Intelligent feeding equipment and method for a panel assembly machine

Abstract

This invention discloses an intelligent feeding device and method for a panel splicing machine. The device includes a PC, a PLC control system, a material lifting and loading mechanism, a spider robotic arm, a vacuum adsorption mechanism, and a vision inspection device. The method is as follows: First, the MES order system transmits product order information to the PC. Then, the material lifting and loading mechanism prepares the materials, and the vision inspection device detects the Grade A logs. Products that do not meet the Grade A standard are placed in a waste disposal area, while data on products that meet the Grade A standard are transmitted to the PC. Finally, the spider robotic arm moves the vacuum adsorption mechanism to the center above the Grade A logs, where it picks up the logs and feeds them into the splicing mechanism. The material lifting and loading mechanism automatically executes the lifting procedure to ensure the materials are within the gripping range of the spider robotic arm. This invention features high feeding efficiency, low labor costs, reliable performance, high safety, and convenient operation.

Description

Technical Field

[0001] This invention relates to the field of automation control technology, and in particular to an intelligent feeding device and method for a panel assembly machine. Background Technology

[0002] Grade A log boards refer to thin boards of 1000mm*600*2mm thickness that are cut from the tree body using a rotary cutting machine. Grade A log boards can be spliced, pressed, edge-trimmed, sanded, and veneered to become qualified plywood, which is used to make fireproof boards and various wooden furniture.

[0003] Visual inspection technology utilizes computer vision and image processing algorithms to analyze and process images of Grade A log boards. This allows for the detection and determination of the number and size of holes and the number and size of stains and defects in the boards, thus assessing whether they meet the Grade A board requirements. Visual inspection algorithms include edge detection, texture analysis, shape matching, and depth image training. By acquiring high-resolution image data and combining it with advanced image processing algorithms, high-precision and high-efficiency inspection of Grade A log boards can be achieved.

[0004] The spider robotic arm utilizes robots or automated equipment to intelligently feed A-grade logs into the splicing machine. Through collaboration with a vision system, it can perceive and identify information such as the position and posture of the logs. Based on preset grasping motion rules and algorithms, the spider robotic arm can feed the A-grade logs into the splicing machine in a prescribed manner for assembly, improving production efficiency, reducing manual operation, and ensuring the stability and accuracy of the splicing machine.

[0005] Existing panel splicing machines mostly use manual feeding, and the testing and evaluation of Grade A log boards mainly rely on personal experience. This subjective factor leads to significant differences in the testing and evaluation results of Grade A log boards, which cannot guarantee the stability of product quality. At the same time, manual feeding is inefficient, and the large amount of repetitive work can easily reduce the safety awareness of workers and cause safety accidents. Summary of the Invention

[0006] The purpose of this invention is to provide a panel feeding device and method that is low in cost, high in efficiency, has high stability in product testing and evaluation, and high in safety.

[0007] The technical solution to achieve the purpose of this invention is: an intelligent feeding device for a panel assembly machine, including a PC terminal, a PLC control system, a material lifting and loading mechanism, a spider robotic arm, a vacuum adsorption mechanism, and a vision inspection device;

[0008] The PC terminal is used to receive product order information, process the data, and send the processed data to the PLC control system.

[0009] The PLC control system is the control center of the intelligent feeding system, controlling the operation of each component;

[0010] The material lifting and loading mechanism is used to lift the Grade A logs to the working area where the spider robotic arm will grasp them.

[0011] The spider robotic arm is equipped with a vacuum adsorption mechanism at its end, which uses vacuum adsorption to transport the Grade A log board from the work area to be grasped to the board splicing mechanism.

[0012] The visual inspection device is used to inspect Grade A log boards, obtain the inspection results and visual positioning data of Grade A log boards, determine whether Grade A log boards are qualified, and provide data support for the operation of the spider robotic arm and vacuum adsorption mechanism.

[0013] A method for intelligent feeding of a panel assembly machine includes the following steps:

[0014] Step 1: The MES order system obtains product order information and transmits it to the PC.

[0015] Step 2: The PLC control system controls the material lifting and loading mechanism to prepare materials;

[0016] Step 3: The visual inspection device detects scratches, dents, holes, stains and other defects on the surface of the Grade A wood board material to obtain the inspection results and visual data.

