A double-layer plastic blow molding detection method and system

Abstract

The present invention discloses a double-layer plastic blow molding detection method and system, which relates to the field of plastic product production, and includes: obtaining the specification characteristics of a plastic sheet to be tested, and screening out differential grid regions based on a pressure distribution heat map; obtaining a plurality of connected regions composed of each differential grid region, screening out a target connected region from the connected regions, and extracting the average pressure gradient as the pressure characteristic of the plastic sheet to be tested; collecting an echo signal set of the target connected region, obtaining the average echo density of the target connected region according to the echo signal set, and using the average echo density as the ultrasonic characteristic of the plastic sheet to be tested; performing feature-level fusion on the specification characteristic, the pressure characteristic, and the ultrasonic characteristic of the plastic sheet to be tested to obtain multi-modal characteristics; inputting the multi-modal characteristics into a convolutional neural network to obtain the wall thickness detection result of the plastic sheet to be tested; the present invention can achieve efficient and accurate wall thickness detection on plastic sheets of different sizes and different materials.

Description

Technical Field

[0001] The present invention relates to the field of plastic product production, and particularly to a double-layer plastic blow molding detection method and system. Background Art

[0002] Plastic sheets are widely used in multiple industries such as packaging, automobile manufacturing, and building materials. Their wall thickness uniformity directly affects the mechanical properties, durability, and service life of products. However, during the production process of plastic sheets, due to the instability of process parameters such as extrusion, blow molding, and hot pressing, local thickness non-uniformity is likely to occur in plastic sheets, such as over-thick areas caused by excessive local material accumulation, or under-thin areas caused by uneven material stretching. Such wall thickness deviation not only affects the quality stability of plastic sheets but may also cause problems such as fracture and deformation during the actual use of products.

[0003] Existing plastic sheet wall thickness detection methods mainly include contact measurement, ultrasonic thickness measurement, optical measurement, etc. Among them, contact measurement methods (such as micrometers and thickness gauges) require manual operation, have a limited measurement range, and it is difficult to cover the global thickness distribution of the entire sheet. Although the ultrasonic thickness measurement method can achieve non-contact detection, it can only measure at a single point or in a local area and is difficult to comprehensively reflect the thickness change of the entire plastic sheet. In addition, the optical measurement method has good effects on some transparent or semi-transparent plastic sheets, but has poor adaptability to different materials and is easily affected by surface roughness and lighting conditions. Therefore, existing detection technologies are difficult to balance comprehensiveness, accuracy, and applicability and cannot meet the quality detection requirements of plastic sheets for large-scale, automated, and high-precision production.

[0004] In view of this, there is an urgent need for a new wall thickness detection method that can improve the comprehensiveness and automation of detection while ensuring detection accuracy. Summary of the Invention

[0005] The purpose of the present invention is to provide a double-layer plastic blow molding detection method and system to solve the problems in the above background art.

[0006] To achieve the above purpose, the present invention provides the following technical solutions:

[0007] In a first aspect, the present invention provides a double-layer plastic blow molding detection method, including:

[0008] Obtaining the specification characteristics and pressure distribution heat map of the plastic sheet to be measured, and screening out the differential grid areas based on the pressure distribution heat map, where the pressure distribution heat map contains multiple grid areas and the measured average pressure of each grid area;

[0009] Obtain multiple connected regions composed of each differential grid region, screen out the target connected region from the connected regions, and extract the pressure average gradient of the target connected region as the pressure feature of the plastic sheet to be measured;

[0010] Collect the echo signal set of the target connected region, obtain the average echo density of the target connected region according to it, and use the average echo density as the ultrasonic feature of the plastic sheet to be measured;

[0011] Perform feature-level fusion on the specification feature, pressure feature, and ultrasonic feature of the plastic sheet to be measured to obtain multi-modal features;

[0012] Input the multi-modal features into a pre-trained convolutional neural network that performs wall thickness detection tasks to obtain the wall thickness detection result of the plastic sheet to be measured.

