Method for monitoring a one-piece weaving

Abstract

One aspect of the invention relates to a method (100) for monitoring the weaving of a piece by an electronic weaving machine, comprising: Obtaining (110) configuration and weaving instruction data supplied to the electronic weaving machine to weave the piece; Generating (120) a digital model of the electronic weaving machine by aggregating the obtained configuration and instruction data; Providing (140) the generated digital model of the electronic weaving machine (120) to the machine learning model in order to obtain a prediction of the voltage applied to the individual motors during the weaving of the piece; and Monitoring (150) the weaving of the piece by comparing the predicted voltage applied to an individual motor with the voltage measured by the voltage sensor associated with said individual motor. Figure to be published with the abstract: Figure 1

Description

Title of the invention: Method for monitoring the weaving of a single piece. TECHNICAL FIELD OF THE INVENTION

[0001] The technical field of the invention is that of the monitoring and management of weaving of pieces.

[0002] The present invention relates to a method for monitoring the weaving of a piece, the weaving of the piece being carried out by an electronic weaving machine. TECHNOLOGICAL BACKGROUND OF THE INVENTION

[0003] Weaving is a process of assembling threads carried out by a weaving machine, making it possible to obtain a fabric. It consists of interlacing threads stretched lengthwise along the weaving machine, called warp threads, and threads stretched widthwise across the weaving machine, called weft threads.

[0004] Current weaving machines, which we will refer to in this application as "electronic weaving machines," are mechanized and computer-controlled. Thanks to the use of these machines, it is now possible to produce, in a limited time, complex woven composite parts with specific technical properties, particularly mechanical ones. Woven composite parts are used in various fields, such as aeronautics, for example, in the manufacture of engine blades.

[0005] An electronic weaving machine uses a harness to hold guide bars in place and return springs to maintain the tension of the warp threads. The guide bars and return springs constitute the loom of the weaving machine. In addition, heddles are attached to the harness. The harness of heddles holds the threads in place to create complex patterns. The heddles attached to the harness are driven by individual motors equipped with tension sensors. The tension sensors measure the tension applied to each motor. When the warp threads are lifted by the harness, a space, called the sheaf, is created between the warp threads. The sheaf is thus the result of the warp threads being separated into two layers: a top and a bottom layer.

[0006] The main anomalies affecting the weaving of a piece are breaks or jams in the warp threads, heddle breaks, or collisions between a spear, which guides the weft threads, and certain warp threads. These weaving anomalies, which can negatively impact the quality of the woven piece or reveal a machine malfunction, are difficult to detect, given the complexity of the process. Controls such as X-ray tomography are certainly applied to the part after it has been made, but they are unreliable and time-consuming to make, and cannot be applied in real time.

[0007] Some electronic weaving machines have a very large number of motors equipped with tension sensors. For example, the Staubli Unival 100 Jacquard machine is equipped with approximately 9,000 motors, each motor having a sensor that quantifies its force on a scale of 0 to 100%. Monitoring a weave produced by such a machine is therefore complex.

[0008] It is known to monitor and manage the weaving of a piece using an alarm system directly implemented on the electronic weaving machine. This alarm system stops the weaving when the value returned by one of the motor sensors is outside a predetermined range. This alarm system has several drawbacks. For example, this system does not take into account the specific characteristics of the piece and / or the setup. Furthermore, this system may generate a large number of false alarms and / or fail to detect a large number of anomalies.

[0009] There is therefore a need to improve the monitoring and management of a one-piece weaving produced by an electronic weaving machine. Summary of the invention

[0010] The invention provides a solution to the problems mentioned above, in particular by enabling the monitoring of a weaving process performed by an electronic weaving machine that takes into account the specific characteristics of the piece and the setup. Thus, the monitoring method according to the invention processes the data acquired during the weaving of a piece and uses a model of the loom capable of predicting the response of the machine sensors to the weaving instructions. The response predicted by the model is successively compared to new data from the ongoing weaving. Discrepancies between the model's prediction and the new data can be reported as defects and located on a map of the piece.

