An automatic welding method for automotive sound-absorbing cotton

Abstract

The present application relates to the field of image processing technology and discloses an automatic welding method for automotive sound-absorbing cotton. The method includes: preprocessing the sound-absorbing cotton raw material through a precision cutting system and high-temperature treatment; performing positioning through a robotic arm and visual recognition to obtain coordinate data; planning a welding path based on the coordinate data; executing welding using an ultrasonic device; evaluating the quality through three-dimensional scanning and acoustic detection; and implementing precise repair for defects to obtain a finished component. The present application realizes the full automation and high-quality control of the sound-absorbing cotton welding process by integrating precise positioning, intelligent path planning, dual-frequency ultrasonic welding, multi-dimensional quality detection, and precise defect repair, improving the first-pass yield of the product and the stability of the acoustic performance.

Description

Technical Field

[0001] This application relates to the field of image processing technology, and particularly to an automatic welding method for automotive sound-absorbing cotton. Background Art

[0002] At present, the automotive industry has increasingly strict requirements for in-vehicle noise control. As a key material for in-vehicle sound insulation and noise reduction, the installation quality of sound-absorbing cotton directly affects the acoustic environment inside the vehicle. Traditional methods for installing automotive sound-absorbing cotton mainly include manual pasting, mechanical fixing, etc. These methods have problems such as low positioning accuracy, uneven connection strength, and low construction efficiency. With the development of automation technology, some manufacturers have begun to use hot melt welding technology to fix sound-absorbing cotton. However, existing welding equipment is mostly designed with a single frequency, making it difficult to adapt to sound-absorbing materials with different thicknesses and densities. Moreover, the detection of welding quality still mainly relies on manual sampling inspection, making it difficult to achieve comprehensive and accurate quality assessment and timely repair.

[0003] The deficiencies in the prior art are mainly manifested in the following aspects: First, it is difficult to accurately adjust welding parameters according to material characteristics, resulting in unstable welding strength; second, there is a lack of real-time monitoring and closed-loop control of the welding process, and the welding quality fluctuates greatly; third, the means for evaluating the quality of the welded product are single, making it difficult to comprehensively detect the geometric shape and acoustic performance of the product; fourth, there is a lack of effective local repair methods after defects are found, often resulting in the scrapping of the whole piece, causing material waste and increased production costs. These problems seriously restrict the improvement of the welding quality and production efficiency of automotive sound-absorbing cotton. Summary of the Invention

[0004] This application provides an automatic welding method for automotive sound-absorbing cotton, which is used to achieve full automation and high-quality control of the sound-absorbing cotton welding process by integrating technologies such as precise positioning, intelligent path planning, dual-frequency ultrasonic welding, multi-dimensional quality detection, and precise defect repair, and improve the first-pass rate of products and the stability of acoustic performance.

[0005] In a first aspect, the present application provides an automatic welding method for automotive sound-absorbing cotton. The automatic welding method for automotive sound-absorbing cotton includes: preprocessing a polyester fiber sound-absorbing cotton raw material through a precision cutting system and a high-frequency hot air circulation furnace to obtain a preprocessed sound-absorbing cotton material; performing modular positioning on the preprocessed sound-absorbing cotton material through a six-axis robotic arm in cooperation with visual recognition to obtain a three-dimensional space coordinate data packet; planning and optimizing a welding path based on the three-dimensional space coordinate data packet to obtain a digital welding instruction packet; performing a welding operation based on the digital welding instruction packet through a dual-frequency ultrasonic generator and a dual-closed-loop feedback control system to obtain a sound-absorbing cotton component after welding; performing quality inspection on the sound-absorbing cotton component after welding through a three-dimensional laser scanning and acoustic feature analysis system to obtain a quality assessment report; and performing defect repair on the basis of the quality assessment report through a spot strengthening welding technique and a nano-level polymer repair agent to obtain a finished automotive sound-absorbing cotton assembly.

[0006] In the technical solution provided by the present application, preprocessing through a precision cutting system and a high-frequency hot air circulation furnace not only ensures the accurate size and shape of the sound-absorbing cotton raw material, but also enables the material to reach the optimal welding state through high-frequency hot air treatment, significantly improving the stability of the subsequent welding quality. The modular positioning technology combining a six-axis robotic arm and visual recognition realizes high-precision spatial positioning of the flexible sound-absorbing cotton material, effectively overcoming the problem that traditional positioning methods are difficult to cope with material deformation and providing an accurate three-dimensional space coordinate data packet for welding path planning. In the welding path planning stage, this solution uses an intelligent optimization algorithm to perform multi-dimensional optimization on the path, considering factors such as material properties, welding strength, and energy consumption efficiency, generating an optimal digital welding instruction packet and greatly improving the welding efficiency and quality. The application of a dual-frequency ultrasonic generator enables the system to automatically select the most suitable ultrasonic frequency according to the material thickness, and the combination of a dual-closed-loop feedback control system realizes real-time monitoring and parameter adjustment of the welding process, effectively preventing over-welding and under-welding phenomena and ensuring the consistency of welding quality. The introduction of a three-dimensional laser scanning and acoustic feature analysis system realizes comprehensive quality inspection of the welded product, which can not only evaluate the geometric accuracy of the product but also test the sound-absorbing performance. The application of a spot strengthening welding technique and a nano-level polymer repair agent realizes precise repair of the detected defects, avoiding the scrapping of the whole piece, greatly reducing the production cost and improving the material utilization rate. This solution has made an innovative contribution especially in the application of intelligent algorithms, including an accurate modeling and correction of material deformation by a flexible material deformation compensation algorithm, an intelligent balance of multi-dimensional parameters by a welding path optimization algorithm, a dynamic response to the real-time welding state by an adaptive welding control algorithm, and an accurate evaluation of the sound-absorbing performance by an acoustic characteristic analysis algorithm in quality assessment, realizing the intelligent and high-quality control of the sound-absorbing cotton welding process. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] To more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the accompanying drawings required for description in the embodiments. Obviously, the accompanying drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other accompanying drawings can be obtained based on these drawings.

[0008] Figure 1 It is a schematic diagram of an embodiment of the automatic welding method of automotive sound-absorbing cotton in the embodiments of the present application;

[0009] Figure 2 It is a schematic process diagram of preprocessing polyester fiber sound-absorbing cotton raw materials by a precision cutting system and a high-frequency hot air circulation furnace in the embodiments of the present application. Specific embodiments

[0010] The embodiments of the present application provide an automatic welding method for automotive sound-absorbing cotton. The terms "first", "second", "third", "fourth", etc. (if any) in the specification, claims and above-mentioned accompanying drawings of the present application are used to distinguish similar objects, and do not necessarily need to describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances so that the embodiments described here can be implemented in an order other than that illustrated or described here. In addition, the terms "comprising" or "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device comprising a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0011] For ease of understanding, the following describes the specific process of the embodiments of the present application. Please refer to Figure 1 An embodiment of the automatic welding method of automotive sound-absorbing cotton in the embodiments of the present application includes:

[0012] Step S101: Preprocess the polyester fiber sound-absorbing cotton raw materials through a precision cutting system and a high-frequency hot air circulation furnace to obtain preprocessed sound-absorbing cotton materials;

[0013] Step S102: Modularly position according to the preprocessed sound-absorbing cotton materials through a six-axis robotic arm in cooperation with visual recognition to obtain a three-dimensional space coordinate data packet;

[0014] Step S103: Plan and optimize the welding path according to the three-dimensional space coordinate data packet to obtain a digital welding instruction packet;

[0015] Step S104: Execute the welding operation according to the digital welding instruction packet through a dual-frequency ultrasonic generator and a double-closed-loop feedback control system to obtain a sound-absorbing cotton component with welding completed;

[0016] Step S105: Use a three-dimensional laser scanning and acoustic feature analysis system to conduct quality inspection on the completed sound-absorbing cotton component after welding, and obtain a quality assessment report.

[0017] Step S106: According to the quality assessment report, use spot strengthening welding technology and nano-level polymer repair agent to repair defects, and obtain a finished automotive sound-absorbing cotton component.

[0018] It can be understood that the execution subject of this application can be an automatic welding system for automotive sound-absorbing cotton, or a terminal or a server. Specifically, it is not limited here. In this embodiment of the application, the server is taken as the execution subject for illustration.

[0019] Specifically, the polyester fiber sound-absorbing cotton raw material is pretreated by a precision cutting system and a high-frequency hot air circulation furnace. During the pretreatment process, the computer-aided design system accurately calculates the geometric dimensions of the sound-absorbing cotton raw material, and generates sound-absorbing cotton geometric parameters suitable for the vehicle body structure. The precision cutting system cuts the raw material into a specific shape according to these parameters, and the high-frequency hot air circulation furnace heats the cut sound-absorbing cotton to 120 ± 5 °C and keeps it for 8 - 12 minutes to make the fiber material reach the molten state. The vacuum suction device then removes the moisture impurities in the material, reduces the moisture content to below 0.5%, and evenly coats the surface of the sound-absorbing cotton with a modified polyvinyl alcohol and nano-level titanium dioxide composite reinforcing agent through a coating device to form a pretreated sound-absorbing cotton material. The six-axis robotic arm cooperates with the vision recognition system for modular positioning. The high-definition CCD camera captures the surface feature points of the pretreated sound-absorbing cotton material, generates material surface feature information, and the feature matching algorithm matches this information with the preset three-dimensional model to obtain the initial position data. The flexible material deformation compensation algorithm corrects the deformation error for the initial position data, calculates the corrected position information, and the accuracy can reach ±0.2 mm. The distributed micro pneumatic suction cup array applies a negative pressure of 15 - 20 kPa to fix the material to ensure stable positioning of the material. At the same time, the thermal imaging sensor monitors the surface temperature distribution of the material and generates temperature parameter feedback data. Integrate the corrected position information and the temperature parameter feedback data to generate a three-dimensional space coordinate data packet.

[0020] Based on the three-dimensional space coordinate data packet, welding path planning and optimization are carried out. The three-dimensional space coordinate data packet is combined with the structure parameters of the vehicle model sound-absorbing cotton to calculate the preliminary welding path. The system uses historical welding quality data to perform multi-dimensional optimization on the preliminary welding path, considering factors such as welding strength and energy consumption efficiency, and generates the optimized welding path. Subsequently, the system decomposes the overall welding task into multiple sub-regions to form welding sub-region data. Each sub-region is further divided into main welding points and auxiliary welding points to form a welding point distribution map. The system differentially calibrates the energy input of each welding point according to the material thickness parameter, increasing the energy input by 20% for areas with a thickness greater than 8 mm and reducing the energy input by 15% for areas less than 3 mm, forming a welding energy distribution curve. The system sets intermittent cooling points to mark the heat control of the welding energy distribution curve, setting a cooling point every 25 cm of welding and cooling for 3 seconds to form a complete digital welding instruction packet.

[0021] According to the digital welding instruction packet, the dual-frequency ultrasonic generator and the double closed-loop feedback control system perform the welding operation. Select the 20 kHz or 40 kHz ultrasonic frequency according to the material thickness parameter in the welding instruction packet to generate the welding frequency configuration parameter. The titanium alloy transducer converts and conducts the ultrasonic energy to the welding point. The real-time impedance matching technology collects the acoustic impedance data of the welding point once every millisecond to form a welding impedance data stream. Based on this, the system dynamically adjusts the ultrasonic power output (150 - 450 W) and pressure (0.3 - 0.8 MPa) to obtain the adaptive welding control value. The temperature monitor and the deformation sensor perform double closed-loop detection on the welding process. When the temperature exceeds 160 °C or the deformation exceeds the standard, the system automatically reduces the power output or increases the cooling time. After welding, the rapid cooling device conducts directional cooling treatment on the welding area to form a welded sound-absorbing cotton component with a high-strength structure.

