Method and device for measuring wear on aircraft brake discs and associated computer program

Abstract

The invention relates to a method for measuring wear of the discs of an aircraft brake provided with a wear indicator (6) mounted movably on a fixed part of the aircraft brake and moving as the discs wear, the method comprising steps of: - acquiring an image of a part of the discs of the aircraft brake, - identifying the wear indicator (6) in the acquired image, - identifying at least three reference points (A1, B1, C2) on the wear indicator (6), and - determining at least one wear measurement of the discs of the aircraft brake by determining at least one dimension (L, W) associated with the wear indicator (6) from the three identified reference points (A1, B1, C2). Figure to be published with the abstract: Figure 8

Description

Title of the invention: Method and device for measuring wear on aircraft brake discs and associated computer program TECHNICAL FIELD OF THE INVENTION

[0001] The technical field of the invention is that of the control of parts of an aircraft, and in particular the control of the discs of an aircraft brake.

[0002] The present invention relates to a method and a device for measuring wear on the discs of an aircraft brake. It also relates to an associated computer program.

[0003] The present invention finds an advantageous application for the control of aircraft, such as airplanes or helicopters. TECHNOLOGICAL BACKGROUND OF THE INVENTION

[0004] Aircraft brakes conventionally comprise a stack of discs (also called "heat sinks") which are pressed against each other by means of pistons so as to generate a braking moment making it possible to slow down the wheel with which the brake is associated. The discs are liable to wear with each braking. It is therefore important to be able to monitor their state of wear in order to plan their replacement (in particular so that the braking of the aircraft is not affected).

[0005] In this regard, aircraft brakes are generally equipped with a wear indicator which is slidably mounted relative to a fixed part of the brake. This wear indicator is arranged so as to remain in contact with one of the faces of the stack of discs in such a way that the wear indicator moves as the discs wear. The position of the wear indicator is therefore indicative of the state of wear of the stack of discs.

[0006] Operators responsible for maintenance therefore regularly check the position of the wear indicator (typically weekly) to check the state of wear of the discs, and replace them with new discs if the wear indicator reaches a maximum wear position. This manual check is not optimal, in particular because it requires the operator to approach the stack of discs which has a very high temperature (several hundred degrees Celsius) following a braking operation. In addition, the accuracy of the wear indicator measurement remains quite limited, with an error of the order of a millimeter.

[0007] Document FR3068098 discloses a method for measuring wear on aircraft brake discs based on acquiring a photograph of the wear indicator. This photograph is combined with known information about the aircraft, and in particular about the number of flights made by the aircraft, to determine an estimate of the position of the wear indicator.

[0008] However, this method does not make it possible to explicitly and precisely obtain the length of the wear indicator, which would then make it possible to concretely characterize the state of wear of the aircraft brake discs. Summary of the invention

[0009] The present invention proposes to improve the determination of the wear of the discs of an aircraft brake by explicitly quantifying this wear by means of an optical measurement of the wear indicator.

[0010] More particularly, the invention proposes a method for measuring the wear of the discs of an aircraft brake provided with a wear indicator mounted movably on a fixed part of the aircraft brake and moving as the discs wear, the method comprising steps of: - acquisition of an image of part of the aircraft brake discs, - identification of the wear indicator in the acquired image, - identification of at least three reference points on the wear indicator, and - determination of at least one measurement of wear of the aircraft brake discs by determining at least one dimension associated with the wear indicator from the three identified reference points.

[0011] Thus, advantageously according to the invention, the wear measurement method makes it possible to evaluate the state of wear of the aircraft brake discs remotely on the basis of an image acquired from the wear indicator. The analysis of this image then makes it possible to quantify the state of wear of the aircraft brake discs by measuring at least one dimension associated with this wear indicator. This measurement is carried out on the basis of at least three reference points, which makes it possible to obtain a reliable and precise measurement of the characteristics of the wear indicator.

[0012] Furthermore, no material modification needs to be made to the aircraft brake to enable the wear condition of the aircraft brake discs to be assessed. No additional material cost is therefore incurred and the method according to the invention can be implemented on any aircraft.

