Wing deformation measurement method and device

Abstract

The present invention proposes a wing deformation measurement method and device, which belongs to the field of flight testing. Set strain measurement points and displacement measurement points; collect wavelength data, calculate high-precision strain data of each strain measurement point; collect displacement data of each displacement measurement point; use the high-precision strain data as input variables, use the displacement data as output variables, screen out the input variables with the greatest correlation with the output variables, and each input variable constitutes a variable set with the greatest correlation; use the variable set with the greatest correlation as the input value of the BP network, construct the BP network, and obtain the BP neural network of each displacement measurement point; use the BP neural network training results to process the real-time high-precision strain data of each detection point, obtain the displacement information of each displacement measurement point, and form multiple deformation curves after the displacement spline difference of each point to obtain the real-time deformation information of the wing. It can realize real-time and accurate measurement of the wing deformation during the flight of the aircraft with fewer sensors.

Description

Technical Field

[0001] The invention belongs to the field of flight testing, and in particular relates to a wing deformation measurement method and device. Background Art

[0002] With the development of modern industry and science and technology, shape sensing technology has received more and more attention and has been widely used in various fields such as medicine, industrial robots, aerospace, etc. In the aviation field, for flexible deformable aircraft, shape sensing technology can be used to monitor the real-time deformation of its wings, so as to feedback the control system to adjust and optimize the wing shape to adapt to different mission environments. On the other hand, for sensor aircraft equipped with wing-integrated conformal radar systems, shape sensing technology is used to obtain wing deformation information in real time, and the phase error of each antenna module is compensated during beamforming to improve the radar detection performance of the sensor aircraft.

[0003] There are two methods for measuring the shape of the wing surface: non-contact method and contact method.

[0004] The first method is non-contact method, which uses stereo cameras, laser radar and other auxiliary devices to obtain the surface image of the measured structure to achieve high-precision deformation measurement. In the actual measurement process, this type of measurement method has many auxiliary measurement devices, which are large in size and poor in portability; the auxiliary measurement devices are inconvenient to install and have a limited measurement range. It is difficult to measure the deformation of the wing in the entire field, and it is impossible to measure the deformation of the wing in high precision and real time.

[0005] The second method is the contact method. Generally, various sensor arrays are set on the surface of the wing skin or inside the wing skin to measure discrete strain or acceleration data, and the algorithm is used to reconstruct the wing deformation. Among many sensors, fiber Bragg Grating (FBG) sensors are popular among researchers for their advantages such as anti-electromagnetic interference, light weight, and miniaturization. Especially in the field of shape sensing, the soft shape and integrated sensing / sensing of optical fiber enable it to fit the measured structural surface, and with the high-density fiber grating measurement points in the fiber core, a small amount of cables (relative to electrical sensors) can be used to cover a large range of structural surfaces, and shape reconstruction algorithms can be used to realize shape sensing functions.

[0006] At present, there are reports both at home and abroad on the use of FBG sensing technology to measure wing deformation. The core part of this technology is the curve reconstruction algorithm based on discrete strain data, mainly including the reconstruction algorithm based on the Frenet framework, the KO algorithm, etc. However, these methods all involve the interpolation and coordinate transfer of discrete data, and there are problems of error accumulation, which require a large number of high-precision sensor points. In 2008, NASA used the KO algorithm to test the wing shape of the IKhana UAV. 1440 sensor points were arranged on each wing to meet the requirements of high-precision shape sensing. In addition, the spacing of all measurement points must be accurately known in order to reconstruct the curve. For the FBG sensor array installed on the aircraft, it is difficult to accurately measure its spacing, especially for the optical fiber sensor placed inside the wing skin. In addition, the large number of FBG measurement points need to cooperate with the OFDR large-capacity demodulation system, but it cannot take into account high real-time performance (the wing conformal radar requires millisecond-level real-time measurement capabilities). On the other hand, when using this type of method, a large number of fiber grating sensor measurement points need to be calibrated one by one after installation, which is a large workload and is not conducive to measuring the deformation of the wing with high precision and high real-time performance. Summary of the invention

[0007] In view of the above problems, the present invention proposes a wing deformation measurement method and device, which at least partially solves the problems existing in the prior art and can achieve real-time and accurate measurement of the wing during the flight of the aircraft with fewer sensors.