[0017] Step 4: Place the products that fail to meet the Grade A standard for raw wood into the waste disposal area.

[0018] Step 5: Transmit the data information of products rated as Grade A log boards from the test results to the PC.

[0019] Step 6: Based on visual data positioning, the spider robotic arm drives the vacuum adsorption mechanism to move to the middle position above the Grade A log board;

[0020] Step 7: The vacuum adsorption mechanism controls the vacuum pressure and vacuum area based on the visual inspection data to pick up the Grade A wood boards.

[0021] Step 8: The spider-like robotic arm drives the vacuum adsorption mechanism to feed the Grade A log into the splicing mechanism;

[0022] Step 9: The material lifting and loading mechanism automatically executes the lifting program to ensure that the material is within the grasping range of the spider robotic arm.

[0023] Furthermore, the product order information in step 1 includes product model, product quantity, customer information, and machine type information.

[0024] Furthermore, in step 2, the PLC control system controls the material lifting and loading mechanism to prepare materials, as detailed below:

[0025] Step 2.1: Use a forklift to feed the Grade A logs into the material lifting and loading mechanism;

[0026] Step 2.2: The material lifting and loading mechanism is controlled by a servo motor to lift the Grade A logs to the work area to be grabbed.

[0027] Step 2.3: After arriving at the work area to be grabbed, the left and right alignment mechanisms perform a clamping action to press the Grade A wood board material firmly, so as to prevent the bottom material from being scattered due to vibration when grabbing the material.

[0028] Furthermore, in step 3, the visual inspection device detects scratches, dents, holes, and stains on the surface of the Grade A log board material, obtaining inspection results and visual data, as follows:

[0029] Step 3.1: Use a visual inspection device to capture multiple frames of images of Grade A log boards, and combine the images into one to reduce image noise;

[0030] Step 3.2: Image preprocessing, including image denoising, image enhancement, and image smoothing, to improve image quality and reduce the impact of noise on detection results;

[0031] Step 3.3: Feature extraction. Features in the image are extracted through edge detection, texture analysis, and color analysis.

[0032] Step 3.4: Use region-based convolutional neural networks, fast region convolutional neural networks, region convolutional neural networks, or single-stage detectors to identify holes and blemishes in the image and their location information.

[0033] Step 3.5: Use semantic segmentation, instance segmentation or panorama segmentation algorithms to segment the image into different regions so that each region can be analyzed independently, separating the holes and stains of the Grade A log boards from the background and providing their size and shape information;

[0034] Step 3.6: Train the deep learning model using a large dataset of labeled samples; the dataset includes images of normal wooden boards and wooden boards with holes and stains of different sizes and numbers, and uses pixel-level annotations or bounding box annotations to mark the location and size of defects;

[0035] Step 3.7: Use data augmentation techniques to transform, rotate, and flip the original image to generate more samples, increasing the diversity and generalization ability of the dataset;

[0036] Step 3.8: Using the prepared dataset, select a suitable network structure, loss function and optimization algorithm, and make appropriate hyperparameter adjustments. Train the object detection or segmentation model to learn to identify the size and number of holes and blemishes.

[0037] Step 3.9: Apply the trained model to the real-time detection task. By inputting the image of the wooden board surface into the model, the prediction results are obtained, including the location, size, and number of holes and stains.

[0038] Furthermore, in step 6, based on visual data positioning, the spider robotic arm drives the vacuum adsorption mechanism to move to the center position above the Grade A log board, as detailed below:

[0039] Step 6.1: Using object detection and localization algorithms, identify the location and center point of the Grade A log board through the machine learning model trained in Step 3;

[0040] Step 6.2: The PC calculates the path and distance the spider robotic arm needs to move and sends it to the PLC control system to move the vacuum adsorption mechanism to the center of the Grade A log board.

[0041] Furthermore, in step 7, the vacuum adsorption mechanism controls the vacuum pressure and vacuum area based on the data from visual inspection, as follows:

[0042] Step 7.1: Separate the holes and stains of the Grade A wood boards detected by the vision system from the background, and divide the areas with dense holes according to the provided size and shape information;

[0043] Step 7.2: Transmit the data of densely porous areas to the PLC control system for calculation, control the vacuum adsorption mechanism to avoid densely porous areas, calculate the magnitude of vacuum pressure to avoid crushing the Grade A wood board due to excessive vacuum pressure, and use the calculated vacuum pressure to pick up the Grade A wood board.

