Method and apparatus for predicting sliding angle of material surface, storage medium, and detection system

Abstract

The present application discloses a method and apparatus for predicting the slip angle of the surface of a material to be measured having a plurality of textures, as well as a storage medium and a detection system. The method for predicting the slip angle of the material surface includes obtaining the coordinate values of the textures within the target area of the surface of the material to be measured, inputting the coordinate values of the textures within the target area into a slip angle prediction model with the coordinate values of the textures as input labels and the slip angle on the texture as output labels to obtain the output value of the slip angle prediction model, and obtaining the slip angle of the surface of the material to be measured based on the output value of the slip angle prediction model.

Description

Technical Field

[0001] <Cross - Reference to Related Applications> This application claims priority based on a Chinese patent application with application number 202310368483.X, titled "Method and Apparatus for Predicting Sliding Angle of Material Surface, Memory Medium, and Detection System", filed with the China National Intellectual Property Administration on April 7, 2023, and all of its content is incorporated herein by reference. This application relates to the field of materials technology, and particularly to a method for predicting the sliding angle of a material surface, an apparatus for predicting the sliding angle of a material surface, a computer - readable storage medium, and a sliding angle detection system.

Background Art

[0002] Wettability is one of the important properties of a material surface and can be represented by the static contact angle and the sliding angle. Since the influencing factors of wettability mainly include the chemical composition and microstructure of the material surface, the wettability of the material surface can be changed by surface modification and surface texturing.

[0003] Among them, the construction of a surface with intelligent controllable wettability is a hot spot in the research of material surfaces. Therefore, the accurate quantitative evaluation of the wettability of a material surface by its microstructure has important significance in material wettability applications, especially in fields where strict accuracy is required.

Summary of the Invention

Problems to be Solved by the Invention

[0004] This application aims to solve at least one of the technical problems existing in the prior art. Therefore, the first objective of this application is to provide a method for predicting the sliding angle of a material surface that can predict the sliding angle based on the influence of the texture position of the material surface on the sliding angle. According to this method, the accuracy of the predicted sliding angle is high.

[0005] The second objective of this application is to provide an apparatus for predicting the sliding angle of a material surface.

[0006] The third object of the present application is to provide a computer-readable storage medium.

[0007] The fourth object of the present application is to provide a slip angle detection system.

Means for Solving the Problems

[0008] In order to achieve the above object, the method for predicting the slip angle of the material surface according to the first aspect of the present application has a plurality of textures on the surface of the material to be measured. The method includes obtaining coordinate values of textures within a target area on the surface of the material to be measured, inputting the coordinate values of the textures within the target area into a slip angle prediction model, which is a neural network prediction model constructed with the coordinate values of the textures as input labels and the slip angle at the textures as output labels, obtaining an output value of the slip angle prediction model, and obtaining the slip angle at the textures within the target area on the surface of the material to be measured based on the output value of the slip angle prediction model.

[0009] The method for predicting the slip angle of the material surface according to the embodiment of the present application takes into account the influence of the position distribution of the material surface texture on the slip angle of the material surface, obtains the coordinate values of the material surface texture, that is, the position distribution of the texture, inputs the coordinate values as input labels into the slip angle prediction model, and can obtain the slip angle at the texture based on the output value of this model. This method is simple and rapid to implement, and the accuracy of the slip angle prediction model is high.

[0010] In some embodiments, before inputting the coordinate values of the textures within the target area into the slip angle prediction model, the method further includes performing principal component analysis on the coordinate values of the textures within the target area on the surface of the material to be measured in order to reduce the feature amount of the target surface. Thereby, the calculation speed of the slip angle prediction model can be improved.

[0011] In some embodiments, the number of surface feature amounts of the texture is reduced to 300 to 400.

[0012] In some embodiments, obtaining the slip angle of the texture in the target region based on the output value of the slip angle prediction model includes grouping the output values of the slip angle prediction model and determining the slip angle of the texture in the target region based on the grouping result of the output values of the slip angle prediction model. Through the grouping process, the influence of the measuring instrument on the slip angle prediction result can be reduced, and the accuracy of the slip angle prediction result can be improved.

[0013] In some embodiments, determining the slip angle of the texture in the target region based on the grouping result of the output values of the slip angle prediction model includes obtaining a predetermined slip angle group to which the output value of the slip angle prediction model belongs, and using the slip angle corresponding to the target slip angle group as the slip angle of the texture in the target region.

