Tire internal bubble defect detection method and system based on tactile perception

Abstract

The invention relates to the technical field of tire detection, in particular to a tire internal bubble defect detection method and system based on tactile perception. The method comprises the steps that a tire to be detected is fixed, space coordinates are calibrated, a multi-axis movement mechanism is used for driving a three-way force sensor to conduct scanning line by line along the surface of the tire, stress data in the X direction, the Y direction and the Z direction are collected under different applied normal forces, noise filtering, zero drift correction and interpolation processing are conducted on the data, and a position-mechanical characteristic data set is constructed. By constructing a multi-force-level mechanical fingerprint map and comparing the multi-force-level mechanical fingerprint map with a standard tire mechanical fingerprint database, an abnormally low-rigidity or deformation area is automatically identified, and the bubble size and the defect severity are estimated. The method overcomes the defects that traditional X-ray and ultrasonic detection equipment is high in cost, complex in operation and low in detection efficiency, has the advantages of being high in detection precision, good in real-time performance and high in adaptability, and is suitable for internal defect online detection and quality control in tire production, retreading and maintenance.

Description

Technical Field

[0001] The present invention relates to the technical field of tire detection, and in particular to a method and system for detecting air bubble defects inside a tire based on tactile perception. Background Art

[0002] Tires are one of the most important components in modern transportation, and their safety directly affects driving safety. Tire quality control, especially the detection of internal defects in tires, is an important part of ensuring tire performance and safety. During the manufacturing, use and maintenance of tires, defects such as internal bubbles may occur due to uneven materials, process problems or external factors. These defects may not be easy to detect in the early stage, but during the use of tires, especially under high-speed driving or high-load conditions, they may cause structural damage to the tires and even cause traffic accidents. Therefore, accurate and effective detection of internal defects in tires is the key to improving tire quality management and safety in use.

[0003] Traditional tire internal defect detection methods mainly rely on technologies such as X-ray imaging, ultrasonic testing, and optical imaging. Although these methods can provide relatively accurate detection results, they are usually limited by high costs, complex operations, low detection efficiency, and the need for specialized equipment. In addition, methods such as X-rays and ultrasonics can only provide relatively fuzzy images or data on the internal structure and defects of tires, and have a greater impact on the complex surface morphology of tires, resulting in low detection accuracy and sensitivity.

[0004] As an emerging non-destructive testing technology, tactile sensing technology has the advantages of high sensitivity, real-time and convenience. The Chinese invention patent application (publication number: CN116818172A, publication date: 2023-09-29) discloses a flexible interface detection method and device based on tactile perception, which includes the following steps: controlling the three-dimensional displacement stage to move downward along the Z-axis direction, obtaining the Z-direction force value of the three-axis force sensor in real time, stopping the movement when the force value reaches the preset threshold, and establishing a reference plane for the sensor to scan the flexible interface; on the reference plane, realizing the synchronous control of the three-axis force sensor data acquisition and the displacement stage based on the hard trigger method; scanning the flexible interface line by line based on the data acquisition method, collecting the force change data of the sensor and the flexible interface in the X, Y and Z directions as three-dimensional tactile information; fusing and calculating the three-dimensional force data to achieve high-precision, high-resolution, multi-dimensional tactile scanning imaging of the flexible interface, and finally realizing flexible interface detection based on tactile perception.

[0005] However, the current tactile sensing technology used in tire internal defect detection is still immature. In particular, there are still some urgent problems to be solved in terms of how to combine scans with different force levels, how to accurately identify internal bubble defects, and how to efficiently process data. Summary of the invention

[0006] In order to solve the above-mentioned technical problems, the purpose of the present invention is to provide a method for detecting bubble defects inside tires based on tactile perception. The method combines multi-force-level scanning with data processing, uses a three-axis force sensor to obtain the force information on the tire surface, and then constructs a multi-force-level mechanical fingerprint map and compares it with a standard fingerprint library to accurately determine whether there are defects such as bubbles inside the tire, thereby overcoming the shortcomings of traditional methods and providing a high-precision and efficient solution for the field of tire detection.

[0007] In order to achieve the above-mentioned purpose, the present invention adopts the following technical solutions:

[0008] A method for detecting air bubble defects inside a tire based on tactile perception, the method comprising the following steps:

[0009] 1) Tire fixing: install the tire to be tested on a rotatable or clampable positioning device, and calibrate the tire's spatial coordinate system;

[0010] 2) Initial loading: Control the multi-axis motion mechanism to drive the three-axis force sensor to move in the vertical direction and contact the outer surface of the tire. When the normal force of the sensor reaches the first preset threshold, a detection reference surface is established;

[0011] 3) Multi-level scanning:

[0012] 3.1) With the detection reference plane as a reference, a scanning path is set on the outer surface of the tire, and normal forces of different magnitudes are applied to the same scanning point in sequence, and the force data in the three directions of X, Y, and Z output by the three-axis force sensor under each level of force loading is recorded;

[0013] 3.2) Repeat the above steps to scan the outer surface of the tire line by line or partition by partition to collect three-dimensional force data of multiple force levels;

[0014] 4) Data preprocessing: noise filtering, zero drift correction and interpolation processing are performed on the collected three-dimensional force data, and the force changes of the same scanning point under different force loading are integrated to generate the corresponding position-mechanical characteristic data set;

[0015] 5) Defect identification:

[0016] 5.1) constructing a multi-force-level mechanical fingerprint based on the position-mechanical feature data set;

[0017] 5.2) comparing the multi-force-level mechanical fingerprint with a pre-established standard mechanical fingerprint library, and when a local area is found to exhibit abnormally low stiffness or deformation characteristics, it is determined that the area has internal bubble defects;

[0018] 5.3) Output the location and defect degree information of the bubble defect area.