[0022] The three-dimensional laser scanning and acoustic feature analysis system performs quality inspection on the welded sound-absorbing cotton component. The multi-angle positioning device rotates the component 360 degrees. The laser emitter projects light onto the surface of the component, and the optical receiver collects the data of the surface reflected light beam to form the original reflected light signal. The triangulation method calculates the surface point depth based on the original reflected light signal to form a surface coordinate data set. The sparse point filtering algorithm eliminates outliers to obtain the optimized coordinate set and reconstructs it into a three-dimensional point cloud model of the component. The system compares this model with the theoretical model to generate deformation deviation data. The acoustic feature analyzer emits sound waves with preset frequencies ranging from 500 Hz to 8000 Hz to the component, receives the reflected wave data, and calculates the sound absorption coefficient (in the range of 0.7 - 0.95) through spectrum analysis to generate the sound absorption function parameter. The micro tensile sensor array applies a test force of 25 - 50 N to the key welding points to obtain the welding strength data. The system comprehensively analyzes the deformation deviation data, the sound absorption function parameter, and the welding strength data to generate a quality assessment report.

[0023] Defect repair is carried out according to the quality assessment report. The parser classifies and analyzes the defect information in the report to generate a defect type location database. For the area with insufficient welding strength, a high-energy ultrasonic focusing head with a small diameter (2 mm) is used for local reinforcement, increasing the welding strength by 20 - 30% to form a strength reinforcement area. For the area with microcracks, a precision spraying device sprays a nano-level polymer repair agent, which can be cured within 60 seconds to form a crack repair area. For the area with slight deformation, the shape memory thermal adjustment technology uses hot air at 80 - 100 °C in combination with vacuum negative pressure to restore the material to its ideal shape, forming a shape restoration area. After the system trims the edges and removes dust from the surface of the material, an acoustic performance enhancement coating (a composite material of nano-porous silica and elastomer, with a thickness of 0.1 - 0.2 mm) is coated to improve the sound absorption performance of the product by 5 - 8%, forming a finished automotive sound-absorbing cotton component.

[0024] Taking the sound-absorbing cotton of the inner door panel of a certain model SUV vehicle as an example, the original polyester fiber sound-absorbing cotton is cut into an irregular shape of 85 cm × 60 cm. After being treated in a high-frequency hot air circulation furnace, the moisture content drops from the original 2.8% to 0.42%. The six-axis robotic arm locates with a positioning accuracy of ±0.18 mm by identifying 8 preset feature points on the material. The welding path planning divides the overall task into three sub-areas: the front door, the rear door, and the pillar, with a total of 57 welding points set. Among them, the differential adjustment of the energy input in the area with thickness change increases the welding strength by 22.4%. During the welding process, the real-time impedance matching technology dynamically adjusts the ultrasonic power, with an average fluctuation range of ±35 W, ensuring the welding stability. Quality inspection finds 3 microcracks and 1 point with insufficient welding strength in the rear door area. After nano-level polymer repair, the sound absorption coefficient of this area is restored to 0.82, meeting the consistency requirement of the overall sound absorption performance. The average sound absorption coefficient of the manufactured finished automotive sound-absorbing cotton component in the frequency band of 100 - 3000 Hz reaches 0.87, meeting the NVH performance standard of this vehicle model.

[0025] In the embodiments of the present application, pretreatment is carried out through a precision cutting system and a high-frequency hot air circulation furnace, which not only ensures the accurate size and shape of the sound-absorbing cotton raw material, but also makes the material reach the best welding state through high-frequency hot air treatment, significantly improving the stability of the subsequent welding quality. Combining the modular positioning technology of a six-axis robotic arm and visual recognition, high-precision spatial positioning of the flexible sound-absorbing cotton material is achieved, effectively overcoming the problem that traditional positioning methods are difficult to cope with material deformation, and providing an accurate three-dimensional space coordinate data packet for welding path planning. In the welding path planning stage, this solution uses an intelligent optimization algorithm to optimize the path in multiple dimensions, considering factors such as material properties, welding strength, and energy consumption efficiency, to generate an optimal digital welding instruction packet, greatly improving the welding efficiency and quality. The application of a dual-frequency ultrasonic generator enables the system to automatically select the most suitable ultrasonic frequency according to the material thickness, and combined with a dual-closed-loop feedback control system, real-time monitoring and parameter adjustment of the welding process are realized, effectively preventing over-welding and under-welding phenomena, and ensuring the consistency of welding quality. The introduction of a three-dimensional laser scanning and acoustic feature analysis system realizes the comprehensive quality inspection of the welded products, which can not only evaluate the geometric accuracy of the products, but also test the sound-absorbing performance. The application of spot strengthening welding technology and nano-level polymer repair agents realizes the precise repair of defects found in the inspection, avoiding the scrapping of the whole piece, greatly reducing the production cost and increasing the material utilization rate. This solution has made an innovative contribution especially in the application of intelligent algorithms, including the precise modeling and correction of material deformation by the flexible material deformation compensation algorithm, the intelligent balance of multi-dimensional parameters by the welding path optimization algorithm, the dynamic response to the real-time welding state by the adaptive welding control algorithm, and the precise evaluation of the sound-absorbing performance by the acoustic feature analysis algorithm in quality assessment, realizing the intelligent and high-quality control of the sound-absorbing cotton welding process.

[0026] In a specific embodiment, the process of executing step S101 may specifically include the following steps:

[0027] (1) Accurately calculate the size of the polyester fiber sound-absorbing cotton raw material through a computer-aided design system to obtain the geometric parameters of the sound-absorbing cotton;

[0028] (2) Cut the shape of the polyester fiber sound-absorbing cotton raw material according to the geometric parameters of the sound-absorbing cotton through a precision cutting device to obtain a cut sound-absorbing cotton that matches the vehicle body structure;

[0029] (3) Treat the cut sound-absorbing cotton at a preset temperature for a predetermined time through a high-frequency hot air circulation furnace to obtain a molten state sound-absorbing cotton;

[0030] (4) Remove moisture and impurities from the molten state sound-absorbing cotton through a vacuum suction device to obtain a dry sound-absorbing cotton with a moisture content lower than 0.5%;

[0031] (5) The surface of the dried sound-absorbing cotton is evenly coated with a composite enhancer of modified polyvinyl alcohol and nano-sized titanium dioxide through a coating device to obtain a pre-treated sound-absorbing cotton material.

[0032] Specifically, Figure 2 This is a schematic process diagram for pre-treating polyester fiber sound-absorbing cotton raw materials through a precision cutting system and a high-frequency hot air circulation furnace in the embodiments of this application. The geometric parameters of the sound-absorbing cotton are obtained by precisely calculating the size of the polyester fiber sound-absorbing cotton raw materials through a computer-aided design system. The computer-aided design system performs precise matching calculations based on the interior three-dimensional model in the vehicle type structure feature database, converting the body interior space data into the precise size parameters required for the sound-absorbing cotton. The calculation system obtains the body interior CAD model, extracts the coordinates of the key installation points, and then considers the elastic deformation characteristics of the sound-absorbing material to perform compensation calculations, generating the geometric contour line and thickness distribution map of the sound-absorbing cotton. The system will specially mark the edge area, fixed point position, and functional area of the sound-absorbing cotton, and the marking information is included in the geometric parameters of the sound-absorbing cotton to ensure accurate identification of the key areas during the subsequent processing. According to the calculated geometric parameters of the sound-absorbing cotton, the precision cutting equipment cuts the shape of the polyester fiber sound-absorbing cotton raw materials to obtain a cut sound-absorbing cotton that matches the body structure. The precision cutting equipment reads the contour line data in the geometric parameters, converts it into a cutting trajectory instruction, and drives the cutting head to move along the specified path. During the cutting process, the real-time vision monitoring system tracks the cutting state to ensure that the cutting accuracy reaches within ±0.5 mm. The system will also make dynamic adjustments according to the tension state of the raw materials to avoid cutting deformation. For areas with different thicknesses and densities, the cutting equipment will adjust the cutting speed and pressure to ensure that the edges are smooth and flat. After the cutting is completed, the equipment checks the size of the cutting result and records the deviation data between the actual cutting size and the designed size, providing a reference for the subsequent processes.

[0033] The cut sound-absorbing cotton is subjected to a preset temperature treatment for a predetermined time in a high-frequency hot air circulation furnace to obtain molten state sound-absorbing cotton. The high-frequency hot air circulation furnace sets the treatment temperature at 120 ± 5 °C and the treatment time at 8 - 12 minutes according to the material characteristic parameter table. The specific time is automatically adjusted according to the material thickness and density. Multiple temperature sensors are installed inside the hot air circulation furnace to form a temperature field distribution network, which monitors the heating condition of the material in real time to ensure uniform overall temperature. The circulating air system blows hot air on the material from multiple angles according to the preset air flow pattern to avoid local overheating or underheating. During the treatment process, the temperature control algorithm continuously collects temperature data and adjusts the heating power through the PID control method to control the temperature fluctuation within the set range. During the heating process, the polyester fiber on the material surface begins to soften and melt, forming a surface state suitable for subsequent welding, while the internal structure of the material maintains its original sound-absorbing characteristics. After the temperature treatment, the molten state sound-absorbing cotton is then subjected to moisture and impurity removal by a vacuum suction device to obtain dry sound-absorbing cotton with a moisture content lower than 0.5%. The vacuum suction device adopts a sub-region adsorption method. According to the thickness and density distribution map of the material, different intensities of negative pressure are applied to different regions to ensure uniform suction effect. During the suction process, the moisture content detection sensor monitors the humidity change in the discharged gas in real time, and calculates the moisture content in the material through the humidity change rate. When it is detected that the moisture content drops to a stable value and is lower than 0.5%, the system automatically ends the suction process. At the same time, the suction device is equipped with a particulate filtration system to capture and remove the tiny impurity particles in the material, improving the material purity. The negative pressure intensity of the suction device is dynamically adjusted as the moisture content decreases to avoid excessive compression deformation of the material structure.

[0034] The surface of the dry sound-absorbing cotton is evenly coated with a composite enhancer of modified polyvinyl alcohol and nano-level titanium dioxide by a coating device to obtain a pretreated sound-absorbing cotton material. The coating device mixes modified polyvinyl alcohol and nano-level titanium dioxide in a ratio of 1:2 to form a composite enhancer. The coating process uses precision spraying technology. Through multiple independently controllable nozzle arrays, different amounts of enhancer are applied to different regions according to the surface profile and thickness distribution map of the material. The spray control system adjusts the spraying pressure and flow rate of the nozzles in real time to ensure that the enhancer forms a uniform coating on the material surface, with the thickness controlled within the range of 0.1 - 0.2 mm. After the coating is completed, the infrared scanning system detects the coating uniformity, identifies the uneven coating regions, and the system automatically marks these regions and performs supplementary coating. After the coated sound-absorbing cotton material undergoes short-time low-temperature curing, a film with good adhesiveness and durability is formed on the surface, thus completing the preparation of the pretreated sound-absorbing cotton material.

[0035] In a specific embodiment, the process of executing step S102 may specifically include the following steps:

[0036] (1) Capture the feature points of the pre-treated sound-absorbing cotton material through a high-definition CCD camera to obtain the surface feature information of the material;

[0037] (2) Perform three-dimensional model matching on the pre-treated sound-absorbing cotton material based on the material surface feature information through a feature matching algorithm to obtain initial position data;

[0038] (3) Correct the deformation error of the initial position data through the flexible material deformation compensation algorithm to obtain the corrected position information;

[0039] (4) applying negative pressure to the pre-treated sound-absorbing cotton material through a distributed micro-pneumatic suction cup array to obtain a material in a stable fixed state;

[0040] (5) Monitor the surface temperature distribution of the material in a stable fixed state through a thermal imaging sensor to obtain temperature parameter feedback data;

[0041] (6) The coordinates are calculated and integrated based on the corrected position information and temperature parameter feedback data to obtain a three-dimensional space coordinate data packet.

