System for determining electrical power

The system uses decision trees based on real consumption data to predict electrical power needs, addressing imprecision in existing methods and minimizing network oversizing and costs by aligning with actual consumption patterns.

FR3117283B1Active Publication Date: 2024-11-29ELECTRICITE DE FRANCE
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
FR2020012722
Authority / Receiving Office
FR · FR
Patent Type
Patents
Current Assignee / Owner
Filing Date
2020-12-04
Publication Date
2024-11-29
Estimated Expiration
2040-12-04

AI Technical Summary

Technical Problem

Existing methods for determining the maximum electrical power drawn by a home connected to an electrical network are imprecise, leading to potential oversizing of the network and unnecessary reinforcement costs, as they rely on estimated or summed nominal powers without considering actual consumption patterns.

Method used

A system using decision trees generated from real consumption data of similar homes to predict the maximum electrical power drawn by a new home, incorporating information on electrical equipment and property details, ensuring a more precise determination.

Benefits of technology

This approach reduces the risk of network oversizing and associated costs by providing a more accurate estimation of electrical power needs, aligning with actual consumption patterns and reducing the need for unnecessary network reinforcement.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 00000018_0000
    Figure 00000018_0000
  • Figure 00000018_0001
    Figure 00000018_0001
  • Figure 00000018_0002
    Figure 00000018_0002
Patent Text Reader

Abstract

The present invention relates to a system (5) comprising: - a memory (6) configured to receive and store: information concerning secondary housings (3), and information concerning a first housing (2), and - a processor (7) configured to: generate a plurality of decision trees (A) stored by the memory (6), and determine the maximum value of electrical power that can be drawn from an electrical network (1) by the first housing (2) by means of the plurality of decision trees (A), and information concerning a first housing (2). Figure for the abstract: Fig. 2
Need to check novelty before this filing date? Find Prior Art

Description

Title of the invention: System for determining an electrical power FIELD OF THE INVENTION

[0001] The present invention relates to the dimensioning of an electrical network, more particularly an electricity transport and / or distribution network to which a plurality of dwellings are connected.

[0002] More specifically, the present invention relates to the determination of a power to be subscribed to by a dwelling connected to an electrical network in order to optimize the sizing of the electrical network. STATE OF THE ART

[0003] When a dwelling is connected to an electrical network, such as an electricity transport and / or distribution network, it is necessary, for safety reasons, to determine the maximum value of the electrical power that the electrical equipment of this dwelling is likely to draw, at the same time, from the electrical network.

[0004] Indeed, if the electrical power drawn from the electrical network by the dwelling is greater than the sum of the nominal electrical powers of the electrical equipment in the dwelling, this may be a sign of a fault in at least one of the electrical equipment in the dwelling. In this case, it is essential to cut off the electrical current flowing between the electrical network and the dwelling. Otherwise, a fire could, for example, break out in the dwelling.

[0005] The electrical network is partly sized according to the maximum values ​​of the electrical power likely to be drawn by the dwellings connected to the electrical network. In particular, the diameter of the cables connecting the different nodes of the electrical network depends directly on these values. Consequently, any change in the properties of the dwellings connected to the electrical network, which would impact these values, must be able to be detected. If necessary, work to reinforce the electrical network must be carried out.

[0006] It is therefore essential to regularly and accurately determine the maximum value of the electrical power that can be drawn from the electricity network by each of the dwellings connected to it. In fact, if this value is overestimated, lengthy and costly work to reinforce the electricity network could be carried out, although it is not necessary.

[0007] The updating of these maximum values ​​of the electrical powers likely to be drawn from the electrical network by the dwellings connected to it, is in particular implemented each time a new occupant of a dwelling connected to the electricity network subscribes to an electricity supply offer.

[0008] Indeed, during such a subscription, the new occupant generally provides information relating to the properties of the accommodation and / or the electrical equipment installed in the accommodation. In addition, the electricity network manager has the value of the electrical power subscribed to by the previous occupant of the accommodation.

[0009] From this, methods for determining a maximum value of electrical power likely to be drawn from an electrical network by a dwelling during a given period have already been developed in the state of the art.