[0019] Preferably, the multi-force level scanning in step 3) includes first performing a preliminary scan with a relatively small force value, and then performing a secondary fine scan of the suspicious area with a higher force value to distinguish the deformation difference caused by tiny surface protrusions and deep bubbles.

[0020] Preferably, the data preprocessing in step 4) further includes filtering or compensating for periodic fluctuations generated by the tire surface pattern to reduce interference of the tread pattern on internal defect identification.

[0021] Preferably, the signal output by the three-axis force sensor is collected synchronously by hardware triggering by an FPGA or SoC platform, specifically including counting the pulse signal of the motion mechanism and triggering sampling when the count reaches a preset value, thereby ensuring the accurate correspondence between the spatial coordinates and the force data in different force loading stages.

[0022] Preferably, the defect identification step 5) utilizes a multi-task deep fingerprint network to realize intelligent defect recognition based on multi-force level scanning data by combining defect area segmentation and bubble size regression tasks; specifically includes an encoder-decoder structure, an attention module and a dual-branch output design, and adopts a multi-task loss function for joint training; in the inference stage, the network can automatically extract defect features from the preprocessed mechanical fingerprint image, and output the defect area, bubble size and defect severity.

[0023] Furthermore, the present invention also provides a tire internal bubble defect detection system based on tactile perception, which is used in the method described above, and the system comprises:

[0024] A tire positioning device, used to fix the tire to be tested and provide a rotation or clamping function;

[0025] A multi-axis motion mechanism, arranged opposite to the tire positioning device, having at least one axis moving in a horizontal direction and one axis moving in a vertical direction;

[0026] A three-axis force sensor, installed at the end of the multi-axis motion mechanism, for detecting the force values ​​in the X, Y and Z directions when in contact with the tire surface;

[0027] A force loading unit, cooperating with the three-axis force sensor, is used to apply multiple levels of normal forces of different magnitudes to the tire surface during scanning;

[0028] A data acquisition control unit, connected to the three-axis force sensor, for acquiring three-axis force data based on a hard trigger method and synchronizing it with the spatial position of the multi-axis motion mechanism;

[0029] The data processing and defect recognition module communicates with the data acquisition control unit, and is used to pre-process the collected multi-force-level three-dimensional force data and execute a defect recognition algorithm to output a tire internal defect judgment result.

[0030] Preferably, the force loading unit comprises a screw mechanism or a cylinder mechanism and is equipped with a force closed-loop controller, which can dynamically adjust the force applied by the three-axis force sensor within a specified force value range.

[0031] Preferably, the three-axis force sensor has a replaceable wear-resistant probe or anti-skid pad to adapt to tire surfaces of different specifications and materials and reduce probe wear.

[0032] Preferably, the data acquisition control unit includes an FPGA or SoC platform, configured with a pulse counter and a high-speed AD conversion circuit, and triggers the sampling of the three-axis force sensor by counting the pulse signals of the stepping or servo motor of the multi-axis motion mechanism and reaching a predetermined pulse value, thereby achieving real-time synchronization of the scanning position and the force data.

[0033] Preferably, the data processing and defect recognition module establishes a standard tire mechanical fingerprint library, and compares the multi-force level scanning data of tires of different models and specifications under normal conditions. When the detection data differs from the standard data within a preset threshold range, it is determined that there are defects such as bubbles or delamination inside the tire, and a visual report of the defects is output.

[0034] Furthermore, the present invention also provides a computer-readable storage medium on which a computer program or instruction is stored. When the computer program or instruction is executed by a processor, step 4) to step 5) of the method are implemented.

[0035] Furthermore, the present invention also provides a computer program product, comprising a computer program or instructions, which implements step 4) to step 5) of the method when executed by a processor.

[0036] The present invention adopts the above technical solution and innovatively combines multi-level scanning, data preprocessing and defect recognition algorithm to achieve efficient and accurate detection of bubble defects inside tires, overcomes the shortcomings of the prior art, and has the following significant technical effects:

[0037] 1. Improve the detection accuracy of bubble defects inside the tire: The present invention applies normal forces of different sizes on the outer surface of the tire, gradually scans and collects the three-dimensional force data of the three-axis force sensor under different force loadings, thereby obtaining multi-force-level mechanical responses. Through multi-force-level scanning, it is possible to more carefully analyze the slight deformation of the tire surface and the stiffness changes caused by internal bubbles, and achieve accurate identification of bubble defects inside the tire. Compared with traditional X-ray, ultrasonic and other detection methods, the present invention can better distinguish between surface micro-protrusions and deep bubbles, significantly improving the detection sensitivity and accuracy of bubble defects.

[0038] 2. Reduce the interference of surface patterns on detection: When the tire surface pattern is relatively complex, the traditional detection method may be interfered by the pattern, resulting in inaccurate detection results. The present invention effectively reduces the interference of the pattern on internal defect identification and improves the reliability of the data by filtering or compensating for the periodic fluctuations generated by the pattern in the data preprocessing stage. This technology effectively solves the problem that traditional tactile detection cannot eliminate pattern interference, so that accurate internal defect detection can still be achieved when the tire surface pattern is relatively complex.