[0011] Thus, an electronic weaving machine compatible with the process according to the invention has an encoder and a motor for each of the eyelets. The motor's role is to position the eyelets according to the instruction provided by the user. The data captured comes from each of the motors. The process according to the invention therefore consists of identifying the signals from the motors that are not consistent with the instruction given to the machine. In order for this operation to be carried out efficiently, it was necessary to identify the physical quantities most correlated with the signal recorded by the motor. The quantities identified include, in particular, the position of the eyelets for each insertion and the setting of the loom. Regarding the position of the eyelets for each insertion, when the eyelet is in a position that does not comply with the instruction, the motor reacts by trying to bring it back to the position that complies with the instruction. This reaction will induce a rapid change in the motor torque, which is normally captured by the acquisition system. The position of the eyelets can be retrieved, for example, from the cross-processing of the loom card and the warp opening instructions. The loom adjustment concerns, in particular, the parameters relating to the fabric feed, commonly called "feed," and the parameters of the guide bars. The feed of the preform can generate overtension in the warp threads, which may prevent the motor from correctly feeding the yarn. The adjustment of the guide bars has a direct impact on the geometry of the warp openings and therefore on the balance condition of the warp threads. Incorrect positioning can indeed lead to the warp threads hitting the sling. This type of event could potentially introduce overtensions detected by the system. Both of this information can be stored in the loom card.

[0012] A first aspect of the invention relates to a computer-implemented method for monitoring the weaving of a piece by an electronic weaving machine, the weaving machine comprising individual motors, each individual motor among the individual motors being equipped with a voltage sensor measuring the voltage exerted at said individual motor, the method comprising: • Obtaining configuration and weaving instruction data provided to the electronic weaving machine to weave the piece, the configuration and weaving instruction data provided to the electronic weaving machine including: • data from a weaving card, the weaving card providing positioning for each of the threads during insertion in the weaving of the piece, and • Sizing data describing warp thread diameters and the distribution of said warp threads between rows and columns of the loom harness, and • Data from a crowd opening instruction describing high and low positions for the rails based on an insertion number and a position in the harness, • Generate a digital model of the electronic weaving machine by aggregating the configuration and setpoint data obtained, • Obtain a trained machine learning model that takes as input a digital model of the electronic weaving machine and is configured to provide as output a prediction of the voltage exerted at the individual motor among the individual motors of the electronic weaving machine during the weaving of the piece, the digital model of the weaving machine being a digital data structure aggregating the configuration data and weaving instructions for the electronic weaving machine, • Provide the generated digital model of the electronic weaving machine to the machine learning model in order to obtain a prediction of the tension exerted on the individual motors of the electronic weaving machine during the weaving of the piece, and • Monitor the weaving of the piece by comparing the predicted tension exerted at an individual motor among the individual motors with the tension measured by the tension sensor associated with said individual motor.

[0013] Thanks to the invention, the monitoring method takes into account the specific characteristics of the weaving machine by integrating, for example, data describing the diameters of the different warp yarns and their distribution between the rows and columns of the loom harness, the harness definition, the type(s) of yarn used, the spring stiffness, and / or the geometry of the sheaf opening. The monitoring method also takes into account the specific characteristics of the woven piece by integrating, for example, data concerning the thickness and density of the weft yarn used for weaving the piece.

[0014] In addition to the characteristics mentioned in the preceding paragraph, the method of monitoring the weaving of a piece by an electronic weaving machine may have one or more additional characteristics from among the following, considered individually or according to all technically possible combinations: • the method further includes modifying the operating conditions of the weaving machine when a voltage difference between the predicted voltage exerted at the individual motor among the individual motors and the voltage measured by the voltage sensor associated with said individual motor exceeds a predetermined threshold voltage value, • The modification of the conditions of use of the weaving machine includes at least one of the following actions: • Identify a component failure in the weaving machine, • Stop the weaving machine, and • Save the position of the weaving defect on the woven piece, • The electronic weaving machine takes as input various warp yarns and includes guide bars, return springs and a harness to which heddles are attached, the heddles being driven by individual motors and in which the configuration and weaving instruction data obtained, the process further including data relating to the bottom loom of the electronic weaving machine describing displacements of the guide bars and stiffnesses of the return springs located under the heddles, • The part woven by the electronic weaving machine is an aeronautical part, • Obtaining the trained machine learning model includes the following sub-steps: • Obtain a history of data sets from weavings of parts performed by the electronic weaving machine, each data set in the historical data set corresponding to data collected during the weaving of the part and including configuration and weaving instruction data supplied to the electronic weaving machine and a voltage measured by the voltage sensor at said individual motor among the individual motors of the electronic weaving machine, • Generate, for each dataset in the dataset history, a digital model of the electronic weaving machine, said digital model being a digital data structure aggregating the configuration and weaving instruction data of the electronic weaving machine, • Obtain a machine learning model that takes as input the digital model of the electronic weaving machine and is configured to provide as output a prediction of the voltage applied to an individual motor of the electronic weaving machine during the weaving of the piece, and • Train the machine learning model using a regression algorithm that penalizes, for each dataset in the historical datasets, the difference between the measured voltage and the predicted output voltage from the machine learning model. • The regression algorithm is a gradient reinforcement algorithm.