[0014] In some embodiments, inputting the coordinate values of the texture in the target region into the slip angle prediction model and obtaining the output value of the slip angle prediction model includes inputting the coordinate values of the texture on the surface of the material to be measured (for example, the coordinate values of a target number of textures) into the input layer of a trained probabilistic neural network, where the number of input nodes and the number of neurons are equal to the dimensionality of the coordinate values of the texture, and obtaining an input feature vector through the input layer; inputting the input feature vector into the pattern layer of the probabilistic neural network, and obtaining the similarity between the input feature vector and the texture of each sample in the training set by the pattern layer; inputting the similarity into the addition layer of the probabilistic neural network, and performing weighted addition for each similarity by the addition layer to obtain a weighted addition score; inputting the weighted addition score into the output layer of the probabilistic neural network, and outputting, through the output layer, the slip angle label with the highest weighted addition score as the output value of the slip angle prediction model.

[0015] In some embodiments, the method includes obtaining detection information of deposits around the texture within the target region, and processing the deposits within the target region based on the detection information such that the deposits around the texture within the target region match.

[0016] In some embodiments, the texture is a dimple. The target number of dimples may be 539 in an area of 1 mm × 1 mm.

[0017] Embodiment 2 of the present application provides an apparatus for predicting the material surface slip angle (hereinafter also referred to as the "apparatus for predicting the material surface slip angle"). The apparatus includes at least one processor and a memory communicably connected to the at least one processor, and a computer program executed by the at least one processor is stored in the memory. When the at least one processor executes the computer program, the method for predicting the material surface slip angle is realized.

[0018] The apparatus for predicting the material surface slip angle according to the embodiment of the present application can realize the method for predicting the material surface slip angle according to the above embodiment by at least one processor. Based on the influence of the texture position on the material surface on the slip angle, the constructed slip angle prediction mode can simply and quickly predict the slip angle at the texture of the material surface to be measured, and further, the accuracy of the predicted slip angle is high.

[0019] Embodiment 3 of the present application provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, the method for predicting the material surface slip angle described in the above embodiment is realized.

[0020] Embodiment 4 of the present application provides a slip angle detection system. The slip angle detection system includes an imaging device for acquiring image information of the surface of a material to be measured, and a device for predicting the slip angle of the surface of the material connected to the imaging device. The image information includes at least an image of the texture within the target area of the surface of the material to be measured.

[0021] The slip angle detection system according to the embodiment of the present application uses the distribution of the texture on the surface of the material to be measured as an input feature quantity, and directly obtains the corresponding slip angle through the constructed slip angle prediction model by the device for predicting the slip angle of the surface of the material. The method is simple and fast, and can improve the prediction accuracy of the slip angle.

[0022] Some of the additional aspects and advantages of the present application will be described below, some will be apparent from the following description, or some will be understood from the implementation of the present application.

Brief Description of the Drawings

[0023] The above and / or additional aspects and advantages of the present application will become apparent and be easily understood from the description of the embodiments with reference to the following attached drawings.

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Embodiments for Carrying Out the Invention

[0024] Hereinafter, embodiments of the present application will be described in detail. However, the embodiments described with reference to the accompanying drawings are exemplary.

[0025] The applicant has discovered that the positional distribution of the microtexture on the material surface affects the wettability of the material surface. By processing microtextures with different distribution methods on the material surface and measuring the numerical values of the contact angle and sliding angle in this microtexture, the correlation between the positional distribution of the microtexture and the contact angle and sliding angle is determined, and a sliding angle prediction model for predicting the sliding angle of the random microtexture surface is constructed. The method for predicting the sliding angle of the material surface in the embodiment of the present application can predict the sliding angle of the material surface based on this sliding angle prediction model. The prediction result is more accurate.

[0026] Hereinafter, the construction process of the sliding angle prediction model in the embodiment of the present application will be described.

[0027] In the above embodiment, taking the case where the microtexture on the material surface is a dimple as an example, the construction process of the sliding angle prediction model will be described.

[0028] First, prepare the equipment. The equipment required for sample preparation and measurement preparation of the sliding angle includes a nanosecond UV laser marker, a wire cut electrical discharge machine, a metal graphic sample grinding and polishing machine, an ultrasonic cleaner, a laboratory ultrapure water device, a vertical high-temperature electric furnace, a contact angle measuring instrument, a measuring cup, and a volumetric flask. Of course, other equipment that achieves the same effect may be adopted, or the equipment may be reduced or added.