[0008] In a first aspect, an embodiment of the present application provides a wing deformation measurement method, comprising:

[0009] The FBG sensor array is packaged on the wing, and the strain measurement points are set, while the displacement measurement points are also set;

[0010] Collect the central wavelength data corresponding to each FBG sensor, calculate the high-precision strain data of each strain measurement point; collect the displacement data of each displacement measurement point;

[0011] Using the high-precision strain data as input variables and the displacement data as output variables, respectively, screen out the input variables with the greatest correlation with the output variables, and each input variable constitutes a variable set with the greatest correlation;

[0012] Taking the variable set with the largest correlation as the input value of the BP network, constructing the BP network, and obtaining the BP neural network of each displacement measuring point;

[0013] The BP neural network training results are used to process the real-time high-precision strain data of each detection point to obtain the displacement information of each displacement measurement point. After the displacement spline difference of each point is calculated, multiple deformation curves are formed to obtain the real-time deformation information of the wing.

[0014] According to a specific implementation of the embodiment of the present application, the FBG sensor array is packaged inside the wing in an embedded packaging manner or on the wing in an adhesive packaging manner.

[0015] According to a specific implementation of the embodiment of the present application, packaging the FBG sensor array on the wing surface by adhesive packaging includes:

[0016] The outer surface of the wing is locally roughened;

[0017] The cleaning fluid removes grease from the roughened outer surface of the wing;

[0018] The FBG sensor array is bonded to the cleaned outer surface of the wing to complete the adhesive-coated packaged FBG sensor array.

[0019] According to a specific implementation of the embodiment of the present application, the bonding length at the strain measuring point is greater than or equal to 10 cm.

[0020] According to a specific implementation of the embodiment of the present application, packaging the FBG sensor array in an embedded packaging manner inside the wing includes:

[0021] The FBG sensor array is buried between a first ply and a second ply close to the surface along the wing span.

[0022] According to a specific implementation of the embodiment of the present application, a sleeve is provided on the root optical fiber of the FBG sensor array.

[0023] According to a specific implementation of the embodiment of the present application, each group of the FBG sensor array includes a plurality of FBG strain gauges arranged at equal intervals.

[0024] According to a specific implementation of the embodiment of the present application, each group of the FBG sensor arrays further includes a FBG thermometer disposed on the tail optical fiber of the group of FBG sensor arrays.

[0025] According to a specific implementation of the embodiment of the present application, the collecting of the central wavelength data corresponding to each FBG sensor to calculate the high-precision strain data of each strain measuring point; and the collecting of the displacement data of each displacement measuring point include:

[0026] In the initial state, the central wavelength values ​​of all FBG sensors are recorded;

[0027] Setting a plurality of displacement measuring points in the wing span direction and the chord direction, applying a plurality of groups of uniformly distributed loads in the wing span direction, monitoring the displacement data of the plurality of displacement measuring points, and recording the measured displacement data of the plurality of displacement measuring points;

[0028] At the same time, the central wavelength data of all FBG sensors are recorded.

[0029] According to a specific implementation of an embodiment of the present application, setting multiple displacement measurement points in the span direction and the chord direction of the wing includes:

[0030] The wing is divided into a units in the span direction and b units in the chord direction. The outer surface of the wing forms an a*b grid, and the nodes of the grid are the displacement measurement points.

[0031] According to a specific implementation of the embodiment of the present application, during the uniform load loading process, the measured strain value of each strain measuring point on the array surface is calculated as follows:

[0032]

[0033] in:

[0034] It represents the measured strain value of the jth strain measurement point in the i-th FBG sensor array, in units of με;

[0035] The change in the central wavelength measured at the strain measurement point, in nm;

[0036] ΔT is the temperature change in the temperature compensation area k where the strain measurement point is located, in °C;

[0037] is the temperature sensitivity coefficient of the strain measurement point, in pm / ℃;

[0038] is the strain sensitivity coefficient of the strain measuring point, in units of pm / με;

[0039] Δλ k is the change in the central wavelength of the temperature measuring point in the temperature compensation area k where the strain measuring point is located, in nm;

[0040] K kT is the temperature sensitivity coefficient of the temperature measuring point, in pm / ℃.

[0041] According to a specific implementation of the embodiment of the present application, based on the high-precision strain data, the input variables with the greatest correlation with the output variables of the high-precision strain data are screened out, and the input variables constitute a variable set with the greatest correlation, including:

[0042] Using the MIV algorithm Filter out the set of input variables that have the greatest correlation with the output variables of high-precision strain data

[0043] According to a specific implementation method of the embodiment of the present application, the constructing of the BP network further includes: training the BP network in real time according to the real-time update of the variable set with the greatest correlation, and constructing a BP network with optimized variables for each displacement measurement point.

