Computing unit and associated hue determination apparatus, fluid dispensing system and dispensing method

Abstract

The invention discloses a computing unit and an associated hue determination apparatus, a fluid dispensing system and a dispensing method. The invention comprises a computing unit (50) of a device (14) comprising:-at least one light source (22) configured to emit light towards a fluid; -a light sensor (24) configured to receive a light signal representative of a hue of a fluid flowing in the fluid distribution system (10), the light sensor (24) further configured to emit a parameter representative of a light intensity of the light signal for a given color channel; the computing unit includes an estimation module (52) configured to determine a variable representative of a hue of the fluid via an artificial intelligence model, each parameter representative of the light intensity being an input variable of the model, an output variable of the artificial intelligence model being a variable representative of the hue of the fluid.

Description

Technical Field

[0001] The present invention relates to a computing unit, and associated hue determination device, fluid distribution system and distribution method. Background Technology

[0002] In the context of coating product application installation, it is known to determine the hue of a fluid (such as paint or cleaning liquid) circulating in a fluid distribution system to determine whether the correct paint hue is circulating in the system, or whether the cleaning liquid is pure, in order to understand the cleanliness of the fluid distribution system. For example, during a hue-changing operation, it is known to perform a visual inspection of the fluid sprayed into the distribution system to verify that the hue-changing operation has been completed and that the new paint hue is circulating in the system without mixing with the old paint.

[0003] For the example from FR3127281A1, it is also known to use an optical sensor to measure the hue of a fluid. For this purpose, it is known to use an optical fiber to transmit an optical signal to the optical sensor, which represents the hue of the fluid. However, the use of an optical fiber causes attenuation of the optical signal. The value of the attenuation depends on the length of the optical fiber and the wavelength of the optical signal. Therefore, it is necessary to manually calibrate the optical sensor to account for the attenuation caused by the optical fiber in order to correctly determine the hue of the fluid, which is a lengthy and complex process. Summary of the Invention

[0004] Therefore, the object of the present invention is to overcome these disadvantages and to determine the hue of a fluid more simply and reliably.

[0005] Therefore, the present invention relates to a computing unit for a hue determination device for a fluid distribution system, the device comprising:

[0006] - At least one light source, which is configured to emit polychromatic or monochromatic light toward the fluid in the measurement area;

[0007] - An optical sensor configured to receive light signals that have been reflected or transmitted through a fluid, the light signals corresponding respectively to light reflection or transmission by the fluid of polychromatic or monochromatic light emitted toward the fluid, the light signals representing the hue of the fluid flowing in the fluid distribution system; the optical sensor is also configured with emission parameters, each parameter representing the light intensity of the light signal for a given color channel; and

[0008] - Acquisition fiber, which is configured to transmit optical signals to an optical sensor;

[0009] The computing unit includes an estimation module configured to receive parameters emitted by the optical sensor.

[0010] According to the invention, the estimation module is further configured to determine variables representing the hue of the fluid via a pre-trained artificial intelligence model, wherein each parameter representing the light intensity of a light signal for a given color channel is an input variable of the model, and the output variable of the artificial intelligence model is a variable representing the hue of the fluid.

[0011] Among them, the variable representing the hue of the fluid is:

[0012] - An index of the purity of the fluid flowing in a fluid distribution system; or

[0013] - Category indicators among multiple categories, where each category corresponds to a predefined hue of the fluid.

[0014] With this invention, determining the hue of a fluid does not require manual calibration of the light sensor. Furthermore, the use of an artificial intelligence model enables reliable results to be obtained without requiring additional optical signal processing steps and by limiting the complexity of the computations performed (which consume resources and time). Determining the hue of the fluid is therefore simpler for the user.

[0015] According to other advantageous aspects of the invention, the computing unit includes one or more of the following features, either individually or in all technically possible combinations:

[0016] - The purity index is a decimal number, preferably between 0 and 1;

[0017] - The purity index is a binary number.

[0018] The purity index is preferably a category indicator between a first category corresponding to clean fluid conditions and a second category corresponding to dirty fluid conditions;

[0019] - The fluid is a coating liquid or cleaning liquid or an aerosol composed of solid colored particles suspended in a gas.