[0029] In some embodiments, as shown in FIG. 1, a random dimple texture with a volume fraction of 19% can be randomly generated in a 1 mm × 1 mm sub-region by a MATLAB® program, arranged in the horizontal and vertical directions by CAD, and a 10 mm × 10 mm sample region can be formed. Using a nanosecond UV laser marker, based on the generated texture drawing, a random texture of the point group distribution shown in FIGS. 2(a), (b), and (c) is processed on the sample surface. The numerical values of the contact angle and the sliding angle of the random dimple surface are measured with a contact angle measuring instrument, and further, the influence of the dimple distribution on the contact angle and the sliding angle of the sample is examined to establish a neural network prediction model (i.e., a sliding angle prediction model) capable of predicting the surface sliding angle.

[0030] Here, regarding sample preparation, as shown in FIG. 3, sample preparation includes sample material preparation, surface treatment, and surface fluorination. For example, sample material preparation includes manufacturing 447 samples by sample cutting and wire cutting. Surface treatment includes surface grinding and polishing, rough grinding with sandpaper #800, and fine grinding with sandpaper #1200. Surface treatment includes surface ultrasonic cleaning, cleaning with absolute ethanol for, for example, 10 min, cleaning with deionized water for, for example, 10 min, and drying. After drying, it is ready for use. Surface fluorination may include grouping the samples so that each group contains 18 samples, adding 200 ml of fluorosilane to each group, then immersing for a certain period of time, for example, 10 h, feeding it into an electric furnace, curing it, and holding it at, for example, 150°C for 60 min.

[0031] Specifically, in order to ensure the consistency of the material during the construction of the sliding angle prediction model, all the samples adopted are wire-cut and processed from the same batch of materials. In order to ensure that randomly adjacent dimples do not overlap, the MATLAB program generates random dimple coordinates with a volume fraction of 19% in a 1 mm × 1 mm sub-region, introduces the generated point coordinates into CAD, arranges them in the horizontal and vertical directions, and a 10 mm × 10 mm sample area can be formed. The drawing is input into a computer controlled by a laser, and the computer program controls the laser light to etch vertically (in the X and Y directions) with a blade made of 3Cr13 stainless steel on a fixed workbench. The nanosecond UV laser processing parameters are respectively a processing speed of 500 mm / s, a frequency of 40 kHz, a pulse width of 10 μs, a dotting time of 0.3 s, and a processing number of 10 times. Finally, the etched sample is ultrasonically cleaned in deionized water and absolute ethanol for 10 minutes each, and the one dried with a blower is prepared.

[0032] JPEG2025523708000002.jpg16170Furthermore, an image of the sample surface (shown in Figures 4(c) and (d)) is taken with a digital CCD camera, and the contact angle is automatically obtained by performing elliptical fitting on the captured image with a contact angle measuring instrument, and the sliding angle is obtained by the inclination of the stage. In order to ensure the accuracy of the results, the contact angle is measured by a 5-point sampling method for each sample, and the average value is taken. From the experimental results, as shown in Figure 5, the sliding angle of the sample is in the range of 2° to 20°, and the contact angle is in the range of 145° to 160°.

[0033] Through the above experiments, a significant influence of the position distribution of the texture on the surface of the material on the wettability of the sample was confirmed. A method for predicting the contact angle and sliding angle of the sample from the position distribution of the texture is possible.

[0034] In an embodiment, a slip angle prediction model is constructed based on the texture position distribution and the slip angle. When constructing the slip angle prediction model, the training set and the test set can be randomly divided. For example, taking the 447 samples prepared above as an example, 70% of the samples may be used as the training set (313), and 30% of the samples may be used as the test set (134). In order to ensure the reproducibility of the model, the random number seed may be set to 10.

[0035] In some embodiments, the data can be trained using the Probabilistic Neural Network (PNN) of the optimization algorithm. PNN is derived from the Radial Basis Neural Network and is a parallel algorithm based on the Bayesian classification rule and the estimation method of the probability density function of the Parzen window. The structure of the Probabilistic Neural Network is simple, the algorithm design is easy, it can realize the function of the non-linear learning algorithm by the linear learning algorithm, and it is widely used in the field of pattern classification. The function newpnn provided by MATLAB can be used to easily design the Probabilistic Neural Network. The Probabilistic Neural Network is composed of an input layer, a hidden layer, an addition layer, and an output layer as shown in FIG. 6.

