Electronic device and method for assisting in the monitoring of an electrical cabinet, electrical current distribution system comprising such a device, and associated computer program

Abstract

Electronic device and method for monitoring an electrical cabinet, electrical current distribution system comprising such a device, and associated computer program. The present invention relates to an electronic device for monitoring an electrical cabinet (10), comprising an acquisition module, configured to acquire values ​​of quantities representative of an intensity and respectively of a voltage of an electrical current, measured by an intensity sensor (41, 42, 43, 44) and respectively a voltage sensor (51, 52, 53, 54), an estimation module, configured to calculate an estimated state of the electrical cabinet (10) via a neural network and an emission module, configured to emit a message representative of the estimated state in one form among a visual form and an audible form.According to the invention, the acquisition module is further configured to acquire a value of at least one additional quantity, and the estimation module is configured to calculate the estimated state of the electrical cabinet (10) via the neural network, further based on at least one additional quantity. Figure 1 for the abstract.

Description

Title of the invention: Electronic device and method for assisting in monitoring an electrical cabinet, system for distributing an electric current comprising such a device, and associated computer program

[0001] The present invention relates to an electronic device and method for assisting in monitoring an electrical cabinet, a system for distributing an electric current comprising such a device, and an associated computer program.

[0002] To monitor an electrical cabinet, and in particular to detect a risk of malfunction, such as for example a fire in the electrical cabinet, it is known to carry out occasional monitoring operations on the cabinets. However, it is not always possible to identify electrical or structural defects in the cabinet during the monitoring operations. It is also known to use sensors, in particular current, voltage and temperature sensors, located inside the electrical cabinet, and to emit an alarm signal if a value measured by one of these sensors crosses a predefined threshold.

[0003] CN 115 880 848 A proposes to use an artificial intelligence algorithm to detect an electrical fire in an electrical cabinet, using data relating to a temperature of components inside the electrical cabinet, a leakage current, an intensity and a voltage of an electric current flowing in the electrical cabinet.

[0004] However, it is necessary to make detection more efficient.

[0005] The aim of the invention is then to propose an electronic monitoring assistance device making it possible to resolve these drawbacks and improve detection.

[0006] For this purpose, the invention relates to an electronic device for assisting in monitoring an electrical cabinet, the electrical cabinet comprising an input terminal and several output terminals, and being capable of being connected to a source via the input terminal and to a plurality of loads via the output terminals and configured to distribute an electric current from the source to the plurality of loads, the electrical cabinet comprising a disconnector connected to the input terminal and a main bus connected between the disconnector and the output terminals, the device comprising: - an acquisition module, configured to acquire values ​​of quantities representative of an intensity and respectively of a voltage of an electric current flowing between the input terminal and the output terminals, measured by an intensity sensor and respectively a voltage sensor; - an estimation module, configured to calculate an estimated state of the cabinet electrical via a neural network, each quantity acquired by the acquisition module being an input variable of the neural network, the estimated state of the electrical cabinet being an output variable of the neural network; and - a transmission module, configured to transmit a message representative of the estimated state in one of a visual form and an audio form.

[0007] According to the invention, the acquisition module is further configured to acquire a value of at least one additional quantity measured by at least one additional sensor among a gas sensor and a humidity sensor, the or each quantity being representative respectively of a quantity of gas particles and of a humidity inside the electrical cabinet. Also according to the invention, the estimation module is then configured to calculate the estimated state of the electrical cabinet via the neural network as a function further of the at least one additional quantity, the at least one additional quantity being an additional input variable of the neural network.

[0008] Thanks to the invention, the estimation module calculates an estimated state of the electrical cabinet using at least one additional quantity, measured by at least one additional sensor. This additional quantity makes it possible to have more relevant variables as input to the neural network, in order to improve the reliability of the calculation of the estimated state of the electrical cabinet.

[0009] According to other advantageous aspects of the invention, the device comprises one or more of the following characteristics, taken individually or in all technically possible combinations:

[0010] -The acquisition module is further configured to acquire the values ​​of the additional quantities measured by the humidity sensor and by the gas sensor.

[0011] - The acquisition module is further configured to acquire a value of a quantity measured by a temperature sensor, the measured quantity being representative of a temperature inside the electrical cabinet.

[0012] - The acquisition module is further configured to acquire values ​​of quantities measured by a plurality of temperature sensors, the quantities being representative of the temperature inside the electrical cabinet, each temperature sensor being configured to measure one of: a temperature in the vicinity of the input terminal, a temperature in the vicinity of a respective output terminal, a temperature of the main bus and a temperature of an electrical protection device, connected between the main bus and a respective output terminal.

[0013] - The acquisition module is further configured to acquire a value of a representative quantity of vibrations of the electrical cabinet, measured by a vibration sensor.

[0014] - The acquisition module is further configured to acquire a value of a representative quantity of noise in the electrical cabinet, measured by a noise sensor.

[0015] - The estimated state of the electrical cabinet is a state from the group consisting of: a normal state, abnormal state and critical state.

