A welding quality monitoring method and system of a plastic liner welding machine

Abstract

This invention discloses a welding quality monitoring method and system for a plastic liner welding machine, belonging to the field of plastic liner welding technology. It addresses the technical problem of inconsistent welding quality in plastic liners, making it difficult to quickly adjust parameters to the optimal level for improving product welding pass rate, thus affecting production efficiency. The method includes: constructing and training an image detection model and a welding quality detection model based on X-ray images; performing basic quality inspections on each welded product of the plastic liner welding machine using the image detection model; performing X-ray inspection on the current welded product when the basic quality inspection meets preset conditions to obtain an X-ray image; inputting the X-ray image into the welding quality detection model to obtain the welding strength of the current welded product; and determining the welding nodes to be adjusted and the corresponding adjustment parameters based on the basic quality inspection results and welding strength inspection results within a preset time period.

Description

Technical Field

[0001] This invention relates to the field of plastic liner welding technology, and in particular to a welding quality monitoring method and system for a plastic liner welding machine. Background Technology

[0002] Hydrogen is a green and efficient petroleum energy source. Type IV hydrogen storage cylinders have become the main field of hydrogen energy research and application due to their light weight, fatigue resistance, and high hydrogen storage density per unit mass. However, due to the limitations of existing processing technology, thin-walled plastic liners with large aspect ratios and large capacity can hardly be processed into shape in one step. Therefore, it is necessary to perform multiple two-stage welding on the injection-molded liners of Type IV cylinders through heating and welding to finally form a complete liner.

[0003] Currently, there are relatively mature plastic liner welding machines and welding production lines. However, the welding of plastic liners is affected by factors such as welding temperature, heating time, heating method, and welding pressure, resulting in inconsistent welding quality and unsatisfactory product welding pass rates. Although the product welding pass rate can be improved by manually adjusting the parameters of each welding step, this method is inefficient and cannot quickly adjust each welding step to the most favorable level for improving the product welding pass rate. Repeated adjustments are required to obtain better adjustment parameters, during which many unqualified welded products will be produced, affecting production efficiency and leading to profit losses. Summary of the Invention

[0004] This invention provides a welding quality monitoring method and system for a plastic liner welding machine to solve the following technical problems: the welding of plastic liners is affected by a variety of factors, resulting in inconsistent welding quality. Furthermore, it requires repeated adjustments to the parameters of each welding step, making it impossible to quickly adjust the parameters to the level most beneficial to improving the product welding qualification rate, thus affecting the production efficiency of plastic liners.

[0005] The embodiments of the present invention adopt the following technical solutions:

[0006] On one hand, embodiments of the present invention provide a method for monitoring the welding quality of a plastic liner welding machine, the method comprising: constructing an image detection model and a welding quality detection model based on X-ray images and training them;

[0007] The basic quality inspection of each welded product of the plastic liner welding machine is carried out by an image detection model; wherein the basic quality inspection includes at least the detection of misalignment and the detection of weld width.

[0008] If the basic quality inspection meets the preset conditions, X-ray inspection is performed on the current welded product to obtain an X-ray image;

[0009] The X-ray image is input into the welding quality detection model to obtain the welding strength of the current welded product;

[0010] Based on the basic quality test results and welding strength test results of all welded products tested within a preset time period, the welding nodes to be adjusted and the corresponding adjustment parameters are determined.

[0011] The welding nodes to be adjusted are adjusted accordingly using the adjustment parameters to improve welding quality.

[0012] In one feasible implementation, an image detection model and a welding quality detection model based on X-ray images are constructed and trained, specifically including:

[0013] A dataset of X-ray images of the welded areas of the plastic liner was constructed, and defect annotations were performed on the images in the X-ray image dataset.

[0014] A densely connected convolutional network with a preset number of convolutional layers is constructed, and a transition layer is added between each convolutional layer to form an initial welding quality detection model; wherein, the transition layer includes a batch normalization operation and an average pooling layer.

