A method and device for locating the impact probability of a composite material

Abstract

The present application provides a method and device for locating the impact probability of a composite material. The method includes: meshing a square plate of the composite material to be measured at a first spacing to obtain a plurality of first grids; after an impact object impacts the square plate of the composite material to be measured, determining the time difference of the impact signals collected by each pair of sensors; using the Gaussian process regression model corresponding to each pair of sensors to process the time difference and the position of each first grid, to obtain the likelihood value of each first grid under each pair of sensors; performing weighted summation on the multiple likelihood values of each first grid to obtain the positioning probability of the first grid; and determining the center point position of the first grid corresponding to the maximum positioning probability as the impact position of the square plate of the composite material to be measured. The present application improves the impact positioning accuracy of the composite material, thereby improving the damage detection accuracy caused by the impact of the aircraft on the composite material, and has the advantages of low cost, time saving and labor saving.

Description

Technical Field

[0001] This application relates to the technical field of impact damage location detection, and in particular, to a method and device for locating the impact probability of composite materials. Background Art

[0002] Currently, in the identification of damage caused by aircraft impacts, non-destructive testing methods are mainly used, such as ultrasonic testing, eddy current testing, and magnetic particle testing. These methods often require expensive and bulky equipment and are carried out on disassembled structures. In addition, these methods rely on the careful planning of ground personnel and are carried out regularly. Therefore, they have many problems such as a large workload, a long time consumption, and difficulty in ensuring the safety of the aircraft during operation.

[0003] In the impact monitoring of composite materials, accurate identification of the impact location helps to determine the location of possible aircraft damage, thereby accelerating the detection and maintenance of key parts of the aircraft and effectively reducing the time cost and economic cost of structural maintenance.

[0004] For the accurate identification of the impact location of composite materials, no effective solution has been proposed yet. Summary of the Invention

[0005] In view of this, this application provides a method and device for locating the impact probability of composite materials to solve the above technical problems.

[0006] In a first aspect, an embodiment of this application provides a method for locating the impact probability of composite materials, which is used to locate the impact on a square plate of a composite material to be measured, and a plurality of sensors are arranged around the square plate of the composite material to be measured; the method includes:

[0007] Grid the square plate of the composite material to be measured according to a first spacing to obtain a plurality of first grids, and determine the position of the center point of each first grid as the position of the first grid;

[0008] After an impact object impacts the square plate of the composite material to be measured, determine the time difference between the impact signals collected by each pair of sensors;

[0009] Use the Gaussian process regression model corresponding to each pair of sensors to process the time difference and the position of each first grid to obtain the likelihood value of each first grid under each pair of sensors; perform weighted summation on the multiple likelihood values of each first grid to obtain the positioning probability of the first grid;

[0010] Determine the position of the center point of the first grid corresponding to the maximum positioning probability as the impact position of the square plate of the composite material to be measured.

[0011] In a possible implementation, the time difference and the position of each first grid are processed using the Gaussian process regression model corresponding to each pair of sensors to obtain the likelihood value of each first grid under each pair of sensors, including:

[0012] For the j-th Gaussian process regression model corresponding to the j-th pair of sensors, at the center point position of the first grid The likelihood value Satisfies:

[0013]

[0014] Wherein, Is the time difference of the j-th pair of sensors; Is the optimal estimated value of the hyperparameter vector of the j-th pair of sensors Is the center point position of the first grid The predicted value at the place, Is The mean value of, Is The variance of.

[0015] In a possible implementation, the method further includes:

[0016] The square plate of the composite material to be measured is meshed according to the second pitch to obtain a plurality of second grids, and the position of the center point of each second grid is determined as the position of the second grid; wherein, the second pitch is greater than the first pitch;

[0017] The steel balls are successively dropped from a preset height onto each second grid of the square plate of the composite material to be measured, and the time difference of the j-th pair of sensors for each second grid is obtained;

[0018] Calculate the log-likelihood value :

[0019]

[0020] Wherein, N is the number of second grids, Is the time difference vector, and the n-th component is the time difference of the j-th pair of sensors for the n-th second grid, Is the time difference vector The kernel matrix of, Is the center point position vector of N second grids; Is the hyperparameter vector of the j-th pair of sensors to be estimated, including: noise variance 、Signal variance And length scale parameter ;

[0021] The optimal estimated value of the hyperparameter vector of the j-th pair of sensors Is:

[0022]

[0023] Search for the optimal estimate through a search algorithm to minimize the negative log-likelihood value.