[0016] The invention also relates to a system for distributing an electric current between a source and a plurality of loads, the system comprising: - an electrical cabinet, comprising: • an input terminal, capable of being connected to the source, • several output terminals, capable of being connected to the plurality of charges, • a disconnector, connected to the input terminal, and • a main bus, connected between the disconnector and the plurality of output terminals,

[0017] the electrical cabinet being configured to distribute the electrical current from the source to the plurality of loads; - an intensity sensor and respectively a voltage sensor, configured to measure values ​​of quantities representative of an intensity and respectively of a voltage of an electric current flowing between the input terminal and the output terminals of the electrical cabinet; - an electronic device for assisting in monitoring the electrical cabinet.

[0018] According to the invention, the system further comprises at least one additional sensor from among a gas sensor and a humidity sensor, configured to acquire a value of at least one additional quantity, respectively representative of a quantity of gas particles and of a humidity inside the electrical cabinet. Also according to the invention, the electronic device for assisting in monitoring is according to any one of the preceding claims.

[0019] The invention also relates to a method for assisting in monitoring an electrical cabinet, the electrical cabinet comprising an input terminal and several output terminals, and being able to be connected to a source via the input terminal and to a plurality of loads via the output terminals and configured to distribute an electric current from the source to the plurality of loads, the electrical cabinet comprising a disconnector connected to the input terminal and a main bus connected between the disconnector and the output terminals, the method being implemented by an electronic monitoring assistance device, the method comprising the following steps: - acquisition of values ​​of quantities measured by an intensity sensor and respectively a voltage sensor, the quantities being representative of an intensity and respectively a voltage of the electric current flowing between the input terminal and the output terminals of the electrical cabinet; - calculation, via a neural network, of an estimated state of the electrical cabinet, each measured quantity being an input variable of the neural network, the estimated state being an output variable of the neural network; and - transmission, in one of a visual form and an audio form, of a message representative of the estimated state.

[0020] According to the invention, the method further comprises a step of acquiring a value of at least one additional quantity measured by at least one additional sensor among a gas sensor and a humidity sensor, the or each quantity being representative respectively of a humidity and a quantity of gas particles inside the electrical cabinet. Also according to the invention, during the calculation step the at least one additional quantity is an additional input variable of the neural network.

[0021] The invention also relates to a computer program comprising software instructions which, when executed by a computer, implement a method for assisting in monitoring an electrical cabinet as defined above.

[0022] The invention will appear more clearly on reading the description which follows, given solely by way of non-limiting example, and made with reference to the drawings in which: - [Fig.l] [Fig.l] is a schematic representation of a system for distributing an electric current according to the invention; - [Fig.2] [Fig.2] is a schematic representation of a device for assisting monitoring of an electrical cabinet according to the invention; - [Fig.3] [Fig.3] is a flowchart of a monitoring assistance process of an electrical cabinet according to the invention.

[0023] [Fig.l] is a schematic representation of a system 1 for distributing an electric current between a source 3 and a plurality of loads 5, the distribution system 1 being intended to be connected to the source 3 and to the plurality of loads 5.

[0024] The source 3 is for example a transformer, such as a medium voltage transformer or a medium voltage to low voltage transformer. Alternatively, the electrical source is another system for distributing an electric current. The electrical source 3 is capable of supplying an electric current to the distribution system 1 which is configured to distribute it between the loads 5.

[0025] The loads 5 consume the electric current supplied by the source 3. The loads 5 are for example domestic equipment, servers, charging stations, etc.

[0026] The distribution system 1 comprises an electrical cabinet 10. The electrical cabinet 10 is located for example inside a building and rests for example on a floor. The electrical cabinet 10 comprises a medium voltage panel and is advantageously configured to distribute a three-phase electric current of voltage between 1000 V and 60 kV. In this case, the electrical cabinet 10 is typically made up of this medium voltage panel.

[0027] Alternatively, the electrical cabinet 10 comprises a low-voltage panel, and is configured to distribute a three-phase electric current of voltage less than or equal to 400V. In this alternative, the electrical cabinet 10 is typically made up of the low-voltage panel.

[0028] Alternatively, the electrical cabinet 10 comprises a low voltage panel and a medium voltage panel. In this alternative, the electrical cabinet 10 is typically made up of these medium voltage and low voltage panels.

[0029] The electrical cabinet 10 comprises an enclosure 11, inside which most of its components are arranged. The enclosure 11 is for example in the form of a box, not shown, capable of being completely closed, and typically comprising four side walls, a rear wall mechanically secured to the four side walls, and a front wall. The front wall generally forms a removable door or a door that can move relative to the side walls between a closed configuration in which the box is closed and an open configuration allowing access to the components arranged inside the box.

[0030] The electrical cabinet 10 is configured to distribute the electric current from the source 3 to the plurality of loads 5. For this purpose, the electrical cabinet 10 comprises an input terminal 12 and several output terminals 13. When the distribution system 1 is in operation, the input terminal 12 is connected to the source 3 and the output terminals 13 are each connected to a respective load 5.

[0031] The electrical cabinet 10 also comprises a disconnector 16 and a main bus 17. The disconnector 16 is connected between the input terminal 12 and the main bus 17. The disconnector 16 is configured to interrupt a flow of electric current flowing from the source 3 in the electrical cabinet 10, in particular to protect the loads 5 in the event of an electrical fault in the electrical cabinet 10 or upstream of the electrical cabinet 10, for example in the event of an electrical fault at the source 3.