[0013] Finally, thanks to the present invention, the state of wear of the aircraft brake discs can be determined without risk (i.e. without getting too close to the heat sink whose temperature is very high), in an automated manner and in a precise manner thanks to the marks identified on the wear indicator.

[0014] In addition to the characteristics which have just been mentioned in the preceding paragraph, the wear measurement method according to one aspect of the invention may have one or more additional characteristics among the following, considered individually or according to all technically possible combinations:

[0015] - in the step of determining the wear measurement of the aircraft brake discs, it is provided for the determination of the width and length of the wear indicator from the three identified reference points; - there is provided, before the step of identifying the at least three reference points, a step of delimiting a reduced part of the acquired image and comprising the wear indicator, the identification of the at least three reference points being implemented from this reduced part of the acquired image; - the identification of the wear indicator in the acquired image is implemented by means of an artificial neural network, the acquired image being provided as input to said artificial neural network so as to obtain, as output to said artificial neural network, an area of ​​the acquired image on which the wear indicator is identified; - the identification of at least three reference points is implemented by analyzing the area of ​​the acquired image obtained at the output of the artificial neural network; - the artificial neural network is a convolutional artificial neural network; - prior to the image acquisition stage, a stage of learning the artificial neural network from a set of basic images is planned; - the identification of the at least three reference points on the wear indicator is implemented by means of another artificial neural network, the acquired image in which the wear indicator has been identified being provided as input to said other artificial neural network so as to obtain, as output to said other artificial neural network, coordinates of the at least three reference points; - the other artificial neural network is a convolutional artificial neural network; - the reduced part of the acquired image is provided as input to the other artificial neural network so as to obtain, as output from said other artificial neural network, the coordinates of the at least three reference points; - prior to the image acquisition step, a step of learning the other artificial neural network from another set of basic images is planned; - before the step of determining the wear measurement of the aircraft brake discs, a step of identifying at least one other reference point is provided, the determination of the wear measurement of the aircraft brake discs being implemented taking into account this other reference point; and - the at least three reference points correspond to visible ends of the wear indicator.

[0016] The invention also relates to a device for measuring wear on the discs of an aircraft brake provided with a wear indicator mounted movably on a fixed part of the aircraft brake and moving as the discs wear, the device including: - a module for acquiring an image of part of the aircraft brake discs, - a module for identifying the wear indicator in the acquired image, - an identification module of at least three reference points on the wear indicator, and - a module for determining at least one measurement of wear of the aircraft brake discs by determining at least one dimension associated with the wear indicator from the three identified reference points.

[0017] The invention finally relates to a computer program comprising instructions executable by a processor and designed to implement a method as defined previously when these instructions are executed by the processor.

[0018] The invention and its various applications will be better understood upon reading the following description and examining the accompanying figures. BRIEF DESCRIPTION OF THE FIGURES

[0019] The figures are presented for information purposes only and in no way limit the invention.

[0020] [Fig.l] schematically represents an aircraft brake as considered in the present invention,

[0021] [Fig.2] is a partial schematic representation of the aircraft brake of [Fig.l] emphasizing a wear indicator included in this aircraft brake,

[0022] [Fig.3] represents, in functional form, an example of a measuring device wear of the discs of an aircraft brake designed to implement a method for measuring wear of the discs of an aircraft brake in accordance with the invention,

[0023] [Fig.4] represents, in the form of a flowchart, an example of a measurement method wear of the aircraft brake discs according to the invention,

[0024] [Fig.5] illustrates step E2 of the wear measurement method shown in [Fig.4],

[0025] [Fig.6] illustrates step E4 of the wear measurement method shown in [Fig.4],

[0026] [Fig.7] illustrates step E6 of the wear measurement method shown in [Fig.4], and

[0027] [Fig.8] illustrates steps E8 and E10 of the wear measurement method shown in the [Fig.4], DETAILED DESCRIPTION

[0028] The present invention is placed in the context of monitoring the brakes of an aircraft. More particularly, the invention aims to enable the monitoring of the wear of the aircraft brake discs. In particular again, the invention aims to quantify, that is to say measure, the wear of the aircraft brake discs, in particular to be able to predict their replacement.