[0010] A known method consists of reusing the value of the power subscribed by the previous occupant of the dwelling. However, this method is not satisfactory because there is no guarantee that this value has not been overestimated. Consequently, such a method does not prevent oversizing of the electrical network. In addition, the new occupant may pay an electricity bill equivalent to that of his predecessor even though he has lower electricity consumption.

[0011] Another known method consists of determining this value by adding together the nominal powers of the electrical equipment installed in the dwelling and possibly applying an expansion coefficient. This method is also not entirely satisfactory. Indeed, it is rare for all the electrical equipment in a dwelling to operate at the same time. Furthermore, the expansion coefficient may be poorly chosen. Here again, such a method does not guarantee that there is no oversizing of the electrical network and no unnecessary additional financial costs for the new occupant.

[0012] Another known method consists of using a model based on the electrical consumption of the main electrical equipment in the home, in particular the heating and the domestic hot water production system. However, the known models are very simple and imprecise, so that such a method is not entirely satisfactory either.

[0013] There is therefore a need to overcome at least one of the drawbacks of the state of the art. Statement of the invention

[0014] An aim of the invention is to improve the determination of the maximum value of the electrical power likely to be drawn from an electrical network by a dwelling in order to limit the oversizing of the electrical network and the additional costs associated with work to reinforce the electrical network.

[0015] To this end, according to one aspect of the invention, a system is proposed for determining a maximum value of an electrical power capable of being drawn from a electrical network by a first dwelling during a first period, the system comprising: - a memory configured to receive and store: for each second housing of a plurality of second housings distinct from the first housing: information concerning electrical equipment of the second housing, information concerning a property of the second housing and a maximum value of electrical power drawn from the electrical network by the second housing during a second period prior to the first period, and information concerning electrical equipment of the first housing and information concerning a property of the first housing, and - a processor configured to: generating a plurality of decision trees, each decision tree providing a distribution of the maximum values ​​of an electrical power drawn from the electrical network by each second dwelling of at least a portion of the second dwellings during the second period, based on information relating to electrical equipment of each second dwelling of this portion of the second dwellings and / or information relating to a property of each second dwelling of this portion of the second dwellings, the plurality of decision trees being stored by the memory, and determining the maximum value of the electrical power likely to be drawn from the electrical network by the first dwelling during the first period by means of the plurality of decision trees, the information relating to electrical equipment of the first dwelling and the information relating to a property of the first dwelling.

[0016] When generating the plurality of decision trees, the system according to the invention uses all the information available to the electricity supplier and / or the electricity network manager each time a new occupant of the dwelling subscribes to an electricity supply contract. This makes the determination of the maximum value of the electrical power likely to be withdrawn from the electricity network by the dwelling much more precise. In addition, the system according to the invention uses the maximum value of the electrical power actually withdrawn from the electricity network by other dwellings connected to it. This actual value can in particular be provided by means of meters installed in the other dwellings connected to the electricity network. By using the actual value for the other dwellings, and not the value estimated during previous subscriptions, the system offers a more precise determination.From there, the value of the power subscribed by the new occupant of the accommodation corresponds much better to the reality of his needs, which reduces the risk of oversizing the electrical network and carrying out reinforcement work. premature, even unnecessary, as well as the risk of additional financial costs for the new occupant.