[0015] The initial welding quality inspection model is trained using the labeled X-ray image dataset to obtain a welding quality inspection model based on X-ray images.

[0016] Construct a general image dataset of the welded areas of the plastic liner;

[0017] A basic image detection model is constructed based on the YOLO model, and the model is trained using the aforementioned ordinary image dataset.

[0018] In one feasible implementation, an X-ray image dataset of the welded areas of the plastic liner is constructed, and defect annotation is performed on the images in the X-ray image dataset, specifically including:

[0019] X-ray images of several welded plastic liner samples were acquired using an X-ray imaging device, and the images were expanded to form the X-ray image dataset.

[0020] Tensile tests were performed on the welded samples of the plastic inner liner, and the weld fracture time of each welded sample of the plastic inner liner was recorded.

[0021] Based on the ratio between the weld fracture time and the standard value, the weld fracture time is converted into a welding strength value;

[0022] Using annotation software, defects are annotated in each X-ray image in the X-ray image dataset; wherein the types of defects annotated include at least: pores, cracks, inclusions, foreign matter, and unfused areas;

[0023] The weld strength values ​​are correlated with the labeled X-ray images to form the final X-ray image dataset.

[0024] In one feasible implementation, basic quality inspections are performed on each welded product of the plastic liner welding machine using an image detection model, specifically including:

[0025] On the plastic inner liner welding production line, an image acquisition device installed at a preset position scans the welding product 360 degrees in a circumferential direction to obtain images of the welding parts of the product.

[0026] The images of the welded parts are input one by one into the image detection model for basic quality detection to obtain basic quality detection results; wherein, the basic quality detection results include at least the misalignment amount and the weld width.

[0027] In one feasible implementation, if the basic quality inspection meets preset conditions, X-ray inspection is performed on the current welded product to obtain an X-ray image, specifically including:

[0028] Calculate the first difference between the misalignment of the current welded product and the normal misalignment, and the second difference between the weld width of the current welded product and the normal weld width.

[0029] If either the first difference or the second difference exceeds its corresponding preset threshold, an X-ray detection is triggered on the current welded product, and an X-ray image of the current welded product is acquired by an X-ray imaging device installed on the plastic liner welding production line.

[0030] In one feasible implementation, based on the basic quality inspection results and welding strength inspection results of all welded products inspected within a preset time period, the welding nodes to be adjusted and the corresponding adjustment parameters are determined, specifically including:

[0031] Based on the basic quality inspection results of all welded products inspected within a preset time period, calculate the average misalignment and the average weld width.

[0032] The average welding strength is calculated based on the welding strength test results of the welded products that were triggered by X-ray detection within a preset time period.

[0033] Calculate the differences between the average value of the misalignment, the average value of the weld width, and the average value of the weld strength and their corresponding normal values;

[0034] If any difference is greater than the set welding machine adjustment threshold, the adjustment parameters for the corresponding welding node are determined.

[0035] In one feasible implementation, if any difference is greater than a set welding machine adjustment threshold, the adjustment parameters for the corresponding welding node are determined, specifically including:

[0036] If the difference between the average value of the misalignment and the normal value of the misalignment is greater than the first adjustment threshold, then based on the misalignment detection results within the preset time period, the first misalignment direction and the second misalignment direction of the welding machine correction mechanism are determined; wherein, the first misalignment direction is an inward misalignment or an outward misalignment; and the second misalignment direction is an upward misalignment or a downward misalignment.

[0037] Calculate the average horizontal misalignment in the first misalignment direction and the average vertical misalignment in the second misalignment direction;

[0038] Based on the difference between the average value of the horizontal misalignment and the normal value of the horizontal misalignment, the first reference position adjustment value of the welding machine correction mechanism is generated;

[0039] The second reference adjustment value of the welding machine correction mechanism is generated based on the difference between the average value of the vertical misalignment and the normal value of the vertical misalignment.