[0020] The present invention also relates to a hue determination device for a fluid distribution system, the device comprising:

[0021] - At least one light source, which is configured to emit polychromatic or monochromatic light toward the fluid in the measurement area;

[0022] - An optical sensor configured to receive light signals that have been reflected or transmitted through a fluid, the light signals corresponding to light reflection or transmission by the fluid toward a polychromatic or monochromatic light emitted toward the fluid, respectively, the light signals representing the hue of the fluid flowing in the fluid distribution system, the optical sensor also configured to emit parameters, each parameter representing the light intensity of the light signal for a given color channel;

[0023] - A data acquisition fiber optic cable configured to transmit optical signals to an optical sensor; and

[0024] - The computing unit as previously described.

[0025] Advantageously, given color channels as red channel, green channel, blue channel and clear channel, the parameters represent the light intensity of the light signal of the red channel, green channel, blue channel and clear channel, respectively.

[0026] The present invention also relates to a fluid distribution system comprising a fluid flow loop and a hue determination device as previously described, connected in series to the fluid flow loop.

[0027] The present invention also relates to a method for determining the hue of a fluid flowing in a fluid distribution system, the method being implemented by the previously described hue determination device, the method comprising the following steps:

[0028] -Emitting polychromatic or monochromatic light toward the fluid from at least one light source;

[0029] - The optical sensor receives light signals that are reflected or transmitted through the fluid;

[0030] - The estimation module receives parameters representing the light intensity of the light signal for a given color channel; and

[0031] - The estimation module determines the variables representing the hue of the fluid.

[0032] According to other advantageous aspects of the invention, the hue determination method includes one or more of the following features, either individually or according to all technically possible combinations:

[0033] -The method also includes the following steps:

[0034] - A fluid change operation is detected when the variable representing the hue of the fluid is modified;

[0035] as well as

[0036] -Once a fluid change operation has been detected, the fluid change operation is determined to have ended when the variable representing the fluid's hue remains substantially constant over a predetermined duration.

[0037] The fluid alteration operation is preferably a system flushing or a color change of the fluid flowing in the system;

[0038] - The fluid is a coating liquid or a cleaning liquid. Attached Figure Description

[0039] The invention will become clearer after reading the following description, which is given only as a non-limiting example and with reference to the accompanying drawings, wherein:

[0040] -[ Figure 1 ] Figure 1 This is a schematic diagram of the fluid distribution system according to the present invention; and

[0041] -[ Figure 2 ] Figure 2 yes Figure 1 The fluid distribution system is a perspective view and cross-sectional view corresponding to a portion of detail A in this figure;

[0042] -[ Figure 3 ] Figure 3 This is a flowchart of the method according to the present invention.

[0043] -[ Figure 4 ] Figure 4 The fluid distribution system according to the second embodiment of the present invention corresponds to Figure 1 A perspective view of a portion of detail A. Detailed Implementation

[0044] The invention is described below in the context of a coating fluid distribution system (e.g., for paints) used in the application and installation of coated products (particularly by spraying). The paint is presented in the form of, for example, a liquid or aerosol, which consists of, for example, solid colored particles suspended in a gas. However, this does not constitute a limitation of the invention to that particular application; the invention can be implemented in any fluid distribution system suitable for conveying fluids, where identifying the hue of the fluid is advantageous, particularly for monitoring operations or phenomena involving changes in the hue of the fluid. The fluid can be liquid or gaseous, and specifically in the following description, the term fluid includes solid particles carried by a gaseous flow or suspended in a gas (particularly an air flow). Solid particles are, for example, colored powders.

[0045] In the following description, the expression "basically equal to" is defined as an equality relationship of plus or minus 10%, more preferably an equality relationship of plus or minus 5%.

[0046] Figure 1 This is a schematic diagram of a fluid distribution system 10. The fluid distribution system 10 is, for example, a paint distribution system, and includes at least one paint component in a reservoir, a projector 11 (also referred to as a sprayer), and a fluid flow loop 12.