[0015] The substeps that can be included in obtaining the machine learning model constitute a learning process. The learning process is designed to allow the machine learning model to generalize weaving monitoring to new parts. In other words, the trained machine learning model is configured to be used in a process for monitoring weaves of parts different from those woven in the training data. Similarly, the trained machine learning model is also configured to be used in a process for monitoring weaves of parts made with new configurations of the electronic weaving machine, that is, with configurations of the electronic weaving machine that were not in the training data.

[0016] A second aspect of the invention relates to a system configured to implement the steps of the process of monitoring the weaving of a piece by an electronic weaving machine.

[0017] A third aspect of the invention relates to a computer program product comprising instructions which, when the program is executed on a computer, lead the latter to implement the steps of the process of monitoring the weaving of a piece by an electronic weaving machine.

[0018] A fourth aspect of the invention relates to a computer-readable recording medium comprising instructions which, when executed by a computer, lead the computer to implement the steps of the process of monitoring the weaving of a piece by an electronic weaving machine.

[0019] The invention and its various applications will be better understood by reading the following description and examining the accompanying figures. BRIEF DESCRIPTION OF THE FIGURES

[0020] The figures are presented for illustrative purposes only and are in no way limiting of the invention. • Fig. 1 shows a flowchart of an example of a method for monitoring a weaving of a piece according to the invention. • Fig. 2 shows a flowchart of an example of one step in the process of monitoring a weaving of a piece according to the invention. • Fig. 3 shows a graphical representation of an example of molting instructions that can be provided to an electronic weaving machine. • Fig. 4 shows an example of the evolution of the height of the guide bars during the weaving of a piece by an electronic weaving machine. • Fig. 5 shows a graph including an example of prediction of the voltage exerted at an individual motor of the electronic weaving machine during the weaving of a piece and an example of voltage measured at an individual motor among the individual motors of the electronic weaving machine during the weaving of the piece. • Fig. 6 shows a graphical representation of data describing the diameters of the different warp yarns and their distribution between the rows and columns of the loom harness.

[0021] Unless otherwise specified, the same element appearing on different figures has a unique reference. DETAILED DESCRIPTION

[0022] The figures are presented by way of illustration and in no way limit the invention.

[0023] Figure 1 shows a flowchart of an example of a process according to the invention. The essential steps of this example process are indicated by a rectangle with solid lines and the optional step is indicated by a rectangle with dotted lines.

[0024] The invention relates to a method 100 for monitoring the weaving of a piece, the weaving of the piece being carried out by an electronic weaving machine.

[0025] The process 100 is implemented by a computer or processor. "Computer-implemented" means that the steps, or virtually all of the steps, are executed by at least one computer, processor, or similar system. Thus, steps are performed by the computer, possibly fully automatically or semi-automatically. In some examples, the triggering of at least some of the process steps can be achieved through user-computer interaction. The level of user-computer interaction required may depend on the intended level of automation and be balanced against the need to implement the user's requirements. In some examples, this level may be user-defined and / or predefined.

[0026] A typical example of computer implementation of a process consists of executing the process with a system adapted for this purpose. The system may include a processor coupled with memory and a graphical user interface (GUI), the memory having stored a computer program containing instructions for implementing the process. The memory may also store a database. Memory is any hardware adapted for such storage, possibly comprising several distinct physical parts.

[0027] A first step 110 of the process 100 consists of obtaining configuration and weaving instruction data supplied to the electronic weaving machine to weave the piece. The term "obtain" may be defined in this application as "receive" or "calculate".

[0028] The weaving configuration and instruction data obtained in step 110 include data from a weaving chart. The weaving chart provides the positioning of each of the yarns, including the upper and lower sheaves, during each insertion in the weaving of the piece.