[0017] Advantageously, but optionally, the system according to the invention may comprise at least one of the following characteristics, taken alone or in combination: - information relating to electrical equipment of the first dwelling and / or the second dwelling includes information relating to the presence and / or ownership of electric heating, a domestic hot water production system, air conditioning, a dishwasher, a tumble dryer, a swimming pool, a refrigerator, a freezer and / or electric cooking equipment of the first dwelling and / or the second dwelling, - information relating to a property of the first dwelling and / or the second dwelling includes information relating to a location of the first dwelling and / or the second dwelling, to a type of the first dwelling and / or the second dwelling, to a surface area of ​​the first dwelling and / or the second dwelling and / or to a number of persons occupying the first dwelling and / or the second dwelling, - the information relating to the type of the first dwelling and / or the second dwelling includes whether the first dwelling and / or the second dwelling belongs to a category among an apartment and a house, - for each second housing, the memory is configured to store information concerning electrical equipment of the second housing, information concerning a property of the second housing and the maximum value of electrical power drawn from the electrical network by the second housing during the second period, in the form of a vector, - the processor is configured to generate each decision tree from a sample of vectors, the sample being obtained by randomly drawing a plurality of observations with replacement made from the vectors stored by the memory, a value of a probability of each observation being the inverse of the number of observations made, - the processor is configured to generate each decision tree using a CART algorithm, - each decision tree includes: a plurality of sheets, each sheet corresponding to a maximum value of electrical power drawn from the electrical network by a second dwelling of this part of the second dwellings during the second period, and a plurality of branches, each branch providing information regarding electrical equipment of each second dwelling of that part of the second dwellings and / or information regarding a property of each second dwelling of this part of the second dwellings by means of which the distribution of the maximum values ​​of an electrical power drawn from the electrical network by each second dwelling of this part of the second dwellings during the second period, is carried out by the processor, a sheet located at the end of a branch providing the maximum value of an electrical power drawn from the electrical network by a second dwelling of the part of the second dwellings, the branch providing the information concerning an electrical equipment of the second dwelling of this part of the second dwellings and / or the information concerning a property of the second dwelling of this part of the second dwellings, - the processor is configured to: for each decision tree, choose a leaf located at one end of a branch, the branch providing information concerning electrical equipment of a second dwelling of this part of the second dwellings and / or information concerning a property of a second dwelling of this part of the second dwellings, which is identical, respectively, to the information concerning electrical equipment of the first dwelling and / or the information concerning a property of the first dwelling, determining the maximum value of the electrical power likely to be drawn from the electrical network by the first dwelling during the first period by means of a calculation of an average of the maximum values ​​of electrical power drawn from the electrical network by the second dwellings of this part of the second dwellings during the second period which correspond to the chosen sheets, and - the processor is configured to generate a value of an electrical power to be subscribed to from an electrical operator for the first dwelling during the first period, from the maximum value of the electrical power likely to be drawn from the electrical network by the first dwelling during the first period that the processor has determined. DESCRIPTION OF FIGURES

[0018] Other characteristics, aims and advantages of the invention will emerge from the following description, which is purely illustrative and non-limiting, and which must be read in conjunction with the appended drawings in which:

[0019] [Fig.l] schematically illustrates an electrical network to which dwellings comprising electrical equipment are connected.

[0020] [Fig.2] schematically illustrates a system according to one embodiment of the invention.

[0021] [Fig. 3] illustrates an example of a decision tree capable of being generated by a processor of a system according to an embodiment of the invention.

[0022] Throughout the figures, similar elements bear identical references. DETAILED DESCRIPTION OF THE INVENTION Electrical network

[0023] With reference to [Fig. 1], an electrical network 1, such as an electricity transmission and / or distribution network, comprises a plurality of electrical cables 10 connected together. Such an electrical network 1 makes it possible to supply electricity to a first dwelling 2 as well as to a plurality of second dwellings 3, distinct from the first dwelling 2, all connected to the electrical network 1. The structure of the electrical network 1, and in particular the dimensions of the cables 10, typically their diameter, depends on the electrical power drawn from the electrical network 1 by the first dwelling 2 and the second dwellings 3.

[0024] Each of the first housing 2 and the second housings 3 comprises at least one piece of electrical equipment 20, 30, that is to say equipment operating by means of electricity supplied by the electrical network 1. Examples of such electrical equipment 20, 30 are an electric heater, an electric water heater or domestic hot water production system, at least one air conditioner, a dishwasher, a clothes dryer, a swimming pool, a refrigerator, a freezer, and electric cooking equipment.

[0025] Furthermore, at least a portion of the second housings 3 comprises a sensor 32, typically a digital meter, configured to collect, in real time, the value of the electrical power drawn from the electrical network 1 by each of the second housings 3. These values ​​can then be communicated to a server 4, by means of communication means such as a computer communication bus, or a wireless communication network, for example an internet network, a Wifi network or a Bluetooth network. The server 4 is, for its part, configured to determine a maximum value of an electrical power drawn from the electrical network 1 by the second housing 3 during a given period from the values ​​collected in real time that the sensor 32 will have transmitted to it.