[0042] Specifically, the feature points of the pre-treated sound-absorbing cotton material are captured by a high-definition CCD camera to obtain the surface feature information of the material. The high-definition CCD camera is a charge-coupled device image sensor with high resolution and low noise characteristics, and can capture tiny surface feature changes. During the capture process, the camera scans the surface of the sound-absorbing cotton from multiple angles to obtain feature data such as surface texture, edge contour, preset marking points, etc. The image processing unit performs noise reduction, contrast enhancement and edge sharpening on the original image to convert the blurred image into a clear feature point data set. The feature point extraction algorithm (such as Harris corner point detection or SIFT algorithm) identifies key feature points from the processed image, including material edge intersection points, reserved hole center points and surface texture feature points, to form a material surface feature information data set. According to the material surface feature information, the pre-treated sound-absorbing cotton material is matched with a three-dimensional model by a feature matching algorithm to obtain the initial position data. The feature matching algorithm is a calculation method that matches the captured feature points with the preset three-dimensional model, and mainly adopts a matching strategy combining the RANSAC (random sample consistency) algorithm and the ICP (iterative closest point) algorithm. The algorithm selects stable and reliable feature points from the feature information, and then matches them with the corresponding points in the three-dimensional model to calculate the transformation matrix. The RANSAC algorithm selects the best matching point set and removes abnormal points through multiple random sampling, while the ICP algorithm continuously optimizes the transformation matrix through iteration to minimize the error between the actual point cloud and the ideal model. The matching process involves the calculation of the rotation matrix and the translation vector to obtain the initial position data of the material in space, including the coordinate origin position, posture angle and spatial distribution information of the feature points.

[0043] The initial position data is corrected for deformation error through a flexible material deformation compensation algorithm to obtain the corrected position information. The flexible material deformation compensation algorithm is a compensation method specifically designed for the non-rigid deformation generated by soft materials (such as sound-absorbing cotton) during the positioning process. This algorithm is based on finite element analysis and material mechanics models, taking into account the elastic properties of the material, the influence of gravity, and the support state to correct the initial position data. The mathematical model of the deformation compensation algorithm can be expressed as follows:

[0044]

[0045] Where, represents the corrected position coordinates, represents the initial position coordinates, is the weight coefficient of the j-th deformation factor, is the deformation displacement function, E is the material elastic modulus tensor, G is the gravity field vector, S is the support state matrix. Regarding the thickness and density differences in different regions of the sound-absorbing cotton, the algorithm calculates the deformation compensation amount in each region, forms a global deformation compensation field, and outputs the corrected position information.

[0046] A negative pressure is applied to the pretreated sound-absorbing cotton material through a distributed micro pneumatic suction cup array to obtain a stably fixed material. The distributed micro pneumatic suction cup array is a fixing device composed of multiple independently controlled small vacuum suction cups, which is arranged according to the shape characteristics and rigidity distribution of the material. The diameter of each suction cup is usually 5 - 15 mm, and it is connected to the central vacuum system through an independently controlled valve. The control unit calculates the optimal suction cup layout and negative pressure distribution scheme based on the corrected position information, and applies different magnitudes of negative pressure (usually in the range of 15 - 20 kPa) to different regions of the material. The pressure distribution follows the principle of high pressure in the rigid region and low pressure in the flexible region of the material to avoid material deformation. The activation sequence of the suction cups is also optimized. First, the key support points are fixed, and then it gradually expands to the surrounding areas to ensure that the material does not produce wrinkles or tensile deformation during the fixing process. When all the suction cups reach the set negative pressure value and stabilize for a period of time, the material enters the stable fixed state.

[0047] The surface temperature distribution of the stably fixed material is monitored by a thermal imaging sensor to obtain temperature parameter feedback data. A thermal imaging sensor is a device that can detect the infrared radiation intensity on the surface of an object and convert it into a temperature distribution map, with a resolution typically reaching 0.05°C. The sensor scans the entire surface of the fixed sound-absorbing cotton material to obtain the temperature field distribution data on the material surface. The data processing unit converts the original thermal image into a standardized temperature matrix, records the temperature value and spatial position of each pixel point, and forms a temperature gradient map. The temperature parameter feedback data includes information such as the average temperature on the material surface, the position of the highest temperature point, the position of the lowest temperature point, the magnitude of the temperature gradient, and the temperature field uniformity index. These data reflect the thermal state and internal structure characteristics of the material and have important guiding significance for the setting of subsequent welding parameters. According to the corrected position information and temperature parameter feedback data, coordinate calculation and integration are performed to obtain a three-dimensional space coordinate data packet. The coordinate calculation and integration process is a process of fusing the geometric position data and thermal characteristic data of the material. Different-dimensional information is integrated into a unified three-dimensional coordinate system through a data fusion algorithm. The fusion process establishes a mapping relationship between the geometric coordinate system and the thermal imaging coordinate system, and then marks the temperature data as an additional attribute of each spatial point of the material. The integrated data includes multi-dimensional information such as the spatial coordinates, temperature values, material thickness, and material density of each feature point. During the data sorting process, the centroid position, principal axis direction, and boundary coordinates of the key functional area of the material are also calculated. These information together constitute a three-dimensional space coordinate data packet, which serves as the basic data for subsequent welding path planning.

[0048] Taking the positioning of the sound-absorbing cotton under the rear seats of a certain type of SUV as an example, a high-definition CCD camera scans the preprocessed sound-absorbing cotton from multiple angles, identifying 142 edge contour points, 8 reserved hole center points, and 326 surface texture feature points, forming complete material surface feature information. The feature matching algorithm compares these points with the standard 3D model in the vehicle model database. After 50 iterations, the RANSAC algorithm selects 429 best matching point sets. Based on this, the ICP algorithm performs 9 rounds of optimization, calculating the initial position data with a rotation angle of 2.3° on the X-axis, 1.8° on the Y-axis, 0.4° on the Z-axis, and a translation vector of (12.5mm, -8.3mm, 5.2mm). The flexible material deformation compensation algorithm detects that the material has sagged by approximately 4.2mm in the middle region due to its own weight. According to the elastic modulus E = 0.28MPa and the influence of the gravity field, the deformation compensation amount is calculated, and the initial position data is corrected to obtain the accurate position coordinates after compensation. The distributed micro pneumatic suction cup array activates 32 suction cups according to the corrected position information, applying a negative pressure of 18.5kPa on average to the material, with a slightly lower pressure of 16kPa in the edge region and a higher pressure of 19.5kPa in the central region, fixing the material in the ideal position. The thermal imaging sensor scans the fixed material and finds that the average surface temperature of the material is 26.3°C, the highest temperature of 28.7°C is located in the region with sufficient heat treatment before, and the lowest temperature of 24.1°C is located in the weak edge region. The temperature field distribution map shows that the internal structure of the material is uniform. The coordinate calculation and integration process fuses the geometric position data and the temperature field data to generate a 3D space coordinate data packet containing multi-dimensional attributes, marking the boundary coordinates and the best welding entry points of 8 key welding regions.

[0049] In a specific embodiment, the process of executing step S103 may specifically include the following steps:

[0050] (1) Perform a preliminary path calculation on the welding region according to the 3D space coordinate data packet in combination with the sound-absorbing cotton structure parameters of the vehicle model to obtain a preliminary welding path;

[0051] (2) Perform multi-dimensional optimization on the preliminary welding path through historical welding quality data to obtain an optimized welding path;

[0052] (3) Perform a partitioned progressive division on the overall welding task according to the optimized welding path to obtain welding sub-region data;

[0053] (4) Mark the main weld points and auxiliary weld points for each sub-region according to the welding sub-region data to obtain a weld point distribution map;

[0054] (5) Differentially calibrate the energy input of each weld point according to the weld point distribution map in combination with the material thickness parameters to obtain a welding energy distribution curve;

[0055] (6) Heat control marks are set on the welding energy distribution curve through intermittent cooling point settings to obtain a digital welding instruction package.

[0056] Specifically, based on the three-dimensional space coordinate data packet and the structure parameters of the vehicle model's sound-absorbing cotton, a preliminary path calculation is performed on the welding area to obtain a preliminary welding path. The three-dimensional space coordinate data packet is the precise spatial position information generated in the previous modular positioning link, including the position, attitude, and feature point distribution of the sound-absorbing cotton. The structure parameters of the vehicle model's sound-absorbing cotton are the material specification information from the vehicle design database, including data such as thickness distribution maps, density distribution maps, stress area identifications, and welding requirements. During the path calculation process, the standard welding point requirements for the sound-absorbing cotton of this vehicle model are extracted from the vehicle database, and then these theoretical welding points are mapped to the actually positioned sound-absorbing cotton. Considering the position deviation and deformation factors of the actual material, the optimal connection path between the welding points is calculated through the shortest path algorithm (such as the improved algorithm). The improved algorithm, compared with the standard algorithm, increases the weight consideration of material characteristics, avoids easily deformable areas and high-stress areas, and generates a welding path more suitable for soft materials. In the embodiment of the present application, the standard algorithm is a heuristic pathfinding algorithm that determines the shortest path through the evaluation function f(n) = g(n) + h(n), where g(n) represents the actual cost from the starting point to the current node, and h(n) represents the estimated cost from the current node to the target node. This algorithm mainly considers spatial distance factors and searches for the geometrically shortest path in a grid or graph structure. The improved algorithm in the present application is based on the standard Based on the algorithm, critical improvements have been made to meet the special requirements of soft materials. The algorithm modifies the evaluation function and introduces a material property weight factor, enabling it to consider not only spatial distance but also the elastic modulus, thickness distribution, and stress conditions of the material. This improvement allows the algorithm to intelligently avoid deformable areas and high-stress areas, even if these areas may offer shorter paths geometrically. It can generate welding paths more suitable for the material properties of soft sound-absorbing cotton, effectively reducing the risks of material deformation and stress concentration during welding. Secondly, by comprehensively considering multi-dimensional factors such as path length, material thickness, and heat accumulation, a balance between welding quality and efficiency is achieved. It can adaptively adjust the path strategy according to the material properties of different regions, making the welding process more stable and reliable. The initially formed welding path contains a sequence of welding point coordinates and connection line information. The initial welding path is optimized multi-dimensionally using historical welding quality data to obtain the optimized welding path. Historical welding quality data refers to the historical welding records and quality assessment results of sound-absorbing cotton for the same or similar vehicle models. These data are stored in a welding quality database and contain multi-dimensional information such as welding parameters, environmental conditions, material status, and quality scores. The multi-dimensional optimization process uses machine learning methods. By analyzing the correlation between various parameter combinations and welding quality in historical data, the best welding strategy is found. The optimization algorithm considers multiple evaluation dimensions such as welding strength, energy consumption efficiency, welding time, and the impact on sound-absorbing performance, and comprehensively scores each candidate path. The artificial neural network model will predict the quality results of different path selections based on the input initial welding path and current material status parameters and give optimization suggestions. The optimization process not only adjusts the order and path of welding points but also fine-tunes the welding parameters in specific regions based on historical experience. After multiple rounds of iterative optimization, the optimized welding path with the best comprehensive performance in multiple evaluation dimensions is obtained.

[0057] The overall welding task is partitioned and progressively divided according to the optimized welding path to obtain welding sub-region data. Partitioning and progressive division is to decompose the overall welding task into several relatively independent sub-regions to achieve more precise welding control and heat management. The division process considers factors such as the geometric shape, thickness variation, stress conditions, and functional requirements of the material, and uses region growing algorithms or watershed segmentation algorithms to divide sub-regions with clear boundaries and relatively uniform interiors. The material properties within each sub-region are similar and suitable for using similar welding parameters. Progressive division is to plan the welding sequence between regions, usually following the order principle from the inside to the outside, from thick to thin, and from key structures to secondary regions, to avoid heat accumulation and material deformation during welding. The welding sub-region data includes information such as the boundary coordinates, area size, average thickness, material density, and adjacent region relationships of each sub-region, and also includes the welding priority ranking and recommended welding parameters for each region.