[0039] 3. Improve the synchronization and real-time performance of data acquisition and processing: The present invention uses an FPGA or SoC platform to perform hardware-triggered synchronous acquisition of the output signal of the three-axis force sensor, ensuring that the spatial coordinates and force data can be accurately matched at different force loading stages. This hard-triggered synchronous acquisition method not only improves the real-time performance of data acquisition, but also greatly reduces the possibility of data mismatch, ensuring high precision and high efficiency during the detection process. Compared with the traditional sampling method that relies on software control, the present invention has significant advantages in speed and accuracy.

[0040] 4. Automatic identification and intelligent judgment of internal defects of tires: In the defect identification process, the present invention uses machine learning or deep learning models to extract features from multi-force level mechanical fingerprints, which can automatically determine the size and distribution of bubbles inside the tire and output the severity score of the defect. This intelligent defect judgment mechanism greatly improves the automation level and accuracy of tire detection, and reduces the errors and subjective interference of manual judgment. This technology not only improves the detection efficiency, but also realizes the intelligence and automation of the tire detection process to a certain extent.

[0041] 5. Reduce detection costs and improve detection efficiency: The present invention uses tactile sensing technology to detect tire defects, without the need for expensive X-ray equipment or ultrasonic detection equipment, significantly reducing the equipment cost of detection. The tactile sensing system has a simple structure, is easy to integrate and deploy, and can achieve rapid detection, making it particularly suitable for large-scale tire production lines or tire retreading plants. While improving detection efficiency, it can also provide high-precision defect detection results to meet the needs of industrial applications.

[0042] 6. Strong adaptability, able to cope with different types of tire detection requirements: The detection method and system of the present invention have strong adaptability and can cope with different types, specifications and patterns of tire detection requirements. Whether it is a car tire, a truck tire or an engineering machinery tire, the present invention can achieve accurate detection by adjusting parameters such as the scanning path, loading force, and scanning accuracy. At the same time, the system can also be adaptively adjusted according to the different characteristics and defect types of the tire surface, further improving the flexibility and accuracy of the detection process.

[0043] In summary, the present invention realizes efficient and accurate detection of bubble defects inside tires through the innovative application of tactile sensing technology, and has high detection accuracy, real-time performance, adaptability and intelligence. Compared with the prior art, the present invention has significant technical advantages in improving tire quality control capabilities, reducing detection costs, and improving detection efficiency, and is an important technological breakthrough in the field of tire manufacturing, retreading and maintenance. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 It is a flowchart of the method of the present invention.

[0045] Figure 2 It is a flowchart of the multi-level scanning method of the present invention.

[0046] Figure 3 This is a flowchart of data preprocessing of the present invention.

[0047] Figure 4 This is a flowchart of defect identification according to the present invention. DETAILED DESCRIPTION

[0048] In order to enable those skilled in the art to understand and implement the technical solution of the present invention, a specific implementation method and system for detecting bubble defects inside tires based on tactile perception is described in detail below. Through the introduction of this implementation, technicians can achieve the technical effects of the present invention and perform efficient and accurate detection of bubble defects inside tires. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0049] Example 1

[0050] like Figure 1 A method for detecting air bubble defects inside a tire based on tactile perception is shown, the method comprising the following steps:

[0051] 1. Tire fixation and spatial coordinate calibration

[0052] First, the tire to be tested is mounted on the test platform. The test platform includes a tire positioning device to fix the tire and ensure that it remains stable during the scanning process. The positioning device may include a clamping system or a turntable system that allows the tire to rotate during the scanning process so that the sensor can scan the entire surface of the tire. In order to ensure the accurate spatial position of the tire, a three-dimensional positioning sensor is used to calibrate the spatial coordinates.

[0053] 2. Initial loading and datum surface establishment

[0054] When the tire is fixed, the system starts the multi-axis motion mechanism, which includes multiple motion axes, at least one axis moves in the horizontal direction and another axis moves in the vertical direction. At this time, the three-axis force sensor is driven to move in the vertical direction and contact the tire surface. The normal force of the sensor gradually increases. When the normal force reaches the preset first threshold, the system stops the vertical movement of the sensor and uses this position as the reference plane to start the subsequent scan of the tire.

[0055] 3. Multi-level scanning

[0056] After establishing the reference plane, the system starts multi-level scanning. Figure 2 As shown, the specific steps are:

[0057] Step 3.1: Based on the set scanning path, the system will apply normal forces of different magnitudes and record the three-dimensional force data of the sensor (force values ​​in the X, Y, and Z directions) at each force level. This can fully capture the mechanical response of the tire surface and its interior.

[0058] Step 3.2: The sensor scans the tire surface line by line along the set scanning path. During each scan, the sensor collects multiple sets of three-dimensional force data according to the loading of different force levels. During the scanning process, the sensor can carefully capture the tiny deformation of the tire surface, especially the changes in mechanical characteristics caused by bubbles inside the tire.

[0059] 4. Data preprocessing and fusion

[0060] During the multi-force level scanning process, the collected three-dimensional force data may contain noise, so the data needs to be preprocessed. Figure 3 As shown, the preprocessing steps include:

[0061] Noise filtering: Remove sensor noise through digital filtering algorithms.

[0062] Zero drift correction: Correct the initial zero drift of the sensor to ensure the accuracy of the force value.

[0063] Interpolation processing: Interpolation processing is performed on the interval data in the scanning path to ensure the smoothness and continuity of the data.

[0064] The processed data are fused into corresponding position-mechanical feature data sets, which are the basis for subsequent defect identification.

[0065] 5. Defect identification and bubble detection

[0066] After completing data preprocessing, the system begins to enter the defect identification stage. Figure 4 As shown, the specific steps are:

[0067] Step 5.1: Based on the position-mechanical feature data set, construct a multi-force level mechanical fingerprint map. These mechanical fingerprint maps are graphical representations of the force response at different locations on the tire surface, which can reflect the mechanical properties and potential defect areas of the tire.