[0026] In order to plan work to reinforce the electrical network 1, it is useful to determine a maximum value of an electrical power likely to be drawn from the electrical network 1 by the first dwelling 2 during a first period. Insofar as the electrical consumption of a dwelling is, most of the time, cyclical, it is preferable to choose the first period as the smallest period of the consumption cycle of the first dwelling 2, typically a year, for example the coming year.

[0027] This determination can in particular be carried out when a new occupant takes possession of the first dwelling 2, and seeks to subscribe to an electricity supply contract with a distributor ensuring the management of the electricity network 1. This is not, however, limiting since this determination can also be implemented during information campaigns on the electricity network 1 carried out by the distributor. Determination system

[0028] With reference to [Fig.2], a system 5 for determining a maximum value of an electrical power likely to be drawn from the electrical network 1 by the first housing 2 during the first period comprises a memory 6 and a processor 7.

[0029] The memory 6 is configured to receive and store information, typically in computer form.

[0030] The processor 7 is configured to perform a certain number of actions on the basis of the information stored by the memory 6.

[0031] The processor 7 and the memory 6 are capable of exchanging information, for example by means of communication means such as a computer communication bus 8, or a wireless communication network, for example an internet network, a Wifi network or a Bluetooth network.

[0032] Advantageously, the system 5 comprises a first device 71 configured to collect information concerning electrical equipment of the first dwelling 2 and / or information concerning a property of the first dwelling 2. The first device 71 may be a mobile terminal, provided with a memory and a user interface. Where appropriate, the memory of the first device 71 may store information transmitted by the new occupant by means of the user interface of the first device 71. An example of a user interface is a dedicated mobile application configured to ask questions to the new occupant concerning electrical equipment 20 of the first dwelling 2 and / or a property of the first dwelling 2 when subscribing to an electricity supply contract, and to record the answers in the memory of the first device 71.This is not, however, limiting, since an employee of the electricity supplier can also ask the questions directly to the new occupant, for example during a telephone communication or exchange of computer messages, and record the information collected himself on the memory 6 of the system 5. Of course, the first device 71 can also operate automatically, without the intervention of the new occupant and / or the employee. For example, the first device 71 can be a sensor installed within the first dwelling 2 and configured to determine the presence of electrical equipment 20 of the first dwelling 2 on the basis of . the evolution of the electricity consumption of the first housing 2 during a given period. In any event, the first device 71 is configured to transmit the information it collects to the memory 6 of the system 5.

[0033] Advantageously, the system 5 comprises at least one second device 72 configured to collect information concerning electrical equipment 30 of a second housing 3 and / or information concerning a property of the second housing 3. The second device 72 may take the form and / or have the same properties as those already described concerning the first device 71.

[0034] In one embodiment, the system 5 comprises a server 4 configured to ensure the exchange of information between the first device 71 and / or the second device 72, and the memory 6. This server 4 can communicate with the first device 71 and / or the second device 72 by means of communication means such as a computer communication bus, or a wireless communication network, for example an internet network, a Wifi network or a Bluetooth network. Advantageously, this is the same server 4 as the one which receives, in real time, from sensors 32 of second housings 3, the value of the electrical power drawn from the electrical network 1 by each of the second housings 3.

[0035] In any event, the architecture of the system 5 can be distributed in several separate locations or gathered in a given location, as long as the different elements that compose it are able to exchange information efficiently. Memory 6

[0036] Memory 6 is configured to receive and store: - for each second housing 3 of a plurality of second housings 3 distinct from the first housing 2: information concerning electrical equipment 30 of the second housing 3, information concerning a property of the second housing 3 and a maximum value of electrical power drawn from the electrical network 1 by the second housing 3 during a second period prior to the first period, and - information concerning electrical equipment 20 of the first housing 2 and information concerning a property of the first housing 2.

[0037] Preferably, the second period is of the same duration as the first period. Typically, the second period corresponds to the year elapsed at the time when the system 5 makes the determination, while the first period corresponds to the year to come.

[0038] Preferably, the information concerning an electrical equipment 30 of the second housing 3 and the information concerning a property of the second housing 3 are collected by the first device 71, which transmits them to the memory 6, possibly via the server 4. Similarly, the maximum value of an electrical power drawn from the electrical network 1 by the second housing 3 during the second period is advantageously determined by the server 4, on the basis of the values ​​collected in real time by the sensor 32 installed in the second housing 3.