[0036] JPEG2025523708000003.jpg40170JPEG2025523708000004.jpg22170JPEG2025523708000005.jpg32170JPEG2025523708000006.jpg11170JPEG2025523708000007.jpg29170

[0037] The above is an example of a PNN (Probabilistic Neural Network). The training speed of the PNN is fast, only slightly exceeding the data reading time. No matter how complex the classification problem is, if there is sufficient training data, it is guaranteed that the optimal solution according to Bayes' rule can be obtained. When increasing or decreasing the training data, it is not necessary to conduct long-term training, and individual error samples can be tolerated. When the number of iterations reaches 100, the calculation results will converge stably. It should be understood that in order to obtain the slip angle prediction model of the present application, training can also be carried out based on other applicable neural network models.

[0038] Based on the slip angle prediction model, the method for predicting the slip angle of the material surface in the embodiment of the present application will be described below with reference to FIGS. 7 to 9.

[0039] FIG. 7 is a flowchart of the method for predicting the slip angle of the material surface according to an embodiment of the present application. As shown in FIG. 7, the method for predicting the slip angle of the material surface includes the following steps S1 to S3.

[0040] In S1, the coordinate values of the texture within the target area on the surface of the material to be measured are obtained.

[0041] Here, examples of the area of the target area include 1 mm × 1 mm, 1.5 mm × 1.5 mm, 2 mm × 2 mm, etc., but it is not particularly limited.

[0042] Specifically, an image of the surface of the material to be measured is captured by an imaging device. The surface of the material to be measured has a plurality of textures, for example, random dimples. By identifying the textures in the image of the surface of the material to be measured and obtaining the coordinates of the textures, for example, the horizontal and vertical coordinate values (x, y), the position distribution of the textures on the surface of the material to be measured can be obtained.

[0043] In S2, the coordinate values of the texture within the target area are input into the slip angle prediction model, and the output value of the slip angle prediction model is obtained.

[0044] Here, the slip angle prediction model is trained with the coordinate values of the texture as input features and the slip angle on the texture as the output label. For the process of obtaining this slip angle prediction model, reference can be made to the description of the above embodiment.

[0045] In S3, based on the output value of the slip angle prediction model, the slip angle on the texture within the target area on the surface of the material to be measured is obtained.

[0046] Specifically, the coordinate values of the texture within the target area on the surface of the material to be measured are input into the slip angle prediction model as input features. This model is trained with the coordinate values of the texture as input and the slip angle as output. Therefore, the output value of this model may be the slip angle on the texture.

[0047] The method for predicting the slip angle of the material surface according to the embodiment of the present application considers the influence of the position distribution of the texture on the material surface on the slip angle of the material surface, obtains the coordinate values of the texture on the material surface, that is, the position distribution of the texture, inputs the coordinate values as input features into the slip angle prediction model, and can obtain the slip angle on the texture based on the output value of this model. Such a method is simple and fast to implement, and furthermore, the accuracy of slip angle prediction can be improved.

[0048] Furthermore, when obtaining the slip angle by the slip angle prediction model, the abscissa and ordinate of the generated random points can be read as input features of the slip angle prediction model and formed into a one-dimensional array, and the measured contact angle and slip angle can also be read as output labels of the slip angle prediction model in a one-dimensional array. However, since a single point group contains, for example, the input coordinate values of 539 dimples (that is, 1078 features are included), if the number of features is too large, it will greatly affect the prediction result of the slip angle prediction model.

[0049] Therefore, in some embodiments of the present application, in order to reduce the feature amount of the target surface, the coordinate values of the texture in the target area of the surface of the material to be measured are subjected to principal component analysis (PCA) and dimensionality reduction processing. Thereby, the computing power of the slip angle prediction model can be improved.

[0050] For example, the analysis result by principal component analysis is shown in FIG. 8. It can be seen from FIG. 8 that when the number of feature amounts after dimensionality reduction is 300, the cumulative explained variance ratio can reach 90%. This means that most of the information volume of this data group can be reflected by 300 to 400 feature amounts. In some embodiments, the number of feature amounts of the target number of textures by the principal component analysis dimensionality reduction process may be, for example, 300, 350, or 400 as the number of feature amounts after PCA dimensionality reduction. Thereby, the computing power of the slip angle prediction model can be improved.