[0024] In a possible implementation, the search algorithm adopts the whale optimization algorithm;

[0025] Search for the optimal estimate through a search algorithm to minimize the negative log-likelihood value, including:

[0026] Initialize M search particles of the population, and use the initialized M search particles as the M search particles for the first iteration; where the position of each search particle is represented as the hyperparameter vector of the jth pair of sensors;

[0027] Determine the search particle with the smallest fitness value among the initialized M search particles as the optimal search particle for the first iteration;

[0028] Repeat the following steps:

[0029] Update the positions of the M search particles in the previous iteration to obtain the positions of the M search particles in the current iteration;

[0030] Substitute the position of each search particle into the fitness function to obtain the fitness value of each search particle; determine the search particle with the smallest fitness value as the optimal search particle for the current iteration;

[0031] Based on the optimal search particle in the previous iteration and the optimal search particle in the current iteration, update the optimal search particle in the current iteration;

[0032] Until the current iteration reaches the preset maximum number of iterations;

[0033] Determine the optimal search particle with the maximum number of iterations as the optimal estimate .

[0034] In a second aspect, an impact probability positioning device for composite materials provided by an embodiment of the present application is used to perform impact positioning on a square plate of a composite material to be measured, and a plurality of sensors are arranged around the square plate of the composite material to be measured; including:

[0035] A grid unit, configured to grid the square plate of the composite material to be measured according to a first spacing to obtain a plurality of first grids, and determine the position of the center point of each first grid as the position of the first grid;

[0036] A determination unit, configured to determine the time difference between the impact signals collected by each pair of sensors after an impact object impacts a square plate of a composite material to be measured;

[0037] A processing unit, configured to process the time difference and the position of each first grid by using a Gaussian process regression model corresponding to each pair of sensors, to obtain a likelihood value of each first grid under each pair of sensors; perform weighted summation on the multiple likelihood values of each first grid to obtain the positioning probability of the first grid;

[0038] A positioning unit, configured to determine the center point position of the first grid corresponding to the maximum positioning probability as the impact position of the square plate of the composite material to be measured.

[0039] In a possible implementation, processing the time difference and the position of each first grid by using a Gaussian process regression model corresponding to each pair of sensors to obtain a likelihood value of each first grid under each pair of sensors includes:

[0040] For the j-th Gaussian process regression model corresponding to the j-th pair of sensors, at the center point position of the first grid the likelihood value satisfies:

[0041]

[0042] wherein, is the time difference of the j-th pair of sensors; is the optimal estimated value of the hyperparameter vector of the j-th pair of sensors is the predicted value at the center point position of the first grid, is the mean of, is the variance of.

[0043] In a possible implementation, the device further includes a training unit, specifically configured to:

[0044] Perform grid division on the square plate of the composite material to be measured according to a second pitch to obtain a plurality of second grids, and determine the position of the center point of each second grid as the position of the second grid; wherein, the second pitch is greater than the first pitch;

[0045] Sequentially drop steel balls from a preset height onto each second grid of the square plate of the composite material to be measured, and obtain the time difference of the j-th pair of sensors for each second grid;

[0046] Calculate the log-likelihood value :

[0047]

[0048] where N is the number of the second grids, is the time difference vector, and the n-th component is the time difference of the j-th pair of sensors in the n-th second grid, is the time difference vector of the kernel matrix, is the central point position vector of N second grids; is the hyperparameter vector of the j-th pair of sensors to be estimated, including: noise variance , signal variance and length scale parameter ;

[0049] The optimal estimated value of the hyperparameter vector of the j-th pair of sensors is:

[0050]

[0051] The optimal estimated value searched by the search algorithm is obtained to minimize the negative log-likelihood value.

[0052] In a possible implementation, the search algorithm adopts the whale optimization algorithm;

[0053] The optimal estimated value searched by the search algorithm is obtained to minimize the negative log-likelihood value, including:

[0054] Initialize M search particles of the population, and use the initialized M search particles as the M search particles in the first iteration; where the position of each search particle is represented as the hyperparameter vector of the j-th pair of sensors;

[0055] Determine the search particle with the minimum fitness value among the initialized M search particles as the optimal search particle in the first iteration;

[0056] Repeat the following steps:

[0057] Update the positions of the M search particles in the previous iteration to obtain the positions of the M search particles in the current iteration;

[0058] Substitute the position of each search particle into the fitness function to obtain the fitness value of each search particle; determine the search particle with the minimum fitness value as the optimal search particle in the current iteration;

[0059] Update the optimal search particle in the current iteration based on the optimal search particle in the previous iteration and the optimal search particle in the current iteration;

[0060] Until the current iteration number reaches the preset maximum iteration number;

[0061] Determine the optimal search particle with the maximum number of iterations as the optimal estimated value 。

[0062] In a third aspect, an embodiment of the present application provides an electronic device, including: a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, the method of the embodiment of the present application is implemented.

[0063] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium storing computer instructions, and when the computer instructions are executed by a processor, the method of the embodiment of the present application is implemented.