[0047] exist Figure 1In this example, the fluid distribution system 10 includes three reservoirs 1, 2, and 3 for different paint hues, and a reservoir 4 for a cleaning liquid (e.g., an aqueous solution, also referred to as a solvent). Reservoirs 1 to 4 are grouped in a dedicated chamber 6, typically referred to as the paint chamber. There are no restrictions on the number and distribution of reservoirs 1 to 4 within the paint chamber. These reservoirs are adjusted according to the intended use of the fluid distribution system 10.

[0048] The color-changing device 8 is supplied by different reservoirs 1 to 4 and enables the selection of which product, paint, or cleaning liquid flows in the fluid flow circuit 12 to supply the injector 11.

[0049] The sprayer can be, for example, a manually operated spray gun, such as... Figure 1 As shown in the figure. In a variant not shown, it is a pneumatic or rotary automatic injector mounted on a robot's multi-axis or reciprocating arm.

[0050] Advantageously, the injector 11 can be of the electrostatic type.

[0051] The fluid flow circuit 12 includes pipes 122 and 124, and fluidly connects the hue-changing device 8 (and thus one of the reservoirs 1 to 4) to the injector 11 via pipes 122 and 124.

[0052] The fluid distribution system 10 also includes a hue determination device 14.

[0053] A portion of the hue determination device 14 is connected in series to the fluid flow loop 12, located between pipes 122 and 124.

[0054] Advantageously, the hue determining device 14 includes a flange 13 (advantageously opaque) that defines a chamber 15 (in Figure 2 (As can be seen in the image), fluid flow vein 16 is arranged through this chamber. Flange 13 is arranged to connect to either side of the pipe in the fluid flow loop 12. Fluid flow vein 16 in Figure 2 The arrow indicates the direction.

[0055] The hue determination device 14 includes at least one light source 22. It also includes a light sensor 24.

[0056] exist Figure 1In the example, the hue determination device 10 includes only one light source 22. The light source 22 is configured to emit polychromatic light (i.e., including at least two different wavelengths) or monochromatic light (i.e., including a single wavelength). The polychromatic light preferably consists of white light. The light source 22 is formed, for example, by several monochromatic sources of different corresponding wavelengths, such as a red laser source, a green laser source, and a blue laser source, or several light-emitting diodes or LEDs of different colors. Alternatively, the light source 22 is a light source emitting a continuous spectrum, such as an incandescent lamp. Furthermore, one or more light sources may also be configured to emit pulsed light.

[0057] Alternatively, when the light source 22 is configured to emit monochromatic light, the light source 22 may be formed, for example, by one or more monochromatic sources (e.g., lasers or monochromatic LEDs) of the same wavelength.

[0058] Light source 22 is configured to face towards Figure 2 The fluid (here, the paint) in the measurement area 23, indicated by dashed lines and formed within chamber 15, emits light. Specifically, the light illuminates the fluid present within the flow veins 16 in the measurement area 23.

[0059] The light sensor 24 is configured to receive light signals and emit parameters, each representing the light intensity of a light signal in a given color channel. The given color channels are, for example, the red, green, and blue channels. Alternatively, the given color channels may be the red, green, blue, and alpha channels, where the alpha channel represents the light intensity of the light signal as a whole. In this case, the parameters emitted by the light sensor 24 represent the light intensities of the light signals in the red, green, blue, and alpha channels, respectively. In other words, the light sensor 24 is configured to emit parameters representing the light intensity in an RVBC or red-green-blue alpha system.

[0060] Light intensity is represented, for example and in a known manner, by a number between 0 and 255, where 0 corresponds to zero intensity of the channel under consideration and 255 corresponds to maximum intensity of the channel under consideration.

[0061] The light sensor 24 and / or light source 22 can be positioned near the measurement area 23. However, the quality of the dispensed fluid or its environment may often include constraints on compliance with European ATEX regulations (for “ATmosphère explosive”) for use in explosive atmospheres, as in the context of a paint dispensing system. However, using a light sensor 24 and / or light source 22 that complies with these constraints may prove expensive, and integrating them into a tone measurement device may be complex or even impossible.