[0051] In the embodiment, through the error analysis of the model prediction result, it can be seen from the experimental data that the difference in the slip angle is relatively small, and the error of the measuring instrument is enlarged, for example, it becomes difficult to distinguish between 2° and 3°. Therefore, the method for predicting the slip angle of the material surface in the embodiment of the present application groups the initial slip angle, which is the output value of the slip angle prediction model, to reduce the error.

[0052] Specifically, the output value of the slip angle prediction model is grouped, and based on the grouping result of the output value of the slip angle prediction model, the slip angle of the texture in the target area of the material to be measured is determined. That is, the slip angle values within a certain range are used as a group, and this group corresponds to the final slip angle, thereby reducing the prediction result error caused by the measuring instrument.

[0053] In some embodiments, the probability that the output value of the slip angle prediction model falls within a predetermined slip angle group is obtained, and the target slip angle group is determined from this probability. For example, the slip angle group with the highest probability of falling within is set as the target slip angle group. Each slip angle group corresponds to a slip angle, and the slip angle corresponding to the target slip angle group is set as the slip angle on the texture within the target region.

[0054] For example, taking the range of 2° to 20° of the slip angle detected above as an example, it is divided into five groups: an ultra-low slip angle group, that is, ultra-low (2°, 3°, 4°, 5°), a low slip angle group, that is, low (6°, 7°, 8°, 9°), a medium slip angle group, that is, media (9°, 10°, 11°, 12°), a high slip angle group, that is, high (13°, 14°, 15°, 16°), and an ultra-high slip angle group, that is, ultra-high (17°, 18°, 19°, 20°). All samples are from the same batch and have undergone surface fluorination treatment, so the influence of material and surface chemical properties on the slip angle is the same, and the factor affecting the surface falling performance is the random position of the dimples. The prediction results after model convergence are shown in Figure 9, and the prediction accuracy can reach 90.2%.

[0055] For example, if the output values of the obtained slip angle prediction model are 13°, 13°, 11°, 15°, 14°, etc., respectively, and the probability of falling within the high slip angle group is the highest, finally, the slip angle on the texture is determined to be the high slip angle. By grouping the output values of the slip angle prediction model, the error of the slip angle measured by the measuring instrument can be reduced, and the accuracy of the prediction result can be improved.

[0056] Furthermore, the method for predicting the slip angle of the material surface according to the embodiment of the present application realizes the prediction of the falling performance of a random texture by predicting the slip angle on the random texture based on the constructed slip angle prediction model. However, it can be seen from the results of the model that there is a certain error in this slip angle prediction model. It can be understood from the analysis that the main causes of the error are the following two points. 1. In some embodiments, the maximum protrusion of a single dimple is about 10 μm, the minimum protrusion is about 5 μm, and the depth of the dimple is about 40 μm. When the protrusions around the dimple are small, the influence of the texture height on wettability is even greater. Since the influence of sediment on the bending and length change of the contact line can be almost ignored, the solid-liquid contact surface can be simplified to a two-dimensional plane. From the continuity analysis of the three-phase contact line, it can be seen that the more dimples there are on the three-phase contact line, the more the number of segmented segments, and the less continuous the three-phase contact line is, the more prominent the pinning effect of the dimple, and the more difficult it is for the droplet to roll. When the texture, such as the sediment around the dimple, affects the continuity of the three-phase contact line, the movement of the three-phase contact line is hindered, a difference occurs between the experimental contact angle and the predicted contact angle, and the accuracy of the prediction model is affected. Therefore, in some embodiments of the present application, by obtaining the detection information of the sediment around the texture in the target area on the surface of the material to be measured and processing the sediment around the texture based on the detection information, the consistency of the sediment around the texture in the target area on the surface of the material to be measured is maintained. For example, image information on the surface of the material to be measured is obtained to remove the influencing factors of the sediment around the texture on the prediction of the sliding angle. 2. Errors occurring in the measurement process of the sliding angle, such as insufficient accuracy of the instrument, different positions of the droplet, etc. In this regard, the accuracy of the model can be improved by increasing the number of samples.