[0068] Furthermore, based on the preprocessed position-mechanical characteristic data set, the three-dimensional force data (X, Y, and Z directions) of each scanning position is taken as a data point. By arranging these data in two-dimensional space, a multi-force level force distribution diagram is generated. Each row represents the scanning path of a tire (scanning along the X-axis), and each column represents the mechanical response of a certain position on the tire surface. Each corresponding point is the force value in the three directions of X, Y, and Z, and different force intensities can be displayed through color gradients (blue represents low force and red represents high force).

[0069] Generation of mechanical fingerprint: The force data (force in the X direction) of each scan path is mapped into a thermal map. The force data in the X, Y, and Z directions can be plotted into separate mechanical images (X force map, Y force map, and Z force map). These images can be generated from the scan results at different force levels. Data at multiple force levels can be spliced ​​together to form a three-dimensional mechanical fingerprint, where each dimension corresponds to a different force condition.

[0070] Step 5.2: Compare the constructed mechanical fingerprint with the standard tire mechanical fingerprint library. The standard tire mechanical fingerprint library includes multi-force level scanning data of different types of tires under normal conditions. Through comparison, the system can determine which areas have significant differences in force characteristics from normal tires.

[0071] Standard mechanical fingerprint library: The standard fingerprint library contains multi-force level scanning data of different types of tires under normal conditions. Each standard tire provides a mechanical fingerprint image, which represents the force response of this type of tire under different scanning force levels. The standard fingerprint library is usually collected by tire manufacturers or laboratories based on a large amount of normal tire data. The dimension of each fingerprint image is the same as the dimension of the mechanical fingerprint image of the actual tested tire.

[0072] Furthermore, the comparison method adopted by the present invention is to compare the mechanical fingerprint of the current tire with the fingerprint in the standard tire fingerprint library through an image matching algorithm. Common image matching algorithms include:

[0073] Mean Square Error (MSE): Calculate the mean square error between two tensiometric fingerprints. Areas with large errors usually indicate anomalies.

[0074] Correlation coefficient: By calculating the correlation between two images, we can determine the similarity. Low correlation areas may indicate defects.

[0075] Local feature matching: When comparing images, you can selectively compare local areas in the image to find areas where there are differences in the tire surface.

[0076] Output comparison results: If the comparison results show that the force characteristics of certain areas are significantly different from those of normal tires (such as large normal force, friction or deformation), the system will mark these areas as potential defect areas, especially those that may have problems such as bubbles or delamination.

[0077] Step 5.3: When the comparison results show that a local area exhibits abnormally low stiffness or deformation characteristics, the system determines that there are internal bubbles or other defects in the area. The system further outputs information such as the location of the defect, bubble size, and degree of the defect for subsequent quality assessment and processing.

[0078] Furthermore, the defect feature identification of the present invention is as follows: when the comparison results show that the stress characteristics (low stiffness, deformation characteristics) of a local area are significantly different from those of a normal tire, the system will automatically identify the area as a potential defect through threshold judgment.

[0079] Low stiffness: The bubble region generally exhibits lower stiffness, meaning that the region will deform more under the same applied force.

[0080] Abnormal deformation: By comparing the changes in normal force and lateral force, if the deformation of a certain area is more significant than other areas, it can be determined that there are internal bubbles.

[0081] Furthermore, the output defect information of the present invention is as follows: the system outputs information such as defect location, bubble size, defect degree, etc., and generates a visual report. The report will show the detected defect area (through a heat map or annotated area) and provide detailed information of the defect (such as the radius, depth or deformation amplitude of the bubble).

[0082] Furthermore, the present invention also establishes the relationship between the bubble size and the force change. The following is a simplified model based on the tire material stiffness and bubble size:

[0083] Effect of bubbles: It is assumed that the bubbles inside the tire have a linear effect on the mechanical response of the tire surface, that is, the size of the bubble is proportional to the change in force.

[0084] Bubble size estimation: 1) A batch of tire samples are selected, in which the internal bubble size (e.g., diameter) is known manually or by means of X-ray detection, and the normal force values ​​of the normal area and the bubble area in these samples under the same loading conditions are measured.

[0085] 2) Under the same standard applied load (10N, 50N, 100N), record for each sample: Normal force F in normal area normal ; Normal force F in the bubble region defect ; Calculate the normal force difference in this area: ΔF=F defect −F normal .

[0086] 3) Establish a calibration model Statistical analysis was performed on multiple sets of data (ΔF and the corresponding bubble diameter D), and the empirical formula was obtained by fitting using linear regression and other methods:

[0087]

[0088] Where a and b are constants obtained from experimental fitting.

[0089] Furthermore, in order to improve the accuracy of bubble detection, the present invention first performs a preliminary scan with a smaller force value, and then performs a secondary fine scan of the suspicious area with a higher force value. This method can effectively distinguish the difference between deep deformation caused by bubbles and tiny surface protrusions, thereby improving the recognition accuracy of bubble defects.

[0090] In order to further improve the intelligent level of detection, the present invention adopts machine learning or deep learning models in defect identification. Specifically, deep neural networks can be used to automatically extract the features of multi-force level mechanical fingerprints and determine the size and distribution of bubbles inside the tire based on these features. This intelligent defect judgment mechanism reduces manual intervention and improves the automation and accuracy of detection.