[0064] The present application improves the impact positioning accuracy of the composite material, thereby improving the damage detection accuracy caused by the impact of the aircraft on the composite material, and has the advantages of low cost, time saving and labor saving. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] In order to more clearly illustrate the specific embodiments of the present application or the technical solutions in the prior art, the following will briefly introduce the drawings required for the description of the specific embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0066] Figure 1 It is a flowchart of the composite material impact probability positioning method provided by the embodiment of the present application;

[0067] Figure 2 It is a schematic diagram of the experimental device provided by the embodiment of the present application;

[0068] Figure 3 It is a schematic diagram of the arrival times of signals received by four sensors provided by the embodiment of the present application;

[0069] Figure 4 It is a schematic diagram for comparing five search algorithms provided by the embodiment of the present application;

[0070] Figure 5 It is the time difference of arrival at each position collected in the experiment provided by the embodiment of the present application;

[0071] Figure 6 It is a schematic diagram of the mapping of more dense impact positions and time differences of arrival learned through Gaussian process provided by the embodiment of the present application;

[0072] FIG. 7(a) is a schematic diagram of the likelihood of the time difference of arrival at each position collected under sensor 1 for an impact event provided by the embodiment of the present application;

[0073] Figure 7(b) is a likelihood schematic diagram of the time difference of arrival collected under sensor 2 at various positions under an impact event provided by an embodiment of the present application;

[0074] Figure 7(c) is a likelihood schematic diagram of the time difference of arrival collected under sensor 3 at various positions under an impact event provided by an embodiment of the present application;

[0075] Figure 7(d) is a likelihood schematic diagram of the time difference of arrival collected under sensor 4 at various positions under an impact event provided by an embodiment of the present application;

[0076] Figure 8 It is a schematic diagram of the positioning prediction result provided by an embodiment of the present application;

[0077] Figure 9 It is a schematic diagram provided by an embodiment of the present application for comparing the predicted position (the most likely position) of each observation sample in the test set with the true position;

[0078] Figure 10 It is a schematic diagram of a square composite material plate with a grid spacing of 100 mm provided by an embodiment of the present application;

[0079] Figure 11 It is a functional structure diagram of a composite material impact probability positioning device provided by an embodiment of the present application;

[0080] Figure 12 It is a structure diagram of an electronic device provided by an embodiment of the present application. Detailed implementation manners

[0081] To make the objectives, technical solutions, and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are some, rather than all, of the embodiments of the present application. Usually, the components of the embodiments of the present application described and illustrated herein can be arranged and designed in various different configurations.

[0082] Therefore, the following detailed description of the embodiments of the present application provided in the accompanying drawings is not intended to limit the scope of the claimed present application, but merely represents selected embodiments of the present application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts shall fall within the protection scope of the present application.

[0083] First, a brief introduction to the design concept of the embodiments of the present application will be given.

[0084] Composite materials are new materials composed of two or more materials with different properties, shapes, or qualities through a certain composite process. They have the advantages of light weight, high specific strength, high specific stiffness, and chemical corrosion resistance, and have been widely used in engineering fields such as aerospace, ships, and automobiles. The use of composite materials can simplify the forming process of structures, greatly reduce the weight of aircraft to improve flight performance, and is also an effective way to directly reduce operating costs. The increasingly widespread use of composite materials in the aerospace field has led to more stringent requirements for material safety.

[0085] Composite materials have the disadvantages of low transverse interface strength, brittle fibers and resin matrices, and are sensitive to impact loads. When the impact energy reaches a certain threshold, internal damage such as matrix cracking, panel delamination, and fiber fracture is likely to occur within the material, significantly reducing the overall strength and reliability of the structure. Different from traditional metal materials, such internal damage in composite materials is difficult to detect by the naked eye in the initial stage. Especially for low-velocity impacts, if not addressed, the internal damage will gradually expand during the subsequent long-term service process, causing a sharp decline in structural stability, and even weakening the structural strength by 35% - 40%, which may lead to catastrophic consequences. However, during the service process of aircraft, it is inevitable to be subjected to external impacts, such as bird strikes, hail impacts, sand impacts during takeoff and landing, and tool drop impacts during maintenance. These impacts may not cause obvious damage immediately, but are extremely likely to form hidden damage inside the structure, and this hidden damage is more dangerous to a certain extent than obvious damage.

[0086] Aiming at the impact location problem of composite materials with a highly nonlinear wave field, this application provides a method for probabilistic impact location of composite materials. This method uses relative position vectors and corresponding time differences of arrival to learn the structural information of composite materials, and uses the Whale Optimization Algorithm (WOA) to optimize the hyperparameters of GPR to achieve probabilistic location of the impact position for the given observed time differences of arrival.

[0087] The method provided by the embodiments of this application can provide effective technical means for the safety and stability of aircraft during use, adopt a condition-based maintenance strategy instead of the traditional plan-based maintenance strategy during maintenance, reduce maintenance costs, and has important engineering application value. In impact monitoring, accurate identification of the impact position helps to determine the possible damaged locations, and then accelerate the inspection and maintenance of key parts, which can effectively reduce the time cost and economic cost of structural maintenance. Estimation of the impact effect is an important way to determine the damage degree and damage mode of composite material structures, and is of great significance to structural integrity.

[0088] After introducing the application scenarios and design concepts of the embodiments of the present application, the technical solutions provided by the embodiments of the present application will be described below.