[0057] Generally, the method for predicting the sliding angle of the material surface according to the embodiments of the present application uses the distribution of the texture on the surface of the material to be measured as the input feature quantity, and directly obtains the corresponding sliding angle by the constructed sliding angle prediction model. The method is simple and fast, and furthermore, the prediction accuracy of the sliding angle can be improved. Also, by grouping the output values of the sliding angle prediction model, the error of the sliding angle prediction by the measuring instrument can be reduced, and the prediction accuracy of the sliding angle can be further improved. Also, considering the influence of the sediment around the texture on the sliding angle, the predicted sliding angle value is compensated, thereby further improving the prediction accuracy of the sliding angle.

[0058] Based on the method for predicting the sliding angle of the material surface according to the above embodiment, Embodiment 2 of the present application provides an apparatus for predicting the sliding angle of the material surface.

[0059] FIG. 10 is a block diagram of an apparatus for predicting the sliding angle of the material surface according to an embodiment of the present application. As shown in FIG. 10, the apparatus 10 for predicting the sliding angle of the material surface includes at least one processor 11 and a memory 12 communicably connected to the at least one processor 11.

[0060] Here, the memory 12 stores a computer program executed by the at least one processor 11. When the at least one processor 11 executes the computer program stored in the memory 12, the method for predicting the sliding angle of the material surface according to the above embodiment can be realized. For the method for predicting the sliding angle of the material surface, reference can be made to the description of the above embodiment.

[0061] The apparatus 10 for predicting the sliding angle of the material surface according to the embodiment of the present application can realize the method for predicting the sliding angle of the material surface according to the above embodiment by at least one processor 11. Based on the influence of the texture position of the material surface on the sliding angle, the sliding angle prediction mode is constructed to predict the sliding angle of the texture on the surface of the material to be measured. It is simple and fast, and furthermore, the accuracy of the predicted sliding angle is high.

[0062] Embodiment 3 of the present application further provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, the method for predicting the sliding angle of the material surface according to the above embodiment is realized.

[0063] Here, in the embodiment, the computer-readable storage medium may be any medium that can include, store, communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. A non-exhaustive list of more specific examples of computer-readable media includes electrical connection parts (electronic devices) having one or more wirings, portable computer disks (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber devices, portable compact disk read-only memory (CD-ROM). Also, the computer-readable medium can be, for example, paper or other media that are optically scanned, electronically acquired, then, if necessary, compiled, interpreted, and processed in other appropriate ways, and stored in the computer's memory. Therefore, the computer-usable / computer-readable medium may be paper or other suitable media on which a program is printed.

[0064] Embodiment 4 of the present application provides a slip angle detection system.

[0065] FIG. 11 is a block diagram of a slip angle detection system according to an embodiment of the present application. As shown in FIG. 11, the slip angle detection system 100 includes an imaging device 20 and a prediction device 10 for the slip angle of the material surface according to the above embodiment.

[0066] Here, the imaging device 20 acquires image information of the surface of the material to be measured. The image information includes at least an image of the texture within the target area of the surface of the material to be measured. Thereby, by performing image recognition, relevant information on the texture and the surrounding deposits can be identified.

[0067] The prediction device 10 for the sliding angle of the material surface is connected to the imaging device 20. The prediction device 10 for the sliding angle of the material surface can include a computer device, a data processing device, a terminal device, or the like. The prediction device 10 for the sliding angle of the material surface can execute the method for predicting the sliding angle of the material surface according to the above embodiment.

[0068] The sliding angle detection system 100 according to the embodiment of the present application uses the distribution of the texture on the surface of the material to be measured as an input feature amount, and directly obtains the corresponding sliding angle by the sliding angle prediction model constructed by the prediction device 10 for the sliding angle of the material surface. The method is simple and fast, and furthermore, the prediction accuracy of the sliding angle can be improved.

[0069] In addition, the prediction device 10 for the sliding angle of the material surface can further group the output values of the sliding angle prediction model, thereby reducing the error of the sliding angle prediction by the measuring instrument and further improving the accuracy of the sliding angle prediction. In addition, the prediction device 10 for the sliding angle of the material surface can further compensate the predicted sliding angle value in consideration of the influence of the deposits around the texture on the sliding angle, thereby further improving the accuracy of the sliding angle prediction.