[0091] 6. System hardware implementation

[0092] The system hardware of the present invention adopts the flexible interface detection method of the Chinese invention patent application (publication number: CN116818172A, publication date: 2023-09-29), which specifically includes the following key components:

[0093] Tire positioning device: used to stably fix the tire to be tested and support rotation or clamping operations.

[0094] Multi-axis motion mechanism: includes multiple adjustable axes, used to drive the three-axis force sensor to scan the tire surface.

[0095] Three-axis force sensor: used to collect three-axis force data when the tire surface contacts the sensor in real time.

[0096] Force loading unit: used to apply normal forces of different magnitudes to the tire surface during the scanning process to ensure that all areas that need to be inspected are covered.

[0097] Data acquisition control unit: responsible for synchronizing the output signals of the three-axis force sensor and accurately matching the data with the spatial position of the tire scan.

[0098] Data processing and defect recognition module: used to pre-process the collected three-dimensional force data, execute the defect recognition algorithm, and finally output the judgment result of the internal defects of the tire.

[0099] 7. Adaptability and scalability

[0100] The detection method and system of the present invention have strong adaptability and can be adaptively adjusted according to the specifications, patterns and surface characteristics of different tires. The system can flexibly configure parameters such as scanning path, loading force, scanning accuracy, etc. to meet the detection requirements of different types of tires. For example, car tires, truck tires or engineering machinery tires can all be detected with high precision by appropriately adjusting parameters.

[0101] To better understand the application of the present invention, a specific application example of a tire internal bubble defect detection method based on tactile perception is provided below. In this example, a passenger car tire that is known to have internal bubble defects during the production process is selected. It is difficult to find internal bubbles in this tire in traditional testing (manual tapping), but there are potential safety hazards in actual use. To verify the effectiveness of the method of the present invention, a multi-force level scanning detection system based on tactile perception is used to detect the tire, and the detection results are compared with the images collected by the X-ray detection equipment.

[0102] 1. Detection system and experimental conditions

[0103] 1) System configuration:

[0104] Tire positioning device: A fixed clamping device is used to firmly fix the tire, and a rotating table is configured to allow the tire to rotate in the horizontal direction to achieve a comprehensive scan.

[0105] Multi-axis motion mechanism: This mechanism controls the three-axis force sensor to move along the X, Y (or row and column) directions on the tire surface to ensure a uniform scanning path.

[0106] Three-axis force sensor: A high-precision sensor that can simultaneously collect force data in the X, Y and Z directions; the sensor has been pre-noise filtered and zero-point calibrated.

[0107] Force loading unit: A cylinder or precision screw mechanism is used to apply normal forces of different sizes (10N, 50N, 100N) to the tire in different scanning areas to capture multi-force level response data.

[0108] Data acquisition and processing unit: Based on the FPGA / SoC platform, it synchronously collects and processes data in real time, constructs a position-mechanical feature data set through preprocessing, interpolation, and normalization, and generates a multi-dimensional mechanical response map (such as RGB fusion map or thermal map).

[0109] Defect Identification Module: Utilizes the pre-established standard tire mechanical fingerprint library and simplified model to compare and automatically identify the scanned data, output the defect area, estimate the bubble size and defect severity.

[0110] 2) Experimental conditions:

[0111] Ambient temperature: 25°C, relative humidity: 60%

[0112] Scanning resolution: approximately one sample per 1 mm, with a total of 100 scan position data collected

[0113] Normal force loading scheme: First perform a preliminary scan with 10N, and then apply 50N and 100N for secondary scans on suspected defect areas

[0114] 2. Detection steps and data collection

[0115] Step 1: Tire Fixing and Coordinate Calibration

[0116] The tire is fixed on the positioning device and the spatial coordinates are calibrated using a laser alignment system to ensure that the tire does not shift during the scanning process.

[0117] Step 2: Initial loading and datum surface establishment

[0118] The multi-axis motion mechanism is controlled to drive the three-axis force sensor to move in the vertical direction so that the sensor contacts the outer surface of the tire for the first time. When the sensor detects that the normal force reaches 10N, the reference plane is established.

[0119] Step 3: Multi-level scanning

[0120] Preliminary scan: Apply a load of 10N, collect data from 100 scanning positions along the tire surface, and record the force values ​​in the X, Y, and Z directions at each position.

[0121] Fine scanning: For areas where abnormal force is detected in the initial scan (for example, the normal force at some locations is significantly higher than the surrounding normal data), a secondary scan is performed at 50N and 100N to collect more detailed multi-force level data. Some of the tire scanning data is shown in Table 1: Table 1 Scanning data after tire pretreatment

[0122]

[0123] For scanning position 52, the preliminary scanning data showed anomalies, and then fine scanning was performed on this position under loads of 50N and 100N to obtain more detailed data.

[0124] Step 4: Data preprocessing and construction of multidimensional mechanical response maps

[0125] All collected data are subjected to noise filtering, zero drift correction and interpolation processing to form a complete position-mechanical characteristic data set. The data matrix is ​​smoothed and normalized by Python, and the data in the X, Y, and Z directions are plotted as heat maps or RGB fusion is used to generate multidimensional mechanical response maps.

[0126] Step 5: Defect Identification

[0127] Constructing mechanical fingerprint: The processed data is used to generate multi-force level thermal maps, especially the response map in the Z direction (normal force). It is observed that at all loading levels, the force in the Z direction of scanning position 52 is significantly higher than that of the surrounding area, and the image color shows obvious abnormalities.