[0013] As a preferred technical solution of the first aspect of the present invention, the specification features include but are not limited to the size, type, and designed thickness of the plastic sheet to be measured;

[0014] Among them, the specific calculation formula for the measured pressure mean value of each grid region is as follows: ; In the formula: is the measured pressure mean value of the grid region in the i-th row and j-th column; is the force collected by the r-th pressure sensor corresponding to the grid region in the i-th row and j-th column; is the grid region area of the grid region in the i-th row and j-th column, is the correction factor, and R is the total number of pressure sensors corresponding to each grid region.

[0015] As a preferred technical solution of the first aspect of the present invention, the correction factor is calculated according to the distance between the pressure sensor and the grid center and the specific calculation formula is , is a constant.

[0016] As a preferred technical solution of the first aspect of the present invention, the screening out of the differential grid regions includes:

[0017] Calculate the global pressure mean value and standard deviation of the plastic sheet to be measured;

[0018] Among them, the expression of the global pressure mean value is as follows:

[0019] ;

[0020] Among them, the expression of the standard deviation is as follows:

[0021] ;

[0022] Generate a pressure threshold interval based on the global pressure mean and standard deviation ;

[0023] Compare the measured pressure mean of each grid area with the pressure threshold interval;

[0024] If or , then determine that the corresponding grid area is a differential grid area;

[0025] If and , then determine that the corresponding grid area is a non-differential grid area;

[0026] Repeat the above steps until each grid area is traversed to obtain all differential grid areas.

[0027] As a preferred technical solution of the first aspect of the present invention, the screening of the target connected area from the connected areas includes:

[0028] Calculate the pressure average gradient of each connected area to obtain a plurality of pressure average gradients;

[0029] Among them, the calculation formula of the pressure average gradient is as follows:

[0030] ;

[0031] In the formula: is the pressure average gradient of the connected area , is the gradient value of the grid , is the total number of differential grid areas within the connected area;

[0032] Among them, the calculation formula of the gradient value of the grid is as follows:

[0033] ;

[0034] In the formula: is the measured pressure mean of the grid , is the measured pressure mean of the grid , is the measured pressure mean of the grid , is the measured pressure mean of the grid , is the width of the grid , is the grid height;

[0035] Sort multiple pressure average gradients in descending order of numerical value, and mark the connected region corresponding to the first sorted pressure average gradient as the target connected region.

[0036] As a preferred technical solution of the first aspect of the present invention, the echo signal concentration includes several ultrasonic echo signals;

[0037] Among them, obtaining the average echo density of the target connected region includes:

[0038] Obtain the number of effective echo signals in the echo signal concentration and the area of the target connected region;

[0039] Calculate the average echo density of the target connected region according to the number of effective echo signals and the area of the target connected region;

[0040] Among them, the specific calculation formula of the average echo density is as follows:

[0041] ;

[0042] In the formula: is the average echo density, is the number of effective echo signals in the target connected region, is the area of the target connected region.

[0043] As a preferred technical solution of the first aspect of the present invention, the training method of the convolutional neural network for performing wall thickness detection tasks is as follows:

[0044] Obtain historical wall thickness detection training data, and divide the historical wall thickness detection training data into a wall thickness detection training set and a wall thickness detection test set; the historical wall thickness detection training data includes wall thickness detection features and their corresponding annotation labels;

[0045] Among them, the wall thickness detection features include specification features, pressure features, and ultrasonic features;

[0046] Among them, the annotation labels include "0" and "1", where "0" indicates that the wall thickness is unqualified, and "1" indicates that the wall thickness is qualified;

[0047] Construct a classifier for performing classification tasks based on the convolutional neural network, use the wall thickness detection features in the wall thickness detection training set as the input of the classifier, and use the annotation labels as the output of the classifier to train the classifier to obtain an initial classification network;

[0048] Verify the model of the initial classification network using the wall thickness detection test set, and output the initial classification network whose test accuracy is greater than or equal to the preset test accuracy threshold as the convolutional neural network for performing the wall thickness detection task.