[0044] In a second aspect, a wing deformation measuring device is provided, the device comprising:

[0045] The FBG sensing module is used to collect the central wavelength data of the FBG sensor array and calculate the high-precision strain data of each strain measurement point;

[0046] A displacement detection module is used to collect displacement data of each displacement measurement point;

[0047] A screening module, used to screen out input variables with the greatest correlation with output variables, each input variable forming a variable set with the greatest correlation, wherein the output variable is the high-precision strain data and the input variable is the displacement data;

[0048] A training module is used to use the variable set with the greatest correlation as the input value of the BP network, construct the BP network, and obtain the BP neural network of each displacement measurement point;

[0049] The operation module is used to obtain the displacement information of each displacement measurement point by using the BP neural network according to the wavelength data of the FBG sensor array collected in real time, and to form multiple deformation curves after taking the displacement spline difference of each point to obtain the real-time deformation information of the wing.

[0050] According to a specific implementation of an embodiment of the present application, the FBG sensor module is a FBG sensor array and a matching FBG wavelength demodulator, and the FBG sensor array is packaged inside the wing in an embedded packaging manner and / or packaged on the outer surface of the wing in an adhesive packaging manner.

[0051] According to a specific implementation of the embodiment of the present application, the FBG sensor array includes at least one group of FBG sensor groups, and each group of FBG sensors includes:

[0052] A number of FBG strain gauges are arranged on the optical fiber at equal intervals;

[0053] A FBG thermometer is arranged on the tail optical fiber of the FBG sensor array;

[0054] The sleeve is sleeved on the optical fiber at the root of the FBG sensor array.

[0055] The wing deformation measurement method and device of the present invention adopt FBG sensing technology, and a small and light FBG sensing network is arranged on the wing, so that the strain information measured by the optical fiber sensor can be directly associated with the real deformation information of the wing, thereby avoiding the error accumulation caused by coordinate recursion in the traditional algorithm. At the same time, a small number of sensor measuring points are used to achieve high-precision wing deformation measurement, reducing the complexity of the system. Other features and advantages of the present invention will be described in the subsequent description, and partly become obvious from the description, or can be understood by implementing the present invention. The objects and other advantages of the present invention can be achieved and obtained through the structures indicated in the description and the drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0057] Figure 1 A schematic diagram of the process of measuring wing deformation according to an embodiment of the present invention is shown;

[0058] Figure 2 It shows a partial structural schematic diagram of the FBG sensor array and wing arrangement of an embodiment of the present invention;

[0059] Figure 3 A schematic diagram showing the structural arrangement of the temperature measurement point in an embodiment of the present invention is shown;

[0060] Figure 4 A schematic diagram of a uniform load partitioning method obtained from wing modeling simulation results in an embodiment of the present invention is shown;

[0061] Figure 5 A schematic structural diagram of a wing deformation training device in an experimental environment according to an embodiment of the present invention is shown;

[0062] Figure 6 A schematic structural diagram of a wing deformation training device according to an embodiment of the present invention is shown;

[0063] Figure 7 A schematic diagram of a strain measurement point screening method based on the MIV algorithm according to an embodiment of the present invention is shown;

[0064] Figure 8 The results of the wing deformation measurement method according to the embodiment of the present invention and the results of the conventional deformation measurement method are shown; and

[0065] Fig. 9A schematic structural diagram of a wing deformation measuring device according to an embodiment of the present invention is shown. DETAILED DESCRIPTION

[0066] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0067] It should be noted that various aspects of the embodiments within the scope of the appended claims are described below. It should be apparent that the aspects described herein can be embodied in a wide variety of forms, and any specific structure and / or function described herein is merely illustrative. Based on the present application, it should be understood by those skilled in the art that an aspect described herein can be implemented independently of any other aspect, and two or more of these aspects can be combined in various ways. For example, any number of aspects described herein can be used to implement the device and / or practice the method. In addition, other structures and / or functionalities other than one or more of the aspects described herein can be used to implement this device and / or practice this method.

[0068] It should also be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present application. The drawings only show components related to the present application rather than being drawn according to the number, shape and size of components in actual implementation. In actual implementation, the type, quantity and proportion of each component may be changed arbitrarily, and the component layout may also be more complicated.

[0069] Additionally, in the following description, specific details are provided to facilitate a thorough understanding of the examples. However, it will be understood by those skilled in the art that the aspects described may be practiced without these specific details.