[0058] Based on the welding sub-region data, each sub-region is marked with primary welding points and auxiliary welding points to obtain a welding point distribution map. The primary welding point refers to the key welding position that bears the main connection strength, usually set at the parts with large stress and important structure; the auxiliary welding point is the secondary welding position that aids in fixation and enhances the overall stability. During the welding point marking process, the key stress points are identified based on the stress analysis results of the material and marked as primary welding points; then, according to the material deformation prediction model, auxiliary welding points are added in the areas prone to deformation or separation. The spatial distribution of the primary and auxiliary welding points follows the principle of "primary point skeleton, auxiliary point filling", ensuring that the primary welding points form the skeleton structure of the overall connection, and the auxiliary welding points fill the gaps between the skeletons to provide uniform support. The welding point density distribution is dynamically adjusted according to the regional importance, with a higher density in important functional areas and a lower density in general areas. The welding point distribution map forms a complete data set containing the spatial coordinates, type identification (primary / auxiliary), priority, and recommended welding parameters of each welding point.

[0059] Based on the welding point distribution map and combined with the material thickness parameters, the energy input for each welding point is differentially calibrated to obtain a welding energy distribution curve. The material thickness parameter refers to the thickness distribution data of the sound-absorbing cotton in different regions, usually provided by the design specifications or measured through previous scans. Differential calibration means that different welding energy parameters are assigned to each welding point according to the material characteristics of different regions. The calibration process uses a thickness-energy mapping model. Areas with a larger thickness require a higher energy input to ensure full penetration, while areas with a smaller thickness require a reduced energy to avoid burn-through. The calibration algorithm takes into account various factors such as the thermal conductivity, melting point, and density of the material to calculate the most suitable energy input value. For areas with a thickness greater than 8 mm, the energy is increased by 20% based on the standard energy; for areas with a thickness less than 3 mm, the energy is reduced by 15% based on the standard energy. The welding energy distribution curve is an energy control curve that changes with the welding process, describing the energy input values of each welding point at each time point during the welding process, including parameters such as power magnitude, duration, and waveform characteristics.

[0060] Heat control marking is performed on the welding energy distribution curve through intermittent cooling point settings to obtain a digital welding instruction package. An intermittent cooling point refers to a short pause point inserted during the welding process, aiming to allow heat dissipation and avoid overheating damage to the material. The heat control marking process predicts the temperature changes in each area during the welding process through a heat accumulation model, and then inserts cooling points before the temperature is about to exceed the safety threshold. Usually, a cooling point is set after every 25 cm of welding length, and the cooling time is 3 seconds. The distribution of cooling points also considers the thermal conductivity and heat dissipation conditions of the material, and increases the cooling point density in areas where heat accumulates quickly. After the heat control marking is completed, all welding parameters are integrated into a digital welding instruction package, which contains complete welding execution data, such as the welding point coordinate sequence, welding path, energy input curve, cooling point position and duration, welding speed adjustment, etc. The digital welding instruction package is a structured data file that can be directly read and executed by the welding execution system.

[0061] Taking the welding of the sound-absorbing cotton in the trunk of a certain car model as an example, the three-dimensional space coordinate data package contains the accurate position information of the positioned sound-absorbing cotton, and a total of 145 feature points are marked. Combining the structural parameters of the sound-absorbing cotton in the trunk of this model in the vehicle model database (the material thickness varies from 6 to 14 mm, the edge area is thinner, and the central load-bearing area is thicker), the preliminary welding path connecting all necessary points is calculated through an improved A* algorithm, with a total length of 267 cm and 68 welding points. Nearly 200 groups of welding records of the same type of sound-absorbing cotton are extracted from the historical quality database. Through multi-dimensional optimization algorithm analysis, it is found that under the current thickness distribution, the welding energy in the central area is low, resulting in a high frequency of insufficient strength problems, and over-welding is prone to occur in the edge area. Based on this, the preliminary path is optimized and adjusted. The welding sequence is adjusted to expand from the center outwards, and the welding parameters of key points are slightly adjusted to form an optimized welding path. The overall welding task is divided into three main sub-areas: the front, middle, and rear. Each area is further divided into 2-3 secondary areas, forming a total of 8 welding sub-areas. Each area is sorted by importance and assigned a welding priority. During the welding point marking process, 13 main welding points are marked in the stress area at the bottom of the trunk, and 42 auxiliary welding points are assigned to the surrounding and connecting areas to form a complete welding point distribution map. When calibrating the energy distribution, considering that the average thickness of the bottom area is 12 mm, the welding energy is set to 118% of the standard value; the average thickness of the edge area is only 7 mm, and the welding energy is reduced to 90% of the standard value to generate a complete energy distribution curve. Through heat accumulation analysis, 11 intermittent cooling points are set on the welding path, mainly distributed in the central area where heat is not easily dissipated, and the cooling time varies from 3 to 5 seconds, generating a digital welding instruction package containing all parameters.

[0062] In a specific embodiment, the process of performing step S104 may specifically include the following steps:

[0063] (1) Select the frequency of the dual - frequency ultrasonic generator according to the material thickness parameter in the digital welding instruction package to obtain the welding frequency configuration parameter;

[0064] (2) Convert the ultrasonic energy through the titanium alloy transducer according to the welding frequency configuration parameter to obtain the welding energy conduction data;

[0065] (3) Collect the acoustic impedance of the welding point at the millisecond level through the real - time impedance matching technology to obtain the welding impedance data stream;

[0066] (4) Dynamically adjust the ultrasonic power output and pressure parameters according to the welding impedance data stream to obtain the adaptive welding control value;

[0067] (5) Perform double - closed - loop detection on the welding state through the temperature monitor and the deformation sensor to obtain the real - time state information of the welding process;

[0068] (6) Perform directional cooling treatment on the welded area through the rapid cooling device to obtain the sound - absorbing cotton component after welding.

[0069] Specifically, select the frequency of the dual - frequency ultrasonic generator according to the material thickness parameter in the digital welding instruction package to obtain the welding frequency configuration parameter. The dual - frequency ultrasonic generator is a device that can output ultrasonic waves of 20 kHz and 40 kHz simultaneously or selectively, providing the most suitable welding energy for sound - absorbing cotton materials with different thicknesses and densities. The digital welding instruction package contains the material thickness data of each welding point position, and the best welding frequency is determined through the thickness - frequency mapping rule. Generally speaking, for areas with a thickness greater than 10 mm, 20 kHz low - frequency ultrasonic waves are selected, which have strong penetration and large welding depth; for areas with a thickness less than 5 mm, 40 kHz high - frequency ultrasonic waves are selected, which have concentrated energy and high precision; for medium - thickness areas of 5 - 10 mm, the appropriate frequency or a combination of the two frequencies is selected based on the material density and fiber direction. The frequency selection algorithm also considers the functional importance and stress conditions of the welding point, and adopts a more conservative frequency selection strategy for key functional points. The generated welding frequency configuration parameter contains detailed information such as the frequency selection, power level, waveform characteristics, and duration of each welding point.

[0070] The ultrasonic energy is converted by a titanium alloy transducer according to the welding frequency configuration parameters to obtain welding energy conduction data. The titanium alloy transducer is a device that converts electrical energy into mechanical vibration energy. The use of titanium alloy material is because of its excellent acoustic characteristics and durability. The piezoelectric ceramic composite drive unit inside the transducer generates high-frequency vibrations under the action of an electric field, amplifies the tiny vibrations to a practical level through a mechanical amplifier, and then conducts the energy to the material contact surface through a welding head. During the energy conversion process, the frequency control module precisely adjusts the power supply output frequency according to the welding frequency configuration parameters, and the power control module adjusts the output power according to the material characteristics (usually in the range of 150 - 450W). During the welding energy conduction process, multiple sensors monitor the working state of the transducer in real time, including parameters such as amplitude size, frequency stability, output power, and temperature. These data together constitute the welding energy conduction data, reflecting the transfer efficiency and quality of energy from the generator to the material.

[0071] The acoustic impedance of the welding point is collected at the millisecond level through real-time impedance matching technology to obtain a welding impedance data stream. Acoustic impedance refers to the degree of obstruction of a material to the propagation of sound waves, which changes dynamically during the welding process as the molten state of the material changes. The real-time impedance matching technology collects the acoustic impedance data of the welding point once every millisecond through a special sensor installed on the welding head, forming an impedance change curve with high time resolution. The data acquisition system uses a high-speed A / D converter to convert the analog signal into a digital signal, and after noise reduction and signal enhancement processing, it forms a welding impedance data stream. These data reflect the dynamic changes in the internal state of the material during the welding process, including the characteristic impedance values at different stages such as the start of melting, the melting process, and cooling and solidification. The impedance data processing unit will perform feature extraction and pattern recognition on the original data to identify key impedance change points, such as the melting start point, the optimal welding point, and the solidification completion point, providing a basis for subsequent parameter adjustment.

[0072] Dynamically adjust the ultrasonic power output and pressure parameters based on the welding impedance data stream to obtain an adaptive welding control value. Adaptive welding control is a process of dynamically adjusting welding parameters according to the real-time welding state, mainly based on the characteristic changes in the impedance data stream. The control algorithm analyzes the slope, fluctuation range, and stability of the impedance change curve to determine whether the current welding state is overheating, insufficient, or appropriate. When a sudden increase in impedance is detected, indicating that the material is not fully melted, the algorithm increases the power output (floating within the range of 150 - 450 W); when the impedance drops rapidly, indicating that the material may be overheated, the power output is reduced. At the same time, the pressure control system also dynamically adjusts the welding pressure (within the range of 0.3 - 0.8 MPa) according to the impedance data to ensure good material contact without excessive compression. The adaptive control algorithm adopts a PID (Proportional-Integral-Derivative) control strategy, comprehensively considering the deviation between the current impedance value and the ideal impedance curve, the change rate, and the cumulative error, to calculate the optimal power and pressure adjustment amounts. The adjusted parameter values form the adaptive welding control value, which is updated every millisecond to achieve precise control of the welding process. The welding state is detected by a double closed-loop system using a temperature monitor and a deformation sensor to obtain real-time state information of the welding process. Double closed-loop detection refers to a detection method that simultaneously monitors two key parameters, temperature and deformation, providing more comprehensive and reliable state information than single-parameter monitoring. The temperature monitor uses a high-precision infrared sensor or thermocouple to measure the surface temperature and temperature distribution of the welding area in real time, and the data sampling rate is usually 100 Hz. The deformation sensor uses a laser displacement sensor or a linear variable differential transformer (LVDT) to monitor the thickness change and planar deformation of the material during welding, with an accuracy of up to 0.01 mm. The double closed-loop detection system synchronizes the temperature data and deformation data in time and registers them in space to form a complete state matrix. When the detected temperature exceeds 160 °C (the safety upper limit of the sound-absorbing cotton material) or the deformation amount exceeds the preset threshold, the control system triggers a protection mechanism to automatically reduce the power output or increase the cooling time. The real-time state information of the welding process also includes data such as the welding head position, welding speed, completed welding length, and estimated remaining time, providing full-round monitoring for the entire welding process.

[0073] The welded area of the sound-absorbing cotton component is subjected to directional cooling treatment by a rapid cooling device to obtain the welded sound-absorbing cotton component. The rapid cooling device is a device that can precisely cool a specific area, mainly composed of a cooling air flow generator, a temperature control unit, and a directional flow guiding structure. After welding is completed, the cooling device preferentially cools the high-temperature area according to the position and temperature distribution of the welding points. During the cooling process, the temperature monitor continuously tracks the temperature change of the welding area, and the cooling control algorithm adjusts the cold air flow rate and direction according to the target cooling curve to ensure that the temperature reduction rate is within the set range (usually 10 - 15 °C / second), avoiding the accumulation of internal stress and deformation caused by too fast cooling. The direction and intensity of the cooling air flow are dynamically adjusted according to the material shape and thickness distribution to ensure cooling uniformity. When the temperature of the welding area drops to the safety threshold (usually 40 - 50 °C), the cooling device stops working, and the completed welding points form a stable connection structure. The welded area subjected to directional cooling treatment has high strength and stability, and the welded sound-absorbing cotton component has the shape, strength, and functional characteristics required by the design.