[0049] In a second aspect, the present invention provides a double-layer plastic blow molding detection system, which is implemented based on the double-layer plastic blow molding detection method described above, and includes:

[0050] A data acquisition module, configured to acquire the specification features and the pressure distribution heat map of the plastic sheet to be tested, and screen out the differential grid regions according to the pressure distribution heat map, where the pressure distribution heat map includes a plurality of grid regions and the measured pressure mean value of each grid region;

[0051] A first feature extraction module, configured to obtain a plurality of connected regions composed of the respective differential grid regions, screen out the target connected region from the connected regions, and extract the pressure average gradient of the target connected region as the pressure feature of the plastic sheet to be tested;

[0052] A second feature extraction module, configured to collect the echo signal set of the target connected region, obtain the average echo density of the target connected region according to it, and use the average echo density as the ultrasonic feature of the plastic sheet to be tested;

[0053] A feature fusion module, configured to perform feature-level fusion on the specification features, pressure features, and ultrasonic features of the plastic sheet to be tested to obtain multi-modal features;

[0054] A blow molding detection module, configured to input the multi-modal features into the convolutional neural network pre-trained for performing the wall thickness detection task to obtain the wall thickness detection result of the plastic sheet to be tested.

[0055] In a third aspect, the present invention provides an electronic device, including a memory, a processor, and a computer program stored on the memory and running on the processor, where when the processor executes the computer program, the double-layer plastic blow molding detection method described in any one of the above is implemented.

[0056] In a fourth aspect, the present invention provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed, the double-layer plastic blow molding detection method described in any one of the above is implemented.

[0057] In the above technical solutions, the technical effects and advantages provided by the present invention:

[0058] The present invention measures the pressure distribution of plastic sheets in a grid pattern through a multi-point pressure sensor array, and accurately screens out abnormal wall thickness areas by combining connected component analysis, avoiding misjudgment and insufficient detection coverage caused by single-point measurement methods; compared with existing methods, this application can identify local anomalies through the statistical characteristics of the global pressure distribution, and further verify the thickness of the target connected area by combining ultrasonic detection, making the wall thickness detection more accurate and stable;

[0059] Moreover, by first screening out target connected areas with drastic wall thickness changes, the high-density measurement of the entire plastic sheet is reduced, the computational complexity is lowered, and the detection accuracy is ensured without being affected; in addition, through a feature-level fusion method, the specification features, pressure features, and ultrasonic features of the plastic sheet are fused, and a deep learning model is used for intelligent wall thickness analysis, making the detection results have stronger adaptability and generalization ability, reducing human intervention, and improving the degree of automation;

[0060] Compared with the prior art, the present invention can achieve efficient and accurate wall thickness detection on plastic sheets of different sizes and materials, ensuring the comprehensiveness, stability, and computational efficiency of the detection, and providing more reliable technical support for the quality control of plastic product production. Brief Description of the Drawings

[0061] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required to be used in the embodiments. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those of ordinary skill in the art, other drawings can also be obtained based on these drawings.

[0062] Figure 1 It is a flowchart of a double-layer plastic blow molding detection method of the present invention;

[0063] Figure 2 It is a framework diagram of a double-layer plastic blow molding detection system of the present invention;

[0064] Figure 3 It is a schematic structural diagram of an electronic device provided by the present invention. Detailed Embodiments

[0065] Now, the exemplary embodiments will be described more fully with reference to the accompanying drawings. However, the exemplary embodiments can be implemented in various forms and should not be construed as limited to the examples set forth herein; rather, these exemplary embodiments are provided so that the present disclosure will be more complete and comprehensive, and will fully convey the concept of the exemplary embodiments to those skilled in the art. The drawings are only schematic illustrations of the present disclosure and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and thus their repeated description will be omitted.