[0029] The aircraft is for example an airplane or a helicopter.

[0030] [Fig.l] represents an aircraft brake 1. The aircraft brake 1 itself does not does not constitute the core of the present invention. It is therefore not described in detail here.

[0031] Essentially, this aircraft brake 1 comprises a torque tube on which a stack 2 of discs is stacked. This stack 2 of discs is commonly called a “heat sink”.

[0032] The stack 2 of disks comprises, in a conventional manner, an alternation of stator disks 2A and rotor disks 2B.

[0033] This stack 2 of discs is arranged so as to extend between a hydraulic crown 3 and a plate 5 fixed to the other end of the torque tube. The hydraulic crown 3 is provided with a plurality of pistons 4 positioned between this hydraulic crown 3 and the stack 2 of discs.

[0034] In practice, the aircraft brake 1 thus formed is positioned on an axle of an aircraft landing gear (not shown). This aircraft landing gear receives a wheel whose rim comprises bars for driving the rotor disks 2B in rotation.

[0035] In order to characterize the wear of the discs of the aircraft brake 1, this brake 1 is provided with a wear indicator 6. This wear indicator 6 is more particularly shown in [Fig.2]. Preferably, at least two wear indicators 6 are present in the aircraft brake 1.

[0036] As shown more particularly in [Fig. 2], this wear indicator 6 is here in the form of a cylindrical metal rod. This metal rod is slidably mounted on the hydraulic crown 3. It is required to remain in contact with the stack 2 of discs, so that the wear indicator 6 is moved towards the stack 2 of discs as the discs wear.

[0037] For this, in practice, the wear indicator 6 is for example driven by a plate (not shown in the figures) positioned between the stack 2 of discs and the hydraulic crown 3. As a variant, a spring system can be used so as to keep the wear indicator 6 in contact with the stack 2 of discs.

[0038] As indicated previously, the present invention aims to quantify, that is to say measure, the wear of the aircraft brake discs. In practice, the state of wear of the aircraft brake discs 1 can be quantified by determining the dimensions of the wear indicator 6 protruding from the hydraulic crown 3. Indeed, the more the aircraft brake discs 1 are worn, the more the wear indicator 6 is moved towards the stack 2 of discs. The length L of the wear indicator 6, outside the hydraulic crown 3 (that is to say opposite the stack 2 of discs) is therefore reduced as the aircraft brake discs 1 wear. The present invention therefore aims to determine these dimensions of the wear indicator 6 so as to quantify the state of wear of the aircraft brake discs 1.

[0039] In practice, this is implemented by means of a device 10 for measuring the wear of the aircraft brake discs 1.

[0040] [Fig. 3] represents, in functional form, the device 10 for measuring the wear of the discs of the aircraft brake 1 (also referred to as device 10 in the following).

[0041] As shown in [Fig.3], the device 10 comprises a processor 20 and a storage device 24.

[0042] The storage device 24 is for example a hard disk or a memory. The storage device 24 is here designed to store reference values, such as for example a maximum wear length used when implementing a wear measurement method as described below.

[0043] The device 10 also comprises a set of functional modules. It comprises for example an image acquisition module 12, a first identification module 14, a second identification module 16 and a module 18 for determining a wear measurement.

[0044] Each of the different modules described is for example implemented by means of computer program instructions stored in the storage device 24 and designed to implement the module concerned when these instructions are executed by the processor 20 of the wear measuring device 10.

[0045] However, as a variant, at least one of the aforementioned modules can be produced by means of a dedicated electronic circuit, for example an application-specific integrated circuit.

[0046] In other words, the wear measurement method according to the invention and described in the remainder of the document is here implemented by computer. Preferably, the wear measurement device 10 is for example embedded in a smartphone (or “smartphone” according to the commonly used English term). This allows in particular a more practical implementation of the wear measurement method according to the invention (because the aircraft brake 1 is often in a restricted space to which access is rather limited).

[0047] The processor 20 is also designed to implement a first artificial neural network NN1 and a second artificial neural network NN2, involved in the method for measuring wear of the aircraft brake discs 1 described below. This first artificial neural network NN1 and this second artificial neural network NN2 are presented in more detail below.