[0032] Alternatively, the disconnector 16 is a circuit breaker, for example an air circuit breaker.

[0033] The main bus 17 is connected between the disconnector 16 and the output terminals 13. The Main bus 17 is usually connected to output terminals 13 indirectly.

[0034] Indeed, as shown in [Fig.l], the electrical cabinet 10 comprises a first branch bus 18, a second branch bus 19, a first electrical protection device 22 and several second electrical protection devices 23.

[0035] The main bus 17 is connected to the first 18 and second 19 branch buses.

[0036] The first branch bus 18 is connected to the first protection device 22. The first protection device 22 is for example an air circuit breaker, or a miniature circuit breaker, or even a differential circuit breaker.

[0037] The second branch bus 19 is connected to the second protection devices 23. In the example of [Fig.l], four second protection devices 23 are shown. The second protection devices 23 are advantageously miniature circuit breakers, or differential circuit breakers.

[0038] As a variant, not shown, the electrical cabinet 10 comprises additional branch buses, not shown, each additional branch bus being connected to at least one additional electrical protection device.

[0039] The protection devices 22, 23 are configured to interrupt a flow of the electric current flowing from the electrical cabinet 10 to the loads 5, in particular in the event of an electrical fault downstream of the electrical cabinet 10, in particular between the output terminals 13 and the loads 5, or in the event of an electrical fault in the loads 5, in order to protect the distribution system 1, as well as the loads 5.

[0040] The electrical cabinet 10 advantageously comprises a drawer 25. The second protection devices 23 are then advantageously housed in the drawer 25. The drawer 25 is typically accessible by a user, for example in order to manipulate the second protection devices 23, to replace them, to add them, or when the second protection devices 23 are circuit breakers, to reset them following an electrical fault.

[0041] According to a variant not shown, the electrical cabinet 10 comprises several drawers, in which additional electrical protection devices are received.

[0042] The electrical cabinet 10 comprises a plinth 27. The plinth 27 advantageously receives connection cables, for example between the source 3 and the disconnector, or between the first protection device 22 and the load 5. According to a particularly advantageous arrangement, not shown, the plinth receives all of the cables connected to the protection devices and to the loads.

[0043] The distribution system 1 also comprises sensors. More specifically, in the example of [Fig.l], the distribution system 1 comprises current sensors 41, 42, 43, 44; voltage sensors 51, 52, 53, 54; gas sensors 56, 57; humidity sensors 58, 59; temperature sensors 61, 62, 63, 64, 65; a vibration sensor 67 and a noise sensor 69.

[0044] In the example of [Fig.l], there are four intensity sensors 41, 42, 43 and 44, namely a first intensity sensor 41, a second intensity sensor 42, a third intensity sensor 43 and a fourth intensity sensor 44, and are configured to measure values ​​of a quantity representative of an intensity of the electric current flowing in the electrical cabinet 10. More precisely, the intensity sensors 41 to 44 are here configured to measure an electric intensity flowing respectively in the input terminal 12 by the first intensity sensor 41, in the first branch bus 18 by the second intensity sensor 42, in the main bus 17 by the third intensity sensor 43 and in the second branch bus 19 by the fourth intensity sensor 44.

[0045] Advantageously, the intensity sensors 41 to 44 are three-phase intensity sensors, that is to say configured to measure the intensity of each phase of the electric current.

[0046] In a variant not shown, the distribution system 1 comprises additional intensity sensors configured to measure an intensity of the electric current further flowing in the protection devices 22 and 23 and the output terminals 13.

[0047] In a particularly advantageous manner, the distribution system 1 comprises intensity sensors measuring the intensity of the electric current in each of the elements mentioned above, and furthermore, comprises intensity sensors measuring the intensity of the electric current at different positions of the main bus 17 and of the first 18 and second 19 secondary buses, at each output terminal 13 and at each protection device 23.

[0048] In the example of [Fig. 1], the voltage sensors 51, 52, 53 and 54 are four in number, namely a first voltage sensor 51, a second voltage sensor 52, a third voltage sensor 53 and a fourth voltage sensor 54, and are configured to measure values ​​of quantities representative of a voltage of the electric current flowing in the cabinet 10. More precisely, the voltage sensors 51 to 54 measure a potential difference between the electric current flowing in the electrical cabinet 10 and a potential reference. In the example of [Fig.l], the voltage sensors 51 to 54 are configured to measure a voltage of the electric current flowing respectively in the input terminal 12 by the first voltage sensor 51, in the first branch bus 18 by the second voltage sensor 52, in the main bus 17 by the third voltage sensor 53 and in the second branch bus 19 by the fourth voltage sensor 54.

[0049] Advantageously, the voltage sensors 51 to 54 are three-phase voltage sensors, i.e. configured to measure a voltage of each phase of the electric current.

[0050] In a variant not shown, the distribution system 1 comprises additional voltage sensors configured to measure a voltage of the electric current circulating in the protection devices 22 and 23 and the output terminals 13.

