictions to new input data.

[0270] According to one embodiment, the method comprises several training and evaluation cycles to optimize the first function Fi.

[0271] According to one embodiment, the training of the first function Fi comprises a parameter adjustment step. These are, for example, adjustable parameters defined before the start of training. The parameters comprise, for example, hyperparameters. Examples of hyperparameters include a learning rate, a number of epochs (or iterations), a size of the training data batch, or an activation function. This advantageously makes it possible to optimize the performance of the model.

[0272] According to one example, the output values ​​of the first function Fi comprise output vectors comprising a classification of the electrical consumption of the CTN container or the SFD cold room, for example of a GF cold group.

[0273] The output values ​​of the first function Fi include, for example, a time component.

[0274] In another example, a random forest is implemented and comprises several decision trees, which are created from subsets of training data each trained on a different sample of data, which are for example randomly selected from a training data set. Each decision tree calculates for example a predicted value based on the input variables.

[0275] Event detection / characterization function

[0276] The method comprises a step of detecting DTCi at least one event associated with the CTN container or the SFD cold room. Examples of events that may be associated with the CTN container or the SFD cold room include:

[0277] - A predicted deviation, for example by means of a history of historical data evolution, for example internal temperature data of the CTN container or the SFD cold room. In the latter case, portions of temperature evolutions of a history can be characteristic of singular events, such as a poorly closed door, a temporary opening of the door for example during a customs check for the case of CTN containers, evolutions linked to micro-cuts linked to excessive energy consumption of a set of equipment, a power cut, etc. A database comprising a characterization of the events by families of curves can be used. The historical data can include curve profiles or characteristic points of a curve, such as thresholds, or even a particular dynamic of the curve;

[0278] - A predicted deviation, for example by means of a mathematical model and historical data, of an interior temperature of the CTN container or the SFD cold room; the mathematical model may result from a function learned from a machine learning model. The mathematical model may correspond to a function characterizing the evolution of a curve of a physical parameter such as the evolution of the temperature in a given context. For this purpose, the coefficients of the function may be learned from historical data;

[0279] - an open door of the CTN container or the SFD cold room;

[0280] - loss of sealing of the CTN container or the SFD cold room;

[0281] - a fault in a refrigeration circuit of the CTN container or the SFD cold room, for example a compressor or evaporator fault, or a refrigerant leak; - a fault in the power supply of the CTN container or the SFD cold room.

[0282] The event is detected among a set of event classes by means of a characterization function Fc.

[0283] The characterization function Fc automatically identifies and characterizes at least one deviation between an expected operating state of the CTN container or the SFD cold room and an actual operating state of the CTN container or the SFD cold room. An expected operating state corresponds, for example, to an expected energy consumption for the CTN container or the SFD cold room over a given period.

[0284] The characterization function Fc identifies and characterizes a deviation between an expected operating state of the CTN container or the SFD cold room, and an actual operating state of the CTN container or the SFD cold room, for example by comparing the operating data Dtfct with the predicted consumption data C pr . This is for example a comparison between an actual consumption data Cre and the predicted consumption data Cpr. In another case, it is a comparison between a data relating to an operating mode of the CTN container or the SFD cold room or equipment of the CTN container or the SFD cold room, such as a data relating to an operating regime of a compression system of the container, and the predicted consumption data C prSuch a comparison is possible, for example, when the operating data Dtfct makes it possible to deduce an electrical consumption from another parameter, for example an operating time or an operating regime of equipment in the CTN container or the SFD cold room, such as a GF cold group.

[0285] In one embodiment, an Evm event is identified from a set of event classes if a deviation between a predicted value and an actual value exceeds a predefined threshold. The threshold may also be a variable value depending on one or more parameters.

[0286] For example, the parameters may include the position of the vehicle or any other physical parameter characterizing the state of the container, and / or the state of the vehicle such as its energy state, and / or a variable related to its operation. The characterization of the deviation may be related to a one-time variation of one or more parameters compared to one or more predicted values.