[0070] The embodiment of the present application provides a wing deformation measurement method, including: installing an FBG sensor array seal 200 on a wing 100, setting strain measurement points and displacement measurement points; collecting central wavelength data corresponding to each FBG sensor, and calculating high-precision strain data of each strain measurement point; collecting displacement data of each displacement measurement point; using the high-precision strain data as input variables and the displacement data as output variables, respectively, screening out input variables with the greatest correlation with the output variables, and each input variable constitutes a variable set with the greatest correlation; using the variable set with the greatest correlation as input values ​​of a BP network, constructing a BP network, and obtaining a BP neural network for each displacement measurement point; using the BP neural network training results to process the real-time high-precision strain data of each detection point, obtaining displacement information of each displacement measurement point, and forming multiple deformation curves after taking the displacement spline difference of each point to obtain real-time deformation information of the wing 100. By adopting FBG sensing technology, a small and lightweight FBG sensor network is deployed on the wing 100. Compared with non-contact measurement methods such as camera method, it has less impact on the shape of the wing 100, and it is easier to directly associate the strain information measured by the optical fiber sensor with the real deformation information of the wing 100, avoiding the error accumulation caused by coordinate recursion in the traditional algorithm, and realizing high-precision deformation measurement of the wing 100 with a smaller number of sensor measurement points, thereby reducing the complexity of the system.

[0071] Next, with reference to the accompanying drawings, the wing deformation measurement method according to an embodiment of the present application is described in detail.

[0072] Please refer to Figure 1 , Figure 1 FIG. 1 is a schematic diagram showing a flow chart of a method for measuring deformation of a wing 100 according to an embodiment of the present invention. Figure 1 As shown, the method comprises the following steps:

[0073] Step S100, encapsulating the FBG sensor array on the wing, setting strain measurement points, and setting displacement measurement points.

[0074] Furthermore, the FBG sensor array 200 is packaged on the wing 100 in an embedded package and / or adhesive package manner. That is, the FBG sensor array 200 is arranged on the surface of the wing 100 or inside the skin of the wing 100, and multiple FBG sensor arrays 200 are arranged in parallel in the span direction of the wing 100. At the same time, multiple displacement measurement points are set in the span direction and chord direction of the wing.

[0075] More specifically, a FBG sensor array 200 is arranged on the surface of the wing 100, and a plurality of FBG sensor arrays 200 are arranged in parallel in the span direction of the wing 100. Figure 2 , Figure 2 FIG. 2 shows a schematic diagram of a partial structure of the FBG sensor array 200 and the wing 100 of an embodiment of the present invention after the arrangement is completed. Figure 2 As shown, the FBG sensor array 200 includes a number of FBG strain gauges 201 equidistantly arranged on the optical fiber 207, and the location where the FBG strain gauge 201 is arranged corresponds to the strain measurement point. The FBG sensor array 200 also includes a FBG thermometer 202 arranged on the tail optical fiber 207 of the group of FBG sensor arrays 200, and the location where the FBG thermometer 202 is arranged corresponds to the temperature measurement point and a sleeve 203 sleeved on the root optical fiber of the FBG sensor array 200 to play a protective role, ensuring that the optical fiber can be safely led out from the wing 100 part. The sleeve 203 in this embodiment is a nylon sleeve. The FBG sensor array 200 is bonded to the surface of the wing 100 using X120 two-component epoxy adhesive. Before bonding, the surface of the wing 100 needs to be roughened. In this embodiment, the surface of the wing 100 is roughened by sandpaper polishing, and acetone is used to clean and remove surface grease. For each strain measurement point, the bonding length should be more than 10 cm. It is worth noting that the strain measurement points are a series of polyimide coated FBG arrays (excluding the FBG at the tail end) written on an optical fiber 207 at equal intervals, and the multiple optical fibers 207 are distributed in the span direction of the wing 100. Please refer to Figure 3 , Figure 3 FIG. 2 shows a schematic diagram of the structure arrangement at the temperature measurement point of an embodiment of the present invention. Figure 3 As shown, the processing method of the temperature measurement point is:

[0076] The last grating 204 of the FBG sensor array 200 is inserted into the capillary steel tube 205 for packaging. The grating 204 is suspended inside the capillary steel tube 205. The optical fiber 207 extends to the outside along the axial direction of the capillary steel tube 205. The root of the optical fiber 207 and the end of the capillary steel tube 205 is connected. The root is glued and cured to form a fixed protective member 206. The inner diameter of the capillary steel tube 205 is greater than 0.5 mm.