[0089] As Figure 1 shown, an impact probability positioning method for composite materials is provided in an embodiment of the present application, which is used to perform impact positioning on a square plate of a composite material to be measured, and a plurality of sensors are arranged around the square plate of the composite material to be measured; including:

[0090] Step 101: Grid the square plate of the composite material to be measured according to a first pitch to obtain a plurality of first grids, and determine the position of the center point of each first grid as the position of the first grid;

[0091] Step 102: After an impact object hits the square plate of the composite material to be measured, determine the time difference between the impact signals collected by each pair of sensors;

[0092] Step 103: Process the time difference and the position of each first grid by using the Gaussian process regression model corresponding to each pair of sensors to obtain the likelihood value of each first grid under each pair of sensors; perform weighted summation on the multiple likelihood values of each first grid to obtain the positioning probability of the first grid;

[0093] Step 104: Determine the position of the center point of the first grid corresponding to the maximum positioning probability as the impact position of the square plate of the composite material to be measured.

[0094] Preferably, the number of sensors is 4, which are respectively arranged at the four corners of the square plate of the composite material to be measured.

[0095] In some embodiments, using the Gaussian process regression model corresponding to each pair of sensors to process the time difference and the position of each first grid to obtain the likelihood value of each first grid under each pair of sensors includes:

[0096] For the j-th Gaussian process regression model corresponding to the j-th pair of sensors, the likelihood value at the position of the center point of the first grid satisfies:

[0097]

[0098] where is the time difference of the j-th pair of sensors; is the optimal estimated value of the hyperparameter vector of the j-th pair of sensors is the position of the center point of the first grid is the predicted value at the position is the mean value of is the variance of

[0099] In some embodiments, the method further includes:

[0100] Meshing the square plate of the composite material to be measured according to a second pitch to obtain a plurality of second grids, and determining the position of the center point of each second grid as the position of the second grid; wherein, the second pitch is greater than the first pitch;

[0101] Sequentially dropping steel balls from a preset height onto each second grid of the square plate of the composite material to be measured, and obtaining the time difference of the j-th pair of sensors for each second grid;

[0102] Calculating the log-likelihood value :

[0103]

[0104] where N is the number of second grids, is the time difference vector, and the n-th component is the time difference of the j-th pair of sensors for the n-th second grid, is the time difference vector of the kernel matrix, is the position vector of the center points of N second grids; is the hyperparameter vector of the j-th pair of sensors to be estimated, including: noise variance , signal variance and length scale parameter ;

[0105] The optimal estimated value of the hyperparameter vector of the j-th pair of sensors is:

[0106]

[0107] Searching for the optimal estimated value through a search algorithm to minimize the negative log-likelihood value.

[0108] In some embodiments, the search algorithm uses the whale optimization algorithm, including:

[0109] Initializing M search particles of the population, and using the initialized M search particles as the M search particles for the first iteration; wherein, the position of each search particle is represented as the hyperparameter vector of the j-th pair of sensors;

[0110] Determining the search particle with the minimum fitness value among the initialized M search particles as the optimal search particle for the first iteration;

[0111] Repeatedly executing the following steps:

[0112] Update the positions of the M search particles in the previous iteration to obtain the positions of the M search particles in the current iteration;

[0113] Substitute the position of each search particle into the fitness function to obtain the fitness value of each search particle; determine the search particle with the minimum fitness value as the optimal search particle in the current iteration;

[0114] Based on the optimal search particle in the previous iteration and the optimal search particle in the current iteration, update the optimal search particle in the current iteration;

[0115] Until the current iteration reaches the preset maximum number of iterations;

[0116] The following describes the specific implementation process of the present application in combination with a specific application scenario.

[0117] For a glass fiber composite sandwich panel with dimensions of 710mm×540mm×50mm, it consists of two layers of glass fiber reinforced plastic (GFRP) surface layers and a polyvinyl chloride foam (PVC) sandwich layer. The glass fiber fabric is laid in a [0° / 90° / 90° / 0°]2S configuration. The composite panel is divided into grids at intervals of 50mm. The low-velocity impact is generated by a small steel ball with a diameter of 18.3mm and a weight of 30g falling from a height of 10cm. Five impacts are performed at each intersection point, and the sensor and impact point layout is as Figure 2 shown. Four piezoelectric ceramic sensors are attached to the composite panel to capture the acoustic emission signals generated by the low-velocity impact. The preamplifier gain is 40dB, and the sampling frequency is 2.5MHz. The acoustic emission signals of 640 impacts are collected for each channel, and a total of 4 channels of acoustic emission signals are obtained.

[0118] Before training the GPR model, it is first necessary to process the collected data. First, extract the difference-in-time-of-arrival (dTOA) of the impact signal reaching different sensors, Figure 3 shows the arrival times of the signals received by the 4 sensors when an impact occurs at the point (200,300). From Figure 3 it can be found that the acoustic emission signal generated by the impact first reaches sensor 4 and finally reaches sensor 2. However, the impact point is closest to sensor 3 in the structure, but the time for the impact signal to reach sensor 3 is later than the time to reach sensor 4, which also shows that the wave propagation speed in the composite material is inconsistent. In such anisotropic materials, traditional methods are not suitable.