[0066] In addition, the described features, structures, or characteristics may be combined in any suitable manner in one or more example embodiments. In the following description, numerous specific details are provided to give a thorough understanding of the example embodiments disclosed in this application. However, those skilled in the art will realize that one or more of the specific details may be omitted to practice the technical solutions disclosed in this application, or other methods, components, steps, etc. may be employed. In other cases, well-known structures, methods, implementations, or operations are not shown or described in detail to avoid obscuring various aspects of the disclosure of this application.

[0067] Example 1

[0068] As Figure 1 shown, this example discloses a double-layer plastic blow molding detection method, including:

[0069] Step 101: Obtain the specification features and the pressure distribution heat map of the plastic sheet to be tested, and screen out the differential grid areas based on the pressure distribution heat map, where the pressure distribution heat map includes a plurality of grid areas and the measured pressure mean value of each grid area;

[0070] Among them, the specification features include but are not limited to the size of the plastic sheet to be tested (such as 30 cm × 30 cm, 100 cm × 100 cm, etc.), type (such as PET, PP, HDPE, PC, etc.), and design thickness (i.e., the standard thickness formulated before production for the corresponding type of plastic sheet to be tested), etc.;

[0071] Specifically, before obtaining the pressure distribution heat map, it includes:

[0072] Match the corresponding division rule according to the size of the plastic sheet to be tested in the specification features, and divide the plastic sheet to be tested into m × n grid areas according to the corresponding division rule, where m and n are integers greater than zero;

[0073] Mark the serial number for each grid area in turn to form the grid number of each grid area (i.e., the position index of each grid area); for example, number each grid area in turn in the order from left to right and from top to bottom to form a unique grid number, such as: in a 32 × 32 grid division, the grid number in the upper left corner is (1, 1), and the grid number in the lower right corner is (32, 32);

[0074] It should be understood that: there is a corresponding division rule for plastic sheets to be tested with different sizes, which is specifically determined by artificial association by technicians. For example, a size of 30 cm × 30 cm is divided into 32 × 32 equal grids;

[0075] It should be noted that: the measured pressure mean value is obtained based on the acquisition of a multi-point pressure sensor array; the pressure distribution heat map is generated according to existing heat map tools, such as Matplotlib (Python), OpenCV (Python), D3.js (JavaScript), etc.; among them, the measured pressure mean value of each grid area is obtained by calculating the forces collected by each pressure sensor in the array after placing the plastic sheet to be measured flat on the multi-point pressure sensor array.

[0076] Specifically, the specific calculation formula for the measured pressure mean value of each grid area is as follows: ; in the formula: is the measured pressure mean value of the grid area in the i-th row and j-th column; is the force collected by the r-th pressure sensor corresponding to the grid area in the i-th row and j-th column; is the grid area of the grid area in the i-th row and j-th column, is the correction factor, and R is the total number of pressure sensors corresponding to each grid area;

[0077] Among them, the correction factor is calculated according to the distance between the pressure sensor and the grid center and the specific calculation formula is , is a constant to avoid division by zero error;

[0078] It should be noted that: the position of the pressure sensor can be determined by the built-in positioning unit, while the coordinates of the grid center are determined by the size of the plastic sheet to be measured and the grid number;

[0079] Exemplarily, assuming that the size of the plastic sheet is W×H (width×height) and the number of grid divisions is m×n (number of rows×number of columns), then the grid size is =W / m, h=H / n, where, is the grid area in the i-th row and j-th column, and the calculation of the grid center position: , , in the formula, is the width of each grid, is the height of each grid, , then is the grid center coordinate of the grid area in the i-th row and j-th column;

[0080] In implementation, the screening of the differential grid areas includes:

[0081] Calculating the global pressure mean value and the standard deviation ;

[0082] Among them, the global pressure mean value The expression is as follows:

[0083] ;

[0084] Among them, the standard deviation The expression is as follows:

[0085] ;

[0086] Generate a pressure threshold interval based on the global pressure mean and standard deviation ;

[0087] Compare the measured pressure mean of each grid area with the pressure threshold interval;

[0088] If or , then determine that the corresponding grid area is a differential grid area;

[0089] If and , then determine that the corresponding grid area is a non-differential grid area;

[0090] Repeat the above steps until each grid area is traversed to obtain all differential grid areas;

[0091] It can be understood that: when or , it means that there is an over-thick or over-thin situation in the corresponding grid area, that is, due to reasons such as blow molding pressure, there is excessive material accumulation in some areas of the plastic sheet to be measured, or there are defects such as pores in some areas, thus resulting in uneven pressure distribution of the plastic sheet to be measured.