[0062] Therefore, the light source 22 and the light sensor 24 are preferably placed at a distance from the fluid flow loop 12, for example, by grouping them in the processing box 26, which also belongs to the hue determination device 14.

[0063] To ensure that the light emitted by the light source 22 illuminates the fluid in the measurement area 23, the device 14 includes one or more optical fibers 34, referred to as illumination fibers. Advantageously, the number of illumination fibers 34 is four, such as... Figure 2 Visible in the image. The illumination fiber 34 is typically several meters long, for example at least ten meters, and preferably, these fibers are up to twenty-five meters long. The illumination fiber 34 is advantageously connected to the light source 22 at one end of its end. The other end of the illumination fiber 34 is advantageously located in the flange 13.

[0064] In a variant not shown, when the device 10 includes several light sources 22, each illumination fiber 34 is connected to one of the light sources 22.

[0065] The device 14 also includes a light-gathering fiber 36, which is advantageously connected at one end to the light sensor 24 and at the other end in the flange 13. Thus, the optical signal from the measurement area 23 is transmitted to the light sensor 24 via the fiber 36. The length of the light-gathering fiber 36 is the same as the length of the illumination fiber 34.

[0066] Advantageously, components (such as cable seals 38) for holding one or more ends of the illumination fiber 34 and the acquisition fiber 36 are arranged in the flange 13.

[0067] Advantageously, the device 14 also includes an observation window 40, which is arranged tangentially to the fluid flow vein 16 within the flange 13. In particular, the observation window 40 may have a porthole shape, such as... Figure 2 Visible in. In Figure 2 In the example, measurement area 23 corresponds to the portion of fluid flow vein 16 visible through observation window 40.

[0068] The device 14 advantageously includes a retaining element 42 mounted in the flange 13 to hold the observation window 40 in position and ensure a seal for the fluid flow vein 16. The retaining element 42 may, for example, include a sealing gasket surrounding the observation window 40.

[0069] Device 14 also includes a computing unit 50. The computing unit 50 is connected to the light source 22 and the light sensor 24, and is formed, for example, by a memory and a processor (not shown) associated with the memory. The computing unit 50 is advantageously located inside the processing box 26.

[0070] The calculation unit 50 includes an estimation module 52, such as... Figure 1 Visible in. In Figure 1 In the example, the estimation module 52 is implemented as software or a software block executable by the processor. The memory of the computing unit 50 can then store the estimation software, and the processor can then execute the estimation software.

[0071] The estimation module 52 is configured to receive parameters emitted by the light sensor 24 and then determine variables representing the hue of the fluid via a pre-trained artificial intelligence model. This model has each parameter representing the light intensity of a light signal for a given color channel as input variables, and the variable representing the hue of the fluid as an output variable. The variable representing the hue of the fluid represents the color of the fluid, or alternatively or additionally, the opacity of the fluid. Specifically, when the light emitted by the light source 22 is monochromatic, the variable representing the hue of the fluid represents the opacity of the fluid. When the light emitted by the light source 22 is polychromatic, the variable representing the hue of the fluid represents the opacity and / or hue of the fluid.

[0072] A variable representing the hue of a fluid is, for example, the purity index of the fluid flowing in the fluid distribution system 10. According to this example, the purity index is a decimal number, typically between 0 and 1; or a binary number, typically equal to 0 or 1. When the purity index is a binary number, it advantageously serves as a category indicator between a first category when fluid 10 is clean and a second category when fluid 10 is dirty.

[0073] According to the example, when the purity index is a decimal number, it is directly proportional to the opacity of the fluid.

[0074] Alternatively, the variable representing the hue of a fluid is a category indicator among multiple categories, each corresponding to a predetermined hue of the fluid.

[0075] In a variant not shown, the estimation module 52 is implemented as a programmable logic component, such as an FPGA (Field Programmable Gate Array), or an integrated circuit, such as an ASIC (Application-Specific Integrated Circuit).