[0074] Taking the welding of the front door inner lining sound-absorbing cotton of a medium-sized SUV as an example, the digital welding instruction package contains detailed material thickness parameters, showing that the thickness of the sound-absorbing cotton varies between 3 - 15 mm, with the thickest reaching 15 mm in the door handle area and the thinnest only 3 mm in the edge area. The frequency selection algorithm selects 20 kHz low-frequency ultrasonic waves for the door handle area, 40 kHz high-frequency ultrasonic waves for the edge area, and a dual-frequency alternating mode for the intermediate transition area according to the thickness distribution map, and a total of 53 welding points are generated to form a frequency configuration parameter table. The titanium alloy transducer sets the output power to 420 W in the door handle area, which drops to 180 W in the edge area, and the amplitude is adjusted within the range of 25 - 40 μm according to the frequency configuration parameters. During the welding process, the real-time impedance matching system collects the impedance data of the welding points every millisecond. The recorded initial impedance of the door handle area is 1250 ohms, the impedance drops to 850 ohms during the melting process, and stabilizes at 980 ohms; the edge area starts from 820 ohms, drops to a minimum of 620 ohms. The adaptive control system adjusts the power in real time according to the impedance change. When the impedance in the door handle area drops too fast, the power is dynamically reduced from the initial 420 W to 385 W; while in the edge area, the impedance suddenly rises in the middle of the welding process, and the power is increased to 210 W accordingly to ensure sufficient melting. The highest temperature recorded by the double closed-loop detection system appears in the middle welding area, reaching 152 °C, lower than the safety threshold of 160 °C; the maximum deformation is 0.82 mm, which appears in the transition area between the edge and the middle. After welding is completed, the rapid cooling device conducts directional cooling on the areas with higher temperatures, and the cooling rate is controlled at 12 °C / second. When the temperature drops to 48 °C, the cooling stops. The completed sound-absorbing cotton component perfectly maintains the designed shape, the strength test of the welding points all meets the design requirements, the edge sealing is good, and the core execution stage of the automatic welding of automotive sound-absorbing cotton is completed.

[0075] In a specific embodiment, the process of performing step S105 may specifically include the following steps:

[0076] (1) Perform a full-surface scan of the welded sound-absorbing cotton component using a high-precision three-dimensional laser scanner to obtain a three-dimensional point cloud model of the component;

[0077] (2) Compare and analyze the three-dimensional point cloud model of the component with the theoretical model to obtain deformation deviation data;

[0078] (3) Emit a preset frequency sound wave to the welded sound-absorbing cotton component using an acoustic feature analyzer to obtain acoustic reflection wave data;

[0079] (4) Analyze the sound absorption characteristics of the welded sound-absorbing cotton component based on the acoustic reflection wave data to obtain sound absorption function parameters;

[0080] (5) Conduct a sampling strength test on the key welding points of the welded sound-absorbing cotton component using a micro tensile sensor array to obtain welding strength data;

[0081] (6) Conduct a comprehensive analysis based on the deformation deviation data, sound absorption function parameters, and welding strength data to obtain a quality assessment report.

[0082] Specifically, perform a full-surface scan of the welded sound-absorbing cotton component using a high-precision three-dimensional laser scanner to obtain a three-dimensional point cloud model of the component. The high-precision three-dimensional laser scanner is an optical measurement device capable of collecting the geometric shape data of the object surface, and the resolution can usually reach 0.05 mm. During the scanning process, the multi-angle positioning device rotates the sound-absorbing cotton component by 360 degrees to ensure full coverage without blind spots. The laser emitter projects structured light onto the surface of the component, and the optical receiver collects the reflected light. The spatial position of the surface points is calculated through the principle of triangulation. The original reflected light signal is processed by filtering, noise reduction, and calibration, and converted into a high-precision three-dimensional coordinate point set. The point cloud data processing algorithm eliminates outliers from the collected surface coordinate data group, identifies and deletes outlier points through sparse point filtering technology, and improves the accuracy of the point cloud data. The processed point cloud data generates a complete three-dimensional point cloud model of the component through a three-dimensional reconstruction algorithm, and this model accurately reflects the actual shape, size, and surface characteristics of the sound-absorbing cotton component.

[0083] Comparative analysis is carried out based on the three-dimensional point cloud model and the theoretical model of the component to obtain deformation deviation data. The theoretical model refers to the standard three-dimensional model created during the product design stage, which contains the ideal shape and size of the sound-absorbing cotton component. The comparative analysis uses the Iterative Closest Point (ICP) algorithm to best match and align the actually scanned three-dimensional point cloud model with the theoretical model. The alignment process includes two stages: rough registration and fine registration. Rough registration quickly finds the approximate position based on feature point matching, and fine registration fine-tunes the relative position through iterative optimization to minimize the overall error between the two models. After completion of the alignment, the deviation distance of each point is calculated to form a deviation distribution map. The deformation analysis algorithm identifies key deformation regions and deformation types, including planar bending, thickness compression, edge deformation, etc., and calculates the maximum deviation, average deviation, and deviation standard deviation of each region. The deformation deviation data also includes the overall dimensional change rate, the severity of deformation in key functional regions, and the position markers of regions with excessive deviation. These data together constitute a comprehensive assessment of the shape quality of the product. By emitting preset frequency sound waves to the welded sound-absorbing cotton component with an acoustic feature analyzer, acoustic reflection wave data is obtained. The acoustic feature analyzer is a professional device capable of generating, receiving, and analyzing sound waves, and is used to test the acoustic properties of materials. During the test, the sound wave emitter emits frequency sound waves in the range of 500Hz - 8000Hz to the sound-absorbing cotton component, covering the main frequency bands of automotive noise. The sound wave receiver captures the sound waves reflected from the surface of the component and converts the analog signal into a digital signal. The acoustic data acquisition system records the incident wave and the reflected wave at each test frequency point, and calculates the reflection ratio and phase difference. To ensure measurement accuracy, the background noise in the test environment needs to be controlled below 30dB, and the test positions are evenly distributed according to the geometric shape of the component. Generally, 20 - 30 test points are selected. The original reflection wave data is processed through time-domain filtering and frequency-domain analysis to generate a complete acoustic reflection characteristic data set, which contains the reflection intensity, phase information, and spatial distribution characteristics at each frequency point.

[0084] The sound absorption characteristics of the welded sound-absorbing cotton component are analyzed based on the acoustic reflection wave data to obtain the sound absorption function parameters. The sound absorption characteristics analysis is a process to evaluate the sound insulation performance of the sound-absorbing cotton material. By performing time-frequency domain conversion on the acoustic reflection wave data through a spectrum analyzer, a frequency distribution spectrogram is obtained, which intuitively shows the reflection characteristics of each frequency point. The energy attenuation calculation is based on the energy ratio of the incident sound wave and the reflected sound wave, and the sound absorption coefficient of each frequency band is calculated to reflect the absorption ability of the material for sound waves of different frequencies. The sound absorption coefficient calculation formula is α = 1 - (Ir / Ii), where α is the sound absorption coefficient, Ir is the intensity of the reflected sound wave, and Ii is the intensity of the incident sound wave. The original sound absorption coefficient data is converted into a standardized sound absorption curve through normalization processing, which is convenient for comparison with industry standards. The sound absorption characteristics analysis also includes special analysis of key frequency points (such as 500Hz, 1000Hz, 2000Hz, 4000Hz), and these frequencies are important index points for evaluating the sound absorption performance of automobiles. By performing acoustic impedance calculation on the standardized sound absorption curve and the key frequency sound absorption index, a material sound resistance characteristic map is generated, and these data are integrated into a complete set of sound absorption function parameters for evaluating the acoustic performance of the product.

[0085] The sampling strength test of the key welding points of the welded sound-absorbing cotton component is carried out through a micro tensile sensor array to obtain the welding strength data. The micro tensile sensor array is a test device composed of multiple small force sensors, which can apply standardized tensile tests to multiple welding points simultaneously or sequentially. During the test process, according to the preset sampling scheme, strength tests are carried out at the key welding points, typical welding points and randomly selected points of the product, and the sampling coverage rate is usually 15%-20% of the total welding points. A standard test force of 25-50N is applied to each test point, and the force value and displacement amount when the welding point starts to deform or break are recorded. The test data processing includes maximum bearing force calculation, elastic deformation interval analysis and plastic deformation characteristic analysis, and a strength-deformation curve of each test point is generated. The welding strength data not only includes the absolute strength value, but also includes the strength uniformity score, weak point distribution map and strength margin estimation, comprehensively reflecting the structural strength status of the product.

[0086] Based on the deformation deviation data, sound absorption function parameters, and welding strength data, a comprehensive analysis is conducted to obtain a quality assessment report. The comprehensive analysis uses a multi-index weighted evaluation method. After normalizing various types of data, different weights are assigned according to their importance. The deformation deviation data mainly affects the assembly matching and appearance quality of the product. The sound absorption function parameters are directly related to the realization of the core function of the product. The welding strength data determines the service life and reliability of the product. The data fusion algorithm maps these three types of index data into a unified scoring system, calculates the comprehensive quality score, and conducts a three-level quality assessment of A / B / C based on a preset threshold. The quality assessment report also includes a detailed defect list, marking the areas that need to be repaired and their defect types, such as areas with excessive deformation, areas with insufficient sound absorption performance, and areas with substandard welding strength. For each defect type, the assessment report gives specific repair suggestions, including reinforcement methods, repair materials, and operating parameters.

[0087] In a specific embodiment, the process of executing step S106 may specifically include the following steps:

[0088] (1) Classify and analyze the defect information in the quality assessment report through a parser to obtain a defect type location database;

[0089] (2) Locally reinforce the area with insufficient welding strength in the defect type location database through a small-diameter high-energy ultrasonic focusing head to obtain a strength reinforcement area;

[0090] (3) Spray a nano-level polymer repair agent on the micro-crack area in the welded sound-absorbing cotton component through a precision spraying device to obtain a crack repair area;

[0091] (4) Perform hot air negative pressure treatment on the slightly deformed area in the welded sound-absorbing cotton component through shape memory thermal adjustment technology to obtain a shape recovery area;

[0092] (5) Trim the edges and remove dust from the surface of the welded sound-absorbing cotton component according to the strength reinforcement area, crack repair area, and shape recovery area to obtain a repaired component;

[0093] (6) Perform surface treatment on the repaired component through an acoustic performance enhancement coating to obtain a finished automotive sound-absorbing cotton assembly.

[0094] Specifically, a defect type location database is obtained by classifying and analyzing the defect information in the quality assessment report through a parser. The parser is a software tool dedicated to text and data structured processing, capable of extracting key information from the quality assessment report and converting it into a standardized data format. The parsing process performs natural language processing on the report text, identifies the defect description paragraphs, and then extracts key information such as defect type, location coordinates, severity, and repair suggestions. The defect classification follows a preset classification system, mainly divided into categories such as insufficient welding strength, material cracks, shape deformation, and substandard sound absorption performance. The location information is converted from the coordinate description in the report to the exact location in the standard three-dimensional coordinate system. The parser also performs correlation analysis on adjacent or related defects to identify defect concentration areas and potential systemic problems. The generated defect type location database contains information such as the type identifier, exact location coordinates, influence range, severity, and recommended repair methods of each defect, providing data support for subsequent targeted repairs. According to the defect type location database, the area with insufficient welding strength is locally reinforced through a small-diameter high-energy ultrasonic focusing head to obtain a strength reinforcement area. The small-diameter high-energy ultrasonic focusing head is a specially designed ultrasonic welding tool with a diameter of about 2mm, which can focus ultrasonic energy into a very small area to achieve precise local reinforcement. During the repair process, the areas with insufficient welding strength are screened from the defect database, sorted according to severity, and the points with the largest strength deviation are processed first. The focusing head positioning system accurately locates the point to be repaired according to the defect coordinates, adjusts the welding angle so that the energy direction is perpendicular to the original welding line to increase the connection strength. The welding parameters are dynamically adjusted according to the defect degree, with a 20-30% increase in energy input in the areas with severe strength deficiency and a 10-15% increase in the slightly deficient areas. The reinforcement process uses a pulsed energy input method, with each pulse duration of 0.1-0.3 seconds and an interval of 0.5-1.0 seconds to avoid overheating damage to the material caused by excessive energy concentration. After the reinforcement is completed, the treated area is quickly cooled and shaped to consolidate the welding structure and form a reinforcement area with the required strength.