[0048] The device 10 is designed to implement the method for measuring wear on aircraft brake discs according to the invention.

[0049] [Fig.4] is a flowchart representing an example of a wear measurement method. aircraft brake discs 1 implemented in the context described above.

[0050] This wear measurement method is for example implemented by the processor 20. Generally, this wear measurement method is implemented by computer. As indicated previously, the wear measurement method is for example implemented from an application embedded in a smartphone (or " smartphone”).

[0051] As shown in [Fig.4], the wear measurement method begins at step E2 of acquiring an image Im of a portion of the aircraft brake discs. This image Im is for example acquired by an image acquisition device on board, for example, a smartphone. [Fig.5] represents an example of an image Im acquired during this step E2.

[0052] In step E4, the processor 20 analyzes the image Im acquired (in step E2) and determines whether the wear indicator 6 is present on this image Im. In other words, step E4 aims to identify the presence of the wear indicator 6 on the image Im acquired. [Fig.6] illustrates this step E4.

[0053] In practice, in the example of the wear measurement method described here, the identification of the wear indicator 6 in the image Im (acquired in step E2) is implemented by means of the first artificial neural network NN1.

[0054] The processor 20 then provides, as input to this first artificial neural network NN1, the image Im acquired (in step E2). This first artificial neural network NN1 is designed to provide as output a zone ZI of the image Im acquired on which the wear indicator 6 is identified.

[0055] This first artificial neural network NN1 therefore has a function of detecting objects (here the wear indicator 6) in an image provided as input (here the acquired image Im).

[0056] The first artificial neural network NN1 is here a convolutional artificial neural network. For example, it is a YOLO type artificial neural network (for "You Only Look Once" according to the commonly used Anglo-Saxon term). In a known manner, this type of artificial neural network allows, from an image provided as input, to detect and classify objects present in this image provided as input. More details concerning this YOLO type of artificial neural network can be found in the article by Joseph Redmon et al., "You Only Look Once: Unified, Real-Time Object Detection", 779-788, 10.1109 / CVPR.2016.91, 2016.

[0057] In other words, the first artificial neural network NN1 is designed to detect a zone ZI of the acquired image Im which comprises the wear indicator 6. [Fig.6] represents the zone ZI identified on the image Im acquired in step E2 by the implementation of the first artificial neural network NN1.

[0058] Prior to implementing the wear measurement method, a learning step (not shown in the figures) allows the training of this first artificial neural network NN1. For this, a set of basic images is used. For each basic image of this set on which the wear indicator 6 is present, an area which comprises this wear indicator 6 is identified. Each basic image of this set which does not represent the wear indicator is associated with an in training of absence of this wear indicator (no area is therefore associated with this corresponding basic image). The weights of the neurons are then adjusted as the learning stage progresses.

[0059] If, in step E4, the first artificial neural network NN1 does not provide any area of ​​the acquired image Im, this means that the wear indicator 6 is not visible on this acquired image Im. The wear measurement method then resumes in step E2, with the acquisition of a new image.

[0060] If, on the contrary, in step E4, the first artificial neural network NN1 provides, as output, a zone ZI of the acquired image Im comprising the wear indicator 6 (as is the case in [Fig.6]), the wear measurement method continues in step E6. This therefore means that the wear indicator 6 is visible on the image Im and that it will be possible to analyze the dimensions of this wear indicator 6 to characterize the wear of the discs of the aircraft brake 1.

[0061] As shown in [Fig.4], the wear measurement method continues in step E6. This step E6 is illustrated in [Fig.7].

[0062] During this step, the processor 20 determines a part Iml of the reduced image Im and comprising the identified wear indicator 6. In other words, during this step E6, the processor 20 restricts the image Im to a part Iml comprising the wear indicator 6 and an environment close to it. This part Iml is visible in [Fig.7].

[0063] In practice, the part Iml comprises for example the wear indicator 6 and an additional margin (for example of the order of 25%, in pixels, in all directions) around this wear indicator 6. As a variant, the part Iml may have a predetermined dimension smaller (while comprising the wear indicator 6) than the dimension of the image Im (acquired in step E2).