[0076] When the computing unit 50 is implemented as one or more software programs (i.e., implemented as a computer program, also referred to as a computer program product), the computing unit can also be recorded on a computer-readable medium (not shown). A computer-readable medium is, for example, a medium capable of storing electronic instructions and coupled to a computer system bus. Examples of readable media include optical discs, magneto-optical discs, ROM memory, RAM memory, any type of non-volatile memory (e.g., FLASH or NVRAM), or magnetic cards. The computer program, including the software instructions, is then stored on the readable medium.

[0077] Device 14 advantageously includes a human-machine interface 54 connected to computing unit 50 to display information for user attention. Alternatively, or additionally, the human-machine interface 54 is used by the user to control device 14. The human-machine interface 54 typically includes a display screen, and optionally includes input devices such as a keyboard and mouse (not shown).

[0078] Artificial intelligence models are pre-trained, advantageously, through machine learning, for example, via supervised learning, as is known in itself.

[0079] Training an artificial intelligence model is typically performed from training datasets, each consisting of input training data corresponding to the model's input variables (i.e., parameters representing the light intensity of a given color channel) and output training data corresponding to representative variables of the expected fluid hue (i.e., the target) from this input training data. Model training is then performed, for example, via backpropagation of error gradients, which are calculated from the difference between the output training data and the output variables estimated by the model from the input training data; that is, from the difference between the target's representative variable from the input training data and the estimated representative variable.

[0080] Artificial intelligence models include random forests, support vector machines or SVM models, neural networks, or ANNs (artificial neural networks).

[0081] In the case of neural networks, the model consists of ordered, continuous layers of neurons, each receiving its input from the output of the previous layer.

[0082] More precisely, each layer comprises neurons that receive input from the output of neurons in previous layers or from the input variables of the first layer.

[0083] Alternatively, more complex neural network structures can be considered, where layers can connect to layers further away than the immediate preceding layer.

[0084] Each neuron is also associated with an operation, that is, the type of processing to be performed by the neuron in the corresponding processing layer.

[0085] Each layer is connected to other layers through multiple synapses. A synaptic weight is associated with each synapse, and each synapse forms a connection between two neurons. It is typically a real number with both positive and negative values. In some cases, the synaptic weight is a complex number.

[0086] Each neuron can perform a weighted summation of one or more values ​​received from neurons in previous layers. Each value is multiplied by the corresponding synaptic weight of each synapse or connection between that neuron and neurons in previous layers. An activation function (typically a non-linear function) is then applied to the weighted sum, and the value obtained from the application of the activation function is passed to the neuron's output, specifically to the neurons to which that neuron is connected in the next layer. Activation functions allow the introduction of non-linearity into the processing performed by each neuron. The sigmoid function, hyperbolic tangent function, and heaviside function are examples of activation functions.

[0087] Optionally, each neuron can also additionally apply a multiplication factor (also known as a bias) to the output of the activation function, and the value passed at the output of the neuron is the product of the bias value and the value from the activation function.

[0088] A method for determining the hue of a fluid flowing in a flow loop 12 is now described. This method is implemented by device 14.

[0089] A fluid (e.g., a coating or aerosol) flows in the distribution system 10, specifically in the fluid flow veins 16 of the device 14. The fluid is colored, opaque, or may be translucent, or even transparent.

[0090] During emission step 100, the light source 22 emits polychromatic or monochromatic light toward the fluid in the measurement area 23. Specifically, the light is emitted through the illumination optical fibers 34 to their ends located in the flange 13. Thus, the fluid flowing in the measurement area 23 is illuminated by light passing through the observation window 40.

[0091] The light reflected by the fluid forms an optical signal, which corresponds to the light reflected by the fluid from the light source 22. The optical signal represents the hue of the fluid.

[0092] The optical signal flows from measurement region 23 through acquisition fiber 36 to optical sensor 24, which receives the optical signal during receiving step 102. The optical signal received by optical sensor 24 is attenuated due to optical fibers 34 and 36, which attenuate depending on their length and the wavelength of the light they compose, and also on the optical signal itself. More specifically, the light attenuates as it flows through illumination fiber 34, and the optical signal attenuates as it flows through acquisition fiber 36.

[0093] The opaque flange 13 advantageously enables strong limitation of any possible light pollution from the external environment, and thus ensures that the light signal received by the light sensor 24 during receiving step 102 is as reliable as possible, with minimal external light pollution.