[0128] Comparison with standard fingerprint library: Compare the currently collected mechanical fingerprint image with the pre-established standard tire fingerprint library, and use indicators such as mean square error and correlation coefficient to determine the abnormal area. The comparison result shows that the scanning position 52 has a deviation exceeding the set threshold from the standard data.

[0129] Estimating bubble size: Based on the aforementioned empirical model between normal force variation and bubble size, the bubble size at the scanning position 52 is calculated.

[0130] 1) Data extraction According to Table 1 (pre-processed data), the data at scanning position 52 is: Applied load: 10N Normal area (scanning position 45~51) Z direction force: 45:2.2N 46:2.3N 47:2.4N 48:2.3N 49:2.2N 50:2.3N 51:2.4N; Take the average: ; Force in the defect area (scanning position 52) in the Z direction: F defect =3.7N 2) Calculate the normal force change ΔF Substituting the values: ΔF=3.7 N−2.3 N=1.4 N 3) Build a calibration model and estimate bubble size Based on multiple experimental fittings, a linear calibration model was established: D (mm) = a × ΔF (N) + b in: a=10mm / N b=0 mm Therefore, the bubble diameter at scanning position 52 is estimated to be: D=10×1.4+0=14 mm.

[0131] Step 6: Output the test results

[0132] The inspection system finally outputs the defect area and detailed information.

[0133] 1) Defect location: Scanning location 52 (and surrounding areas that may be affected);

[0134] 2) Bubble size: estimated diameter is about 14mm;

[0135] 3) Defect severity: Due to the obvious abnormal stress, it is judged as a serious defect;

[0136] At the same time, multi-dimensional mechanical response diagrams and thermal maps are generated to intuitively display abnormal areas in graphical form.

[0137] 4.X-ray inspection verification

[0138] To verify the test results based on tactile perception, an X-ray inspection device was selected to scan the same tire. The images provided by the X-ray inspection device show the internal structure of the tire and can directly detect internal bubbles or delamination areas.

[0139] Comparative experimental steps:

[0140] X-ray scanning: Perform X-ray inspection on the tire to collect images of the internal structure.

[0141] Result comparison: Compare the defect area shown in the X-ray image with the scan position 52 and its surrounding area identified by tactile detection.

[0142] If an obvious low-density area (representing bubbles or voids) appears in the X-ray image in the scanning position 52, it proves that the detection result of the method of the present invention is correct.

[0143] Experimental data and verification:

[0144] Tactile detection results: At scanning position 52, the force in the Z direction increases significantly, and the system estimates that the bubble diameter in this area is about 14 mm.

[0145] X-ray detection results: X-ray images showed that there were obvious low-density areas in the corresponding areas, and the bubble size was estimated to be 1.0~2.0cm, which was consistent with the tactile detection results.

[0146] Example 2 In order to further improve the intelligent level of detection, the defect identification in Example 1 of the present invention adopts machine learning or deep learning model. Specifically, deep neural networks can be used to automatically extract the features of multi-force-level mechanical fingerprints, and judge the size and distribution position of bubbles inside the tire based on these features. This intelligent defect judgment mechanism reduces manual intervention and improves the automation and accuracy of detection. A deep learning method is given below for automatically extracting defect features from multi-force-level mechanical fingerprints, and judging the size and distribution position of bubbles inside the tire based on these features. This method adopts a multi-task learning strategy, integrating the two tasks of defect area segmentation and bubble size regression, collectively referred to as "Multi-Task Deep Fingerprint Network (MT-DFN)".

[0147] 1) Input data construction 1.1) Data Source The tire detection system is used to collect pre-processed multi-force level scanning data, and the force data in the three directions of X, Y, and Z of each scanning point are combined into image data. To ensure data diversity, the input image can be a grayscale image (a heat map of a single force direction) or a multi-channel image (RGB channels represent the normalized results of the forces in the X, Y, and Z directions respectively).

[0148] 1.2) Data Labeling Segmentation label: Based on X-ray detection or manual annotation, a binary segmentation map of the defect area is generated for each input image. The defect area is represented by "1" and the normal area is represented by "0".

[0149] Regression label: For the defect area, annotate its true bubble size (diameter, in mm).

[0150] 2) Network structure design 2.1) Encoder Basic architecture: A structure similar to U-Net is used. The input image size is set to 256×256 pixels.

[0151] Residual convolutional blocks: In the encoder, several residual modules (ResBlock) are used, each of which consists of two layers of 3×3 convolutions, batch normalization, and ReLU activation, plus skip connections.

[0152] Downsampling: Use convolution or max pooling layers with a stride of 2 to gradually reduce the size of the feature map (256×256 → 128×128 → 64×64 → 32×32), while increasing the number of channels (from 64 to 128, 256, 512).

[0153] 2.2) Attention Module An attention gate is added after the middle layer of the encoder or each downsampling stage to highlight the local features related to the defect through an adaptive weighting mechanism. This module uses operations such as 1×1 convolution and sigmoid activation to generate an attention map, which is then multiplied pixel by pixel with the feature map to obtain a weighted feature map.

[0154] 2.3) Decoder Upsampling layer: Use deconvolution (Transposed Convolution) or bilinear interpolation to upsample the feature map and gradually restore the size.

[0155] Skip connection: concatenate the feature map of the corresponding layer of the encoder with the upsampled feature map of the decoder to retain the detail information.

[0156] Convolutional fusion: After upsampling and skip connections, the fused features are further processed through a series of convolutional layers to generate high-resolution feature maps.

[0157] 2.4) Multi-task output branch Segmentation branch: At the end of the decoder, a 1×1 convolution is used to map to two channels (defective and normal), followed by Softmax to generate the classification probability of each pixel. The output size of the segmentation branch is the same as the input image (256×256), which is used to locate the bubble defect area.