[0029] The configuration and weaving instruction data obtained in step 110 also include sizing data describing the diameters of the different warp threads and their distribution between the rows and columns of the loom harness. Figure 6 shows a graphical representation of data describing the diameters of the different warp threads and their distribution between the rows and columns of the loom harness. This sizing data makes it possible, in particular, to identify the loom motors whose warp threads are most likely to interact with each other, and therefore to jam each other. In other words, this sizing data makes it possible to group the different Motors arranged in "columns" have chain wires that are more likely to interact with each other, and in particular to jam each other. The 100 process can therefore take this proximity factor between motors into account to refine its predictions. Information regarding the diameters of the various warp threads is also important, as it allows for the identification of the threads most prone to rubbing against neighboring threads and exerting additional weight on the motors that drive them. Indeed, the larger the diameter of a thread, the more it tends to rub against neighboring threads and exert additional weight on the motors that drive them. This snare data can be obtained as Excel files, which can be converted into objects, for example using Python, containing the relationships between each loom motor and its position in the harness. The position is described by the row and column numbers, along with the diameter of the warp threads passing through each snare.

[0030] The configuration and weaving instruction data obtained in step 110 also include sheaf opening instruction data. The sheaf opening instruction data describes the high and low positions of the heddles as a function of an insertion number and a position in the harness. These positions can be calculated from one or more machine instruction files. The weaving machine instruction files can, for example, be at least one of the following: • a file of type "*.tif" containing the information of the weave structure, • a file of type "*.gof" containing the geometry information of the opening of the crowd for each program, • a "*.bmp" file containing the programs used for each engine and each insertion over time, and • a file of type “*.shd” allowing the link between files of type “*.gof” and of type “*.bmp”.

[0031] A representation of the data contained in an example file containing the programs used for each motor and each insertion over time is illustrated in [Fig. 3] for a Staübli Unival 100 type weaving machine. Thus, [Fig. 3] illustrates shedding instructions, commonly called "shedding" in English, which can be provided to a Staübli Unival 100 type weaving machine. The data concerning the sheaf opening allows for a better accuracy in predicting the tension of one or more motors because it determines the amplitude of the movement of the heddles, and therefore indirectly the force applied by the motors.

[0032] In one variant, compatible with the preceding variants, the configuration and weaving instruction data obtained in step 110 also include data relating to the loom bottom, i.e. the guide bars and the reed of the machine Electronic weaving. The data relating to the lower loom includes data describing the displacements of the guide bars and data describing the stiffnesses of the return springs located under the heddles. Figure 4 illustrates an example of the evolution of the guide bar height during the 6000 insertions in the weaving of an aircraft engine blade. The horizontal axis of the graph corresponds to the number of insertions in the weaving of an aircraft engine blade, and the vertical axis corresponds to the height of the guide bars. Curve 410 represents the range between the minimum and maximum heights of the guide bars during an insertion, and curve 420 represents the minimum and maximum heights of the guide bars during an insertion.This data relating to the lower loom improves the accuracy of predicting the tension applied to one or more motors because the movements of the reed and guide bars apply traction to the fabric, which translates into tension in the heddles. Indeed, the height of the guide bars is useful for determining motor torques, because by raising or lowering the fabric, they indirectly exert traction on the warp threads, which is then transmitted to the heddles. Similarly, the higher the loom's forward speed, the greater the induced traction on the heddles. Therefore, the movements of the reed and guide bars apply traction to the fabric, which translates into tension in the motors.

[0033] In an example consistent with the preceding examples, the spring stiffnesses are obtained from torque measurement data on a set of parts. Thus, an example of a procedure for obtaining spring stiffnesses consists of empirically calculating the stiffness and unstretched length of each of the loom springs from the torque measurement data on a set of parts. A first step in this example procedure consists of calculating, for each motor and each insertion, the median of the torque measurements on the set of parts, which eliminates statistical anomalies. A second step in this example procedure consists of filtering the points, i.e., the drive data, for which at least three successive insertions are associated with the same heddle height, thus establishing a case of static equilibrium of forces between the motor and the spring.A third step in this example procedure involves calculating the slope and y-intercept of a linear regression between the torque and the height of the spool for all these points. These two values ​​allow us to characterize the stiffness and the unstretched length of the spring and can therefore be used in the continuation of procedure 100.

[0034] In an example, consistent with the preceding examples, the height of each heddle is obtained before, during, and after each insertion. Indeed, when the electronic weaving machine does not allow continuous measurement of the torque value of the For the motor, this value may depend on the future movement of the slide. For example, when a slide moves from a low position to a high position, the motor will anticipate the upward movement from the first cycle, and consequently, the tension measured during the first insertion will be higher than when the slide was also in the low position during the second cycle.