[0095] The micro-crack areas in the sound-absorbing cotton components after welding are sprayed with nano-scale polymer repair agents through a precision spraying device to obtain crack repair areas. The precision spraying device is a piece of equipment that can precisely control the spraying position, angle, and dosage, and usually includes a high-precision moving platform, a micro-control valve, and a pressure control system. The nano-scale polymer repair agent is a special material with good fluidity and permeability, which can penetrate into micro-cracks and quickly solidify to form an elastic connection with the same performance as the raw material. During the repair process, crack information, including crack position, length, width, and depth, is extracted from the defect database. The spraying device calculates the optimal spraying angle and pressure according to the crack characteristics. Generally, low-pressure and high-angle spraying is selected for deep cracks, and high-pressure and low-angle spraying is selected for surface cracks. The dosage of the repair agent is precisely calculated according to the crack volume to ensure full filling without overflow. After spraying, the curing reaction of the repair agent is triggered by ultraviolet light or thermal catalysis. The curing time is usually controlled within 60 seconds to form a repair area that is tightly bonded to the raw material.

[0096] The slightly deformed areas in the sound-absorbing cotton components after welding are treated with hot air negative pressure through shape memory thermal adjustment technology to obtain a shape recovery area. Shape memory thermal adjustment technology is a technology that uses the plasticity of the sound-absorbing cotton material under heating and the shape memory characteristics after cooling, combined with negative pressure forming to achieve deformation correction. During the treatment process, deformation area information, including deformation type, deformation degree, and ideal shape data, is extracted from the defect database. The thermal adjustment device sets the hot air temperature (80 - 100 °C), air flow velocity, and heating time according to the deformation characteristics, and precisely heats the deformed area to make the material reach a softened but not melted state. At the same time, the negative pressure forming system generates a negative pressure mold or negative pressure distribution according to the ideal shape data, applies an appropriate negative pressure (usually 5 - 15 kPa), and pulls the softened material back to the ideal position. When the material reaches the target shape, it is quickly cooled while maintaining the negative pressure state. After the temperature drops to room temperature, the material solidifies in the corrected position to form an area with the normal shape restored. The shape recovery process will monitor the surface temperature and deformation degree of the material in real time and dynamically adjust the parameters to ensure the correction effect.

[0097] Edge trimming and surface dust removal are performed on the welded sound-absorbing cotton component according to the strength reinforcement area, crack repair area, and shape restoration area to obtain the repaired component. Edge trimming refers to precisely finishing the edges of the component, removing burrs, excess materials, and irregular edges to ensure that the shape of the component meets the design requirements. The trimming tool selects the appropriate trimming cutter, grinding head, or laser trimmer according to the edge type and performs precise trimming with reference to the edge contour data of the 3D model. The dust removal process uses low-pressure air flow combined with electrostatic adsorption technology to remove dust, fiber debris, and processing residues on the surface and inside of the component. During the cleaning process, the air flow pressure and electrostatic strength are adjusted according to the material properties to ensure thorough cleaning without damaging the material. After edge trimming and surface dust removal are completed, a comprehensive inspection is carried out on the component to ensure that all repaired areas are flat and smooth, without obvious seam marks, and the appearance and shape meet the design requirements, forming the repaired sound-absorbing cotton component.

[0098] The surface of the repaired component is treated with an acoustic performance enhancement coating to obtain the finished automotive sound-absorbing cotton assembly. The acoustic performance enhancement coating is a specially developed functional coating material composed of a composite of nano-porous silica and elastomer, which can improve the sound-absorbing performance, durability, and anti-aging ability of the material. The coating thickness is usually controlled within the range of 0.1 - 0.2 mm to ensure enhanced sound-absorbing performance without affecting the elasticity and softness of the material. The coating process uses uniform spraying or dipping methods, and the best coating method is selected according to the geometric shape and functional requirements of the component. The coating system dynamically adjusts the coating formula and thickness according to the acoustic function requirements of different regions, increasing the coating concentration in key sound-absorbing regions and maintaining the standard concentration in general regions. The coating curing uses low-temperature rapid curing technology to avoid damage to the sound-absorbing cotton material caused by high temperature. After coating is completed, the component is subjected to aging tests and sound-absorbing performance tests to ensure that the coating is uniform, firmly adhered, and the functions meet the standards. The finished automotive sound-absorbing cotton assembly will be attached with a complete quality traceability QR code, recording the production parameters, test data, and repair information of the product to achieve quality monitoring throughout the life cycle.

[0099] In a specific embodiment, the process of performing the step of full-surface scanning of the welded sound-absorbing cotton component by a high-precision 3D laser scanner may specifically include the following steps:

[0100] (1) The welded sound-absorbing cotton component is rotated 360 degrees through a multi-angle positioning device to obtain a full-range scanning preparation state;

[0101] (2) According to the full-range scanning preparation state, the surface of the welded sound-absorbing cotton component is irradiated with light through a laser emitter to obtain surface reflected beam data;

[0102] (3) The surface reflected beam data is collected and sorted through an optical receiver to obtain the original reflected light signal;

[0103] (4) Calculate the depth of the surface points of the welded sound-absorbing cotton component through triangulation based on the original reflected light signal to obtain a surface coordinate data set;

[0104] (5) Remove outliers from the surface coordinate data set through a sparse point filtering algorithm to obtain an optimized coordinate set;

[0105] (6) Perform three-dimensional space reconstruction processing based on the optimized coordinate set to obtain a three-dimensional point cloud model of the component.

[0106] Specifically, the welded sound-absorbing cotton component is arranged in a 360-degree rotation through a multi-angle positioning device to obtain an all-round scanning preparation state. The multi-angle positioning device is a mechanical device specially designed for accurately rotating and positioning the welded sound-absorbing cotton component during the scanning process, consisting of a rotating platform and an adjustable fixture, which can firmly fix the component and achieve a smooth 360-degree rotation. The positioning process uses a non-invasive clamping mechanism to install the component on the platform, avoiding deformation of the soft material. The rotational movement is controlled by a high-precision servo motor, usually positioned and rotated at an incremental angle of 10-15 degrees to ensure sufficient overlap between adjacent scanning positions. At the same time, the vertical positioning is adjusted to capture data from different height angles, forming a scanning position matrix covering the entire surface of the component. This comprehensive positioning strategy eliminates blind spots and shadow areas, obtaining an all-round scanning preparation state required for complete geometric data acquisition.

[0107] Based on the all-round scanning preparation state, project light onto the surface of the welded sound-absorbing cotton component through a laser emitter to obtain surface reflected beam data. The laser emitter generates light with an accurate wavelength (usually 650nm) and projects it onto the surface of the component in a predefined pattern - usually a grid line, dot, or special encoding pattern. These patterns are distorted according to the surface geometry, generating distortions directly corresponding to the surface contour. The emission system dynamically adjusts the light intensity according to the reflection characteristics of the material, using higher intensity for darker or more absorptive areas and lower intensity for highly reflective areas. The laser emission sequence follows a predefined pattern to ensure complete coverage of the surface, and the emission angle and pattern density vary according to the complexity of the component and the required resolution. The generated reflected beam data contains encoded information about how the structured light pattern interacts and deforms with the three-dimensional surface of the component.

[0108] The surface reflected beam data is collected and sorted by an optical receiver to obtain the original reflected light signal. The optical receiver consists of high-resolution digital cameras equipped with special optical filters that isolate the laser wavelength from ambient light. These cameras capture the distorted light patterns at resolutions typically between 2 and 5 megapixels and are capable of detecting subtle surface changes. The captured images undergo initial processing, including noise reduction, contrast enhancement, and background elimination, to isolate the structured light pattern. The receiver system is synchronized with the laser emitter to match each captured image with its corresponding emission pattern and position. Multiple cameras located at different angles simultaneously capture the same projection pattern, providing redundant data and improving accuracy and reliability. The processed image data forms the original reflected light signal, which contains the original information about the pattern distortion caused by the geometry of the component surface.

[0109] Based on the original reflected light signal, the depth of the surface points of the welded sound-absorbing cotton component is calculated by triangulation to obtain a set of surface coordinate data. The triangulation calculation process is based on the principle of geometric optics and determines the depth by measuring the triangle formed by the two ends of a known baseline (laser emitter and optical receiver) and the target point. Its mathematical model can be expressed as:

[0110]

[0111] where, represents the depth value of the target point, B is the baseline distance between the emitter and the receiver, f is the focal length of the receiver lens, is the observed pattern displacement, is the angle between the emitted ray and the baseline, is the pattern position coordinate on the reference plane, is the image center coordinate. For the sound-absorbing cotton component with a complex surface, a material reflection characteristic correction factor and the surface normal vector are introduced in the calculation to further optimize the depth calculation accuracy. The calculation process is performed independently for each identified feature point to form a set of surface coordinate data containing three-dimensional coordinates, which directly reflects the geometric characteristics of the surface of the sound-absorbing cotton component.

[0112] Outliers are removed from the surface coordinate data set through a sparse point filtering algorithm to obtain an optimized coordinate set. The sparse point filtering algorithm is specifically designed to handle measurement outliers caused by light scattering, material reflection, or sensor noise. The algorithm divides the entire point cloud into grids, and local statistical analysis is performed within each grid cell to calculate the spatial distribution characteristics of the points. For each point, its local neighborhood characteristics, including average distance, standard deviation, and local curvature, are calculated based on the K-Nearest Neighbor (KNN) method. Points that significantly deviate from the local statistical characteristics are marked as potential outliers. The algorithm adopts a strategy that combines the distance threshold method, density clustering method, and local plane fitting method to comprehensively evaluate the reliability of each point. Points with scores lower than the threshold are removed, while boundary points and feature points are retained through special processing. The filtering intensity is dynamically adjusted according to the material properties and scanning quality to ensure that real details are retained while removing noise. The processed data forms an optimized coordinate set, which has higher geometric accuracy and consistency. Three-dimensional spatial reconstruction processing is performed based on the optimized coordinate set to obtain a three-dimensional point cloud model of the component. The three-dimensional reconstruction process converts discrete point coordinates into a continuous surface model, and spatial indexing organization of the point cloud is carried out through spatial partitioning techniques (such as octree) to improve the efficiency of subsequent processing. The surface reconstruction algorithm uses the Moving Least Squares (MLS) method to smooth local areas and reduce the impact of residual noise. For areas with minor data missing, a surface interpolation algorithm is used to fill the gaps to ensure the integrity of the model. The reconstruction process takes into account the characteristics of the sound-absorbing cotton material, and special edge-preserving processing is adopted for feature areas such as edges and corners to avoid loss of details due to excessive smoothing. The generated three-dimensional point cloud model accurately represents the geometric shape, size, and surface characteristics of the sound-absorbing cotton component, providing basic data for subsequent deformation analysis and quality assessment.