[0287] In one case, a recurrence of a variation of a parameter identified by means of the characterization function Fc makes it possible to characterize the event Evm. For example, a deviation below a threshold value could be considered acceptable taken individually, but the repetition of this variation over a given period makes it possible to identify and characterize an event.

[0288] According to one embodiment, the characterization function Fc comprises an expert system configured to implement an inference engine. This is for example a software component allowing a system to reason on a data set and to deduce information from this data set using a set of rules.

[0289] The inference engine is for example implemented from a knowledge base and from at least one predefined rule. The knowledge base includes for example predicted values ​​of electricity consumption and associated actual values ​​of electricity consumption. An example of an implemented rule can be that any deviation greater than a value X between the predicted value of electricity consumption C prand the actual electricity consumption value Cre is associated with a problematic event. Another example of a rule is that any average deviation greater than a value Y between the predicted electricity consumption value Cpr and the actual electricity consumption value Cre over a time range greater than or equal to Z minutes characterizes an opening of the door of the CTN container or the SFD cold room.

[0290] According to one embodiment, the detection and characterization of an event Evm using the characterization function Fc comprises the association of said event Evm with an event class.

[0291] In one embodiment, an event is automatically relabeled if a recurrence greater than a threshold value of this event is observed over a given time interval. This is for example a variation in energy consumption which, taken punctually, does not characterize an event in itself, but whose recurrence makes it possible to determine that an event has occurred. Characteristic data of a model Dc2 and second trained machine learning function F2

[0292] In one embodiment, with reference to FIG. 6, the second characteristic data D C 2 includes data produced by a second machine learning function F2, called second function F2.

[0293] The second function F2 takes as input, for example, the electrical consumption of the CTN container or the SFD cold room.

[0294] The second function F2 allows, for example, to determine a classification, for example to associate an electricity consumption value, a set of electricity consumption values ​​or even an electricity consumption profile with a given CTN container model or GF refrigeration unit taken from a set of ENSc container or refrigeration unit models.

[0295] According to one embodiment, with reference to FIG. 7, the second function F2 is trained TR3 by means of a third machine learning AM3.

[0296] According to an example of training the second function F2, the method comprises a step of collecting data, for example data characteristic of at least one CTN container or at least one GF refrigeration unit relating to specific container models, such as electrical consumption data associated with given models of CTN containers or GF refrigeration units. The method then comprises, for example, cleaning and preprocessing this data. The data is, for example, divided into several sets, such as a training data set and a validation data set. The model is then trained on the training data set. The second function F2 receives, for example, as input, during training, electrical consumption data, such as consumption values ​​over time and as a function of given parameters, associated with models of CTN containers or GF refrigeration units.The method then comprises a step of defining the learning model, for example a classification model for associating electricity consumption values ​​or electricity consumption profiles with models of CTN containers or GF refrigeration units. The parameters of the second function F2 are for example adjusted to reduce an error function (or cost function), for example by means of an optimization algorithm such as gradient descent, mean square error or binary cross entropy. The model is for example evaluated using data taken from the validation data. Such an evaluation makes it possible in particular to estimate the capacity of the first function F2 to accurately determine a model of CTN container or GF refrigeration unit based on electricity consumption data. The training comprises for example several training and evaluation cycles to optimize the second function F2.

[0297] According to one embodiment, any of the training steps of any of the first, second or third machine learning AMI, AM2, AM3 is directly applicable to another machine learning.

[0298] According to one embodiment, any of the input or output data of any of the first, second or third machine learning AMI, AM2, AM3 can be used as input data of another machine learning.

[0299] System

[0300] According to another aspect, the invention relates to a system configured to implement the steps of the method according to the invention.

[0301] The system further comprises at least one first computer CLC1 configured to receive the first input data Dtini. The first input data Dtini are for example received by the first computer CLC1 following their transmission from a local memory or from a remote memory such as a remote server.

[0302] The first CLC1 calculator is for example configured to generate, by means of a first predictive model of electrical consumption of the MOD container P trained from the first AMI machine learning, at least one predicted value of electricity consumption of the CTN container or the SFD cold room, for example of a GF cold group. According to another example, this is another calculator.