[0035] In one embodiment, compatible with the preceding embodiments, data concerning the behavior of neighboring wires, for example, wires in the same column of the harness, can also be obtained in step 110. Indeed, neighboring wires can interact with each other through friction, become jammed, and thus influence the tension of the other motors in the same column. For example, it is possible to obtain, for each motor in a column, the height of the rail before, during, and after each insertion.

[0036] In an alternative embodiment, compatible with the preceding embodiments, measurement data specific to each part can also be obtained in step 110. The measurement data corresponds to the torque values ​​of each machine motor for each insertion. The measurement data is, for example, contained in files of type "*.Ret".

[0037] In a second step 120, a digital model of the electronic weaving machine is generated by aggregating the configuration and setpoint data obtained.

[0038] In an example, compatible with the preceding examples, the digital model of the electronic weaving machine can be generated 120 using different steps. For example, an image file, which may be of type “*.tif”, containing the weave information, includes, for each motor and each insertion, a binary number indicating when the motor is in the down or up position. In order to retrieve this data, it is thus possible to construct a parser for this type of file that stores this heddle lifting data in the form of an array, for example a “NumPy” type array. Each of the “*.The ".gof" file, containing the crowd opening geometry information for each program, can also be parsed using a syntax analyzer to retrieve the value in millimeters of the elevation of each rail in its upper and lower positions, for each insertion, and for a specific setpoint value. Each ".gof" file represents a different setpoint value. The ".shd" file, which links the ".gof" and ".bmp" files, can be converted by scripts, such as Python scripts, into a lookup table between a setpoint value and a color code. Finally, the ".bmp" file, containing the programs used for each motor and each insertion over time, can also contain a color code for each insertion and each motor.This color code can be stored as an array, for example a "NumPy" type array, using a dedicated script in the language. Python. Once all the information has been retrieved from the files, the first step is to combine the data from the "*.shd" file and the "*.bmp" file to map each "motor, insertion" pair not to a color code, but to a setpoint value. Then, the data from the "*.gof" files is added to map each "motor, insertion" pair not to a setpoint value, but to the elevation values ​​in the upper and lower positions, taken from the "*.gof" file, for the given motor and insertion. In practice, the result of this operation can be stored as two arrays, for example, "NumPy" arrays, one containing the lower positions of the rails for each "motor, insertion" pair, and the other containing the upper positions for each "motor, insertion" pair.Finally, this information can be combined with the "*.tif" file containing the weave structure information to determine, for each "motor, insertion" pair, whether to select the heddle height value in the up or down position. The final result of this series of operations could be, for example, a table, such as a "NumPy" array, containing the height (e.g., the height in millimeters) of the corresponding heddle for each motor and each insertion. This data is useful for predicting motor tension. Each heddle is actuated by both a motor and a return spring. Thus, the more the heddle is pulled upward by the motor, the greater the force applied by the spring, and consequently, the torque applied by the motor also increases.

[0039] In an example consistent with the preceding examples, generation 120 of the digital model of the electronic weaving machine can further include the aggregation of data in a dedicated class of the Python language, named Loom, allowing for efficient data manipulation. Moreover, the use of memory space has been optimized, notably by a choice of numeric types adapted to the different variables, for example, integers stored in memory using 8 or 16 bits.

[0040] In an example consistent with the preceding examples, when data relating to the lower loom of the electronic weaving machine are obtained in step 110, the generation 120 of the digital model of the electronic weaving machine may further include the storage of data relating to the lower loom of the electronic weaving machine. The data relating to the lower loom of the electronic weaving machine is present in a "*.tif" file and indicates, for each insertion, the thickness of the weft yarn used, the speed of the conveyor belt moving the fabric, whether the fabric is moving forward or stopped, and the height of the guide bars. This data can be retrieved and stored using a parser. The storage of this data can be carried out in different arrays, for example in an object constructed in the Python language comprising various arrays of type "NumPy".

[0041] In an example consistent with the preceding examples, when data concerning the behavior of neighboring threads are obtained in step 110, the generation 120 of the digital model of the electronic weaving machine can further include the reduction of the data obtained, for example, using principal component analysis. For example, when 32 motors are present per column, 96 data points can be obtained for each insertion. With principal component analysis, this data can be reduced to 10 synthetic variables that can be used for the generation 120 of the digital model of the electronic weaving machine.