[0040] In one feasible implementation, if any difference is greater than a set welding machine adjustment threshold, the adjustment parameters for the corresponding welding node are determined, specifically including:

[0041] If the difference between the average weld width and the normal weld width is greater than the second adjustment threshold, then the adjustment value of the clamping force of the welding machine clamping mechanism is determined based on the ratio between the difference and the normal weld width, and the current clamping force of the welding machine clamping mechanism.

[0042] In one feasible implementation, if any difference is greater than a set welding machine adjustment threshold, the adjustment parameters for the corresponding welding node are determined, specifically including:

[0043] If the difference between the average welding strength and the normal welding strength is greater than the third adjustment threshold, then based on the ratio between the difference and the normal welding strength, and the current heating temperature of the welding machine's heating mechanism, a heating temperature adjustment value for the welding machine's heating mechanism is determined; and,

[0044] Based on the ratio between the difference and the normal value of welding strength, and the current clamping force of the welding machine clamping mechanism, the clamping force adjustment value of the welding machine clamping mechanism is determined.

[0045] On the other hand, embodiments of the present invention also provide a welding quality monitoring system for a plastic liner welding machine, the system comprising:

[0046] The model building module is used to build and train image detection models and welding quality detection models based on X-ray images;

[0047] The basic quality inspection module is used to perform basic quality inspections on each welded product of the plastic liner welding machine through an image detection model; wherein, the basic quality inspection includes at least misalignment detection and weld width detection;

[0048] The depth quality inspection module is used to perform X-ray inspection on the current welded product when the basic quality inspection meets the preset conditions, and obtain an X-ray image; the X-ray image is then input into the welding quality inspection model to obtain the welding strength of the current welded product.

[0049] The welding node adjustment module is used to determine the welding node to be adjusted and the corresponding adjustment parameters based on the basic quality inspection results and welding strength inspection results of all welding products inspected within a preset time period; and to adjust the welding node to be adjusted accordingly through the adjustment parameters to improve the welding quality.

[0050] Compared with the prior art, the welding quality monitoring method and system for a plastic inner liner welding machine provided in this embodiment of the invention has the following beneficial effects:

[0051] This invention utilizes image processing and X-ray inspection technologies to comprehensively inspect the welding quality of products welded by a plastic liner welding machine. Based on different inspection parameters, it allows for targeted, fine-tuning of various welding mechanisms within the machine. This enables rapid adjustment of each welding step to the level most beneficial for improving the product's weld pass rate, without the need for repeated adjustments. It can quickly identify welding defects based on product quality feedback and make corresponding adjustments, reducing the defect rate of welded products and improving the production efficiency of plastic liners. Attached Figure Description

[0052] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. In the drawings:

[0053] Figure 1 A flowchart illustrating a welding quality monitoring method for a plastic inner liner welding machine, provided in an embodiment of the present invention;

[0054] Figure 2 This is a schematic diagram of the welding quality monitoring system for a plastic inner liner welding machine provided in an embodiment of the present invention. Detailed Implementation

[0055] To enable those skilled in the art to better understand the technical solutions of this invention, the technical solutions of the embodiments of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this invention, and not all embodiments. Based on the embodiments of this specification, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this invention.

[0056] This invention provides a method for monitoring the welding quality of a plastic inner liner welding machine, such as... Figure 1 As shown, the welding quality monitoring method for the plastic inner liner welding machine specifically includes steps S101-S105:

[0057] S101. Construct an image detection model and a welding quality detection model based on X-ray images and train them.

[0058] Specifically, we first construct an X-ray image dataset of the welded parts of the plastic inner liner, and then annotate the images in the X-ray image dataset with defects.

[0059] As a feasible implementation method, the specific construction process of the X-ray image dataset is as follows:

[0060] First, X-ray images of several welded plastic liner samples were acquired using an X-ray imaging device, and then the images were expanded to form an X-ray image dataset. This invention performed rotation, horizontal and vertical movement, shearing transformation, magnification, and horizontal flipping operations on the directly acquired X-ray images. The transformation coefficients for all operations were randomly generated within a certain range. Horizontal flipping, however, has no transformation range and occurs with a probability of 0.5.