[0113] In a specific embodiment, the process of performing the step of analyzing the sound absorption characteristics of the welded sound-absorbing cotton component according to the acoustic reflection wave data may specifically include the following steps:

[0114] (1) Perform time-frequency domain conversion on the acoustic reflection wave data through a spectrum analyzer to obtain a frequency distribution spectrogram;

[0115] (2) Evaluate the energy loss of sound waves in each frequency band through energy attenuation calculation according to the frequency distribution spectrogram to obtain a frequency band attenuation coefficient table;

[0116] (3) Perform standardized conversion on the frequency band attenuation coefficient table through normalization processing to obtain a standardized sound absorption curve;

[0117] (4) Extract and compare the sound absorption capabilities at different frequency points according to the standardized sound absorption curve to obtain key frequency sound absorption indicators;

[0118] (5) Perform acoustic impedance calculation on the key frequency sound absorption indicators through impedance analysis to obtain a material acoustic resistance characteristic map;

[0119] (6) Perform data fusion based on the acoustic impedance characteristic spectrum of the material and the standardized sound absorption curve to obtain the sound absorption function parameters.

[0120] Specifically, perform time-frequency domain conversion on the acoustic reflection wave data through a spectrum analyzer to obtain a frequency distribution spectrogram. A spectrum analyzer is a device dedicated to acoustic signal processing, which can convert the acoustic wave signal in the time domain into a frequency domain representation. The conversion process uses the fast Fourier transform algorithm to decompose the original reflection wave data in the time series into the amplitude and phase information of different frequency components. Before processing, the original data is first processed by a window function to reduce spectral leakage. Commonly used window functions are the Hann window or the Hamming window, and the window length is usually set to 8192 or 16384 points to obtain sufficient frequency resolution. After FFT calculation, a complex spectrum containing frequency points in the range of 500 Hz to 8000 Hz is obtained. Subsequently, the power spectral density is calculated to obtain a frequency distribution spectrogram reflecting the energy distribution of each frequency component. The spectrogram is usually represented on a logarithmic scale, with the horizontal axis being the frequency and the vertical axis being the energy intensity, intuitively showing the reflection characteristics of the material to sound waves of different frequencies.

[0121] Evaluate the energy loss of sound waves in each frequency band through energy attenuation calculation based on the frequency distribution spectrogram to obtain a frequency band attenuation coefficient table. Energy attenuation calculation is the basis for measuring the sound absorption performance of materials. The core is to calculate the energy difference between the incident sound wave and the reflected sound wave. The calculation process divides the frequency range into several frequency bands, usually divided by 1 / 3 octave or 1 / 1 octave, such as standard frequency points like 500 Hz, 630 Hz, 800 Hz, 1000 Hz, etc. In each frequency band, calculate the ratio of the incident sound wave energy to the reflected sound wave energy to obtain the energy reflection coefficient. This ratio is then used to calculate the energy absorption coefficient, that is, the proportion of the energy absorbed by the material to the total energy of the incident sound wave. For complex-shaped sound-absorbing cotton components, the influence of the incident angle also needs to be considered. Usually, multi-angle data is collected and the average absorption coefficient at different incident angles is calculated. A frequency band attenuation coefficient table containing the absorption coefficients of each frequency band is formed to comprehensively reflect the sound absorption characteristics of the material in the acoustic spectrum.

[0122] The frequency band attenuation coefficient table is standardized and transformed through normalization to obtain a standardized sound absorption curve. The purpose of normalization is to eliminate the influence of test condition differences, making the data measured under different batches and different environments comparable. The reference state is selected during the processing, such as the ideal sound absorption performance under standard test conditions or the industry standard curve. Then, linear or non-linear transformation is performed on the original attenuation coefficient to map the data into the 0-1 interval. Commonly used transformation methods include the maximum-minimum normalization method and the Z-score normalization method. For different frequency bands, different weights are assigned according to their importance in acoustic evaluation. Generally, automotive interior sound absorption materials have a higher weight for the mid-high frequency (1000 - 4000 Hz) band. The normalized data is plotted as a continuous curve to form a standardized sound absorption curve. The horizontal axis is the frequency, and the vertical axis is the standardized sound absorption coefficient. The shape of the curve intuitively reflects the sound absorption characteristic distribution of the material in the sound spectrum.

[0123] The sound absorption capabilities at different frequency points are extracted and compared based on the standardized sound absorption curve to obtain the key frequency sound absorption index. The key frequency points refer to the frequencies that are of special importance in the automotive acoustic environment, usually including the human voice frequency band (500 - 2000 Hz), the engine noise frequency band (100 - 500 Hz), the wind noise frequency band (2000 - 6000 Hz), etc. During the index extraction process, the sound absorption coefficient values corresponding to these key frequency points are selected from the standardized sound absorption curve, and then compared and analyzed with the target threshold. The comparative analysis not only focuses on the absolute sound absorption value but also emphasizes the shape characteristics of the sound absorption curve, such as the peak position, bandwidth, slope, etc. These characteristics are crucial for evaluating the sound absorption effect of the material in actual applications. The feature extraction algorithm extracts key feature parameters from the curve by methods such as finding local extreme points and calculating the spectral energy centroid. The formed key frequency sound absorption index is a set of comprehensive evaluation indexes that include the sound absorption capabilities of the characteristic frequency points and the description of the curve characteristics, providing a quantitative basis for the acoustic performance evaluation of the material. Through impedance analysis, the acoustic impedance is calculated for the key frequency sound absorption index to obtain the acoustic impedance characteristic map of the material. Acoustic impedance is a physical quantity that measures the degree of obstruction of the material to the transmission of sound waves and directly determines the reflection and transmission characteristics of sound waves at the material interface. The impedance calculation is based on the key frequency sound absorption index and the principle of sound wave propagation. The surface acoustic impedance is inversely deduced from the sound absorption coefficient, and the characteristic impedance in complex form is calculated, including the real part (acoustic resistance) and the imaginary part (acoustic reactance). The influence of physical parameters such as the density, porosity, and flow resistance of the material on the impedance is considered during the calculation to form the complex impedance values at each frequency point. The acoustic impedance characteristic map uses the frequency as the horizontal axis and the modulus and phase of the complex impedance as the vertical axis to comprehensively describe the impedance characteristics of the material in the acoustic spectrum, reflecting the reflection, absorption, and transmission behaviors of the material to sound waves of different frequencies.

[0124] Data fusion is performed based on the acoustic impedance characteristic spectrum of the material and the standardized sound absorption curve to obtain the sound absorption function parameters. Data fusion is a process of integrating two complementary acoustic characteristic representations into a set of comprehensive evaluation parameters. The key features in the acoustic impedance characteristic spectrum, such as the impedance peak frequency, valley frequency, bandwidth, and uniformity, are extracted during the fusion process. Then, these features are mapped and correlated with the key indicators of the standardized sound absorption curve to establish the correspondence between the acoustic impedance characteristics and the sound absorption performance. The fusion algorithm uses the weighted average method, multi-attribute decision-making method, or fuzzy logic method, comprehensively considering the importance and mutual influence of different acoustic indicators to generate the sound absorption function parameters. These parameters include multi-dimensional indicators such as the average sound absorption coefficient, sound absorption bandwidth, frequency selectivity, impedance matching degree, and acoustic stability, comprehensively reflecting the acoustic performance characteristics of the completed sound-absorbing cotton component after welding.

[0125] In a specific embodiment, the process of performing the step of evaluating the energy loss of sound waves in each frequency band by calculating the energy attenuation according to the frequency distribution spectrogram may specifically include the following steps:

[0126] (1) The frequency distribution spectrogram is segmented by a frequency band divider to obtain multiple independent frequency band intervals;

[0127] (2) Calculate the energy ratio of the incident sound wave to the reflected sound wave for each independent frequency band interval to obtain the original energy ratio data;

[0128] (3) Compare and calibrate the difference between the original energy ratio data and the sound absorption performance of the standard reference material to obtain the target sound absorption ability value;

[0129] (4) Mathematically model the attenuation characteristics of each frequency band according to the target sound absorption ability value to obtain the frequency band attenuation characteristic curve;

[0130] (5) Correct the frequency band attenuation characteristic curve for environmental noise and measurement errors through interference factor compensation to obtain the corrected attenuation curve;

[0131] (6) Discretely sample the sound absorption coefficient of each frequency band according to the corrected attenuation curve to obtain the frequency band attenuation coefficient table.

[0132] Specifically, the frequency distribution spectrogram is segmented by a frequency band divider to obtain multiple independent frequency band intervals. The frequency band divider is a frequency domain processing tool specifically for acoustic data analysis, which can divide broadband acoustic data into narrow frequency bands for easy analysis. The division process uses the standard octave method to divide the frequency range from 500 Hz to 8000 Hz into multiple analysis intervals according to international standards. Common division methods include 1 / 1 octave division (such as 500 Hz, 1000 Hz, 2000 Hz, etc.) and 1 / 3 octave division (such as 500 Hz, 630 Hz, 800 Hz, 1000 Hz, etc.). During the division process, a bank of band-pass filters is applied to the original spectral data, and the center frequency and bandwidth of each filter are strictly set according to acoustic standards. The filtered data is organized into independent but complete-spectrum covering frequency band intervals, and each interval contains the detailed spectral characteristics within that frequency band, laying a foundation for subsequent accurate analysis of the sound absorption performance of each frequency band.

[0133] The energy ratio of the incident sound wave to the reflected sound wave is calculated for each independent frequency band interval to obtain the original energy ratio data. The energy ratio calculation is a direct method for evaluating the sound absorption performance of materials. In the calculation, the energy of the incident sound wave within the frequency band is compared with the energy of the reflected sound wave. During the calculation process, the incident sound pressure level and the reflected sound pressure level of each frequency band are extracted from the test records, the sound pressure levels are converted into energy values, and then the ratio of the two is calculated. For the case of plane wave incidence, the energy ratio is directly proportional to the square ratio of the sound pressure; for complex incidence conditions, the sound field type, the incident angle, and the sound wave propagation characteristics also need to be considered. During the processing, special attention is paid to the data continuity at the boundaries of each frequency band, and an interpolation algorithm is used to ensure smooth data transition at the frequency band junctions. The obtained original energy ratio data is a set of numerical values representing the sound energy reflection in each frequency band, and the lower the value, the better the sound absorption performance of the corresponding frequency band.

[0134] By comparing and calibrating the difference between the original energy ratio data and the sound absorption performance of the standard reference material, the target sound absorption capacity value is obtained. Comparative calibration is a key step to eliminate measurement system errors and improve data reliability. In the calibration process, standard reference materials with known acoustic properties are used as benchmarks, and these materials are usually standard sound-absorbing panels certified by national-level laboratories. During the comparison, the original energy ratio data of the test sample is compared with the data of the reference material under the same test conditions, and the difference values in each frequency band are calculated. This differential comparison eliminates the bias influence of the measurement system itself and highlights the difference in sound absorption performance between the material under test and the standard material. The calibration algorithm considers the importance weights of different frequency bands and assigns higher weights to the key frequency bands of automotive interiors (such as the human voice frequency band of 1000 - 3000 Hz). After this relative comparison process, the obtained target sound absorption capacity value more objectively reflects the sound absorption performance of the material in the actual application environment. Based on the target sound absorption capacity value, a mathematical model is established for the attenuation characteristics of each frequency band to obtain the frequency band attenuation characteristic curve. Mathematical modeling is a process of converting discrete sound absorption data points into continuous function expressions, aiming to describe the sound absorption behavior law of the material in the entire frequency spectrum range. Modeling methods such as polynomial fitting, spline interpolation, or Bezier curves are used, and the most suitable function form is selected according to the distribution characteristics of the data points. During the modeling process, preprocessing is performed on the target sound absorption capacity value, including removing outliers, smoothing, and data normalization. Then, through optimization algorithms such as the least squares method, the best fitting parameters are found to minimize the error between the modeling curve and the measured data points. Special attention is paid to the continuity and smoothness of the curve at the intersections of each frequency band during modeling. When necessary, a segmented modeling method is used to ensure that the curve accurately reflects the change in sound absorption characteristics of the material in different frequency bands. The frequency band attenuation characteristic curve after modeling is a continuous function that can predict the sound absorption performance of the material at any frequency point.