[0042] In an example consistent with the preceding examples, when measurement data is obtained in step 110, the generation 120 of the digital model of the electronic weaving machine may further include the storage of this measurement data. This data can be retrieved and stored using a parser. The storage of this data can be carried out in various arrays, for example, NumPy-type arrays. Information concerning the torque of each motor of the machine for each insertion can be encoded, for example, as a 5-bit integer representing the torque at the time of insertion as a percentage of the maximum torque that the motor can apply.

[0043] The method 100 includes a third step 130 of obtaining a trained machine learning model taking as input a digital model of the electronic weaving machine and configured to provide as output a prediction of the voltage applied to an individual motor of the electronic weaving machine during the weaving of the piece. The digital model of the weaving machine is a digital data structure aggregating the configuration and weaving setpoint data of the electronic weaving machine.

[0044] In an example consistent with the preceding examples, step 130 of process 100 comprises four substeps illustrated in [Fig. 2]. [Fig. 2] shows a flowchart of an example of step 130 comprising four substeps 131 to 134.

[0045] A first substep 131 of step 130 consists of obtaining drive data. The drive data obtained is a history of data sets from the weaving of one or more pieces by an electronic weaving machine. The electronic weaving machine from which the drive data was obtained is preferably of the same type as the electronic weaving machine monitored using method 100. For example, the drive data consists of data collected on several thousand motors of an electronic weaving machine during several thousand insertions. during the weaving of a part. In another example, consistent with the previous one, data was collected during the weaving of one or more aeronautical parts, such as an aircraft engine blade. Each dataset in the history corresponds to the data collected during the weaving of a part by the electronic weaving machine. Each dataset in the history of datasets includes the electronic weaving machine's configuration and weaving setpoint data and one or more voltages measured by voltage sensors during the weaving of a part by the electronic weaving machine. Among the measured voltages, it is possible to identify which voltage measurement corresponds to a particular individual motor.Furthermore, it should be noted that within each dataset, a correlation is possible between the configuration and weaving setpoint data of the electronic weaving machine and the voltage(s) measured by one or more tension sensors during the weaving of a piece by the electronic weaving machine. In other words, it is possible to determine which configuration and weaving setpoint data of the electronic weaving machine correspond to the voltage measured by tension sensors during the weaving of a piece by the electronic weaving machine.

[0046] Preferably, these configuration and setpoint data are of the same type and composition as the configuration and setpoint data used during process 100. Thus, all the data that can be obtained in step 110 of process 100 can also be obtained in substep 131 in order to be used in substeps 132 to 134.

[0047] In a second substep 132 of step 130, a digital model of the electronic weaving machine is generated for each data set in the data set history. The digital model of the electronic weaving machine is a digital data structure that aggregates the configuration and weaving instruction data of the electronic weaving machine for the considered data set. The digital model can also aggregate data relating to the geometry of the woven piece. The digital model allows for easy and efficient manipulation of this data. It can, for example, be a vector of numbers. Furthermore, it should be noted that a correspondence between each generated digital model of the electronic weaving machine and each voltage measured by the voltage sensors at the individual motors of the machine is possible.Thus, in step 134, it is possible to retrieve the voltage(s) measured by the voltage sensor(s) at the individual motors of the machine corresponding to the digital model of the electronic weaving machine provided as input to the machine learning model. Furthermore, the data aggregation and generation of the digital model can preferably be similar to steps 120 and 132. For example, if the... The numerical model generated in step 120 contains data of three types aggregated in a specific order; the numerical model generated in step 132 must also contain data of these three types aggregated in the same order. Preferably, the numerical model is generated in step 132 using the same method as the method used to generate the numerical model in step 120.

[0048] A third substep 133 of step 130 consists of obtaining a machine learning model. A machine learning model can be defined as a mathematical model capable of learning, for example, to predict an outcome, from training data. In the invention, the machine learning model is a model that takes as input a digital model of an electronic weaving machine as generated in steps 120 and 132. The machine learning model is configured to provide as output a prediction of one or more voltages applied to one or more individual motors of an electronic weaving machine during the weaving of the piece.

[0049] In an example consistent with the preceding examples, the machine learning model is a combination of a large number of individual models, for example, decision trees. These are hundreds or thousands of individual models that are highly dependent on each other, with a common objective.