[0092] Step 102: Obtain multiple connected areas composed of each differential grid area, screen out the target connected area from the connected areas, and extract the pressure average gradient of the target connected area as the pressure feature of the plastic sheet to be measured;

[0093] It should be noted that: there are multiple differential grid areas in each connected area; among them, the multiple connected areas are obtained by using the connected component analysis (CCA) method; the CCA method is to regard the differential grid area as the "foreground pixel" in the binary image, set all non-differential grid areas to "0" by creating a binary grid matrix, and set the differential grid area to "1", and then use depth-first search (DFS) or breadth-first search (BFS) to traverse adjacent "1" in the connected area to find multiple connected areas;

[0094] In implementation, the screening of the target connected area from the connected areas includes:

[0095] Calculate the pressure average gradient of each connected region to obtain multiple pressure average gradients; the pressure average gradient (MPG) is a value used to measure the pressure average gradient of all grids within a certain connected region;

[0096] Among them, the calculation formula for the pressure average gradient is as follows:

[0097] ;

[0098] In the formula: is the pressure average gradient of the connected region , is the gradient value of the grid , is the total number of differential grid regions within the connected region;

[0099] Among them, the calculation formula for the gradient value of the grid is as follows:

[0100] ;

[0101] In the formula: is the measured average pressure of the grid , that is, the pressure of the grid on the right relative to the grid , is the measured average pressure of the grid , that is, the pressure of the grid on the left relative to the grid , is the measured average pressure of the grid , that is, the pressure of the grid below relative to the grid , is the measured average pressure of the grid , that is, the pressure of the grid above relative to the grid , is the width of the grid , a normalization factor for calculating the gradient in the X direction, is the height of the grid , a normalization factor for calculating the gradient in the y direction;

[0102] Sort the multiple pressure average gradients from largest to smallest, and mark the connected region corresponding to the first sorted pressure average gradient as the target connected region;

[0103] It should be understood that: the greater the average pressure gradient, the more drastic the pressure change in the corresponding connected region is reflected, which usually corresponds to the part with the most significant wall thickness change; therefore, by determining the target connected region based on the average pressure gradient and using the average pressure gradient of the target connected region as the pressure characteristic of the plastic sheet to be measured, the accuracy of the thickness detection of the plastic sheet to be measured can be significantly improved, and at the same time, the computational resource consumption caused by calculating each local region can be reduced.

[0104] Step 103: Collect the echo signal set of the target connected region, obtain the average echo density of the target connected region according to it, and use the average echo density as the ultrasonic characteristic of the plastic sheet to be measured;

[0105] Among them, the echo signal set includes a number of ultrasonic echo signals;

[0106] It should be noted that: the echo signal set is obtained after scanning the target connected region by an ultrasonic detection device, and the ultrasonic detection device includes but is not limited to an ultrasonic detector, an ultrasonic array detector, an ultrasonic measuring instrument, etc.;

[0107] In implementation, obtaining the average echo density of the target connected region includes:

[0108] Obtain the number of effective echo signals in the echo signal set and the area of the target connected region;

[0109] It should be understood that: the effective echo signals are selected according to a preset effective echo reflection time interval, that is, obtain the echo reflection time of each ultrasonic echo signal in the echo signal set, and compare the echo reflection time with the effective echo reflection time interval , if the echo reflection time is within the effective echo reflection time interval , then mark the corresponding ultrasonic echo signal as an effective echo signal, otherwise, do not mark the corresponding ultrasonic echo signal as an effective echo signal;

[0110] Calculate the average echo density of the target connected region according to the number of effective echo signals and the area of the target connected region;

[0111] Among them, the specific calculation formula of the average echo density is as follows:

[0112] ;

[0113] In the formula: is the average echo density, is the number of effective echo signals in the target connected region, is the area of the target connected region;

[0114] It should be understood that by collecting the pressure characteristics and ultrasonic characteristics of the plastic sheet to be measured, the errors and false detections caused by a single detection method can be eliminated, and at the same time, it is beneficial to improve the wall thickness detection accuracy of the plastic sheet.