[0303] The system further comprises means for acquiring at least one operating data item of the container Dtfct, such as an acquisition device. The system further comprises a computer for detecting, by means of a characterization function Fc, at least one event associated with the container CTN from a set of event classes.

[0304] In one embodiment, the system includes the CTN container.

[0305] In one embodiment, the system comprises the SFD cold room. The SFD cold room comprises, for example, a GF cold group, or a plurality of GF cold groups.

[0306] In one embodiment, the system comprises the vehicle Vhc.

[0307] In one embodiment, the system comprises a terminal connected to the CTN container to receive operating data Dtfct from said CTN container. The CTN container and the terminal are for example connected by wired means or by a wireless link.

[0308] In one embodiment, the system comprises a terminal connected to the cold room SFD to receive operating data Dtfct from said cold room SFD, for example operating data from the cold group GF. The cold room SFD and the terminal are for example connected by wired means or by a wireless link.

[0309] According to another aspect, the invention relates to a system for classifying an Evm event associated with a CTN container on board a Vhc vehicle, or associated with an SFD cold room comprising at least one GF cold unit.

[0310] The system comprises for example a plurality of CTN containers each generating at least one operating data Dtfct, such as actual electricity consumption data Cre.

[0311] The system includes, for example, an administration console for the operation of said CTN containers.

[0312] The administration console communicates with the CTN containers, for example, locally via a wired connection, or remotely via a wireless connection.

[0313] According to another example, the system comprises a plurality of GF refrigeration units each generating at least one operating data Dtfct, such as actual electrical consumption data Cre. The system comprises, for example, a console for administering the operation of said GF refrigeration units.

[0314] The administration console communicates, for example, with the GF cooling units locally via a wired connection, or remotely via a wireless connection.

[0315] The system comprises, for example, at least one calculator configured to implement at least one Evm event classification function.

[0316] According to another aspect, the invention relates to a computer program product comprising instructions which, when the program is executed by a computer, cause the latter to implement steps of the method of the invention.

Claims

CLAIMS 1. Method for detecting an event (Evm) associated with a container (CTN) on board a vehicle (Vhc), the method comprising the following steps: • First reception (RECi), by a first calculator (CLCi), of first input data (Dtini) comprising at least one temperature data item (Dtempi) relating to an internal temperature of the container (CTN); • Second reception (REC2), by said first computer (CLC1) or by another computer, of second input data (Dtin2) comprising at least one operating data (Dtfct) relating to the operation of at least one item of equipment of the container (CTN); • Selection (SLCi) of a first predictive model of electricity consumption (MOD P ) of the container (CTN), said model being trained from a first machine learning (AMI); • Implementation (MO1) of the first predictive model of electricity consumption (MODP ) to determine at least one predicted electrical consumption data of the container (C pr ), called predicted consumption data (C pr ), and characterizing an expected operating state of the container (CTN); • Detection (DTC1) of at least one event associated with the container (CTN) among a set of event classes, by means of a characterization function (Fc) automatically identifying and characterizing at least one deviation between the expected operating state of the container (CTN) and an actual operating state of the container (CTN), by means of the predicted consumption data (C pr ) and the operating data (Dtfct).

2. Method according to claim 1, in which the operating data (Dtfct) comprises actual electrical consumption data of the container (Cre), called actual consumption data (Cre), and in which the characterization function (Fc) identifies and characterizes at least one difference between a value associated with said actual consumption data (Cre) and a value associated with the predicted consumption data (C pr ).

3. Method according to any one of the preceding claims, in which the operating data (Dtfct) comprises at least one piece of data relating to an operating duration and / or an operating regime of at least one piece of equipment of the container (CTN).

4. Method according to any one of the preceding claims, in which the characterization function (Fc) comprises an inference engine implemented from a knowledge base and from at least one predefined rule.