[0050] In a fourth substep 134 of step 130, the machine learning model obtained in substep 133 is trained. The training of the machine learning model comprises a first part, in which the digital model of the electronic weaving machine generated in step 132 is provided as input in order to obtain as output a prediction of one or more voltages exerted on one or more individual motors of the electronic weaving machine during the weaving of the piece. In a second part, this prediction, as well as the voltage(s) measured during the weaving of the piece, are used by a regression algorithm. The objective of the regression algorithm is to penalize, for each set of digital data in the digital data history, a difference between the voltage measured by the voltage sensor and the voltage prediction generated as output by the machine learning model.Thus, the machine learning model is trained using a regression algorithm, based on a historical dataset of weaving patterns produced by the electronic weaving machine, to predict the "normal" signal from the sensors according to the setpoint. For example, it is possible to use extracted measured torque values ​​stored in "*.Ret" files. The term "normal" here refers to the signal measured by the tension sensors in the absence of any anomalies. This algorithm is completely agnostic to the actual operation of the electronic weaving machine.

[0051] In an example, consistent with the preceding examples, the data collected During the monitoring process, 100 can also be used as training data for the example step 130 comprising the four sub-steps 131 to 134. Thus, each monitoring of a weaving of a new piece can participate in the learning and refinement of the machine learning model obtained using the four sub-steps 131 to 134.

[0052] In an example consistent with the preceding examples, the machine learning model is trained using a gradient boosting algorithm. A gradient boosting algorithm compatible with the invention trains a large number of decision trees to obtain a model capable of predicting the stresses exerted on individual motors of the electronic weaving machine during the weaving of the fabric, iteratively improving the predictions through gradient descent optimization. For example, the LightGBM library can be used with a model comprising several hundred or even several thousand decision trees to perform step 134.

[0053] Thus, step 130 makes it possible to obtain a machine learning model capable of predicting the voltages exerted on individual motors of the electronic weaving machine during the weaving of the part. The trained machine learning model can therefore be used in the continuation of the process 100 for monitoring the weaving of a part.

[0054] A fourth step 140 of the process 100 consists of providing the generated electronic weaving machine digital model 130 to the machine learning model in order to obtain a prediction of the tension exerted at the level of one or more individual motors of the electronic weaving machine during the weaving of the piece.

[0055] In a fifth step 150 of the process 100, the weaving of the part is monitored using the prediction, obtained in step 140, of the voltage applied to one or more individual motors of the electronic weaving machine during the weaving of the part and the voltage measured by one or more voltage sensors. The monitoring is performed, for example, by calculating the difference between the prediction, obtained in step 140, of the voltage applied to one or more individual motors of the electronic weaving machine during the weaving of the part and the voltage measured by one or more voltage sensors. In another example, the voltage prediction is performed for a period comprising a predetermined number, for example, between 5 and 20, of insertions, and a squared difference between the predicted voltage and the measured voltage over each period is calculated to monitor the weaving of the part.

[0056] An optional sixth step 160 of the process 100 consists of modifying the operating conditions of the weaving machine when a voltage difference between the predicted voltage exerted at one or more individual motors of the weaving machine electronic and the voltage measured by the voltage sensor(s) is greater than a predetermined threshold voltage value. Figure 5 shows a graph comprising two curves, 510 and 520. The horizontal axis of the graph represents the number of insertions performed, and the vertical axis represents the voltage at one or more individual motors of the electronic weaving machine during the weaving of a piece. Curve 510 is an example of how the predicted voltage applied to one or more individual motors of the electronic weaving machine during weaving the piece can be obtained in step 140. Curve 520 is the voltage measured at one or more individual motors of the electronic weaving machine during weaving the piece. For this example, the predicted voltage and the measured voltage are very close for the first 100 insertions.From the 100th insertion onwards, a significant discrepancy exists between the predicted voltage and the measured voltage. Thus, in this example, an anomaly is detected from the 100th insertion onwards if this discrepancy exceeds the predetermined threshold voltage value. An example of a predetermined threshold voltage value is between 1 and 10 newtons.

[0057] The monitoring method 100 can be used during or after the weaving of a piece. When a defect is detected during weaving, modifying the operating conditions of the weaving machine can involve stopping the electronic weaving machine. The electronic weaving machine can be stopped automatically, semi-automatically, or manually. Alternatively, when a defect is detected during weaving, the position of the weaving defect on the woven piece can be saved. When a defect is detected after weaving, the position of the defect on the woven piece can also be identified. The monitoring method 100 can also be used to identify the degradation of a weaving machine component or the lifespan of the weaving machine components that drive the yarns.