[0064] This step E6 is optional here but it is advantageous because it makes it possible to obtain an image of reduced dimensions including the wear indicator. This then allows a faster implementation of the wear measurement (which is based on an image analysis) and also less expensive in terms of computing resources.

[0065] The wear measurement is implemented by means of an image analysis on which the wear indicator 6 is visible. It is carried out either directly from the image Im acquired (in step E2), or from the part Iml obtained in step E6.

[0066] The wear measurement method then continues with step E8. During this step E8, the processor 20 identifies at least three reference points A1, B1, C2 on the wear indicator 6. These reference points A1, B1, C2 are marks identified on the wear indicator 6 which will make it possible to measure the wear of the discs of the aircraft brake 1. Step E8 therefore aims to identify marks, on the wear indicator 6, which will make it possible to characterize the wear indicator 6 so as to evaluate its state of wear. Step E8 is illustrated in [Fig.8].

[0067] In practice, these reference points A1, B1, C2 correspond for example to specific points of the wear indicator 6 such as its free ends or its ends at the level of the hydraulic crown 3. These are in particular visible ends, on the image Im or the part Iml, of the wear indicator 6. Here, the two reference points A1, B1 are located at the end of the wear indicator 6 positioned at the level of the hydraulic crown 3. The third reference point C2 is located at the level of the free end 6A of the wear indicator 6.

[0068] In the example of the wear measurement method described here, the identification of the at least three reference points Al, Bl, C2 (in the image Im or in the part Iml) is implemented by means of the second artificial neural network NN2.

[0069] The processor 20 then provides, as input to this second artificial neural network NN2, the image Im (acquired in step E2) or the part Iml (determined in step E6) on which the wear indicator 6 is present. This second artificial neural network NN2 is designed to provide as output coordinates of the three reference points A1, B1, C2. These coordinates are for example here expressed in pixels of the image Im or of the part Iml.

[0070] This second artificial neural network NN2 therefore presents here a function of visual recognition of object markers (here the reference points A1, B1, C2 of the wear indicator 6) in an image provided as input (here the image Im acquired in step E2 or the part Iml obtained in step E6).

[0071] The second artificial neural network NN2 is here a convolutional artificial neural network. It is for example an artificial neural network of the HR-Net type (for "High-Resolution Network" according to the commonly used Anglo-Saxon name). As is known, this type of artificial neural network allows, from an image provided as input, to detect objects, classify images or implement a semantic segmentation of the pixels of the image (i.e. to assign a category to each pixel of the image provided as input). More details concerning this HR-Net type of artificial neural network can be found in the article "Deep High-Resolution Representation Learning for Visual Recognition", by Jingdong Wang et al., arXiv: 1908.07919, August 20, 2019.

[0072] Alternatively, the second artificial neural network may be of the YOLO type (as described previously). Alternatively, the second artificial neural network may be of the LiteNet type (for “Lightweight Neural Network” according to the commonly used English term). These variants require less computing power and are therefore more suitable, for example, for implementation in a smartphone.

[0073] Thus, the second artificial neural network NN2 is designed to identify at least three reference points A1, B1, C2 on the wear indicator 6 which will allow to characterize it. More particularly, these reference points Al, Bl, C2 will make it possible to characterize the wear of the brake discs by determining dimensions associated with the wear indicator 6 (and derived from these three reference points Al, Bl, C2).

[0074] Prior to implementing the wear measurement method, a learning step (not shown in the figures) allows the training of this second artificial neural network NN2. For this, another set of basic images is used. The wear indicator 6 is visible on each of the images of this other set. For each basic image of this other set, three reference points on this wear indicator 6 are identified. The weights of the neurons are then adjusted as the learning step progresses.

[0075] At the end of step E8, the processor 20 therefore has three reference points A1, B1, C2 associated with the wear indicator 6 visible on the image Im (or on the part Iml) and which will make it possible to characterize the wear of the aircraft brake discs.

[0076] As shown in [Fig. 4], the wear measurement method continues in step E10. During this step, the processor 20 identifies at least one other reference point C1, A2, B2, A3, B3, C3 on the wear indicator 6. This other reference point C1, A2, B2, A3, B3, C3 is distinct from the three reference points A1, B1, C2 identified in step E8.