5. Method according to any one of the preceding claims, in which the step of selecting (SLCi) the first predictive consumption model (MOD P) is implemented automatically by means of a first function (Fi) whose parameters depend on data taken from the first input data (Dtini) and / or data taken from the second input data (Dtin2).

6. Method according to claim 5, in which the first function (Fi) is a learning function trained by means of a second machine learning (AM2).

7. Method according to any one of the preceding claims, in which the first input data (Dtini) comprises at least one external environment data (Dtext) measured by at least one sensor positioned on the container (CTN) or on the vehicle (Vhc), at least one external environment data (Dtext) being estimated from a position data (Dt pos ) vehicle time stamp (Vhc) received from a tracking device.

8. Method according to any one of the preceding claims, in which the first input data (Dtini) comprises at least meteorological data (Dtmt) received by means of a communication interface and from at least one remote server.

9. Method according to any one of the preceding claims, in which the second input data (Dtin2) comprise at least one first characteristic data item of the container (Dtd), called first characteristic data item (Dtd), and relating to an operating mode of the container (Dd) characterized by a limited interval of internal set temperatures of the container (CTN).

10. Method according to any one of the preceding claims, in which the second input data (Dtin2) comprise at least one second data characteristic of the container (D C 2), called second characteristic data (D C2), and relating to a model and / or a brand of the container (CTN) characterizing technical characteristics specific to the container (CTN).

11. Method according to claims 2 and 10, in which at least one second characteristic data (D C 2) includes data produced by a second function (F2) trained using a third machine learning (AMS), called the second function (F2), taking as input the real consumption data (Cre).

12. Method according to any one of the preceding claims, in which the first machine learning (AMI) receives as input data comprising histories of the same input data (Dtini, Dtin2) relating to different container models or different histories of input data (Dtini, Dtin2) relating to the same container model.

13. System for detecting an event associated with a container (CTN) configured to implement the steps of the method according to any one of claims 1 to 12.

14. System according to claim 13, comprising: • At least one calculator for: i. Receiving (RECi) the first input data (Dtini ) comprising at least one temperature data item (Dtempi ) related to an internal temperature of the container (CTN); ii. Receiving (REC2) the second input data (Dtin2) comprising at least one operating data item (Dtfct) relating to an operation of at least one item of equipment of the container (CTN); iii. Selecting (SLC1) a first predictive model of electrical consumption (MODp) of the container (CTN) from a set of predictive models of electrical consumption (ENSM); iv. Implementing (MO1) the first predictive model of electrical consumption (MODp) to determine at least one predicted electrical consumption data item of the container (Cpr), called predicted consumption data item (Cpr), and characterizing an expected operating state of the container (CTN); v.Detect (DTC1) at least one event (Evm) associated with the container (CTN) from a set of event classes, by means of a characterization function (Fc) automatically identifying and characterizing at least one deviation between the expected operating state of the container (CTN) and an actual operating state of the container (CTN), and at least by means of the predicted electrical consumption value (Cpr) and the operating data (Dtfct).

15. System according to any one of claims 13 to 14, comprising: • The container (CTN); • The vehicle (Vhc); • A terminal connected to said container (CTN).

16. System for classifying an event (Evm) associated with a container (CTN) on board a vehicle (Vhc), said system comprising: • a plurality of containers each generating at least one temperature datum (Dttempi) relating to their internal temperature and at least one operating datum (Dtfct) relating to the operation of at least one item of container equipment (CTN); • an administration console for the operation of the plurality of containers; • at least one calculator configured to: o implement a first predictive model of electricity consumption (MOD P ) to determine at least one predicted electrical consumption data of the container (C pr ) characterizing an expected operating state of the container (CTN); o implement at least one function for classifying an event (Evm) characterized by at least one difference between an expected operating state of the container (CTN) and an actual operating state of the container (CTN), by means of the predicted electrical consumption value (Cpr) and the operating data (Dtfct).

17. Computer program product comprising instructions which, when the program is executed by a computer, cause the latter to implement the steps of the method according to any one of claims 1 to 12.

Images