[0061] Tensile tests were then performed on these welded samples of plastic liners, and the weld fracture time of each sample was recorded. The ratio between the weld fracture time of each sample and the standard value was calculated, and this ratio was used as the weld strength value of the welded sample of the plastic liner.

[0062] Furthermore, using annotation software, defects are annotated in each X-ray image in the X-ray image dataset. The annotated defect types include at least: porosity, cracks, inclusions, foreign matter, and lack of fusion. Finally, the weld strength values ​​are correlated with the annotated X-ray images to form the final X-ray image dataset.

[0063] Furthermore, a densely connected convolutional network with a predetermined number of convolutional layers is constructed, and a transition layer is added between each convolutional layer to form an initial welding quality detection model; wherein, the transition layer includes batch normalization operation and average pooling layer. Then, the initial welding quality detection model is trained using an annotated X-ray image dataset to obtain a welding quality detection model based on X-ray images.

[0064] Furthermore, a general image dataset of the welded areas of the plastic liner is constructed. This general image dataset is created by capturing 360-degree circumferential scan images of the welded areas of plastic liners with different weld qualities.

[0065] Furthermore, based on the YOLO model, a basic image detection model is constructed and trained using a common image dataset.

[0066] S102. Perform basic quality inspection on each welded product of the plastic liner welding machine using an image detection model; the basic quality inspection includes at least the detection of misalignment and the detection of weld width.

[0067] Specifically, on the plastic inner liner welding production line, an image acquisition device installed at a preset position scans the welded products 360 degrees in a circumferential direction to acquire images of the welded parts. These welded part images are then input one by one into an image detection model for basic quality inspection, yielding basic quality inspection results. These results include at least the amount of misalignment and the weld width.

[0068] As a feasible implementation method, the misalignment amount refers to the offset of the edges or surfaces of the two materials being welded during the welding process. This may be due to deviations in the reference position, resulting in horizontal or vertical misalignment between the two materials. Therefore, the misalignment amount here includes both horizontal and vertical misalignment. This invention performs a 360-degree circumferential scan of the misalignment position of the welded product and monitors the maximum misalignment width across the entire misalignment area as the misalignment amount. Simultaneously, the cross-sectional area of ​​the protrusion at the weld seam is also monitored. The cross-sectional area of ​​the protrusion can be combined with the misalignment amount to determine whether the basic weld quality is acceptable.

[0069] S103. Under the condition that the basic quality inspection meets the preset conditions, perform X-ray inspection on the current welded product to obtain an X-ray image. Input the X-ray image into the welding quality inspection model to obtain the welding strength of the current welded product.

[0070] Specifically, the first difference between the misalignment of the current welded product and the normal misalignment is calculated, and the second difference between the weld width of the current welded product and the normal weld width is calculated.

[0071] If either the first difference or the second difference exceeds its corresponding preset threshold, an X-ray inspection is triggered on the current welded product, and an X-ray image of the current welded product is acquired by an X-ray imaging device installed on the plastic inner liner welding production line.

[0072] The X-ray image is then input into the welding quality inspection model to obtain the welding strength of the current welded product.

[0073] S104. Based on the basic quality inspection results and welding strength inspection results of all welded products inspected within a preset time period, determine the welding nodes to be adjusted and the corresponding adjustment parameters.

[0074] Specifically, based on the basic quality inspection results of all welded products inspected within a preset time period, the average value of misalignment and the average value of weld width are calculated.

[0075] Furthermore, based on the welding strength test results of the welded products that were triggered by X-ray detection within a preset time period, the average welding strength is calculated.

[0076] Then, calculate the differences between the average misalignment, average weld width, and average weld strength and their corresponding normal values. If any difference exceeds the set welding machine adjustment threshold, determine the adjustment parameters for the corresponding welding node.