[0077] The root glue curing process can ensure that the measuring point is not affected by the structural strain of the wing 100, but only by the temperature. The temperature measuring points at the tail end of the multiple FBG sensor arrays 200 are evenly distributed in the span direction of the wing 100, and are used to perform real-time temperature compensation on the strain measuring points in their respective temperature compensation areas to obtain accurate strain information. The temperature compensation area adopts the principle of proximity, and selects the temperature measuring point closest to each strain measuring point for compensation.

[0078] Multiple displacement measurement points are set in the span direction and chord direction of the wing 100, including:

[0079] The wing is divided into a units in the span direction and b units in the chord direction. The outer surface of the wing 100 forms an a*b grid, and the nodes of the grid are the displacement measurement points.

[0080] It is worth noting that, under ideal conditions, the distribution of the strain measurement points and the displacement measurement points corresponding to the distribution of the FBG sensors are evenly and densely distributed on the wing 100, and the two overlap.

[0081] Step S200: Modeling and simulating the wing 100 to obtain uniform load conditions on the wing 100 under different overloads during flight.

[0082] Assume that the maximum expected overload of the aircraft during flight is a. After leaving a 20% margin, the maximum overload is set to 1.2a, and the step size is set to 0.05a. 24 sets of overload data are obtained between the initial state and the maximum overload. The area is divided into 5 loading blocks in the span direction. The uniformly distributed loads in the 5 areas in the span direction of the wing 100 under each set of wing 100 overload data are calculated using simulation software. Please refer to Figure 4 , Figure 4 A schematic diagram of a uniform load partitioning method obtained from the wing 100 modeling simulation results in an embodiment of the present invention is shown.

[0083] Based on the above simulation results of the wing 100, an actuator with a controller is used to apply 24 groups of uniformly distributed loads in the span direction of the wing, and each group of states stays for a period of time for system stability and data preservation. At the same time, for multiple displacement measurement points set on the wing, a displacement detector is used to monitor the displacement data of these points, and the data is transmitted to a computer. Exemplarily, the displacement detector in this embodiment uses a laser displacement sensor.

[0084] By using the actuator to load the wing with displacement, the flight condition of the aircraft during normal flight can be effectively simulated. This ensures that the training results are more consistent with the actual results, and thus the detection is more accurate during the actual aircraft flight. This step is completed in the laboratory stage. When measuring the deformation of the wing during the normal operation of the aircraft, this step is not necessary.

[0085] Step S300, collecting the central wavelength data corresponding to each FBG sensor, calculating the high-precision strain data of each strain measurement point; collecting the displacement data of each displacement measurement point.

[0086] Specifically, the wing 100 is kept in a straight state in the span direction, at which time the displacement of each point of the wing 100 is zero, and this state is defined as the initial state, and the central wavelength data values ​​of all FBG sensors measured in this state are recorded; using the simulation results obtained in step S200, a uniform load is applied in the span direction of the wing 100. At the same time, the displacement detection module is used to monitor and collect the displacement data of each displacement measurement point. At the same time, the computer records the central wavelength data of the corresponding FBG sensor in real time.

[0087] For example, please refer to Figure 5 , Figure 5FIG. 2 shows a schematic diagram of the structure of a wing deformation training device implemented in an experimental environment according to the present invention. Figure 5 As shown, in the actual measurement process under the experimental environment, the FBG sensor array 200 is connected to a FBG wavelength demodulator 400 through a flange 300, and the FBG wavelength demodulator 400 is connected to a computer 500. While ensuring the shape measurement accuracy, a high-speed FBG wavelength demodulator based on wavelength division multiplexing is used. The demodulation rate of the FBG wavelength demodulator 400 selected in this embodiment is ≥1kHz, which can achieve a data processing speed of more than kHz (the existing OFDR technology rate is 100Hz), meeting the high real-time deformation application such as the wing 100 conformal radar.

[0088] At room temperature and pressure, the wing 100 is kept in a straight state in the span direction. At this time, the displacement of each point of the wing 100 is zero. This state is defined as the initial state. The computer records the central wavelength data values ​​of all FBGs measured by the FBG wavelength demodulator 400 in this state. Using the simulation results obtained in step S200, an actuator 600 with a controller 601 is used to apply 24 groups of uniformly distributed loads in the span direction of the wing 100. Each group of states stays for at least 10 minutes to ensure system stability and data preservation. At the same time, for multiple displacement measurement points set in the span direction and chord direction of the wing 100, the displacement data of these points are monitored using a laser displacement sensor 700, and the data are transmitted to the computer 500, recorded as It represents the measured displacement data of the displacement measurement point in row a and column b. Similarly, at this time, the computer records the corresponding FBG central wavelength data in real time.