[0039] Preferably, the information relating to electrical equipment 20, 30 of the first housing 2 and / or of the second housing 3 comprises information relating to the presence and / or a property of an electric heater, an electric water heater or domestic hot water (DHW) production system, at least one air conditioner, a dishwasher, a tumble dryer, a swimming pool, a refrigerator, a freezer and / or electric cooking equipment of the first housing 2 and / or of the second housing 3.Typically, the information relating to electrical equipment 20 of the first dwelling 2 takes the following form: “presence of electric heating, absence of electrical production of domestic hot water, presence of air conditioning, nominal power of the air conditioner of 1 kW, presence of two refrigerators, absence of electric cooking equipment”.

[0040] Advantageously, the information relating to a property of the first dwelling 2 and / or the second dwelling 3 comprises information relating to a location of the first dwelling 2 and / or the second dwelling 3, to a type of the first dwelling 2 and / or the second dwelling 3, to a surface area of ​​the first dwelling 2 and / or the second dwelling 3 and / or to a number of people occupying the first dwelling 2 and / or the second dwelling 3. Furthermore, the information relating to the type of the first dwelling 2 and / or the second dwelling 3 preferably comprises a membership of the first dwelling 2 and / or the second dwelling 3 to a category among an apartment and a house. Typically, the information relating to a property of the second dwelling 3 takes the following form: “dwelling located on the French Riviera, surface area of ​​the dwelling of 300 m2, fewer than four people occupying the dwelling, house-type dwelling”.

[0041] In one embodiment, for each second housing 3, the memory 6 is configured to store the information concerning an electrical equipment 30 of the second housing 3, the information concerning a property of the second housing 3 and the maximum value of an electrical power drawn from the electrical network 1 by the second housing 3 during the second period, in the form of a vector. In this way, access to the information concerning the second housings 3 in the memory 6 is facilitated. Processor

[0042] With reference to [Fig. 3], the processor 7 is configured to generate a plurality of decision trees A1, A2, typically several hundred, for example approximately 500 decision trees A1, A2. Each decision tree A1, A2 provides a distribution of the maximum values ​​of an electrical power drawn from the electrical network 1 by each second housing 3 of at least a portion of the second housings 3 for which the memory 6 has stored information, during the second period. In each decision tree A1, A2, this distribution is carried out according to information concerning an electrical equipment 30 of each second dwelling 3 of this part of the second dwellings 3 and / or information concerning a property of each second dwelling 3 of this part of the second dwellings 3. In addition, the plurality of decision trees A1, A2 is stored by the memory 6. The decision trees A1, A2 have the advantage of being easily readable and understandable, as can be seen in [Fig.3].

[0043] More precisely, as visible in [Fig.3], each decision tree A1, A2 comprises a plurality of leaves Fil, F12, F13, F14, F15, F21, F22, F23, F24, F25, each leaf Fil, F12, F13, F14, F15, F21, F22, F23, F24, F25 corresponding to a maximum value of an electrical power drawn from the electrical network 1 by a second housing 3 of this part of the second housings 3, during the second period.In addition, each decision tree A1, A2 comprises a plurality of branches B11, B12, B13, B14, B15, B21, B22, B23, B24, B25, each branch B11, B12, B13, B14, B15, B21, B22, B23, B24, B25 providing the information concerning an electrical equipment 30 of each second dwelling 3 of this part of the second dwellings 3 and / or the information concerning a property of each second dwelling 3 of this part of the second dwellings 3, by means of which the distribution of the maximum values ​​of an electrical power drawn from the electrical network 1 by each second dwelling 3 of this part of the second dwellings 3, during the second period, is carried out by the processor 7.Thus, a sheet Fil, F12, F13, F14, F15, F21, F22, F23, F24, F25 located at the end of a branch Bll, B12, B13, B14, B15, B21, B22, B23, B24, B25 provides the maximum value of an electrical power drawn from the electrical network 1 by a second dwelling 3 of the part of the second dwellings 3, the branch Bll, B12, B13, B14, B15, B21, B22, B23, B24, B25 providing the information concerning an electrical equipment 30 of the second dwelling 3 of this part of the second dwellings 3 and / or the information concerning a property of the second dwelling 3 of this part of the second dwellings 3.