[0094] The light sensor 24 converts the received light signal into parameters, each representing the light intensity of the light signal for a given color channel. In this example, the light sensor 24 converts the light signal into four numbers between 0 and 255, corresponding to the light intensity of each channel in the RVBC system.

[0095] The calculation unit 50 (specifically, the estimation module 52) receives parameters during the receiving step 104.

[0096] During step 106, estimation module 52 determines variables representing the hue of the fluid via an artificial intelligence model. Each parameter representing the light intensity of a light signal for a given color channel is a corresponding input variable of the artificial intelligence model. The output variable of the artificial intelligence model is the variable representing the hue of the fluid.

[0097] Advantageously, after the estimation module 52 determines the variable representing the hue of the fluid, during display step 108, a message representing the variable representing the hue of the fluid is displayed on the human-machine interface 54 in visual and / or possibly audible form. Alternatively, during display step 108, the calculation unit 50 sends the message representing the variable representing the hue of the fluid to a remote terminal, such as a user's computer or mobile phone, which then constitutes the human-machine interface 54.

[0098] Advantageously, the method is implemented by device 14 to perform any one of the following three tasks.

[0099] The first task is to estimate the purity of the fluid flowing in the fluid distribution system 10. This task is performed, for example, by a cleaning liquid during the flushing of system 10. In this case, the variable representing the hue of the fluid is a fluid purity index, and this index is in decimal form. For example, the fluid purity index is a decimal number between 0 and 1 or between 0 and 100. The fluid purity index corresponds to the purity of the fluid flowing in system 10. During the flushing operation, system 10 is initially dirty, and the variable representing the hue of the fluid will, for example, be essentially equal to zero. Then, as flushing proceeds, if the hue determination method is repeated over time, the value of the variable representing the hue of the fluid increases, for example, becoming essentially equal to 1 or 100 when the fluid is pure, corresponding to a completely clean system 10.

[0100] The second task is a binary classification task of fluid purity. In this case, the variable representing the fluid's hue is a fluid purity index, which is in binary form. Preferably, the purity index is a category indicator between a first category corresponding to a clean fluid and a second category corresponding to a dirty fluid. For example, this task is performed during the rinsing of system 10 to determine whether system 10 is clean.

[0101] Specifically, the fluid purity index enables the determination of whether system 10 is clean or dirty, for example, during a flushing operation. In fact, if the fluid is considered dirty, in other words, if the purity index determined by the model indicates a category corresponding to dirty fluid, then system 10 is considered dirty; and if the purity index determined by the model indicates a category corresponding to clean fluid, then system 10 is considered clean.

[0102] Advantageously, in the cases of the first and second tasks, the variable representing the hue of the fluid represents the opacity of the fluid, when the fluid is opaque, the fluid is considered dirty, and when the fluid is transparent, the fluid is considered clean.

[0103] The third task is to determine the hue of the fluid from a predefined color palette. In this case, the variable representing the fluid hue represents the color of the fluid. The variable representing the fluid hue is a category indicator among multiple categories, each corresponding to a predefined fluid hue. This task is performed, for example, to verify that the hue of the fluid flowing in system 10 is indeed the expected hue.

[0104] Advantageously, when computing unit 50 is configured to perform several of the three tasks, separately trained models are used for each task. For example, the user indicates which task they want device 14 to perform, and then computing unit 50 uses the associated model.

[0105] Advantageously, when the user only wishes to perform the first and / or second task, light source 22 is configured to emit only monochromatic light. Specifically, monochromatic light is sufficient to evaluate the opacity of the fluid, and therefore sufficient for the model to reliably determine the variables representing the hue of the fluid. Training the model is also performed using monochromatic light.

[0106] Advantageously, the determination method is executed continuously, and the continuous values ​​of the variable representing the hue of the fluid are stored in the memory of the calculation unit 50.