[0115] Step 104: Perform feature-level fusion on the specification characteristics, pressure characteristics, and ultrasonic characteristics of the plastic sheet to be measured to obtain multi-modal characteristics;

[0116] It should be noted that the feature-level fusion is implemented by either direct feature splicing or weighted feature fusion, and this embodiment does not make too many limitations in the comparison.

[0117] Step 105: Input the multi-modal characteristics into a convolutional neural network pre-trained to perform the wall thickness detection task to obtain the wall thickness detection result of the plastic sheet to be measured;

[0118] Specifically, the training method of the convolutional neural network performing the wall thickness detection task is as follows:

[0119] Obtain historical wall thickness detection training data, and divide the historical wall thickness detection training data into a wall thickness detection training set and a wall thickness detection test set; the historical wall thickness detection training data includes wall thickness detection characteristics and their corresponding annotation labels;

[0120] Among them, the wall thickness detection characteristics include specification characteristics, pressure characteristics, and ultrasonic characteristics;

[0121] Among them, the annotation labels include "0" and "1", where "0" indicates that the wall thickness is unqualified, and "1" indicates that the wall thickness is qualified;

[0122] It should be noted that the wall thickness detection characteristics in the historical wall thickness detection training data are actually collected and recorded by technical personnel according to experimental data, and the annotation labels in the historical wall thickness detection training data are manually annotated by data annotators;

[0123] Construct a classifier for performing a classification task based on a convolutional neural network, use the wall thickness detection characteristics in the wall thickness detection training set as the input of the classifier, and use the annotation labels as the output of the classifier to train the classifier to obtain an initial classification network;

[0124] Use the wall thickness detection test set to perform model verification on the initial classification network, and output the initial classification network greater than or equal to the preset test accuracy threshold as the convolutional neural network for performing the wall thickness detection task.

[0125] Embodiment 2

[0126] As Figure 2 shown, the parts not detailed in this embodiment are as shown in Embodiment 1. This embodiment publicly provides a double-layer plastic blow molding detection system, including:

[0127] A data acquisition module 201 is configured to acquire the specification features and the pressure distribution heat map of the plastic sheet to be tested, and screen out the differential grid regions according to the pressure distribution heat map. The pressure distribution heat map includes a plurality of grid regions and the measured average pressure of each grid region.

[0128] A first feature extraction module 202 is configured to obtain a plurality of connected regions composed of the respective differential grid regions, screen out a target connected region from the connected regions, and extract the pressure average gradient of the target connected region as the pressure feature of the plastic sheet to be tested.

[0129] A second feature extraction module 203 is configured to collect an echo signal set of the target connected region, obtain the average echo density of the target connected region according to the echo signal set, and use the average echo density as the ultrasonic feature of the plastic sheet to be tested.

[0130] A feature fusion module 204 is configured to perform feature-level fusion on the specification features, the pressure features, and the ultrasonic features of the plastic sheet to be tested to obtain multi-modal features.

[0131] A blow molding detection module 205 is configured to input the multi-modal features into a convolutional neural network pre-trained to perform a wall thickness detection task to obtain the wall thickness detection result of the plastic sheet to be tested.

[0132] Embodiment 3

[0133] Please refer to Figure 3 As shown, this embodiment discloses an electronic device, including a memory, a processor, and a computer program stored on the memory and running on the processor. When the processor executes the computer program, the double-layer plastic blow molding detection method provided by any one of the above methods is implemented.