[0089] During the uniform load loading process, the strain at each strain measuring point on the array surface The calculation method is as follows:

[0090]

[0091] in, represents the measured strain value of the jth strain measurement point in the i-th FBG sensor array 200, in units of με; is the change in the central wavelength measured at the strain measuring point, in nm; ΔT is the temperature change in the temperature compensation area k where the strain measuring point is located, in °C; is the temperature sensitivity coefficient of the strain measurement point, in pm / ℃; is the strain sensitivity coefficient of the strain measuring point, which is 1.209 here, and the unit is pm / με; Δλ k K is the change in the central wavelength of the temperature measuring point in the temperature compensation area k where the strain measuring point is located, in nm; kT is the temperature sensitivity coefficient of the temperature measuring point, in pm / ℃.

[0092] in, and K kT It can be obtained through temperature test calibration. The method is to make a small piece of structural parts with the same material as the wing 100, and paste / bury the strain and temperature measuring points on it using the same process. After placing it in the temperature box, adjust the temperature, record the wavelength data of the strain and temperature measuring points at different temperatures, and after linear fitting, the linear term coefficient is and K kT .

[0093] When the aircraft is out of the experimental environment and is operating normally, there is no need to use the actuating device 600 with the controller 601 to apply 24 groups of uniformly distributed loads in the span direction of the wing 100. Specifically, the structural schematic diagram of the wing deformation training device according to the embodiment of the present invention is as follows: Figure 6 shown.

[0094] Step S400, using the high-precision strain data as input variables and the displacement data as output variables, respectively, screen out the input variables with the greatest correlation with the output variables, and each input variable constitutes a variable set with the greatest correlation.

[0095] Based on the above output test data, each displacement measurement point on the wing 100 Using MIV (Mean Impact Value) algorithm from Filter out the set of input variables with the greatest correlation

[0096] Specifically, after the network training is terminated, each independent variable feature in the training sample P is increased / decreased by 10% based on its original value to form two new training samples P1 and P2. P1 and P2 are used as simulation samples and simulated using the built network to obtain two simulation results A1 and A2. The difference between A1 and A2 is calculated, which is the impact value (Impact Value, IV) of the output after changing the independent variable. Finally, the IV is averaged according to the number of observations to obtain the MIV of the independent variable for the dependent variable - network output. According to the above steps, the MIV value of each independent variable is calculated in turn. Finally, the variables are sorted according to the absolute value of the MIV to obtain the relative importance of the variables on the network output, so as to determine the degree of influence of the input features on the network results, that is, to achieve variable screening.

[0097] For example, please refer to Figure 7 , Figure 7 FIG. 2 shows a flow chart of a strain measurement point screening method based on the MIV algorithm according to an embodiment of the present invention. Figure 7 As shown, the specific method is as follows:

[0098] Step S401: For input variables After increasing / decreasing by 10%, a new set is formed and

[0099] Step S402: For each displacement measurement point, a BP neural network is constructed using the original data, and an input layer, a hidden layer and an output layer are set.

[0100] The input data is the measured data value of each strain measurement point under all uniformly distributed loads. The output data is the measured value of the current displacement measurement point under all uniformly distributed loads. Set the BP neural network parameters, including learning rate, number of iterations, error index, and initialize the network.

[0101] Step S403: For each displacement measurement point, use the BP neural network constructed in step S402 to calculate all and As the network output value after input data and Calculate their respective value:

[0102]

[0103] Step S404: Sort the values ​​in descending order and select the top 50% The corresponding strain measurement point set That is, the input variable that has the greatest impact on the displacement measurement point, and it is used as the input value of the subsequent BP network.

[0104] Step S500: taking the variable set with the largest correlation as the input value of the BP network, constructing the BP network, and obtaining the BP neural network of each displacement measuring point.

[0105] Specifically, according to the results of S400, a BP network with optimized variables is constructed for each displacement measuring point, and the variables are updated in real time, while other parameters are kept unchanged. The training is repeated to obtain a BP neural network for each displacement measuring point.

[0106] According to the overload condition of the aircraft wing 100 during actual flight, the approximate deformation distribution law is obtained by simulation calculation, so that the BP neural network training is more targeted, the processing of invalid data is avoided, and the accuracy of the deformation measurement of the wing 100 during actual flight is improved.