[0107] Optionally, the calculation unit 50 detects a fluid change operation when the variable representing the hue of the fluid is modified compared to the variable representing the hue of the fluid previously determined during detection step 110 (especially when the change in the variable representing the hue of the fluid exceeds a predetermined threshold). If the variable representing the hue of the fluid is a category indicator, the variable representing the hue of the fluid is considered modified when the category indicated by the variable representing the hue of the fluid is modified. If the variable representing the hue of the fluid is a binary number, the variable representing the hue of the fluid is considered modified when the value of the binary number changes; or alternatively, if it is a decimal number, the variable representing the hue of the fluid is considered modified when the value of the number changes by at least 10%. Advantageously, the user can select the necessary changes for the variable representing the hue of the fluid to be considered modified. The fluid change operation is advantageously a flushing or hue change of the fluid flowing in system 10.

[0108] Advantageously, if a fluid change operation has already been detected at detection step 110, the calculation unit 50 determines during determination step 112 that the hue change operation or rinsing operation has been completed when the continuous variable representing the hue of the fluid remains substantially constant over a predetermined duration.

[0109] When the variable representing the hue of the fluid is a decimal number, essentially constant means that the variable representing the hue of the fluid, continuously determined by the artificial intelligence model, changes by less than 10% over a predetermined period of time, preferably less than 5% over the predetermined period of time. When the variable representing the hue of the fluid is a binary number or a category indicator, essentially constant means that the variable representing the hue of the fluid, continuously determined by the artificial intelligence model, is the same, that is, unchanged.

[0110] When the variable representing the hue of the fluid remains substantially constant over a predetermined duration, it means that the hue of the fluid flowing in system 10 has stabilized. In the case of a hue change, it means that the desired hue has been achieved, and in the case of flushing, it means that the system has been completely flushed. Advantageously, when the fluid change operation is completed, the computing unit 52 commands a specific display on the human-machine interface 54, or commands the sending of a specific message to a remote terminal.

[0111] The predetermined duration for considering the stability of the variable is, for example, between 0.5 and 5 seconds.

[0112] Alternatively or additionally, device 14 is configured to transmit a variable representing the hue of the fluid to a control unit, which is configured to control the valves or a set of valves of the dispensing system 10 based on the variable representing the hue of the fluid. Thus, for example, it is possible to optimize hue-changing operations or adjust the composition of the fluid to obtain a desired hue.

[0113] In addition to the above description regarding the learning (i.e., training) of the model, the training of the model is performed, for example, in the form of supervised learning. The model receives as input a series of data comprising parameters representing the light intensity of a light signal for a given color channel. This series corresponds to data provided by a light sensor belonging to a hue determination device, which includes illumination and acquisition optical fibers of predefined lengths. Depending on the task on which the model is trained, each series is associated with a category or a decimal number. The model then learns to make predictions and is trained until satisfactory performance is achieved. A performance validation operation of the model is advantageously performed.

[0114] The model is then integrated into the estimation unit of device 14, whose fibers 34 and 36 advantageously have the same length as the fibers used for model training. This enables more reliable predictions to be obtained through the artificial intelligence model.

[0115] Alternatively, the AI ​​model is trained using reinforcement learning. Thus, when the model is used to determine the hue of a fluid, the user indicates, for example, whether the model's prediction is correct, and the user's feedback is taken into account by the model in its future predictions. Reinforcement learning is advantageously performed on a trained model and allows for improvements to the model's predictions.

[0116] Figure 4The flange 13 of the hue determination device 114 according to a second embodiment of the invention is shown. The hue determination device 114 is configured to perform hue measurement by means of the transmission of an optical signal through a fluid. In this case, the flange 13 of the hue measurement device 114 advantageously includes two recesses 138 for insertion into the end of the illumination optical fiber 34 on one side and the end of the acquisition optical fiber 36 on the other side, and two observation windows 140 arranged tangentially to the fluid flow vein 16 across the measurement region 23, each observation window located at the bottom of one of the recesses 138. Advantageously, the observation windows 140 face each other.

[0117] The fluid determination method remains unchanged, except that the optical signal corresponds to the light transmission of light emitted by the fluid from the light source 22 toward the fluid.

[0118] Any feature described above for one embodiment or variant may be implemented for other embodiments and variants described above, provided that it is technically feasible.