[0044] In one embodiment, the processor 7 is configured to generate each decision tree A1, A2 from a sample of vectors. In addition, this sample is obtained by randomly drawing a plurality of observations with replacement made from the vectors containing the information concerning an electrical equipment 30 of each second dwelling 3, the information concerning a property of each second dwelling 3 and the maximum value of an electrical power drawn from the electrical network 1 by each second dwelling 3 during the second period, which are stored by the memory 6. Finally, a value of a probability of each observation is chosen to be the inverse of the number of observations made. This process of determining the sample of vectors, called "bootstrap" in English terminology, ensures good representativeness of the information contained in each decision tree A1, A2.

[0045] Preferably, the processor 7 is configured to generate each decision tree A1, A2 using a CART algorithm (for "Classification And Regression Trees", in English terminology), for example described in the document Classification and Regression Trees Léo Breiman, Jerome Friedman, Charles J. Stone, RA Olshen (1984). Indeed, it is a powerful and efficient algorithm in this regard.Indeed, for each given sample, the decision tree A1, A2 generated by means of the CART algorithm makes it possible to determine the best possible distribution of the maximum values ​​of an electrical power drawn from the electrical network 1 by each second dwelling 3 of at least a part of the second dwellings 3 during the second period, based on information concerning electrical equipment 30 of each second dwelling 3 of this part of the second dwellings 3 and / or information concerning a property of each second dwelling 3 of this part of the second dwellings 3.

[0046] Furthermore, the processor 7 is configured to determine the maximum value of the electrical power likely to be drawn from the electrical network 1 by the first dwelling 2 during the first period by means of the plurality of decision trees A1, A2, the information concerning an electrical equipment 20 of the first dwelling 2 and the information concerning a property of the first dwelling 2.

[0047] In one embodiment, this determination is performed by the processor 7 in the following manner.

[0048] First, for each decision tree A1, A2, the processor 7 chooses a leaf Fil, F12, F13, F14, F15, F21, F22, F23, F24, F25 located at one end of a branch B11, B12, B13, B14, B15, B21, B22, B23, B24, B25. The branch B11, B12, B13, B14, B15, B21, B22, B23, B24, B25 provides information concerning electrical equipment 30 of a second dwelling 3 of the part of the second dwellings 3 concerned by the decision tree A1, A2 and / or information concerning a property of a second dwelling 3 of the part of the second dwellings 3 concerned by the decision tree A1, A2. This choice is made so that this information is identical, respectively, to the information concerning electrical equipment 20 of the first dwelling 2 and / or the information concerning a property of the first dwelling 2.

[0049] Then, the processor 7 determines the maximum value of the electrical power likely to be drawn from the electrical network 1 by the first dwelling 2 during the first period by means of a calculation of an average of the maximum values ​​of an electrical power drawn from the electrical network 1 by the second dwellings 3 of the part of the second dwellings 3 concerned by the decision tree A1, A2 during of the second period, which correspond to the leaves Fil, F12, F13, F14, F15, F21, F22, F23, F24, F25 chosen. Of course, other calculations than that of an average can be implemented, in particular by assigning weight coefficients according to the decision tree Al, A2 considered. Typically, the value retained from a decision tree Al, A2 will have as much weight in the calculation carried out by the processor 7 as the information contained in the decision tree Al, A2 is similar to that received concerning the first housing 2.

[0050] In one embodiment, the processor 7 is configured to generate a value of an electrical power to be subscribed to from an electrical operator for the first dwelling 2 during the first period, from the maximum value of the electrical power likely to be drawn from the electrical network 1 by the first dwelling 2 during the first period that the processor 7 has determined. Typically, this value of an electrical power to be subscribed is the nearest integer greater than the determined value.

[0051] Advantageously, the determined value and / or the value of an electrical power to be subscribed to is transmitted to the memory 6, which stores it, with a view to possible restitution to a user or transmission to the server 4. Returning to [Fig. 2], in one embodiment, the system 5 further comprises a communication interface 9 configured to restore to a user the determined value and / or the value of an electrical power to be subscribed to. This communication interface 9 may be located near the processor 7 or the memory 6 or the server 4. Examples

[0052] Returning to [Fig.3], a specific example of operation of system 5 will now be detailed.