[0119] After calculating the arrival time of the impact signal on each sensor, the time difference of arrival for each pair of sensors is calculated separately. Using 4 sensors can generate 6 different pairing combinations, and a vector containing 6 time differences of arrival will be generated for each impact. After establishing the data set, under each pair of sensors, the Gaussian process representation of the mapping between the impact position and the time difference of arrival is learned. When using the sum composite kernel function of the SE with an anisotropic kernel and the Matérn 3 / 2 covariance function as the kernel function of the Gaussian process model, the parameter range of the model and the results optimized by WOA are shown in Table 1:

[0120] Table 1: Model Parameters and Optimization Results

[0121]

[0122] Figure 4 The optimization performances of WOA are compared with those of the BFGS algorithm, the Scale Conjugate Gradient (SCG) algorithm, the Gray Wolf Optimizer (GWO) algorithm, and the Genetic Algorithm (GA). The parameters of WOA, the Genetic Algorithm, and the Gray Wolf Optimizer are shown in Table 2. From Figure 4 it can be found that compared with the Genetic Algorithm and the Gray Wolf Optimizer, WOA can converge to a better result by virtue of its good global search ability and local search ability. Although the BFGS algorithm and the conjugate gradient algorithm can also converge to a lower value, WOA is more efficient and converges to a better result with fewer iterations.

[0123] Table 2: Parameters of WOA, GA, and GWO Algorithms

[0124]

[0125] Figure 5 and Figure 6 give the learning results of the mapping between the spatial position and the time difference of arrival under the combination of sensor 1 and sensor 4. Among them Figure 5 is the time difference of arrival at each position collected in the experiment, which is used for the learning of the Gaussian process model. Figure 6 is the mapping of the denser impact position and the time difference of arrival learned through Gaussian process. It can be found that the denser mapping is not calculated according to linear interpolation, but obtained according to the training mechanism of the GPR model. The training data is used to learn the complex characteristics of the composite material structure, which is also the principle of high-resolution positioning by GPR. This is also the reason why the first spacing is smaller than the second spacing. When performing impact positioning, the grid of the composite material square plate can be divided smaller to achieve high-resolution positioning.

[0126] Next, location prediction is carried out. In the experiment, the collected data is divided into a training set and a test set, and the hyperparameters of the Gaussian process model corresponding to each pair of sensors are trained using the training set. Figures 7(a), 7(b), 7(c), and 7(d) show a shock event randomly selected from the test set, and the likelihood of the time difference of arrival collected under sensors 1 (Sensor_1), 2 (Sensor_2), 3 (Sensor_3), and 4 (Sensor_4) at various positions. Among them, "×" represents the true position, and the brighter the color, the greater the likelihood value, that is, the greater the possibility that it is the shock position. In each pair of sensors, it can be found that the true position is on the bright strip. However, the area of the bright strip has a large span, and it is difficult to determine the relatively accurate shock position. According to the acoustic emission theory, when determining the acoustic emission source based on the time difference of arrival of the fastest wave, if the acoustic emission source is not on the connection path of two sensors, at least three sensors are required to locate the acoustic emission source.

[0127] Using Figure 2 the data in, combined with six pairs of sensor combinations, calculate the coordinates of the prediction points, and the prediction results are as Figure 8 shown. "×" on the figure represents the position of the true shock, and the brighter the color, the greater the likelihood value. It can be found that there is a region with a large likelihood value near the true shock position, that is, the possibility that the shock occurs in this region is large, which also indicates that the shock position is accurately predicted. In addition, the figure gives the probability of shock occurring at each position of the structure. During the maintenance process, maintenance personnel can select a certain confidence region for maintenance according to experience, thus avoiding the time cost and economic cost brought by large-area maintenance.

[0128] Compare the predicted position (the most likely position) of each observation sample in the test set with the true position, and the results are as Figure 9 shown. The figure respectively gives the region (99.5% confidence region) between the upper 0.005 quantile and the upper bound of the likelihood value for each test point, and the area of this region is 0.5% of the area of the structure. From the results, it can be found that the true position basically falls within the 99.5% confidence range of the prediction result. The area of these prediction regions is only 0.5% of the area of the overall structure, and maintenance personnel can detect the structure at a lower cost. In addition, from the figure, it can be found that there are obvious differences between some predicted positions and the true positions. In this case, estimating a possible region where a shock occurs is more valuable than a definite shock position coordinate.

[0129] To evaluate the overall performance of the algorithm, calculate the root mean square error of the predicted positions of the test samples according to the following formula, where N represents the number of test samples.