[0077] The other reference point C1, A2, B2, A3, B3, C3 has the same characteristics as the three reference points A1, B1, C2 identified in step E8: it also forms a reference mark identified on the wear indicator 6 and which will make it possible to characterize the wear of the discs of the aircraft brake 1.

[0078] The other reference point C1, A2, B2, A3, B3, C3 is also identified here by means of the implementation of the second artificial neural network NN2 described previously.

[0079] Here, as shown in [Fig.8], six other reference points C1, A2, B2, A3, B3, C3 are for example identified. Among these six other reference points C1, A2, B2, A3, B3, C3, two of them A2, B2 correspond for example to reference points associated with the free end of the wear indicator 6.

[0080] Two other reference points A3, B3 correspond to other reference points located at the end of the wear indicator 6 positioned at the level of the hydraulic crown 3. These two other reference points A3, B3 (associated with the hydraulic crown 3) are particularly advantageous because they make it possible to compensate for any bias in the wear measurement which would be associated with the angle of the shot (for the acquired image Im).

[0081] This step E10 is optional but it is advantageous because it makes it possible to have a greater number of reference points for measuring the wear of the aircraft brake discs. This then makes it possible to improve the accuracy of the wear measurement, in particular by limiting the perspective effects associated with the angle of view. view (of the acquired Im image).

[0082] The wear measurement method then continues with a step E12 of determining at least one wear measurement of the discs of the aircraft brake 1. More particularly, this wear measurement is implemented by determining at least one dimension associated with the wear indicator 6.

[0083] The dimension associated with the wear indicator 6 is for example here the length L or the width W of the visible wear indicator 6 identified on the acquired image Im. Preferably, the length L and the width W of the wear indicator 6 are determined in step E12 so as to improve the characterization of the wear of the discs of the aircraft brake 1.

[0084] In practice, the dimension associated with the wear indicator 6 is here determined from the three reference points A1, B1, C2 identified in step E8. More particularly, the reference points A1, B1 make it possible to determine the width W of the wear indicator 6. The middle of the reference points A1, B1 and the reference point C2 make it possible, for their part, to determine the length L of the wear indicator 6.

[0085] In practice, these dimensions are determined in pixels on the acquired image Im or on the part Iml. The processor 20 then converts the dimensions in pixels obtained into dimensions in the metric system.

[0086] For this, the diameter of the wear indicator 6 is used as reference data. This diameter corresponds to the width W of the wear indicator 6 measured on the image Im (or on the part Iml). The diameter of the wear indicator 6 therefore serves here as calibration data to deduce therefrom (by a rule of proportionality) the other dimensions of the wear indicator 6 (and in particular the length L of the wear indicator 6). By way of example, the wear indicator 6 here has a diameter less than 8 millimeters (mm), for example of the order of 7.94 mm.

[0087] Finally, the width W of the determined wear indicator 6 makes it possible to deduce the length L of the wear indicator 6, which is the dimension particularly relevant for evaluating the wear of the aircraft brake discs 1.

[0088] Thus, in step E12, the processor 20 has a dimension (here the length) of the wear indicator 6 visible on the acquired image Im (or on the part Iml) from which the wear of the discs of the aircraft brake 1 will be evaluated.

[0089] Optionally, if step E10 described previously has been implemented, the processor 20 has the three reference points A1, B1, C2 as well as the other reference points C1, A2, B2, A3, B3, C3 to determine the wear measurement of the aircraft brake discs 1.

[0090] The two reference points A1, B1 make it possible to determine a first width W1. The two other reference points A2, B2 make it possible to determine a second width W2.

[0091] The width W of the wear indicator 6 is determined from the first determined width W1 and the second determined width W2. More particularly, the width W of the wear indicator 6 is determined as being the average of the first width W1 and the second width W2.

[0092] The use of the two width values ​​(of the first width W1 and of the second width W2), each determined at one end of the wear indicator 6 is particularly advantageous because it makes it possible to limit the perspective effects associated with the shooting angle (of the acquired image Im). This then makes it possible to improve the determination of the width W of the wear indicator 6 from the acquired image Im (or of the part Iml).