[0134] Since the electronic device introduced in this embodiment is the electronic device adopted for implementing the double-layer plastic blow molding detection method in the embodiments of the present application, based on the double-layer plastic blow molding detection method introduced in the embodiments of the present application, those skilled in the art can understand the specific implementation manners and various variations of the electronic device in this embodiment. Therefore, the specific implementation of how this electronic device implements the method in the embodiments of the present application will not be described in detail herein. As long as those skilled in the art implement the electronic device adopted for the double-layer plastic blow molding detection method in the embodiments of the present application, it falls within the scope of protection of the present application.

[0135] Embodiment 4

[0136] This embodiment publicly provides a computer-readable storage medium, including a memory, a processor, and a computer program stored on the memory and running on the processor. When the processor executes the computer program, it implements any one of the double-layer plastic blow molding detection methods provided by the above-mentioned various methods.

[0137] The above formulas are all dimensionless and take their numerical values for calculation. The formulas are obtained by collecting a large amount of data for software simulation to obtain a formula closest to the actual situation. The preset parameters, weights, and threshold selections in the formulas are set by those skilled in the art according to the actual situation.

[0138] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on the computer, the processes or functions described in the embodiments of the present invention are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via a wired network or a wireless network. The computer-readable storage medium can be any available medium that the computer can access or a data storage device such as a server or a data center that includes one or more collections of available media. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a magnetic tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state drive.

[0139] Only some exemplary embodiments of the present invention have been described by way of illustration above. Undoubtedly, for those of ordinary skill in the art, the described embodiments can be modified in various different ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and descriptions are illustrative in nature and should not be construed as limiting the protection scope of the claims of the present invention.

Claims

1. A double-layer plastic blow molding detection method, characterized in that, Including: Obtain the specification characteristics and the thermal map of the pressure distribution of the plastic sheet to be measured, and screen out the differential grid areas based on the thermal map of the pressure distribution. The thermal map of the pressure distribution contains multiple grid areas and the measured average pressure of each grid area; Among them, the screening out of the differential grid areas includes: Calculate the global pressure mean of the plastic sheet to be measured and standard deviation ; Generate a pressure threshold interval based on the global mean pressure and standard deviation ; Compare the measured average pressure of each grid area with the pressure threshold range; If or , it is determined that the corresponding grid area is a differential grid area, where is the measured average pressure of the grid area in the i-th row and j-th column; Obtain multiple connected areas composed of each differential grid area, screen out the target connected area from the connected areas, and extract the average pressure gradient of the target connected area as the pressure characteristic of the plastic sheet to be measured; Among them, the screening out of the target connected area from the connected areas includes: Calculate the average pressure gradient of each connected area to obtain multiple average pressure gradients; Sort the multiple average pressure gradients from large to small in value, and mark the connected area corresponding to the first sorted average pressure gradient as the target connected area; Collect the echo signal set of the target connected area, obtain the average echo density of the target connected area according to it, and use the average echo density as the ultrasonic characteristic of the plastic sheet to be measured; Among them, the echo signal set includes several ultrasonic echo signals; Among them, the obtaining of the average echo density of the target connected area includes: Obtain the number of valid echo signals in the echo signal set and the area of the target connected area; Calculate the average echo density of the target connected area according to the number of valid echo signals and the area of the target connected area; Perform feature-level fusion on the specification characteristics, pressure characteristics and ultrasonic characteristics of the plastic sheet to be measured to obtain multi-modal characteristics; Input the multi-modal characteristics into a convolutional neural network pre-trained to perform the wall thickness detection task to obtain the wall thickness detection result of the plastic sheet to be measured.

2. The double-layer plastic blow molding detection method according to claim 1, wherein The specification characteristics include but are not limited to the size, type and designed thickness of the plastic sheet to be measured; Among them, the specific calculation formula for the measured pressure mean value of each grid area is as follows: ; In the formula: is the force collected by the r-th pressure sensor corresponding to the grid area in the i-th row and j-th column; is the grid area of the grid area in the i-th row and j-th column, is the correction factor, and R is the total number of pressure sensors corresponding to each grid area.