[0051] In a particularly advantageous manner, the distribution system 1 comprises voltage sensors measuring the voltage of the electric current flowing in each of the elements mentioned above, and furthermore, comprises voltage sensors measuring the voltage of the electric current at different positions of the main bus 17 and of the first 18 and second 19 secondary buses, at each output terminal 13 and in each of the protection devices 23.

[0052] In the example of [Fig.l], the gas sensors 56 and 57 are two in number, namely a first gas sensor 56 and a second gas sensor 57 and are configured to measure a quantity representative of a quantity of gas particles inside the electrical cabinet 10, that is to say inside the casing 11 of the electrical cabinet 10. More precisely, the gas sensors 56 and 57 measure the quantity of particles of at least one given gas. Preferably, the gas sensors 56 and 57 measure the type of gas inside the electrical cabinet, and the quantity of particles for each type of gas.

[0053] As shown in [Fig.l], the first gas sensor 56 is located on the casing 11, and advantageously, at the top of the casing 2, that is to say at a position furthest from the ground. The second gas sensor 57 is located inside the drawer 19, on an internal wall of the drawer 19.

[0054] In a variant not shown, the distribution system 1 comprises additional gas sensors, configured to measure the quantity of gas particles in, or in the vicinity of, the output terminals 13. In the case where the output terminals 13 are in the vicinity of each other, that is to say any output terminal 13 is less than 10 cm from any other output terminal 13, then a single gas sensor 56 or 57 is advantageously sufficient to measure the quantity of gas particles in the vicinity of all the output terminals 13.

[0055] In the example of [Fig. 1], the humidity sensors 58 and 59 are two in number, namely a first humidity sensor 58 and a second humidity sensor 59, and are configured to measure a quantity representative of humidity. In particular, the quantity is representative of the humidity of the air inside the envelope 11.

[0056] In practice, the humidity sensors 58 and 59 are generally also sensors of a quantity representative of ambient temperature, that is to say of a quantity representative of an air temperature in the vicinity of the humidity sensor 58, 59. By "in the vicinity of", is meant located at most a distance of a few centimeters, for example ten centimeters.

[0057] As shown in [Fig.l], the first humidity sensor 58 is located in the vicinity of the input terminal 12, on the casing 11, and is configured to measure a quantity representative of a humidity of the air in the vicinity of the input terminal 12. The second humidity sensor 59 is located inside the drawer 19, on an internal wall of the drawer 19, and is configured to measure a quantity representative of a humidity of the air inside the drawer 19.

[0058] In a variant not shown, the distribution system 1 comprises additional humidity sensors, configured to measure the humidity in, or in the vicinity of, the output terminals 13. In the case where the output terminals 13 are in the vicinity of each other, then a single humidity sensor 58 or 59 is advantageously sufficient to measure the humidity in the vicinity of all the output terminals 13.

[0059] In the example of [Fig.l], the temperature sensors 61, 62, 63, 64 and 65 are five in number, namely a first temperature sensor 61, a second temperature sensor 62, a third temperature sensor 63, a fourth temperature sensor 64 and a fifth temperature sensor 65. The temperature sensors 61 to 65 are configured to measure values ​​of quantities representative of a temperature of the cabinet 10. More particularly, the temperature sensors 61 to 65 are surface temperature sensors and are configured to measure values ​​of quantities representative of a temperature of a surface of the elements of the cabinet 10.

[0060] The temperature sensors 61 to 65 are here configured to respectively measure a temperature of the input terminal 12 by the first temperature sensor 61, of the first branch bus 18 by the second temperature sensor 62, of the main bus 17 by the third temperature sensor 63, of the first protection device 22 by the fourth temperature sensor 64 and of one of the second protection devices 23 by the fifth temperature sensor 65.

[0061] In a variant not shown, the distribution system 1 comprises additional temperature sensors, configured to measure a temperature of five of the following elements: the main bus 17, the second branch bus 19, the protection device 22 and the output terminals 13.

[0062] In a particularly advantageous manner, the distribution system 1 comprises temperature sensors measuring the temperature of each of the elements mentioned above, and furthermore, comprises temperature sensors measuring the temperature at different positions of the main bus 17 and of the first 18 and second 19 secondary buses, and at each protection device 23.

[0063] The vibration sensor 67 is configured to acquire a value of a magnitude representative of vibrations of the electrical cabinet 10. The vibrations of the electrical cabinet 10 are for example due to the passage of small animals such as rodents, or to poor fixing of the components of the electrical cabinet to each other. The vibration sensor 67 is for example an accelerometer or a gyroscope. In the example of [Fig.l], the vibration sensor 67 is placed in the plinth 27. Alternatively, the vibration sensor 67 is located on one of the following elements: the casing 11, the input terminal 12, the first branch bus 18, the main bus 17, the second branch bus 19, the electrical protection devices 22 and 23 and the output terminals 13; and is configured to measure the quantity representative of the vibrations in the vicinity of the corresponding element.

[0064] According to a variant not shown, the distribution system 1 comprises several vibration sensors located on all or part of the previously mentioned elements.

[0065] In a particularly advantageous manner, the distribution system 1 comprises vibration sensors measuring the vibrations of each of the elements mentioned above, and furthermore, comprises vibration sensors measuring the vibrations at different positions of the main bus 17 and of the first 18 and second 19 secondary buses.