[0130]

[0131] The advantage of the probability-based method is that it allows the identification of regions with a probability greater than a certain probability threshold. Additionally, during the detection process, it is more desirable to detect small-area regions. To evaluate the performance of the model from this perspective, two evaluation metrics for the probability localization method, ACCP and Smin, are defined. Among them, ACCP represents the accuracy rate of the true impact position falling into the predicted region with a given threshold p, where , represents the quantile of the prediction result. For example, when p = 0.975, ACC0.975 represents the accuracy rate of the true impact position in the predicted region with a confidence level greater than 97.5% among all test samples. The minimum detection threshold Smin represents the percentage of the detection structure area required to detect the true impact position, that is, the minimum upper quantile for detecting the impact position.

[0132] It can be found that the prediction result is related to the result of the covariance function. Here, a total of 4 covariance functions, SE, Matérn3 / 2, SE + Matérn3 / 2, and SE × Matérn3 / 2, are explored, and the influence of the isotropic kernel and the anisotropic kernel is also considered. Predictions are made for the test points on the test set, and the results are shown in Table 3.

[0133] It can be found from the table that the performance of using the anisotropic kernel function is better than that of the isotropic kernel function, which is due to the anisotropy of the composite material, resulting in asymmetry in each dimension. Additionally, the performance of the squared exponential covariance function SE is the worst, because this strong smoothing assumption is unrealistic for the impact localization modeling of composite materials. The composite covariance functions of the SE and Matérn3 / 2 covariance functions inherit the advantages of these two kernel functions, and the kernel functions formed by their summation have the best performance. The root mean square error on the anisotropic kernel is 8.94 mm. When the threshold is set to 97.5%, the true position of the impact can be identified 100%, and on average, only 0.15% of the structure area needs to be detected to identify the true impact position.

[0134] Table 3: Performance comparison of four kernel functions

[0135]

[0136] In data-driven methods, the amount of data in the training set has an important impact on the performance of the model. To study the sensitivity of the localization algorithm to the number of data points used in the training phase, training sets with different grid spacings were constructed to learn the Gaussian process representation, predict the test points on the test set, and calculate the RMSE for each grid spacing as an evaluation metric. The grid spacing is defined as the distance between each training point in the x and y directions. In addition to the 50 mm grid spacing used in the above analysis, a 100 mm grid spacing was also defined, as Figure 10 shown.

[0137] In addition, to evaluate the impact of the grid on the model, the accuracy of the true impact position falling within the 97.5% confidence interval and the minimum detection threshold were calculated respectively , and the performance of the models trained with each grid spacing is shown in Table 4.

[0138] At a grid spacing of 100 mm, the root mean square error is 31.62 mm, and the accuracy is affected to a certain extent. If traditional coordinate-based positioning methods are used, it is difficult to determine the true impact position. Using the GPR-based method provides the probability results of impact localization. The number of training grid points using a 100 mm grid spacing is 36, which is only 28% of that of a 50 mm grid spacing, but on average, only 1.27% of the structure area needs to be detected at least to find the true impact position. In addition, this probability-based method also gives maintenance personnel a guiding maintenance strategy, which can locate the impact damage position more quickly.

[0139] Table 4: Influence of different grid spacings

[0140]

[0141] In addition to the impact localization problem, this method can also solve the localization of some damages, provided that the damage generates acoustic emission events when it occurs. In addition, except for anisotropic composite materials, this method is also applicable to complex structures with special parts such as holes and bosses. Through the offline learning of the model, the probability localization of impacts or damages can be achieved in real-time processes, providing guiding opinions for further detection and maintenance.

[0142] Based on the same inventive concept, an embodiment of the present application provides a composite material impact probability localization device for performing impact localization on a square plate of a composite material to be measured. A plurality of sensors are arranged around the square plate of the composite material to be measured. Referring to Figure 11 shown, the composite material impact probability localization device 200 provided by the embodiment of the present application at least includes:[[]]

[0143] The grid unit 201 is configured to grid the square plate of the composite material to be measured at a first pitch, obtain a plurality of first grids, and determine the position of the center point of each first grid as the position of the first grid;

[0144] The determination unit 202 is configured to determine the time difference between the impact signals collected by each pair of sensors after the impact object impacts the square plate of the composite material to be measured;

[0145] The processing unit 203 is configured to process the time difference and the position of each first grid by using the Gaussian process regression model corresponding to each pair of sensors to obtain the likelihood value of each first grid under each pair of sensors; perform weighted summation on the multiple likelihood values of each first grid to obtain the positioning probability of the first grid;

[0146] The positioning unit 204 is configured to determine the position of the center point of the first grid corresponding to the maximum positioning probability as the impact position of the square plate of the composite material to be measured.

[0147] It should be noted that the principle of the composite material impact probability positioning device 200 provided in the embodiments of the present application for solving technical problems is similar to the method provided in the embodiments of the present application. Therefore, the implementation of the composite material impact probability positioning device 200 provided in the embodiments of the present application can refer to the implementation of the method provided in the embodiments of the present application, and the repeated parts will not be described again.