[0077] As a specific implementation method, if the difference between the average value of the misalignment and the normal value of the misalignment is greater than the first adjustment threshold, the first misalignment direction and the second misalignment direction of the welding machine correction mechanism are determined based on the misalignment detection results within a preset time period; wherein, the first misalignment direction is inward misalignment or outward misalignment; and the second misalignment direction is upward misalignment or downward misalignment.

[0078] Then, the average horizontal misalignment in the first misalignment direction and the average vertical misalignment in the second misalignment direction are calculated. Based on the difference between the average horizontal misalignment and the normal value of horizontal misalignment, the first reference position adjustment value of the welding machine correction mechanism is generated. Based on the difference between the average vertical misalignment and the normal value of vertical misalignment, the second reference position adjustment value of the welding machine correction mechanism is generated.

[0079] In one embodiment, if the average misalignment exceeds the normal value significantly, it indicates that the alignment of the welding machine's correction mechanism is not precise enough and requires adjustment. Since the inner liner material being welded has both inner and outer surfaces, the misalignment detection results are first used to determine whether material A is shifted inwards or outwards towards material B. If A is shifted inwards towards B, the first misalignment direction is inward; if A is shifted outwards towards B, the first misalignment direction is outward. If the upper and lower edges of A and B are not aligned, if the upper edge of A is higher than B, the second misalignment direction is upward; if the upper edge of A is lower than B, the second misalignment direction is downward. After determining the misalignment direction, the difference between the average misalignment value of welded products exhibiting the same misalignment direction and the normal value is determined as the correction adjustment value for that direction.

[0080] As a second specific implementation, if the difference between the average weld width and the normal weld width is greater than the second adjustment threshold, the adjustment value of the clamping force of the welding machine clamping mechanism is determined based on the ratio between the difference and the normal weld width, and the current clamping force of the welding machine clamping mechanism.

[0081] In one embodiment, if the difference between the average weld width and the normal value is high, the current clamping force of the welding machine clamping mechanism is multiplied by the ratio of the difference to the normal value to obtain the adjustment value of the clamping force that needs to be increased.

[0082] As a third specific implementation, if the difference between the average welding strength and the normal welding strength is greater than the third adjustment threshold, then the heating temperature adjustment value of the welding machine heating mechanism is determined based on the ratio between the difference and the normal welding strength, and the current heating temperature of the welding machine heating mechanism; and the clamping force adjustment value of the welding machine clamping mechanism is determined based on the ratio between the difference and the normal welding strength, and the current clamping force of the welding machine clamping mechanism.

[0083] In one embodiment, the welding strength is related to both the heating temperature and the clamping force during welding. Therefore, by multiplying the actual parameters of the current mechanism by the ratio of the welding strength difference to the normal value, the heating temperature adjustment value of the heating mechanism and the clamping force adjustment value of the clamping mechanism are obtained respectively.

[0084] S105. Adjust the parameters to make corresponding adjustments to the welding nodes to be adjusted in order to improve the welding quality.

[0085] Specifically, after obtaining the adjustment parameters in S104, the parameters of the welding nodes that need adjustment are adjusted. For example, if the average weld width of all welded products inspected within 5 hours exceeds the normal value, the difference between the two is calculated, and the clamping force adjustment value is calculated. Then, based on the current clamping force of the welding machine's clamping mechanism, this clamping force adjustment value is added to obtain the adjusted clamping force, which is then applied to the clamping mechanism to improve the weld width of subsequent products.

[0086] In addition, embodiments of the present invention also provide a welding quality monitoring system for a plastic inner liner welding machine, such as... Figure 2 As shown, the welding quality monitoring system 200 of the plastic liner welding machine specifically includes:

[0087] The model building module is used to build and train image detection models and welding quality detection models based on X-ray images;

[0088] The basic quality inspection module is used to perform basic quality inspections on each welded product of the plastic liner welding machine through an image detection model; wherein, the basic quality inspection includes at least misalignment detection and weld width detection;

[0089] The depth quality inspection module is used to perform X-ray inspection on the current welded product when the basic quality inspection meets the preset conditions, and obtain an X-ray image; the X-ray image is then input into the welding quality inspection model to obtain the welding strength of the current welded product.