[0066] The noise sensor 69 is configured to acquire a value of a quantity representative of noise inside the electrical cabinet 10. The noise inside the electrical cabinet 10 is for example caused by small animals such as rodents, by poor fixing of the elements of the electrical cabinet to each other, or by a foreign body introduced by mistake into the electrical cabinet, such as for example a tool forgotten during maintenance. The noise sensor 69 is for example a microphone. In the example of [Fig.l], the noise sensor 69 is placed in the drawer 25 and is configured to measure the quantity representative of the noise in the drawer 25.Alternatively, the noise sensor 69 is located on or in the vicinity of one of the following elements: the casing 11, the input terminal 12, the electrical protection devices 22 and 23, the output terminals 13 and the plinth 27; and is configured to measure the quantity representative of the noise in the vicinity of the corresponding element.

[0067] According to a variant not shown, the distribution system 1 comprises several noise sensors located on or in the vicinity of all or part of the previously mentioned elements, and are configured to measure the noise in the vicinity of all or part of the previously mentioned elements.

[0068] The distribution system 1 further comprises an electronic device 90 for assisting in monitoring the electrical cabinet 10, also called a monitoring assistance device 90. In the example of [Fig. 1], the monitoring assistance device 90 is an autonomous electronic device located inside the enclosure 11. In a variant not shown, the monitoring assistance device 90 is included in other electronic equipment, for example in a server, then preferably located outside the enclosure 11.

[0069] The monitoring assistance device 90 comprises an acquisition module 92, an estimation module 94 and a transmission module 96, visible in [Fig.2].

[0070] In the example of [Fig.2], the monitoring assistance device 90 comprises an information processing unit 100 formed for example of a memory 102 and a processor 104 associated with the memory 102.

[0071] In the example of [Fig.2], the acquisition module 92, the estimation module 94 and the transmission module 96 are each produced in the form of software, or a software brick, executable by the processor 104. The memory 102 of the monitoring assistance device 90 is then capable of storing acquisition software, estimation software and transmission software. The processor 104 is then capable of executing each of the software among the acquisition software, the estimation software and the transmission software.

[0072] In a variant not shown, the acquisition module 92, the estimation module 94 and the transmission module 96 are each produced in the form of a programmable logic component, such as an FPGA (Field Programmable Gate Array), or an integrated circuit, such as an ASIC (Application Specific Integrated Circuit).

[0073] When the monitoring assistance device 90 is produced in the form of one or more software programs, that is to say in the form of a computer program, also called a computer program product, it is furthermore capable of being recorded on a medium, not shown, that is readable by a computer. The computer-readable medium is, for example, a medium capable of storing electronic instructions and of being coupled to a bus of a computer system. For example, the readable medium is an optical disk, a magneto-optical disk, a ROM memory, a RAM memory, any type of non-volatile memory (for example FLASH or NVRAM) or a magnetic card. A computer program comprising software instructions is then stored on the readable medium.

[0074] The acquisition module 92 is configured to acquire the values ​​of the quantities measured by each of the intensity sensors 41 to 44, each of the voltage sensors 51 to 54, each of the gas sensors 56 and 57, each of the humidity sensors 58 and 59, each of the temperature sensors 61 to 64, the vibration sensor 67 and the noise sensor 69.

[0075] The estimation module 94 is configured to calculate an estimated state of the electrical cabinet 10 via an artificial neural network, or ANN (Artificial Neural Network), hereinafter called a neural network. For this, the estimation module 94 uses the values ​​acquired by the acquisition module 92, such that each quantity acquired by the acquisition module 92 is an input variable of the neural network. The estimated state is an output variable of the neural network.

[0076] The neural network comprises an ordered succession of layers of neurons each of which takes its inputs from the outputs of the previous layer. More precisely, each layer consists of neurons taking their inputs from the outputs of the neurons in the previous layer, or from the input variables for the first layer.

[0077] Alternatively, more complex neural network structures can be envisaged with a layer that can be connected to a layer further away than the immediately preceding layer.

[0078] Each neuron is also associated with an operation, i.e. a type of processing, to be carried out by said neuron within the corresponding processing layer.

[0079] Each layer is connected to the other layers by a plurality of synapses. A synaptic weight is associated with each synapse, and each synapse forms a connection between two neurons. It is often a real number, which takes both positive and negative values. In some cases, the synaptic weight is a complex number.

[0080] Each neuron is capable of performing a weighted sum of the value(s) received from the neurons of the previous layer, each value then being multiplied by the respective synaptic weight of each synapse, or link, between said neuron and the neurons of the previous layer, then applying an activation function, typically a non-linear function, to said weighted sum, and delivering at the output of said neuron, in particular to the neurons of the following layer which are connected to it, the value resulting from the application of the activation function. The activation function makes it possible to introduce a non-linearity into the processing carried out by each neuron. The sigmoid function, the hyperbolic tangent function, the Heaviside function are examples of activation functions.

[0081] As an optional addition, each neuron is also capable of applying, in addition, a multiplicative factor, also called bias, to the output of the activation function, and the value delivered at the output of said neuron is then the product of the bias value and the value from the activation function.