[0107] Step S600, using the BP neural network training results to process the real-time high-precision strain data of each detection point, obtain the displacement information of each displacement measurement point, and form multiple deformation curves after taking the displacement spline difference of each point to obtain the real-time deformation information of the wing 100.

[0108] The real-time wavelength data in this embodiment is collected by a high-speed fiber Bragg grating wavelength demodulator (demodulation rate ≥ 1 kHz), and the deformation curve formed is a deformation curve that is smooth in the span direction.

[0109] The results obtained by using the deformation measurement method of the wing 100 of the present invention are as follows: Figure 8 As shown. It can be seen that compared with the traditional deformation measurement algorithm, it is more consistent with the measured data. Taking the terminal displacement data as an example, due to the small number of measurement points, the error of the traditional algorithm reaches the maximum after cumulative transmission to the tip, which is about 3.7mm different from the true value. The result of the detection using the wing 100 deformation measurement method provided by the present invention is that the error of the wing 100 terminal displacement reached is about 0.6mm, and the accuracy is greatly improved.

[0110] Please refer to Fig. 9 , Fig. 9 FIG. 2 shows a schematic diagram of the structure of a wing deformation measuring device according to an embodiment of the present invention. Fig. 9 As shown, the device comprises:

[0111] The FBG sensor module is used to collect the central wave data of the FBG sensor array and calculate the high-precision strain data of each strain measurement point;

[0112] A displacement detection module is used to collect displacement data of each displacement measurement point;

[0113] A screening module, used to screen out input variables with the greatest correlation with output variables, each input variable forming a variable set with the greatest correlation, wherein the output variable is the high-precision strain data and the input variable is the displacement data;

[0114] A training module is used to use the variable set with the greatest correlation as the input value of the BP network, construct the BP network, and obtain the BP neural network of each displacement measurement point;

[0115] The operation module is used to obtain the displacement information of each displacement measurement point by using the BP neural network according to the wavelength data of the FBG sensor array collected in real time, and to form multiple deformation curves after taking the displacement spline difference of each point to obtain the real-time deformation information of the wing.

[0116] Furthermore, the FBG sensor module is a FBG sensor array 200 and a matching FBG wavelength demodulator, and the FBG sensor array 200 is packaged inside the wing 100 in an embedded packaging manner and / or packaged on the outer surface of the wing 100 in an adhesive packaging manner.

[0117] Furthermore, the FBG sensor array 200 includes: at least one group of FBG sensor groups, each group of FBG sensors includes:

[0118] A plurality of FBG strain gauges 201 are arranged on the optical fiber at equal intervals; an FBG thermometer 202 is arranged on the tail optical fiber of the FBG sensor array 200; and a sleeve 203 is sleeved on the optical fiber at the root of the FBG sensor array 200.

[0119] Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent substitutions for some of the technical features therein; and these modifications or substitutions do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for measuring wing deformation, characterized in that: include: The FBG sensor array is packaged on the wing, and the strain measurement points are set, while the displacement measurement points are also set; Collect the central wavelength data corresponding to each FBG sensor and calculate the high-precision strain data of each strain measurement point; Collect displacement data of each displacement measuring point, including: In the initial state, the central wavelength values ​​of all FBG sensors are recorded; A plurality of displacement measuring points are set in the span direction and the chord direction of the wing, a plurality of groups of uniformly distributed loads are applied in the span direction of the wing, displacement data of the plurality of displacement measuring points are monitored, and the measured displacement data of the plurality of displacement measuring points are recorded; at the same time, the central wavelength data of all FBG sensors are recorded; Using the high-precision strain data as input variables and the displacement data as output variables, respectively, screen out the input variables with the greatest correlation with the output variables, and each input variable constitutes a variable set with the greatest correlation; Taking the variable set with the largest correlation as the input value of the BP network, constructing the BP network, and obtaining the BP neural network of each displacement measuring point; The BP neural network training results are used to process the real-time high-precision strain data of each detection point to obtain the displacement information of each displacement measurement point. After the displacement spline difference of each point is calculated, multiple deformation curves are formed to obtain the real-time deformation information of the wing.

2. The wing deformation measurement method according to claim 1, characterized in that: The FBG sensor array is packaged inside the wing in an embedded packaging manner or on the wing in an adhesive packaging manner.

3. The wing deformation measurement method according to claim 2, characterized in that: The FBG sensor array is packaged on the wing surface by adhesive packaging, including: The outer surface of the wing is locally roughened; The cleaning fluid removes grease from the roughened outer surface of the wing; The FBG sensor array is bonded to the cleaned outer surface of the wing to complete the adhesive-coated packaged FBG sensor array.