[0058] In a preferred embodiment, compatible with the preceding variants, the weaving monitored by process 100 is the weaving of an aeronautical part such as an aircraft engine blade.

Claims

Demands

1. A computer-implemented method (100) for monitoring the weaving of a single piece by an electronic weaving machine, the weaving machine comprising individual motors, each individual motor among the individual motors being equipped with a voltage sensor measuring the voltage exerted at said individual motor, and the method comprising: Obtaining (110) configuration and weaving instruction data supplied to the electronic weaving machine to weave the piece, the configuration and weaving instruction data supplied to the electronic weaving machine including: • Data from a weaving card, the weaving card providing a positioning of each of the yarns during an insertion in the weaving of the piece, and • Sizing data describing warp thread diameters and the distribution of said warp threads between rows and columns of the loom harness, and • Data from a crowd opening instruction describing high and low positions for the rails based on an insertion number and a position in the harness, Generate (120) a digital model of the electronic weaving machine by aggregating the configuration and setpoint data obtained, Obtain (130) a trained machine learning model taking as input a digital model of the electronic weaving machine and configured to provide as output a prediction of a voltage exerted at the individual motor among the individual motors of the electronic weaving machine during the weaving of the piece, the digital model of the weaving machine being a digital data structure aggregating the configuration and weaving setpoint data of the electronic weaving machine, Provide (140) the generated (120) digital model of the electronic weaving machine to the machine learning model in order to obtain a prediction of the voltage exerted at the individual motors of the electronic weaving machine during the weaving of the piece, and - Monitor (150) the weaving of the piece by comparing the prediction of the voltage exerted at the individual motor among the individual motors and the voltage measured by the voltage sensor associated with said individual motor.

2. Method (100) according to claim 1 wherein the method further comprises modifying the operating conditions of the weaving machine (160) when a voltage difference between the predicted voltage exerted at the individual motor among the individual motors and the voltage measured by the voltage sensor associated with said individual motor is greater than a predetermined threshold voltage value.

3. A method (100) according to claim 1 or 2 wherein the modification of the operating conditions of the weaving machine (160) includes at least one action among: - Identifying a degradation of a component of the weaving machine, - Stopping the weaving machine, and - Saving the position of the weaving defect on the woven piece.

4. A method (100) according to any one of the preceding claims, wherein the electronic weaving machine takes as input different warp yarns and comprises guide bars, return springs and a harness to which heddles are attached, the heddles being driven by individual motors and wherein the configuration and weaving instruction data obtained (110) are further comprised of data relating to the bottom loom of the electronic weaving machine describing displacements of the guide bars and stiffnesses of the return springs located under the heddles.

5. Method (100) according to any one of the preceding claims wherein the part woven by the electronic weaving machine is an aeronautical part.

6. Method (100), according to any one of the preceding claims wherein obtaining (130) the trained machine learning model comprises the substeps: - Obtaining (131) a history of data sets from weavings of parts performed by the electronic weaving machine, each data set in the history of data sets corresponding to data collected during the weaving of the part and comprising configuration and weaving instruction data supplied to the electronic weaving machine and a voltage measured by the voltage sensor at said individual motor among the individual motors of the electronic weaving machine, - Generating (132), for each data set in the history of data sets, a digital model of the electronic weaving machine, said digital model being a digital data structure aggregating the configuration and weaving instruction data of the electronic weaving machine,- Obtain (133) a machine learning model taking as input the digital model of the electronic weaving machine and configured to provide as output a prediction of a voltage exerted at the level of an individual motor of the electronic weaving machine during the weaving of the piece, and - Train (134) the machine learning model using a regression algorithm penalizing, for each dataset in the data set history, a difference between the measured voltage and the predicted voltage generated as output by the machine learning model.

7. Method (100) according to the preceding claim wherein the regression algorithm is a gradient reinforcement algorithm.

8. System configured to implement the steps of the process (100) according to any one of the preceding claims.

9. A computer program product comprising instructions which, when the program is executed on a computer, cause the computer to to implement the steps of the process (100) according to any one of claims 1 to 7.

10. A computer-readable recording medium comprising instructions which, when executed by a computer, cause the computer to carry out the steps of the process (100) according to any one of claims 1 to 7.

Images