[0148] As Figure 12 shown, the electronic device 300 provided in the embodiments of the present application at least includes: a processor 301, a memory 302, and a computer program stored on the memory 302 and executable on the processor 301. When the processor 301 executes the computer program, the composite material impact probability positioning method provided in the embodiments of the present application is implemented.

[0149] The electronic device 300 provided in the embodiments of the present application may further include a bus 303 connecting different components (including the processor 301 and the memory 302). Among them, the bus 303 represents one or more of several types of bus structures, including a memory bus, a peripheral bus, a local bus, etc.

[0150] The memory 302 may include a readable medium in the form of a volatile memory, such as a random access memory (RAM) 3021 and / or a cache memory 3022, and may further include a read-only memory (ROM) 3023.

[0151] The memory 302 may also include a program tool 3024 having a set (at least one) of program modules 3025, and the program modules 3025 include, but are not limited to: an operating subsystem, one or more application programs, other program modules, and program data. Each or some combination of these examples may include an implementation of a network environment.

[0152] The electronic device 300 may also communicate with one or more external devices 304 (such as a keyboard, a remote control, etc.), may also communicate with one or more devices that enable a user to interact with the electronic device 300 (such as a mobile phone, a computer, etc.), and / or communicate with any device that enables the electronic device 300 to communicate with one or more other electronic devices 300 (such as a router, a modem, etc.). Such communication may be carried out through an input / output (I / O) interface 305. And, the electronic device 300 may also communicate with one or more networks (such as a local area network (LAN), a wide area network (WAN), and / or a public network, such as the Internet) through a network adapter 306. As Figure 12 shown, the network adapter 306 communicates with other modules of the electronic device 300 through a bus 303. It should be understood that although Figure 12 not shown in the figure, other hardware and / or software modules may be used in combination with the electronic device 300, including but not limited to: microcode, device drivers, redundant processors, external disk drive arrays, redundant arrays of independent disks (RAID) subsystems, tape drives, and data backup storage subsystems, etc.

[0153] It should be noted that Figure 12 the electronic device 300 shown is only an example and should not impose any limitation on the functions and usage scope of the embodiments of the present application.

[0154] The embodiments of the present application also provide a computer-readable storage medium storing computer instructions, and when the computer instructions are executed by a processor, the method for locating the impact probability of a composite material provided by the embodiments of the present application is implemented.

[0155] In addition, although the operations of the method of the present application are described in a specific order in the drawings, this does not require or imply that these operations must be performed in this specific order, or that all the operations shown must be performed to achieve the desired result. Additionally or alternatively, some steps may be omitted, multiple steps may be combined into one step for execution, and / or one step may be decomposed into multiple steps for execution.

[0156] Although the preferred embodiments of the present application have been described, those skilled in the art can make additional changes and modifications once they learn the basic creative concepts. Therefore, the appended claims are intended to be construed as including the preferred embodiments and all changes and modifications falling within the scope of the present application.

[0157] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some or all of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for locating the impact probability of a composite material, which is used for impact location of a square plate of a composite material to be measured, and a plurality of sensors are arranged around the square plate of the composite material to be measured; characterized in that, Including: Meshing the square plate of the composite material to be measured according to the first spacing to obtain a plurality of first meshes, and determining the position of the center point of each first mesh as the position of the first mesh; After the impact object impacts the square plate of the composite material to be measured, determining the time difference of the impact signals collected by each pair of sensors; Processing the time difference and the position of each first mesh by using the Gaussian process regression model corresponding to each pair of sensors to obtain the likelihood value of each first mesh under each pair of sensors; performing weighted summation on the multiple likelihood values of each first mesh to obtain the positioning probability of the first mesh; Determining the position of the center point of the first mesh corresponding to the maximum positioning probability as the impact position of the square plate of the composite material to be measured; Processing the time difference and the position of each first mesh by using the Gaussian process regression model corresponding to each pair of sensors to obtain the likelihood value of each first mesh under each pair of sensors, including: For the j-th Gaussian process regression model corresponding to the j-th pair of sensors, at the center point position of the first grid the likelihood value satisfies: wherein, is the time difference of the j-th pair of sensors; is the optimal estimated value of the hyperparameter vector of the j-th pair of sensors is the center point position of the first grid at the predicted value, is the mean of, is the variance of.

2. The method for locating the impact probability of the composite material according to claim 1, characterized in that, The method further includes: Meshing the square plate of the composite material to be measured according to the second spacing to obtain a plurality of second meshes, and determining the position of the center point of each second mesh as the position of the second mesh; wherein, the second spacing is greater than the first spacing; Sequentially dropping steel balls from a preset height onto each second mesh of the square plate of the composite material to be measured, and obtaining the time difference of the j-th pair of sensors for each second mesh; Calculate the log-likelihood value : where N is the number of the second grids, is the time difference vector, and the n-th component is the time difference of the j-th pair of sensors in the n-th second grid, is the time difference vector of the kernel matrix, is the central point position vector of N second grids; is the hyperparameter vector of the j-th pair of sensors to be estimated, including: noise variance , signal variance and length scale parameter ; Optimal estimated value of the hyperparameter vector of the j-th pair of sensors is as follows: Search for the optimal estimate through a search algorithm to minimize the negative log-likelihood value.