[0082] Alternatively, the neural network is a convolutional neural network. The convolutional neural network is also sometimes called a convolutional neural network or by the acronym CNN which refers to the English term “Convolutional Neural Networks”.

[0083] In a convolutional neural network, each neuron in the same layer has exactly the same connection pattern as its neighboring neurons, but at different input positions. The connection pattern is called a convolution kernel or, more often, a "kernel" in reference to the corresponding English term.

[0084] A fully connected layer of neurons is a layer in which the neurons of said layer are each connected to all the neurons of the layer previous.

[0085] Such a type of layer is more often referred to as “fully connected” and sometimes referred to as “dense layer”.

[0086] The estimated state of the electrical cabinet 10 is advantageously a state from the group consisting of a normal state, an abnormal state and a critical state. The normal state is associated with normal operation of the electrical cabinet 10, that is to say that the neural network estimates that no electrical, mechanical or thermal fault is present in the electrical cabinet 10. The abnormal state is associated with abnormal operation of the electrical cabinet 10, that is to say that the neural network estimates that an electrical, mechanical or thermal fault is present, but that it does not pose a problem requiring immediate action on the part of the user. An abnormal state is for example the detection of noise in the electrical cabinet 10, without modification of the other quantities measured by the sensors.The critical state is associated with a critical fault in the electrical cabinet 10, i.e. the neural network estimates that an electrical, mechanical or thermal fault is present and that it poses a major risk to the operation of the electrical cabinet, requiring for example an emergency maintenance operation, and / or a cut-off of the electrical current flowing in the electrical cabinet 10, i.e. a shutdown of the electrical cabinet 10.

[0087] In other words, the normal state, the abnormal state and the critical state correspond to three successive levels of criticality, the normal state being associated with a first level, the abnormal state being associated with a second level higher than the first level, and finally the critical state being associated with a third level higher than the second level. These states allow a user to estimate whether it is necessary to intervene on the electrical cabinet 10 for a maintenance operation, and the degree of urgency of the maintenance operation.

[0088] The neural network is trained, during a preliminary or prior training phase, so that the neural network can then correctly predict the estimated state during the inference phase, subsequent to this preliminary learning phase. The learning is carried out using training data corresponding to both the input variables and the output variable(s) of the neural network. For example, the training data are examples of sensor measurements, and the estimated state of the electrical cabinet associated with these measurements.An error calculation between the predicted output variable (by the neural network from the input training data) and the corresponding output training data is performed, then an adjustment of the synaptic weights of the neural network according to the calculated error is performed, for example via a gradient backpropagation algorithm and with reiteration of the above in order to minimize the error, i.e. to progressively reduce the error until reaching a minimal threshold. predefined.

[0089] The learning phase is then completed when the error has been reduced so as to reach this threshold. The synaptic weights and any biases used for the inference of the neural network are then those obtained at the end of this learning phase, i.e. corresponding to this minimal error between the predicted output variable and the corresponding output learning data.

[0090] The transmission module 96 is configured to transmit a message representative of the estimated state in one of a visual form and an audible form. The transmission module 96 is for example a human-machine interface, or is connected to a human-machine interface, to display a message representative of the estimated state of the electrical cabinet 10. The message is for example a color code, a color being associated with an estimated state, or an audible message, in particular an alarm if the estimation module 94 calculates that the estimated state is a critical state. Those skilled in the art will understand that each estimated state is associated with a respective visual or audible restitution, and distinct from one estimated state to the other, so that the user can more easily and more quickly perceive which state of the electrical cabinet 10 has been estimated by the monitoring assistance device 90. In other words, each estimated state is associated with a respective color and / or a respective audible message..

[0091] Optionally, the transmission module 96 is configured to transmit additional information relating to the sensors, for example the values ​​of the quantities measured by the sensors.

[0092] In a particularly advantageous manner, the estimation module 94 is configured to estimate the state of the electrical cabinet 10 for several distinct risks, associated with the electrical cabinet 10. A first risk is for example a defect of the electrical cabinet 10, whether electrical, mechanical or thermal; and a second risk is for example a risk of electric arc in the electrical cabinet 10. Advantageously, other risks are estimated by the estimation module 94, such as the risk of the presence of small animals or rodents inside the electrical cabinet 10, the risk of the presence of water inside the electrical cabinet 10, etc.

[0093] In the case where the estimation module 94 is configured to estimate the state of the electrical cabinet 10 for several risks, the estimation module 94 calculates an estimated state for each of the risks. In this case, the output variables of the neural network are an estimated state of the electrical cabinet 10 for each risk, and the transmission module 96 is then configured to transmit a message representative of the estimated state for each risk.

[0094] Advantageously, the estimation module 94 is configured to generate a resulting estimated state equal to the estimated state for the greatest risk at the time the estimation is carried out. The greatest risk is, for example, that having a pro highest occurrence capacity. Alternatively or additionally, the greatest risk is the one with the highest criticality level according to a predefined order of criticality between risks, while having a probability of occurrence greater than a predefined minimum threshold. The transmission module 96 is then configured to transmit a message representative of the resulting estimated state, or even to transmit additional information linked to the risk which is the greatest at the time of the estimation.