[0053] As visible in [Fig.3], two decision trees A1, A2 were generated by the processor 7.

[0054] In the first decision tree A1, the maximum values ​​of an electrical power drawn from the electrical network 1 by each second dwelling 3 of a part of the second dwellings 3 during the second period are distributed, which are the following: 5,123 VA, 8,765 VA, 12,543 VA, 10,189 VA and 5,675 VA. In addition, the processor 7 has carried out this distribution as a function of: - the following information concerning a property of the second dwellings 3 of this part of the second dwellings 3: the surface area of ​​the dwelling less than or not than 50 m2 and the number of occupants greater than or not than 2 people, and - the following information concerning the electrical equipment 30 of the second dwellings 3 of this part of the second dwellings 3: presence or absence of electric heating and presence or absence of a dishwasher.

[0055] In the second decision tree A2, the maximum values ​​of an electrical power drawn from the electrical network 1 by each second dwelling 3 of another part of the second dwellings 3 during the second period are distributed, which are as follows: 11,456 VA, 8,012 VA, 7,634 VA, 5,412 VA and: 4,753 VA. In addition, the processor 7 has carried out this distribution as a function of: - the following information concerning a property of the second dwellings 3 of this part of the second dwellings 3: the dwelling is of the house or apartment type, and - the following information concerning the electrical equipment 30 of the second dwellings 3 of this part of the second dwellings 3: presence or absence of electric heating, presence or absence of electric cooking equipment and presence or absence of a domestic hot water production device.

[0056] Furthermore, the following information was collected concerning a first dwelling 2, whose new occupant wishes to subscribe to an electricity supply contract: - presence of electric heating, electric production of domestic hot water, a refrigerator and a washing machine, and - the first accommodation 2 is located in Corsica, it is a house of 85 m2 and the new occupants are two.

[0057] In the first decision tree A1, the processor 7 chooses the fourth leaf F14, whose branch B14 shares the most information in common with that collected concerning the first housing 2, and which corresponds to a value of 10,189 VA.

[0058] In the second decision tree A2, the processor 7 chooses the first leaf Fil, whose branch B11 shares the most information in common with that collected concerning the first housing 2, and which corresponds to a value of 11,456 VA.

[0059] The processor 7 then determines the maximum value of the electrical power likely to be drawn from the electrical network 1 by the first dwelling 2 during the first period at 10,823 VA, i.e. the average of the values ​​chosen from the two decision trees A1, A2. From there, the processor 7 generates the value of 12kV as being the electrical power to be subscribed to from an electrical operator.

[0060] This value is then transmitted to a manager of the electricity network 1 who will take it into account in the planning of work to reinforce the electricity network 1.

Claims

1. Claims System (5) for determining a maximum value of electrical power likely to be drawn from an electrical network (1) by a first dwelling (2) during a first period, the system (5) comprising: - a memory (6) configured to receive and store: for each second housing (3) of a plurality of second housings (3) distinct from the first housing (2): information concerning electrical equipment (30) of the second housing (3), information concerning a property of the second housing (3) and a maximum value of electrical power drawn from the electrical network (1) by the second housing (3) during a second period prior to the first period, and information concerning electrical equipment (20) of the first dwelling (2) and information concerning a property of the first dwelling (2), and - a processor (7) configured to: generating a plurality of decision trees (A1, A2), each decision tree comprising: - a plurality of sheets (Fil, F12, F13, F14, F15, F21, F22, F23, F24, F25), each sheet (Fil, F12, F13, F14, F15, F21, F22, F23, F24, F25) corresponding to a maximum value of an electrical power drawn from the electrical network (1) by a second housing (3) of at least a portion of the second housings (3) during the second period, and - a plurality of branches (B11, B12, B13, B14, B15, B21, B22, B23, B24, B25), each branch (B11, B12, B13, B14, B15, B21, B22, B23, B24, B25) providing information relating to an electrical equipment (30) of each second dwelling (3) of this part of the second dwellings (3) and / or information relating to a property of each second dwelling (3) of this part of the second dwellings (3) by means of which the distribution of the maximum values ​​of an electrical power drawn from the electrical network (1) by each second dwelling (3) of this part of the second dwellings (3) during the second period, is carried out by the processor (7), a sheet (Fil, F12, F13, F14, F15, F21, F22, F23, F24, F25) located at the end of a branch (Bll, B12, B13, B14, B15, B21, B22, B23, B24, B25) providing the maximum value of an electrical power drawn from the electrical network (1) by a second housing (3) of the part of the second housings (3), the branch (Bll, B12, B13, B14, B15, B21, B22, B23, B24, B25) providing the information concerning an electrical equipment (30) of the second housing (3) of this part of the second housings (3) and / or the information concerning a property of the second housing (3) of this part of the second housings (3) the plurality of decision trees (A1, A2) being stored by the memory (6), and determining the maximum value of the electrical power likely to be drawn from the electrical network (1) by the first dwelling (2) during the first period by means of the plurality of decision trees (A1, A2),information concerning electrical equipment (20) of the first dwelling (2) and information concerning a property of the first dwelling (2).,