[0090] The welding node adjustment module is used to determine the welding node to be adjusted and the corresponding adjustment parameters based on the basic quality inspection results and welding strength inspection results of all welding products inspected within a preset time period; and to adjust the welding node to be adjusted accordingly through the adjustment parameters to improve the welding quality.

[0091] The various embodiments in this invention are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the embodiments of apparatus, devices, and non-volatile computer storage media are basically similar to the method embodiments, so the descriptions are relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0092] The foregoing has described specific embodiments of the present invention. Furthermore, the processes depicted in the drawings do not necessarily require a specific or sequential order to achieve the desired results. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0093] The above description is merely an embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, various modifications and variations can be made to the embodiments of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principle of the embodiments of the present invention should be included within the protection scope of the present invention.

Claims

1. A method of monitoring the quality of welding of a plastic liner welding machine, characterized in that, The method includes: Constructing and training an image detection model and a welding quality detection model based on X-ray images, specifically including: A dataset of X-ray images of welded areas of plastic linerboards is constructed, and defects are annotated in the images of the X-ray image dataset. Specifically, this includes: acquiring X-ray images of several welded samples of plastic linerboards using an X-ray imaging device, and expanding the images to form the X-ray image dataset; conducting tensile tests on the welded samples of the plastic linerboards, and recording the weld fracture time of each sample; converting the weld fracture time into a weld strength value based on the ratio between the weld fracture time and a standard value; annotating defects in each X-ray image of the X-ray image dataset using annotation software; wherein the annotated defect types include at least: porosity, cracks, slag inclusions, foreign matter, and incomplete fusion; and correlating the weld strength value with the annotated X-ray images to form the final X-ray image dataset. A densely connected convolutional network with a preset number of convolutional layers is constructed, and a transition layer is added between each convolutional layer to form an initial welding quality detection model; wherein, the transition layer includes a batch normalization operation and an average pooling layer. The initial welding quality inspection model is trained using the labeled X-ray image dataset to obtain a welding quality inspection model based on X-ray images. Construct a general image dataset of the welded areas of the plastic liner; Based on the YOLO model, a basic image detection model is constructed, and the model is trained using the aforementioned ordinary image dataset. The basic quality inspection of each welded product of the plastic liner welding machine is carried out by an image detection model; wherein the basic quality inspection includes at least the detection of misalignment and the detection of weld width. If the basic quality inspection meets the preset conditions, X-ray inspection is performed on the current welded product to obtain an X-ray image, specifically including: Calculate the first difference between the misalignment of the current welded product and the normal misalignment, and the second difference between the weld width of the current welded product and the normal weld width. If either the first difference or the second difference exceeds its corresponding preset threshold, X-ray detection is triggered on the current welded product, and an X-ray image of the current welded product is acquired by an X-ray imaging device installed on the plastic liner welding production line. The X-ray image is input into the welding quality detection model to obtain the welding strength of the current welded product; Based on the basic quality test results and welding strength test results of all welded products tested within a preset time period, the welding nodes to be adjusted and the corresponding adjustment parameters are determined. The welding nodes to be adjusted are adjusted accordingly using the adjustment parameters to improve welding quality.

2. The welding quality monitoring method for a plastic inner liner welding machine according to claim 1, characterized in that, The basic quality inspection of each welded product from the plastic liner welding machine is performed using an image detection model, specifically including: On the plastic inner liner welding production line, an image acquisition device installed at a preset position scans the welding product 360 degrees in a circumferential direction to obtain images of the welding parts of the product. The images of the welded parts are input one by one into the image detection model for basic quality detection to obtain basic quality detection results; wherein, the basic quality detection results include at least the misalignment amount and the weld width.