[0095] Alternatively, the estimation module 94 calculates an estimated state of the electrical cabinet 10 only for the greatest risk. In this case, the output variable of the neural network is the estimated state of the electrical cabinet 10 for the greatest risk, and the transmission module 96 is then configured to transmit a message representative of the estimated state for said risk.

[0096] In order to best estimate the various risks, certain sensors are advantageously added, or possibly removed. In particular, in the case of an estimation of the risk of electric arc, the vibration 67 and noise 69 sensors are particularly advantageous in order to obtain the most precise and accurate prediction possible, whereas in the case of an estimation of the risk of electrical, mechanical or thermal fault of the electrical cabinet, the vibration 67 and noise 69 sensors can be omitted without this having a significant impact on the accuracy of the calculation carried out by the estimation module 94.

[0097] The risk(s) taken into account for the calculation of the estimated state by the estimation module 94 are advantageously chosen by the user, the sensors being added according to the risk(s) that the user wishes the estimation module 94 to evaluate. Alternatively, the sensors necessary for the calculation of the estimated state for the different risks are all present in the electrical cabinet 10, and the sensors are then activated by the monitoring assistance device 90 according to the risk(s) chosen by the user.

[0098] A method for assisting in monitoring the electrical cabinet 10, implemented by the monitoring assistance device 90, is now described.

[0099] The electrical cabinet 10 is in operation, thus electric current flows through the electrical cabinet 10 from the source 3 to the loads 5. The current sensors 41 to 44, the voltage sensors 51 to 54, the gas sensors 56 and 57, the humidity sensors 58 and 59, the temperature sensors 61 to 64, the vibration sensor 67 and the noise sensor 69 are in operation, is able to carry out measurements according to the risk for which the estimated state must be calculated.

[0100] During a step 201, the monitoring assistance device 90, and more precisely the acquisition module 92, acquires the values ​​of the quantities measured by the intensity sensors 41 to 44, the voltage sensors 51 to 54, the gas sensors 56 and 57, the humidity sensors 58 and 59, temperature sensors 61 to 64, vibration sensor 67 and noise sensor 69.

[0101] These values ​​are transmitted to the estimation module 94, which during a following step 202 calculates, via the neural network, the estimated state of the electrical cabinet 10. As mentioned previously, each quantity measured by the intensity sensors 41 to 44, the voltage sensors 51 to 54, the gas sensors 56 and 57, the humidity sensors 58 and 59, the temperature sensors 61 to 64, the vibration sensor 67 and the noise sensor 69 is an input variable of the neural network, and the output variable of the neural network is the estimated state.

[0102] During a step 203 subsequent to the calculation state 202, the transmission module 96 transmits, either in visual form, or in sound form, or both in visual form and in sound form, a message representative of the state estimated during the calculation step 202. Optionally, the transmission module 96 also transmits additional data, such as the values ​​measured by the sensors.

[0103] Alternatively, during step 202, the estimation module 94 calculates an estimated state for each risk. In this case, during step 203, the transmission module 96 transmits a message representative of the estimated state for each risk.

[0104] According to another variant, if the estimation module 94 is configured to estimate several risks associated with the electrical cabinet, then during step 202, the estimation module 94 generates a resulting estimated state equal to the estimated state for the greatest risk at the time when the estimation is carried out, in addition to the calculation of the estimated state at step 202. In this case, at step 203, the transmission module 96 transmits a message representative of the resulting estimated state, or even additional information linked to the type of risk which is the greatest at the time of the estimation.

[0105] According to a variant not shown, the estimation module 94 is advantageously further configured to control the disconnector 16 in the open position, corresponding to an absence of current flow in the electrical cabinet 10, when the estimated state is equal to the critical state.

[0106] Thus, the monitoring assistance device 90 assists in monitoring the electrical cabinet 10. The possibility of evaluating certain specific risks, such as the risk of electric arc, by adding specific sensors makes it possible to personalize the monitoring of the electrical cabinet 10 in accordance with the user's preferences. The transmission of a message representative of the estimated state facilitates the user's decision-making. Depending on the message transmitted by the transmission module 96, the user decides on the action to be implemented, for example a maintenance operation and a date on which it is carried out, and / or a cut-off of the electric current flowing in the electrical cabinet 10.

[0107] Any feature described for an embodiment or variation in this The foregoing may be implemented for the other embodiments and variants described above, as long as technically feasible.