2. System (5) according to claim 1, in which the information relating to electrical equipment (20, 30) of the first housing (2) and / or of the second housing (3) comprises information relating to a presence and / or a property of an electric heater, a domestic hot water production system, air conditioning, a dishwasher, a clothes dryer, a swimming pool, a refrigerator, a freezer and / or electric cooking equipment of the first housing (2) and / or of the second housing (3).

3. System (5) according to one of claims 1 and 2, in which the information relating to a property of the first dwelling (2) and / or the second dwelling (3) comprises information relating to a location of the first dwelling (2) and / or the second dwelling (3), to a type of the first dwelling (2) and / or the second dwelling (3), to a surface area of ​​the first dwelling (2) and / or the second dwelling (3) and / or to a number of people occupying the first dwelling (2) and / or the second dwelling (3).

4. System (5) according to claim 3, in which the information relating to the type of the first dwelling (2) and / or the second dwelling (3) comprises a membership of the first dwelling (2) and / or the second dwelling (3) in a category among an apartment and a house.

5. System (5) according to one of claims 1 to 4, in which, for each second housing (3), the memory (6) is configured to store the information concerning an electrical equipment (30) of the second housing (3), the information concerning a property of the second housing (3) and the maximum value of an electrical power drawn from the electrical network (1) by the second housing (3) during the second period, in the form of a vector.

6. System (5) according to claim 5, in which the processor (7) is configured to generate each decision tree (A1, A2) from a sample of vectors, the sample being obtained by randomly drawing a plurality of observations with replacement made from the vectors stored by the memory (6), a value of a probability of each observation being inverse to the number of observations made.

7. System (5) according to one of claims 1 to 6, wherein the processor (7) is configured to generate each decision tree (A1, A2) using a CART algorithm.

8. System (5) according to one of claims 1 to 7, wherein the processor (7) is configured to: - for each decision tree (A1, A2), choose a leaf (F1, F12, F13, F14, F15, F21, F22, F23, F24, F25) located at one end of a branch (B11, B12, B13, B14, B15, B21, B22, B23, B24, B25), the branch (B11, B12, B13, B14, B15, B21, B22, B23, B24, B25) providing information concerning electrical equipment (30) of a second dwelling (3) of this part of the second dwellings (3) and / or information concerning a property of a second dwelling (3) of this part of the second dwellings (3), which is identical, respectively, to the information concerning electrical equipment (20) of the first dwelling (2) and / or information concerning a property of the first dwelling (2),-determine the maximum value of the electrical power likely to be drawn from the electrical network (1) by the first dwelling (2) during the first period by means of a calculation of an average of the maximum values ​​of an electrical power drawn from the electrical network (1) by the second dwellings (3) of this part of the second dwellings (3) during the second period which correspond to the sheets (Fil, F12, F13, F14, F15, F21, F22, F23, F24, F25) chosen.,

9. System (5) according to one of claims 1 to 8, in which the processor (7) is configured to generate a value of an electrical power to be subscribed to from an electrical operator for the first dwelling (2) during the first period, from the maximum value of the electrical power likely to be drawn from the electrical network (1) by the first dwelling (2) during the first period that the processor (7) has determined.