[0093] This width W also makes it possible here to serve as a calibration for determining the length L of the wear indicator 6 as described previously (given that the diameter of the wear indicator 6 is known).

[0094] The use of the other reference points C1, A2, B2, A3, B3, C3 makes it possible to refine the determination of the length L of the wear indicator 6. The reference point C2 serves as the first end for determining the length L of the wear indicator 6.

[0095] A second end for determining the length L of the wear indicator 6 is determined from the different reference points A1, B1, C1, A3, B3, C3.

[0096] A first intermediate point Cl is determined as the barycenter of the two reference points Al, BL A second intermediate point C3 is determined as the barycenter of the other two reference points A3, B3.

[0097] The second end for determining the length L of the wear indicator 6 is for example here determined as the middle between the first intermediate point C1 and the second intermediate point C3. The definition of the second end for determining the length L of the wear indicator 6 is particularly advantageous because it makes it possible to limit the perspective effects associated with the viewing angle (of the acquired image Im). Indeed, by using only the first intermediate point C1, a bias could exist if the viewing angle is not exactly implemented at 90 degrees (this would then lead to an overestimation of the length of the wear indicator). This then makes it possible to improve the accuracy of the determined length L of the wear indicator 6.

[0098] The length L is then determined as the distance between, on the one hand, the barycenter of the first intermediate point C1 and the second intermediate point C3, and, on the other hand, the reference point C2. This length L is here obtained in pixels but converted into the metric system thanks to the calibration carried out from the known diameter of the wear indicator 6.

[0099] Thus, at the end of step E12, the processor 20 has a dimension (here the length) of the wear indicator 6 which will make it possible to conclude as to the wear of the disks. of aircraft brake 1.

[0100] More precisely, in step E14, the processor 20 determines the wear of the discs of the brake 1 of the aircraft by comparing the value of the dimension obtained in step E12 with a predefined value. This predefined value is associated with a maximum wear length of the wear indicator 6 (indicating wear of the discs of the aircraft brake 1 requiring the replacement of these discs of the aircraft brake 1). This maximum wear length is for example stored in the storage device 24 of the wear measuring device 10. This maximum wear length is for example here of the order of 1 millimeter (mm).

[0101] Thus, if the length L determined in step E12 is less than or equal to the stored maximum wear length, the processor 20 sends an alert signal indicating the wear of the discs of the brake concerned (step E16). This alert signal is presented for example in the form of a notification displaying a message recommending the replacement of the discs of the aircraft brake 1. For example here, a length L less than or equal to 1 mm leads to a flight ban for the aircraft, notified in the alert signal sent in step E12.

[0102] Otherwise, if the length L determined in step E12 is greater than the maximum wear length stored, the method resumes at step E2.

[0103] The method according to the invention is for example implemented periodically to check the wear of the aircraft brake discs 1.

[0104] Thus, thanks to the invention, the wear measurement method makes it possible to evaluate the state of wear of the aircraft brake discs remotely on the basis of an image acquired from the wear indicator. The analysis of this image then makes it possible to quantify the state of wear of the aircraft brake discs by measuring at least one dimension associated with this wear indicator. This measurement is carried out on the basis of at least three reference points, which makes it possible to obtain a reliable measurement of the characteristics of the wear indicator.

[0105] Advantageously, the present invention does not require any external equipment (such as a ruler for example) to evaluate the wear measurement of the aircraft brake discs 1.

[0106] Furthermore, no material modification needs to be made to the aircraft brake to enable the wear condition of the aircraft brake discs to be assessed. No additional material cost is therefore incurred and the invention can be implemented on any aircraft.

[0107] Finally, thanks to the present invention, the state of wear of the aircraft brake discs can be determined without risk (i.e. without getting too close to the heat sink whose temperature is very high), in an automated manner and in a precise manner thanks to the marks identified on the wear indicator.

[0108] The present invention finds a particularly advantageous application for monitoring the wear of the brakes of an aircraft. This is particularly useful for transmitting information to the user concerning the next replacement of the brake discs.