Claims

1. A computing unit (50) for a hue determination device (14; 114) for a fluid distribution system (10), said device (14; 114) comprising: - At least one light source (22), said at least one light source being configured to emit polychromatic or monochromatic light toward the fluid in the measurement area (23); - A light sensor (24) configured to receive light signals that have been reflected or transmitted through the fluid, the light signals corresponding respectively to light reflection or transmission by the fluid of polychromatic or monochromatic light emitted toward the fluid, the light signals representing the hue of the fluid flowing in the fluid distribution system (10), the light sensor (24) also being configured to emit parameters, each parameter representing the light intensity of the light signal for a given color channel; as well as - Acquisition fiber (36), the acquisition fiber being configured to transmit the optical signal to the optical sensor (24); The computing unit (50) includes an estimation module (52) configured to receive the parameters emitted by the optical sensor (24). The estimation module (52) is characterized in that it is further configured to determine variables representing the hue of the fluid via a pre-trained artificial intelligence model, wherein each parameter representing the light intensity of the light signal for a given color channel is an input variable of the model, and the output variable of the artificial intelligence model is the variable representing the hue of the fluid. The variable representing the hue of the fluid is: - The purity index of the fluid flowing in the fluid distribution system (10); or - Category indicators among multiple categories, each corresponding to a predefined fluid hue.

2. The calculation unit (50) according to claim 1, wherein the purity index is a decimal number.

3. The calculation unit according to claim 1, wherein the purity index is included between 0 and 1.

4. The calculation unit according to claim 1, wherein the purity index is a binary number.

5. The calculation unit according to claim 1, wherein the purity index is a category indicator in a first category corresponding to the case where the fluid (10) is clean and a second category corresponding to the case where the fluid (10) is dirty.

6. The computing unit (50) according to claim 1, wherein the fluid is a coating liquid or a cleaning liquid or an aerosol composed of solid colored particles suspended in a gas.

7. A hue determination device (14) for a fluid distribution system (10); 114), including: - At least one light source (22), said at least one light source being configured to emit polychromatic or monochromatic light toward the fluid in the measurement area (23); - A light sensor (24) configured to receive light signals that have been reflected or transmitted through the fluid, the light signals corresponding respectively to light reflection or transmission by the fluid of polychromatic or monochromatic light emitted toward the fluid, the light signals representing the hue of the fluid flowing in the fluid distribution system (10), the light sensor (24) also being configured to emit parameters, each parameter representing the light intensity of the light signal for a given color channel; - Acquisition fiber (36), the acquisition fiber being configured to transmit the optical signal to the optical sensor (24); and - The computing unit (50) according to any one of the preceding claims.

8. The device (14; 114) according to claim 7, wherein the given color channel is a red channel, a green channel, a blue channel and a transparent channel, and the parameters represent the light intensity of the light signal of the red channel, the green channel, the blue channel and the transparent channel, respectively.

9. A fluid distribution system (10) comprising a fluid flow loop (12) and a hue determination device (14; 114) according to claim 7 connected in series to the fluid flow loop (12).

10. A method for determining the hue of a fluid flowing in a fluid distribution system (10), said method being implemented by a hue determination device (14; 114) according to claim 7, said method comprising the following steps: - The at least one light source (22) emits (100) polychromatic or monochromatic light toward the fluid; - The light signal reflected or transmitted through the fluid by the light sensor (24) (102) is received by the light sensor (24); - The estimation module (52) receives (104) the parameter representing the light intensity of the light signal of the given color channel; and - The variable representing the hue of the fluid is determined (106) by the estimation module (52).

11. The determining method according to claim 10, further comprising the following steps: - When the variable representing the hue of the fluid is modified, a fluid change operation (110) is detected; as well as -If a fluid change operation has been detected, and the variable representing the hue of the fluid is substantially constant for a predetermined duration, it is determined (112) that the fluid change operation has ended. The fluid alteration operation is preferably a flushing of the system or a color change of the fluid flowing in the system (10).

12. The determination method according to claim 10, wherein the fluid is a coating liquid or a cleaning liquid.

Images