[0135] By compensating for interference factors, the environmental noise and measurement error of the frequency band attenuation characteristic curve are corrected to obtain the corrected attenuation curve. Compensating for interference factors is an important part of improving measurement accuracy, mainly targeting influencing factors such as environmental noise, measurement system errors, and sample state fluctuations during the test process. The compensation process identifies the main interference sources, including background noise, reflection surface interference, acoustic wave scattering, and sensor errors. For environmental noise, noise baseline extraction and signal denoising techniques are used to remove the noise components from the original curve. Measurement error compensation is based on the system calibration data to systematically correct the curve. The compensation algorithm also considers the influence of factors such as the surface state of the material, temperature and humidity conditions, and installation methods on the measurement results, and through establishing a correlation model between the interference factors and the measurement deviation, the curve is corrected in a targeted manner. After comprehensive compensation for interference factors, the obtained corrected attenuation curve is closer to the true sound absorption performance characteristics of the material under ideal conditions.

[0136] Discretely sample the sound absorption coefficients of each frequency band according to the corrected attenuation curve to obtain a frequency band attenuation coefficient table. Discrete sampling is to convert a continuous curve function into a practical data table form for subsequent evaluation and application. The sampling process is based on acoustic evaluation criteria, and the sound absorption coefficient values are extracted at standard frequency points (such as 125 Hz, 250 Hz, 500 Hz, 1000 Hz, 2000 Hz, 4000 Hz, 8000 Hz, etc.). When sampling, the importance distribution of the frequency band is considered, and denser sampling points are used in key frequency bands to ensure that the sound absorption characteristics in the important frequency range are accurately expressed. The sampling results are formatted to form a standardized data table containing frequency points and corresponding sound absorption coefficients. This tabular form of data is convenient for comparison with design requirements and is also convenient for presentation and analysis in the quality assessment report. The frequency band attenuation coefficient table becomes an important basis for evaluating the welding quality of the sound-absorbing cotton and directly reflects the influence of the welding process on the sound absorption function of the product.

[0137] As described above, the above embodiments are only used to illustrate the technical solutions of the present application and are not intended to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the various embodiments of the present application.

Claims

1. An automatic welding method for automotive sound-absorbing cotton, characterized in that, The automatic welding method for automotive sound-absorbing cotton includes: Pre-treating the polyester fiber sound-absorbing cotton raw material through a precision cutting system and a high-frequency hot air circulation furnace to obtain pre-treated sound-absorbing cotton material; Performing modular positioning on the pre-treated sound-absorbing cotton material through a six-axis robotic arm in cooperation with visual recognition to obtain a three-dimensional space coordinate data packet, including: capturing feature points of the pre-treated sound-absorbing cotton material through a high-definition CCD camera to obtain material surface feature information; performing three-dimensional model matching on the pre-treated sound-absorbing cotton material through a feature matching algorithm based on the material surface feature information to obtain initial position data; correcting the deformation error of the initial position data through a flexible material deformation compensation algorithm to obtain corrected position information; applying negative pressure fixation to the pre-treated sound-absorbing cotton material through a distributed micro pneumatic suction cup array to obtain a stably fixed state material; monitoring the surface temperature distribution of the stably fixed state material through a thermal imaging sensor to obtain temperature parameter feedback data; performing coordinate calculation and integration based on the corrected position information and the temperature parameter feedback data to obtain the three-dimensional space coordinate data packet; Performing welding path planning and optimization based on the three-dimensional space coordinate data packet to obtain a digital welding instruction packet; Executing welding operations based on the digital welding instruction packet through a dual-frequency ultrasonic generator and a dual-closed-loop feedback control system to obtain a sound-absorbing cotton component after welding; Performing quality inspection on the sound-absorbing cotton component after welding through a three-dimensional laser scanning and acoustic feature analysis system to obtain a quality assessment report; Performing defect repair on the basis of the quality assessment report through a point strengthening welding technique and a nano-level polymer repair agent to obtain a finished automotive sound-absorbing cotton assembly.

2. The automatic welding method of the automotive sound-absorbing cotton according to claim 1, characterized in that, The pre-treating the polyester fiber sound-absorbing cotton raw material through a precision cutting system and a high-frequency hot air circulation furnace to obtain pre-treated sound-absorbing cotton material includes: Performing precise size calculation on the polyester fiber sound-absorbing cotton raw material through a computer-aided design system to obtain sound-absorbing cotton geometric parameters; Performing shape cutting on the polyester fiber sound-absorbing cotton raw material according to the sound-absorbing cotton geometric parameters through a precision cutting device to obtain a cut sound-absorbing cotton that matches the vehicle body structure; Performing preset temperature treatment on the cut sound-absorbing cotton in a high-frequency hot air circulation furnace for a predetermined time to obtain molten state sound-absorbing cotton; Removing moisture and impurities from the molten state sound-absorbing cotton through a vacuum suction device to obtain dry sound-absorbing cotton with a moisture content lower than 0.5%; Uniformly coating the surface of the dry sound-absorbing cotton with a modified polyvinyl alcohol and nano-level titanium dioxide composite enhancer through a coating device to obtain the pre-treated sound-absorbing cotton material.

3. The automatic welding method of the automotive sound-absorbing cotton according to claim 1, characterized in that, The performing welding path planning and optimization based on the three-dimensional space coordinate data packet to obtain a digital welding instruction packet includes: Performing preliminary path calculation on the welding area according to the three-dimensional space coordinate data packet in cooperation with the sound-absorbing cotton structure parameters of the vehicle model to obtain a preliminary welding path; Performing multi-dimensional optimization on the preliminary welding path through historical welding quality data to obtain an optimized welding path; Performing partitioned progressive division on the overall welding task according to the optimized welding path to obtain welding sub-region data; Mark the main solder joints and auxiliary solder joints for each sub-region according to the welding sub-region data to obtain a solder joint distribution map; Differentially calibrate the energy input of each solder joint according to the solder joint distribution map in combination with the material thickness parameters to obtain a welding energy distribution curve; Perform heat control marking on the welding energy distribution curve through the intermittent cooling point setting to obtain the digital welding instruction package.

4. The automatic welding method of the automotive sound-absorbing cotton according to claim 1, characterized in that, Execute the welding operation according to the digital welding instruction package through a dual-frequency ultrasonic generator and a dual-closed-loop feedback control system to obtain a welded sound-absorbing cotton component, including: Select the frequency of the dual-frequency ultrasonic generator according to the material thickness parameters in the digital welding instruction package to obtain welding frequency configuration parameters; Convert the ultrasonic energy through a titanium alloy transducer according to the welding frequency configuration parameters to obtain welding energy conduction data; Collect the acoustic impedance of the welding point at the millisecond level through real-time impedance matching technology to obtain a welding impedance data stream; Dynamically adjust the ultrasonic power output and pressure parameters according to the welding impedance data stream to obtain an adaptive welding control value; Perform dual-closed-loop detection on the welding state through a temperature monitor and a deformation sensor to obtain real-time state information of the welding process; Perform directional cooling treatment on the welded area through a rapid cooling device to obtain the welded sound-absorbing cotton component.

5. The automatic welding method of the automotive sound-absorbing cotton according to claim 1, wherein Perform quality inspection on the welded sound-absorbing cotton component through a three-dimensional laser scanning and acoustic feature analysis system to obtain a quality assessment report, including: Perform a full-surface scan on the welded sound-absorbing cotton component through a high-precision three-dimensional laser scanner to obtain a three-dimensional point cloud model of the component; Compare and analyze the three-dimensional point cloud model of the component with the theoretical model to obtain deformation deviation data; Emit specific frequency sound waves to the welded sound-absorbing cotton component through an acoustic feature analyzer to obtain acoustic reflection wave data; Analyze the sound absorption characteristics of the welded sound-absorbing cotton component according to the acoustic reflection wave data to obtain sound absorption function parameters; Perform a sampling strength test on the key solder joints of the welded sound-absorbing cotton component through a micro tensile sensor array to obtain welding strength data; Perform comprehensive analysis according to the deformation deviation data, the sound absorption function parameters, and the welding strength data to obtain the quality assessment report.

6. The automatic welding method of the automotive sound-absorbing cotton according to claim 1, characterized in that Perform defect repair on the basis of the quality assessment report through spot strengthening welding technology and nano-level polymer repair agent to obtain a finished automobile sound-absorbing cotton component, including: Classify and analyze the defect information in the quality assessment report through a parser to obtain a defect type location database; Perform local reinforcement on the area with insufficient welding strength according to the defect type location database through a small-diameter high-energy ultrasonic focusing head to obtain a strength reinforcement area; Spray the nano-level polymer repair agent on the micro crack area in the welded sound-absorbing cotton component through a precision spraying device to obtain a crack repair area; Perform hot air negative pressure treatment on the area with slight deformation in the welded sound-absorbing cotton component through shape memory thermal adjustment technology to obtain a shape recovery area; Edge trimming and surface dust removal are performed on the welded sound-absorbing cotton component according to the strength reinforcement area, the crack repair area, and the shape restoration area to obtain a repaired component; The surface of the repaired component is treated with an acoustic performance enhancement coating to obtain the finished automotive sound-absorbing cotton assembly.

7. The automatic welding method of the automotive sound-absorbing cotton according to claim 5, characterized in that The welded sound-absorbing cotton component is scanned on the entire surface by a high-precision three-dimensional laser scanner to obtain a three-dimensional point cloud model of the component, including: The welded sound-absorbing cotton component is rotated 360 degrees through a multi-angle positioning device to obtain a full-range scanning preparation state; According to the full-range scanning preparation state, a light beam is projected onto the surface of the welded sound-absorbing cotton component through a laser emitter to obtain surface reflected beam data; The surface reflected beam data is collected and sorted through an optical receiver to obtain an original reflected light signal; According to the original reflected light signal, the depth of the surface points of the welded sound-absorbing cotton component is calculated by the triangulation method to obtain a surface coordinate data set; Outliers are removed from the surface coordinate data set through a sparse point filtering algorithm to obtain an optimized coordinate set; Three-dimensional space reconstruction processing is performed according to the optimized coordinate set to obtain the three-dimensional point cloud model of the component.

8. The automatic welding method of the automotive sound-absorbing cotton according to claim 5, characterized in that The sound absorption characteristics of the welded sound-absorbing cotton component are analyzed according to the acoustic reflection wave data to obtain sound absorption function parameters, including: The acoustic reflection wave data is subjected to time-frequency domain conversion through a spectrum analyzer to obtain a frequency distribution spectrogram; According to the frequency distribution spectrogram, the energy loss of sound waves in each frequency band is evaluated through energy attenuation calculation to obtain a frequency band attenuation coefficient table; The frequency band attenuation coefficient table is subjected to standardization conversion through normalization processing to obtain a standardized sound absorption curve; The sound absorption capabilities at different frequency points are extracted and compared according to the standardized sound absorption curve to obtain key frequency sound absorption indicators; Acoustic impedance calculation is performed on the key frequency sound absorption indicators through impedance analysis to obtain a material sound resistance characteristic map; Data fusion is performed according to the material sound resistance characteristic map and the standardized sound absorption curve to obtain the sound absorption function parameters.

9. The automatic welding method of the automotive sound-absorbing cotton according to claim 8, characterized in that, The energy loss of sound waves in each frequency band is evaluated through energy attenuation calculation according to the frequency distribution spectrogram to obtain a frequency band attenuation coefficient table, including: The frequency distribution spectrogram is segmented through a frequency band divider to obtain multiple independent frequency band intervals; The energy ratio of incident sound waves to reflected sound waves is calculated for each independent frequency band interval to obtain original energy ratio data; The difference between the original energy ratio data and the sound absorption performance of a standard reference material is compared through comparison and calibration to obtain a relative sound absorption ability value; Mathematical modeling of the attenuation characteristics of each frequency band is performed according to the relative sound absorption ability value to obtain a frequency band attenuation characteristic curve; The frequency band attenuation characteristic curve is corrected for environmental noise and measurement error through interference factor compensation to obtain a corrected attenuation curve; The sound absorption coefficient of each frequency band is discretely sampled according to the corrected attenuation curve to obtain the frequency band attenuation coefficient table.

Images