4. The wing deformation measurement method according to claim 3, characterized in that: The bonding length at the strain measuring point is greater than or equal to 10 cm.

5. The wing deformation measurement method according to claim 2, characterized in that: The FBG sensor array is packaged in an embedded package inside the wing including: The FBG sensor array is buried between a first ply and a second ply close to the surface along the wing span.

6. The wing deformation measurement method according to claim 1, characterized in that: A sleeve is provided on the root optical fiber of the FBG sensor array.

7. The wing deformation measurement method according to claim 1, characterized in that: Each group of the FBG sensor arrays includes a number of FBG strain gauges that are equidistantly arranged.

8. The wing deformation measurement method according to claim 7, characterized in that: Each group of the FBG sensor arrays further includes a FBG thermometer arranged on the tail optical fiber of the group of FBG sensor arrays.

9. The wing deformation measurement method according to claim 1, characterized in that: The setting of multiple displacement measuring points in the span direction and the chord direction of the wing includes: The wing is divided into a units in the span direction and b units in the chord direction. The outer surface of the wing forms an a*b grid, and the nodes of the grid are the displacement measurement points.

10. The wing deformation measurement method according to claim 6, characterized in that: During the uniform load loading process, the measured strain values ​​of each strain measuring point on the array surface are calculated as follows: in: It represents the measured strain value of the jth strain measurement point in the i-th FBG sensor array, in units of με; The change in the central wavelength measured at the strain measurement point, in nm; ΔT is the temperature change in the temperature compensation area k where the strain measurement point is located, in °C; is the temperature sensitivity coefficient of the strain measurement point, in pm / ℃; is the strain sensitivity coefficient of the strain measuring point, in units of pm / με; Δλ k is the change in the central wavelength of the temperature compensation area k where the strain measurement point is located, and the unit is nm; K kT is the temperature sensitivity coefficient of the temperature measuring point, in pm / ℃.

11. The wing deformation measurement method according to claim 10, characterized in that: Based on the high-precision strain data, the input variables with the greatest correlation with the output variables of the high-precision strain data are screened out. The variable set with the greatest correlation formed by each input variable includes: Using the MIV algorithm Filter out the set of input variables that have the greatest correlation with the output variables of high-precision strain data 12. The wing deformation measurement method according to claim 1, characterized in that: The constructing of the BP network further comprises: training the BP network in real time according to the real-time updating of the variable set with the greatest correlation, and constructing a BP network with optimized variables for each displacement measuring point.

13. A wing deformation measuring device, characterized in that: The device comprises: The FBG sensing module is used to collect the central wavelength data of the FBG sensor array and calculate the high-precision strain data of each strain measurement point; the displacement detection module is used to collect the displacement data of each displacement measurement point, including: In the initial state, the central wavelength values ​​of all FBG sensors are recorded; A plurality of displacement measuring points are set in the span direction and the chord direction of the wing, a plurality of groups of uniformly distributed loads are applied in the span direction of the wing, displacement data of the plurality of displacement measuring points are monitored, and the measured displacement data of the plurality of displacement measuring points are recorded; at the same time, the central wavelength data of all FBG sensors are recorded; A screening module, used to screen out input variables with the greatest correlation with output variables, each input variable forming a variable set with the greatest correlation, wherein the output variable is the high-precision strain data and the input variable is the displacement data; A training module is used to use the variable set with the greatest correlation as the input value of the BP network, construct the BP network, and obtain the BP neural network of each displacement measurement point; The operation module is used to obtain the displacement information of each displacement measurement point by using the BP neural network according to the wavelength data of the FBG sensor array collected in real time, and to form multiple deformation curves after taking the displacement spline difference of each point to obtain the real-time deformation information of the wing.

14. The wing deformation measuring device according to claim 13, characterized in that: The FBG sensor module is a FBG sensor array and a matching FBG wavelength demodulator. The FBG sensor array is packaged inside the wing in an embedded packaging manner and / or is packaged on the outer surface of the wing in an adhesive packaging manner.

15. The wing deformation measuring device according to claim 13, characterized in that: The FBG sensor array comprises at least one group of FBG sensor groups, each group of FBG sensors comprising: A number of FBG strain gauges are arranged on the optical fiber at equal intervals; A FBG thermometer is arranged on the tail optical fiber of the FBG sensor array; The sleeve is sleeved on the optical fiber at the root of the FBG sensor array.

Images