[0070] In the description of this specification, the description of terms such as "one embodiment", "some embodiments", "example", "specific example" or "some examples" means that the specific configuration, structure, material or feature described in this embodiment or example is included in at least one embodiment or example of the present application. In this specification, the schematic description of the above terms does not necessarily target the same embodiment or example.

[0071] Although the embodiments of the present application have been described and illustrated, various changes, modifications, substitutions and deformations can be made without departing from the spirit of the present disclosure, and it can be understood by those skilled in the art that the scope of the present application is limited by the claims and their equivalents.

Claims

1. A method for predicting the sliding angle of a material surface, the material surface having a plurality of textures, the method comprising: obtaining the coordinate values of the textures within a target region on the surface of the material to be measured; inputting the coordinate values of the textures within the target region into a sliding angle prediction model, which is a neural network prediction model constructed with the coordinate values of the textures as input labels and the sliding angle on the textures as output labels, and obtaining the output value of the sliding angle prediction model; obtaining the sliding angle on the textures within the target region on the surface of the material to be measured based on the output value of the sliding angle prediction model The method for predicting the sliding angle of a material surface is characterized by including the above steps.

2. Before inputting the coordinate values of the textures within the target region into the sliding angle prediction model, further comprising performing principal component analysis on the coordinate values of the textures within the target region on the surface of the material to be measured to reduce the number of input feature quantities on the surface of the textures The method for predicting the sliding angle of a material surface according to claim 1, characterized by the above step.

3. The number of surface feature quantities of the textures is reduced to 300 - 400 The method for predicting the sliding angle of a material surface according to claim 2 or 3, characterized by the above step.

4. Obtaining the sliding angle on the textures within the target region based on the output value of the sliding angle prediction model includes: grouping the output values of the sliding angle prediction model; determining the sliding angle on the textures within the target region based on the grouping result of the output values of the sliding angle prediction model The method for predicting the sliding angle of a material surface according to any one of claims 1 to 3, characterized by including the above steps.

5. Determining the sliding angle on the textures within the target region based on the grouping result of the output values of the sliding angle prediction model includes: obtaining a predetermined sliding angle group corresponding to the output value of the sliding angle prediction model; taking the sliding angle corresponding to the target sliding angle group as the sliding angle on the textures within the target region The method for predicting the sliding angle of a material surface according to claim 4, characterized by including the above steps.

6. Inputting the coordinate values of the textures within the target region into the sliding angle prediction model and obtaining the output value of the sliding angle prediction model includes: Inputting the coordinate values of the texture on the surface of the material to be measured into the input layer of the trained probabilistic neural network, and obtaining an input feature vector by the input layer, wherein the number of input nodes and neurons in the input layer is equal to the number of dimensions of the coordinate values of the texture, and Inputting the input feature vector into the pattern layer of the probabilistic neural network, and obtaining the similarity between the input feature vector and each sample texture in the training set by the pattern layer, and Inputting the similarity into the addition layer of the probabilistic neural network, and performing weighted addition for each similarity by the addition layer to obtain a weighted addition score, and Inputting the weighted addition score into the output layer of the probabilistic neural network, and outputting, as the output value of the slip angle prediction model, the slip angle label with the highest weighted addition score by the output layer The method for predicting the slip angle of the material surface according to claim 2 or 3, comprising the above.

7. Obtaining the size of the texture in the target area and the detection information of the surrounding deposits, and Processing the deposits in the target area so that the deposits surrounding the texture in the target area match based on the detection information The method for predicting the slip angle of the material surface according to any one of claims 1 to 6, further comprising the above.

8. The texture is a dimple The method for predicting the slip angle of the material surface according to any one of claims 1 to 7, characterized by the above.

9. At least one processor, and A memory communicably connected to the at least one processor, wherein A computer program executed by the at least one processor is stored in the memory, and when the at least one processor executes the computer program, the method for predicting the slip angle of the material surface according to any one of claims 1 to 8 is realized The device for predicting the slip angle of the material surface, characterized by the above.

10. A computer program is stored which, when executed by a processor, realizes the method for predicting the slip angle of the material surface according to any one of claims 1 to 8 The computer-readable storage medium, characterized by the above.

11. An imaging device for acquiring image information of the surface of a material to be measured, wherein the image information includes at least an image of the texture within a target region on the surface of the material to be measured. The prediction device for the slip angle of the material surface according to claim 9, which is connected to the imaging device. A slip angle detection system, characterized by comprising the above.

Images