3. The method for locating the impact probability of the composite material according to claim 2, wherein The search algorithm adopts the whale optimization algorithm; Search for the optimal estimated value through a search algorithm , to minimize the negative log-likelihood value, including: Initializing M search particles of the population, and taking the initialized M search particles as the M search particles of the first iteration; wherein, the position of each search particle is represented as the hyperparameter vector of the j-th pair of sensors; Determining the search particle with the minimum fitness value among the initialized M search particles as the optimal search particle of the first iteration; Repeatedly execute the following steps: Updating the positions of the M search particles of the previous iteration number to obtain the positions of the M search particles of the current iteration number; Substituting the position of each search particle into the fitness function to obtain the fitness value of each search particle; determining the search particle with the minimum fitness value as the optimal search particle of the current iteration number; Updating the optimal search particle of the current iteration number based on the optimal search particle of the previous iteration number and the optimal search particle of the current iteration number; Until the current iteration number reaches the preset maximum iteration number; Determine the optimal search particle with the maximum number of iterations as the optimal estimate value .

4. A device for positioning the impact probability of a composite material, which is used for impact positioning of a square plate of a composite material to be measured, and a plurality of sensors are arranged around the square plate of the composite material to be measured; characterized in that, Including: A meshing unit, configured to mesh the square plate of the composite material to be measured according to the first spacing to obtain a plurality of first meshes, and determine the position of the center point of each first mesh as the position of the first mesh; A determining unit, configured to determine the time difference of the impact signals collected by each pair of sensors after the impact object impacts the square plate of the composite material to be measured; A processing unit, configured to process the time difference and the position of each first mesh by using the Gaussian process regression model corresponding to each pair of sensors to obtain the likelihood value of each first mesh under each pair of sensors; Performing weighted summation on the multiple likelihood values of each first mesh to obtain the positioning probability of the first mesh; A positioning unit, configured to determine the position of the center point of the first mesh corresponding to the maximum positioning probability as the impact position of the square plate of the composite material to be measured; Processing the time difference and the positions of each first grid using the Gaussian process regression model corresponding to each pair of sensors to obtain the likelihood value of each first grid under each pair of sensors, including: For the j-th Gaussian process regression model corresponding to the j-th pair of sensors, at the center point position of the first grid the likelihood value satisfies: Among them, is the time difference of the j-th pair of sensors; is the optimal estimated value of the hyperparameter vector of the j-th pair of sensors is the center point position of the first grid the predicted value at, is the mean of, is the variance of.

5. The composite material impact probability positioning device according to claim 4, wherein The device further includes a training unit, specifically used for: Meshing the square plate of the composite material to be measured according to a second pitch to obtain a plurality of second grids, and determining the position of the center point of each second grid as the position of the second grid; wherein, the second pitch is greater than the first pitch; Sequentially dropping steel balls from a preset height onto each second grid of the square plate of the composite material to be measured, and obtaining the time difference of the j-th pair of sensors for each second grid; Calculate the log-likelihood value : where N is the number of second grids, is the time difference vector, and the n-th component is the time difference of the j-th pair of sensors in the n-th second grid, is the time difference vector of the kernel matrix, is the central point position vector of N second grids; is the hyperparameter vector of the j-th pair of sensors to be estimated, including: noise variance , signal variance and length scale parameter ; Optimal estimated value of the hyperparameter vector of the j-th pair of sensors is as follows: Search for the optimal estimate through a search algorithm to minimize the negative log-likelihood value.

6. The composite material impact probability positioning device according to claim 5, characterized in that The search algorithm adopts the whale optimization algorithm; Search for the optimal estimate through a search algorithm , to minimize the negative log-likelihood value, including: Initializing M search particles of the population, and using the initialized M search particles as the M search particles for the first iteration; wherein, the position of each search particle is represented as the hyperparameter vector of the j-th pair of sensors; Determining the search particle with the minimum fitness value among the initialized M search particles as the optimal search particle for the first iteration; Repeatedly execute the following steps: Updating the positions of the M search particles in the previous iteration number to obtain the positions of the M search particles in the current iteration number; Substituting the position of each search particle into the fitness function to obtain the fitness value of each search particle; determining the search particle with the minimum fitness value as the optimal search particle for the current iteration number; Updating the optimal search particle for the current iteration number based on the optimal search particle in the previous iteration number and the optimal search particle in the current iteration number; Until the current iteration number reaches the preset maximum iteration number; Determine the optimal search particle with the maximum number of iterations as the optimal estimate value .

7. An electronic device, characterized in that, Including: A memory, a processor, and a computer program stored on the memory and executable on the processor, and when the processor executes the computer program, the method described in any one of claims 1-3 is implemented.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions, and when the computer instructions are executed by the processor, the method described in any one of claims 1-3 is implemented.

Images