[0158] Regression branch: Perform global average pooling on the feature map at the end of the decoder to obtain a fixed-length feature vector. After several fully connected layers (FC-256→FC-64), a continuous real value is finally output to represent the bubble diameter (or other size indicators) of the detection area. The regression branch can predict each defect area separately or perform overall regression in the entire image. The prediction results are then mapped to specific areas through connected domain analysis.

[0159] 3.) Loss function and training 3.1) Multi-task loss function To jointly train the segmentation and regression tasks, a total loss function is used:

[0160] Where: L seg is the loss of the segmentation task, using Dice Loss, cross entropy loss or joint loss; L reg is the loss of the regression task, usually the mean square error (MSE) loss; λ1 and λ2 are weight coefficients, which are determined through experiments according to the importance of the task.

[0161] 3.2) Training details Optimizer: Adam optimizer is used, the initial learning rate is set to 1e-4, and then gradually decays; Data enhancement: Perform operations such as rotation, translation, scaling, and noise injection on training data to expand the data set; Batch size: Select 32 or 16, and choose the appropriate batch size based on the GPU memory; Training cycle: Usually train for 50~100 epochs until the validation set loss stabilizes.

[0162] 4) Reasoning and post-processing 4.1) Reasoning Phase The preprocessed multi-force level mechanical fingerprint image is input, and the network simultaneously outputs the segmentation map and regression results. The segmentation map is used to extract each defect area using connected domain analysis, and the regression prediction value of each area is averaged or weighted to obtain the final bubble size estimate of the area.

[0163] 4.2) Post-processing Morphological operations (such as opening and closing operations) are performed on the segmented image to further remove noise points and ensure clear boundaries of the defect area. Based on the regression results and the preset threshold judgment, the final defect report is output, including the coordinates of the defect area, bubble diameter, and defect severity score.

[0164] The multi-task deep fingerprint network (MT-DFN) proposed in this paper realizes intelligent defect recognition based on multi-force level scanning data by combining defect area segmentation and bubble size regression tasks. The method specifically includes an encoder-decoder structure, an attention module, and a dual-branch output design, and uses a multi-task loss function for joint training. In the inference stage, the network can automatically extract defect features from the preprocessed mechanical fingerprint image, and output the defect area, bubble size, and defect severity, providing a specific and innovative solution for tire internal defect detection.

[0165] The following is an experimental data to demonstrate the performance of the multi-task deep fingerprint network (MT-DFN) in tire internal bubble defect detection. The experimental data is based on an actual tire dataset, which contains 500 multi-force level mechanical fingerprint images, each image is accompanied by manually annotated defect areas (bubbles) and true size information. The dataset is randomly divided into a training set (400 images) and a validation set (100 images). After 80 epochs of training, the network achieved the following performance indicators: 1. Segmentation Task Performance Dice coefficient: 0.96 IoU (Intersection over Union): 0.92 Pixel-level accuracy: 98%; This shows that the network can accurately segment the bubble defect area inside the tire, and the segmentation result is highly consistent with the defect image obtained by X-ray detection.

[0166] 2. Regression Task Performance Mean absolute error (MAE): 0.25mm Root mean square error (RMSE): 0.35mm R² coefficient: 0.98; These indicators show that the network predicts the bubble diameter very accurately, with an average prediction error of only 0.5 mm and good regression performance.

[0167] 3. Prediction results of some test samples Table 2 shows the comparison between the actual bubble diameters of 10 test samples and the network prediction results. The data are all diameter values ​​in mm and the corresponding absolute errors: Table 2 Comparison of the actual bubble diameter of the test samples and the network prediction results Sample No. True bubble diameter (mm) Predicted bubble diameter (mm) Absolute error (mm) 1 2.5 2.6 0.1 2 3.0 3.1 0.1 3 2.8 2.7 0.1 4 3.2 3.0 0.2 5 2.9 3.0 0.1 6 3.5 3.4 0.1 7 2.7 2.8 0.1 8 3.3 3.5 0.2 9 2.6 2.5 0.1 10 3.1 3.0 0.1 It can be seen from the table that the absolute errors between the predicted bubble diameters and the true values ​​are all within 0.2 mm, showing extremely high prediction accuracy.

[0168] 4. Experimental data statistics and comparison verification In addition, by training and testing a dataset consisting of 500 multi-force-level mechanical fingerprints, the improved network achieved the following overall performance on the validation set: Segmentation task: Dice coefficient reached 0.96, IoU reached 0.92, and pixel accuracy was 98%; Regression task: MAE is 0.25mm, RMSE is 0.35mm, and R² reaches 0.98; Online detection speed: The average inference time for each image is less than 100ms, which is suitable for real-time detection needs.

[0169] Compared with the X-ray detection results, the defect area detected by the present invention has a coincidence rate of more than 97% with the low-density area shown in the X-ray image, which proves the high accuracy and reliability of this method in actual tire quality control.

[0170] Furthermore, the present invention also provides a computer-readable storage medium, including one or more programs for execution by one or more processors of an electronic device, wherein the one or more programs include instructions for executing the method described in this specific embodiment.

[0171] It should be noted that computer-readable media includes permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer-readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer-readable media does not include temporary computer-readable media such as modulated data signals and carrier waves.

[0172] The present invention may be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types. The present invention may also be practiced in distributed computing environments where tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules may be located in local and remote computer storage media, including storage devices.