3. The double-layer plastic blow molding detection method according to claim 2, characterized in that, The correction factor is calculated based on the distance between the pressure sensor and the center of the grid and the specific calculation formula is , where 4. The double-layer plastic blow molding detection method according to claim 3, characterized in that, The screening out of the differential grid areas further includes: If and , it is determined that the corresponding grid area is a non-differential grid area; Among them, the global average pressure has the following expression: ; Among them, the standard deviation has the following expression: ; Where: m is the number of rows and n is the number of columns.

5. The double-layer plastic blow molding detection method according to claim 4, characterized in that, The calculation formula of the average pressure gradient is as follows: ; In the formula: is the pressure average gradient of the connected region , is the gradient value of the grid , is the total number of differential grid regions within the connected region; Among them, the grid The calculation formula for the gradient value is as follows: ; Where: For Grid The measured pressure mean, For Grid The measured pressure mean, For Grid The measured pressure mean, For Grid The measured pressure mean, For Grid The width of For Grid height.

6. The double-layer plastic blow molding detection method according to claim 5, characterized in that, The specific calculation formula of the average echo density is as follows: ; Wherein: is the average echo density, is the number of effective echo signals within the target connected region, is the area of the target connected region.

7. The double-layer plastic blow molding detection method according to claim 6, characterized in that, The training method of the convolutional neural network for performing the wall thickness detection task is as follows: Obtain historical wall thickness detection training data, and divide the historical wall thickness detection training data into a wall thickness detection training set and a wall thickness detection test set; the historical wall thickness detection training data includes wall thickness detection characteristics and their corresponding labeled labels; Among them, the wall thickness detection characteristics include specification characteristics, pressure characteristics and ultrasonic characteristics; Among them, the labeled labels include "0" and "1", where "0" indicates that the wall thickness is unqualified and "1" indicates that the wall thickness is qualified; Construct a classifier for performing a classification task based on a convolutional neural network, use the wall thickness detection characteristics in the wall thickness detection training set as the input of the classifier, and use the labeled labels as the output of the classifier to train the classifier to obtain an initial classification network; Use the wall thickness detection test set to perform model verification on the initial classification network, and output the initial classification network greater than or equal to the preset test accuracy threshold as the convolutional neural network for performing the wall thickness detection task.

8. A double-layer plastic blow molding detection system, which is implemented based on the double-layer plastic blow molding detection method described in any one of claims 1-7, and is characterized in that, Including: A data acquisition module, configured to acquire the specification features and the pressure distribution heat map of the plastic sheet to be tested, and screen out the differential grid regions based on the pressure distribution heat map, where the pressure distribution heat map includes a plurality of grid regions and the measured average pressure of each grid region; A first feature extraction module, configured to obtain a plurality of connected regions composed of the respective differential grid regions, screen out a target connected region from the connected regions, and extract the pressure average gradient of the target connected region as the pressure feature of the plastic sheet to be tested; A second feature extraction module, configured to collect an echo signal set of the target connected region, obtain the average echo density of the target connected region according to the echo signal set, and use the average echo density as the ultrasonic feature of the plastic sheet to be tested; A feature fusion module, configured to perform feature-level fusion on the specification features, the pressure features, and the ultrasonic features of the plastic sheet to be tested to obtain multi-modal features; A blow molding detection module, configured to input the multi-modal features into a convolutional neural network pre-trained to perform a wall thickness detection task to obtain the wall thickness detection result of the plastic sheet to be tested.

9. An electronic device, comprising a memory, a processor, and a computer program stored on the memory and running on the processor, characterized in that When the processor executes the computer program, the double-layer plastic blow molding detection method according to any one of claims 1-7 is implemented.

10. A computer-readable storage medium, characterized in that, A computer program is stored on the computer-readable storage medium, and when the computer program is executed, the double-layer plastic blow molding detection method according to any one of claims 1-7 is implemented.

Images