Claims

Claims

1. Electronic monitoring assistance device (90) for an electrical cabinet (10), the electrical cabinet (10) comprising an input terminal (12) and a plurality of output terminals (13), and being capable of being connected to a source (3) via the input terminal (12) and to a plurality of loads (5) via the output terminals (13) and configured to distribute an electric current from the source (3) to the plurality of loads (5), the electrical cabinet (10) comprising a disconnector (16) connected to the input terminal (12) and a main bus (17) connected between the disconnector (16) and the output terminals (13), the device (90) comprising: - an acquisition module (92), configured to acquire values ​​of quantities representative of an intensity and respectively of a voltage of an electric current flowing between the input terminal (12) and the output terminals (13), measured by an intensity sensor (41, 42, 43, 44) and respectively a voltage sensor (51, 52, 53, 54); - an estimation module (94), configured to calculate an estimated state of the electrical cabinet (10) via a neural network, each quantity acquired by the acquisition module (92) being an input variable of the neural network, the estimated state of the electrical cabinet (10) being an output variable of the neural network; and - a transmission module (96), configured to transmit a message representative of the estimated state in one of a visual form and an audio form; characterized in that the acquisition module is further configured to acquire a value of at least one additional quantity measured by at least one additional sensor among a gas sensor (56, 57) and a humidity sensor (58, 59), the or each quantity being representative respectively of a quantity of gas particles and of a humidity inside the electrical cabinet (10), and in that the estimation module (94) is then configured to calculate the estimated state of the electrical cabinet (10) via the neural network as a function further of the at least one additional quantity, the at least one additional quantity being an additional input variable of the neural network neurons.

2. Device (90) according to claim 1 wherein the acquisition module (92) is further configured to acquire the values ​​of the additional quantities measured by the humidity sensor (56, 57) and by the gas sensor (58, 59).

3. Device (90) according to any one of the preceding claims, wherein the acquisition module (92) is further configured to acquire a value of a quantity measured by a temperature sensor (61, 62, 63, 64, 65), the measured quantity being representative of a temperature inside the electrical cabinet (10).

4. Device (90) according to any one of the preceding claims, wherein the acquisition module (92) is further configured to acquire values ​​of quantities measured by a plurality of temperature sensors (61, 62, 63, 64, 65), the quantities being representative of the temperature inside the electrical cabinet (10), each temperature sensor (61, 62, 63, 64, 65) being configured to measure one of: a temperature in the vicinity of the input terminal (12), a temperature in the vicinity of a respective output terminal (13), a temperature of the main bus (17) and a temperature of an electrical protection device (22, 23), connected between the main bus (17) and a respective output terminal (13).

5. Device (90) according to any one of the preceding claims, in which the acquisition module (92) is further configured to acquire a value of a quantity representative of vibrations of the electrical cabinet (10), measured by a vibration sensor (67).

6. Device (90) according to any one of the preceding claims, in which the acquisition module (92) is further configured to acquire a value of a quantity representative of noise in the electrical cabinet (10), measured by a noise sensor (69).

7. A device (90) according to any preceding claim, wherein the estimated state of the electrical cabinet (10) is one of the group consisting of: a normal state, an abnormal state, and a critical state.

8. System (1) for distributing an electric current between a source (3) and a plurality of loads (5), the system (1) comprising: - an electrical cabinet (10), comprising: • an input terminal (12), capable of being connected to the source (3),

9. • several output terminals (13), capable of being connected to the plurality of loads (5), • a disconnector (16), connected to the input terminal (12), and • a main bus (17), connected between the disconnector (16) and the plurality of output terminals (13), the electrical cabinet (10) being configured to distribute the electric current from the source (3) to the plurality of loads (5); - an intensity sensor (41, 42, 43, 44) and respectively a voltage sensor (51, 52, 53, 54), configured to measure values ​​of quantities representative of an intensity and respectively of a voltage of an electric current flowing between the input terminal (12) and the output terminals (13) of the electrical cabinet (10); - an electronic monitoring assistance device (90) for the electrical cabinet (10), characterized in that the system (1) further comprises at least one additional sensor among a gas sensor (56, 57) and a humidity sensor (58, 59), configured to acquire a value of at least one additional quantity, representative respectively of a quantity of gas particles and of a humidity inside the electrical cabinet (10) and in that the electronic monitoring assistance device (90) is according to any one of the preceding claims. A method for assisting in monitoring an electrical cabinet (10), the electrical cabinet (10) comprising an input terminal (12) and a plurality of output terminals (13), and being capable of being connected to a source (3) via the input terminal (12) and to a plurality of loads (5) via the output terminals (13) and configured to distribute an electric current from the source (3) to the plurality of loads (5), the electrical cabinet (10) comprising a disconnector (16) connected to the input terminal (12) and a main bus (17) connected between the disconnector (16) and the output terminals (13), the method being implemented by an electronic monitoring assistance device (90), the method comprising the following steps: acquisition (201) of values ​​of quantities measured by a intensity sensor (41, 42, 43, 44) and respectively a voltage sensor (51, 52, 53, 54), the quantities being representative of an intensity and respectively of a voltage of the electric current flowing between the input terminal (12) and the output terminals (13) of the electrical cabinet; - calculation (202), via a neural network, of an estimated state of the electrical cabinet (10), each measured quantity being an input variable of the neural network, the estimated state being an output variable of the neural network; and - transmission (203), in one form among a visual form and an audio form, of a message representative of the estimated state, characterized in that the method further comprises a step of acquiring (201) a value of at least one additional quantity measured by at least one additional sensor among a gas sensor (56, 57) and a humidity sensor (58, 59), the or each quantity being representative respectively of a humidity and a quantity of gas particles inside the electrical cabinet (10), and in that during the calculation step (202) the at least one additional quantity is an additional input variable of the neural network.

10. A computer program comprising software instructions which, when executed by a computer, implement a method according to claim 10.

Images