[0173] The above is a description of the embodiments of the present invention. Through the above description of the disclosed embodiments, professionals and technicians in the field can implement or use the present invention. Various modifications to these embodiments will be apparent to professionals and technicians in the field. The general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to these embodiments shown in this article, but will conform to the widest range consistent with the principles and novelties disclosed herein.

Claims

1. A method for detecting air bubble defects inside a tire based on tactile perception, characterized in that: The method comprises the following steps: 1) Tire fixing: install the tire to be tested on a rotatable or clampable positioning device, and calibrate the tire's spatial coordinate system; 2) Initial loading: Control the multi-axis motion mechanism to drive the three-axis force sensor to move in the vertical direction and contact the outer surface of the tire. When the normal force of the sensor reaches the first preset threshold, a detection reference surface is established; 3) Multi-level scanning: 3.1) With the detection reference plane as a reference, a scanning path is set on the outer surface of the tire, and normal forces of different magnitudes are applied to the same scanning point in sequence, and the force data in the three directions of X, Y, and Z output by the three-axis force sensor under each level of force loading is recorded; 3.2) Repeat the above steps to scan the outer surface of the tire line by line or partition by partition to collect three-dimensional force data of multiple force levels; 4) Data preprocessing: noise filtering, zero drift correction and interpolation processing are performed on the collected three-dimensional force data, and the force changes of the same scanning point under different force loading are integrated to generate the corresponding position-mechanical characteristic data set; 5) Defect Identification: 5.1) constructing a multi-force-level mechanical fingerprint map based on the position-mechanical feature data set; 5.2) comparing the multi-force-level mechanical fingerprint with a pre-established standard mechanical fingerprint library, and when a local area is found to exhibit abnormally low stiffness or deformation characteristics, it is determined that the area has an internal bubble defect; 5.3) Output the location and degree of the bubble defect area.

2. The method according to claim 1, characterized in that: The multi-force level scanning in step 3) includes first performing a preliminary scan with a relatively small force value, and then performing a secondary fine scan of the suspicious area with a higher force value to distinguish the deformation difference caused by the tiny surface protrusions and the deep bubbles.

3. The method according to claim 1, characterized in that: The data preprocessing in step 4) also includes filtering or compensating for the periodic fluctuations generated by the tire surface pattern to reduce the interference of the tread pattern on the internal defect identification.

4. The method according to claim 1, characterized in that: The signal output by the three-axis force sensor is collected synchronously by hardware triggering of the FPGA or SoC platform, specifically including counting the pulse signal of the motion mechanism and triggering sampling when the count reaches a preset value, thereby ensuring the accurate correspondence between the spatial coordinates and the force data at different force loading stages.

5. The method according to any one of claims 1 to 4, characterized in that: The defect recognition step 5) uses a multi-task deep fingerprint network to realize intelligent defect recognition based on multi-level scanning data by combining defect area segmentation and bubble size regression tasks; specifically, it includes an encoder-decoder structure, an attention module, and a dual-branch output design, and uses a multi-task loss function for joint training; In the inference stage, the network can automatically extract defect features from the preprocessed mechanical fingerprint image and output the defect area, bubble size and defect severity.

6. A tire internal bubble defect detection system based on tactile perception, used to implement the method according to claims 1-5, characterized in that: The system comprises: A tire positioning device, used to fix the tire to be tested and provide a rotation or clamping function; A multi-axis motion mechanism, arranged opposite to the tire positioning device, having at least one axis moving in a horizontal direction and one axis moving in a vertical direction; A three-axis force sensor, installed at the end of the multi-axis motion mechanism, for detecting the force values ​​in the X, Y and Z directions when in contact with the tire surface; A force loading unit, cooperating with the three-axis force sensor, is used to apply multiple levels of normal forces of different magnitudes to the tire surface during scanning; A data acquisition control unit, connected to the three-axis force sensor, for acquiring three-axis force data based on a hard trigger method and synchronizing it with the spatial position of the multi-axis motion mechanism; The data processing and defect recognition module communicates with the data acquisition control unit, and is used to pre-process the collected multi-force-level three-dimensional force data and execute a defect recognition algorithm to output a tire internal defect judgment result.

7. The system according to claim 6, characterized in that: The force loading unit includes a screw mechanism or a cylinder mechanism and is equipped with a force closed-loop controller, which can dynamically adjust the force applied by the three-axis force sensor within a specified force value range; And / or, the three-axis force sensor has a replaceable wear-resistant probe or anti-skid pad to adapt to tire surfaces of different specifications and materials and reduce probe wear.

8. The system according to claim 6, characterized in that: The data acquisition control unit includes an FPGA or SoC platform, configured with a pulse counter and a high-speed AD conversion circuit, and triggers the sampling of the three-axis force sensor by counting the pulse signals of the stepping or servo motor of the multi-axis motion mechanism and reaching a predetermined pulse value, thereby achieving real-time synchronization of the scanning position and the force data; And / or, the data processing and defect identification module establishes a standard tire mechanical fingerprint library, and compares the multi-force level scanning data of tires of different models and specifications under normal conditions. When the detection data differs from the standard data within a preset threshold range, it is determined that there are bubbles or delamination defects inside the tire, and a visual report of the defects is output.

9. A computer-readable storage medium having a computer program or instruction stored thereon, characterized in that: When the computer program or instruction is executed by a processor, steps 4) to 5) of the method according to any one of claims 1 to 5 are implemented.

10. A computer program product comprising a computer program or instructions, characterized in that When the computer program or instruction is executed by a processor, steps 4) to 5) of the method according to any one of claims 1 to 5 are implemented.

Images