Claims

Claims

1. Method for measuring wear of the discs of an aircraft brake (1) provided with a wear indicator (6) mounted movably on a fixed part of the aircraft brake (1) and moving as the discs wear, the method comprising steps of: - acquiring an image (Im) of a part of the discs of the aircraft brake (1), - identifying the wear indicator (6) in the acquired image (Im), - identifying at least three reference points (A1, B1, C2) on the wear indicator (6), and - determining at least one wear measurement of the discs of the aircraft brake (1) by determining at least one dimension (L, W) associated with the wear indicator (6) from the three identified reference points (A1, B1, C2).

2. Method according to claim 1, in which, in the step of determining the wear measurement of the aircraft brake discs (1), provision is made to determine the width (W) and the length (L) of the wear indicator (6) from the three identified reference points (A1, B1, C2).

3. Method according to claim 1 or 2, in which there is provided, before the step of identifying the at least three reference points (Al, Bl, C2), a step of delimiting a reduced part (Iml) of the acquired image (Im) and comprising the wear indicator (6), the identification of the at least three reference points (Al, Bl, C2) being implemented from this reduced part (Iml) of the acquired image (Im).

4. Method according to any one of claims 1 to 3, in which the identification of the wear indicator (6) in the acquired image (Im) is implemented by means of an artificial neural network (NN1), the acquired image (Im) being provided as input to said artificial neural network (NN1) so as to obtain, as output from said artificial neural network (NN1), an area (Zl) of the acquired image (Im) on which the wear indicator (6) is identified.

5. Method according to claim 4, wherein the identification of the at least three reference points (A1, B1, C2) is implemented by analyzing the area (Zl) of the acquired image (Im) obtained at the output of the artificial neural network (NN1).

6. The method of claim 4 or 5, wherein the artificial neural network (NN1) is a convolutional artificial neural network.

7. Method according to any one of claims 4 to 6, in which there is provided, prior to the step of acquiring the image, a step of training the artificial neural network (NN1) from a set of basic images.

8. Method according to any one of claims 1 to 7, wherein the identification of the at least three reference points (Al, Bl, C2) on the wear indicator (6) is implemented by means of another artificial neural network (NN2), the acquired image (Im) in which the wear indicator (6) has been identified being provided as input to said other artificial neural network (NN2) so as to obtain, as output from said other artificial neural network (NN2), coordinates of the at least three reference points (Al, Bl, C2).

9. The method of claim 8, wherein the other artificial neural network (NN2) is a convolutional artificial neural network.

10. Method according to claim 8 or 9 taken in dependence on claim 3, in which the reduced part (Iml) of the acquired image (Im) is provided as input to the other artificial neural network (NN2) so as to obtain, as output from said other artificial neural network (NN2), the coordinates of the at least three reference points (Al, Bl, C2).

11. Method according to any one of claims 8 to 10, in which there is provided, prior to the step of acquiring the image (Im), a step of training the other artificial neural network (NN2) from another set of basic images.

12. Method according to any one of claims 1 to 11, in which, before the step of determining the wear measurement of the aircraft brake discs (1), a step of identifying at least one other reference point (Cl, A2, B2, A3, B3, C3) is provided, the determination of the wear measurement of the aircraft brake discs (1) being implemented taking into account this other reference point (Cl, A2, B2, A3, B3, C3).

13. Method according to any one of claims 1 to 12, wherein the at least three reference points (A1, B1, C2) correspond to visible ends of the wear indicator (6).

14. Device (10) for measuring the wear of the discs of an aircraft brake (1) provided with a wear indicator (6) mounted movably on a fixed part of the aircraft brake (1) and moving as the discs wear, the device (10) comprising: - an acquisition module (12) of an image (Im) of a part of the aircraft brake discs (1), - an identification module (14) of the wear indicator (6) in the acquired image (Im), - an identification module (16) of at least three reference points (A1, B1, C2) on the wear indicator (6), and - a module (18) for determining at least one measurement of wear of the aircraft brake discs (1) by determining at least one dimension associated with the wear indicator (6) from the three reference points (A1, B1, C2) identified.

15. A computer program comprising instructions executable by a processor (20) and designed to implement a method according to any one of claims 1 to 13 when these instructions are executed by the processor (20).

Images