3. The welding quality monitoring method for a plastic inner liner welding machine according to claim 1, characterized in that, Based on the basic quality inspection results and welding strength inspection results of all welded products inspected within a preset time period, the welding nodes to be adjusted and the corresponding adjustment parameters are determined, specifically including: Based on the basic quality inspection results of all welded products inspected within a preset time period, calculate the average misalignment and the average weld width. The average welding strength is calculated based on the welding strength test results of the welded products that were triggered by X-ray detection within a preset time period. Calculate the differences between the average value of the misalignment, the average value of the weld width, and the average value of the weld strength and their corresponding normal values; If any difference is greater than the set welding machine adjustment threshold, the adjustment parameters for the corresponding welding node are determined.

4. The welding quality monitoring method for a plastic inner liner welding machine according to claim 3, characterized in that, If any difference exceeds the set welding machine adjustment threshold, the adjustment parameters for the corresponding welding node are determined, specifically including: If the difference between the average value of the misalignment and the normal value of the misalignment is greater than the first adjustment threshold, then based on the misalignment detection results within the preset time period, the first misalignment direction and the second misalignment direction of the welding machine correction mechanism are determined; wherein, the first misalignment direction is an inward misalignment or an outward misalignment; and the second misalignment direction is an upward misalignment or a downward misalignment. Calculate the average horizontal misalignment in the first misalignment direction and the average vertical misalignment in the second misalignment direction; Based on the difference between the average value of the horizontal misalignment and the normal value of the horizontal misalignment, the first reference position adjustment value of the welding machine correction mechanism is generated; The second reference adjustment value of the welding machine correction mechanism is generated based on the difference between the average value of the vertical misalignment and the normal value of the vertical misalignment.

5. The welding quality monitoring method for a plastic inner liner welding machine according to claim 3, characterized in that, If any difference exceeds the set welding machine adjustment threshold, the adjustment parameters for the corresponding welding node are determined, specifically including: If the difference between the average weld width and the normal weld width is greater than the second adjustment threshold, then the adjustment value of the clamping force of the welding machine clamping mechanism is determined based on the ratio between the difference and the normal weld width, and the current clamping force of the welding machine clamping mechanism.

6. The welding quality monitoring method for a plastic inner liner welding machine according to claim 3, characterized in that, If any difference exceeds the set welding machine adjustment threshold, the adjustment parameters for the corresponding welding node are determined, specifically including: If the difference between the average welding strength and the normal welding strength is greater than the third adjustment threshold, then based on the ratio between the difference and the normal welding strength, and the current heating temperature of the welding machine's heating mechanism, a heating temperature adjustment value for the welding machine's heating mechanism is determined; and, Based on the ratio between the difference and the normal value of welding strength, and the current clamping force of the welding machine clamping mechanism, the clamping force adjustment value of the welding machine clamping mechanism is determined.

7. A welding quality monitoring system for a plastic liner welding machine, employing the welding quality monitoring method for a plastic liner welding machine as described in any one of claims 1-6, characterized in that, The system includes: The model building module is used to build and train image detection models and welding quality detection models based on X-ray images; The basic quality inspection module is used to perform basic quality inspections on each welded product of the plastic liner welding machine through an image detection model; wherein, the basic quality inspection includes at least misalignment detection and weld width detection; The depth quality inspection module is used to perform X-ray inspection on the current welded product when the basic quality inspection meets the preset conditions, and obtain an X-ray image; the X-ray image is then input into the welding quality inspection model to obtain the welding strength of the current welded product. The welding node adjustment module is used to determine the welding node to be adjusted and the corresponding adjustment parameters based on the basic quality inspection results and welding strength inspection results of all welding products inspected within a preset time period; and to adjust the welding node to be adjusted accordingly through the adjustment parameters to improve the welding quality.

Images