[0044] Furthermore, in step 9, the material lifting and loading mechanism automatically executes the lifting program to ensure that the material is within the grasping range of the spider robotic arm, as detailed below:

[0045] During the feeding process, when the number of Grade A logs decreases to the set threshold, the material lifting and loading mechanism automatically executes the lifting program to ensure that the material is within the grasping range of the spider robotic arm until there is no material in the material box.

[0046] Compared with the prior art, the present invention has the following significant advantages: (1) It realizes the automatic evaluation of the grade of the wood board and automatic feeding of the A-grade wood board, reducing the need for manual intervention and thus improving production efficiency; (2) Automation reduces the risk of human error, as the robotic arm and system will perform tasks according to the predetermined program, reducing errors caused by fatigue or negligence; (3) It realizes precise feeding, improving feeding stability and safety, thereby reducing damage and waste; (4) The vision inspection system can detect scratches and other defects on the surface of the A-grade wood board, automatically evaluate whether the wood board material can reach the A-grade wood board, and ensure that only products that meet the quality standards can enter the splicing machine, improving product quality and customer satisfaction; (5) It realizes automated order processing. Through integration with the MES order system, the factory can realize the real-time transmission and processing of order information, improving the degree of automation of production; (6) It realizes real-time monitoring and reporting. Factory managers can monitor the production process in real time through the system, including order execution status, output, efficiency, etc., and the system can generate production reports, which helps to optimize and improve the production process; (7) It enhances safety. Because automation reduces the need for human contact with heavy materials, it reduces the safety risks associated with operating industrial equipment. Attached Figure Description

[0047] Figure 1 This is a flowchart illustrating an intelligent feeding method for a panel assembly machine according to the present invention. Detailed Implementation

[0048] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0049] This invention provides an intelligent feeding device for a panel assembly machine, comprising a PC, a PLC control system, a material lifting and loading mechanism, a spider robotic arm, a vacuum adsorption mechanism, and a vision inspection device.

[0050] The PC terminal is used to receive product order information, process the data, and send the processed data to the PLC control system.

[0051] The PLC control system is the control center of the intelligent feeding system, controlling the operation of each component;

[0052] The material lifting and loading mechanism is used to lift the Grade A logs to the working area where the spider robotic arm will grasp them.

[0053] The spider robotic arm is equipped with a vacuum adsorption mechanism at its end, which uses vacuum adsorption to transport the Grade A log board from the work area to be grasped to the board splicing mechanism.

[0054] The visual inspection device is used to inspect Grade A log boards, obtain the inspection results and visual positioning data of Grade A log boards, determine whether Grade A log boards are qualified, and provide data support for the operation of the spider robotic arm and vacuum adsorption mechanism.

[0055] Combination Figure 1 A method for intelligent feeding of a panel assembly machine includes the following steps:

[0056] Step 1: The MES order system obtains product order information and transmits it to the PC.

[0057] Step 2: The PLC control system controls the material lifting and loading mechanism to prepare materials;

[0058] Step 3: The visual inspection device detects scratches, dents, holes, stains and other defects on the surface of the Grade A wood board material to obtain the inspection results and visual data.

[0059] Step 4: Place the products that fail to meet the Grade A standard for raw wood into the waste disposal area.

[0060] Step 5: Transmit the data information of products rated as Grade A log boards from the test results to the PC.

[0061] Step 6: Based on visual data positioning, the spider robotic arm drives the vacuum adsorption mechanism to move to the middle position above the Grade A log board;

[0062] Step 7: The vacuum adsorption mechanism controls the vacuum pressure and vacuum area based on the visual inspection data to pick up the Grade A wood boards.

[0063] Step 8: The spider-like robotic arm drives the vacuum adsorption mechanism to feed the Grade A log into the splicing mechanism;

[0064] Step 9: The material lifting and loading mechanism automatically executes the lifting program to ensure that the material is within the grasping range of the spider robotic arm.

[0065] As a specific example, in step 1, the MES order system obtains product order information and transmits the product order information to the PC.

[0066] Customer order information includes product model, quantity, customer information, and machine type. The MES order system transmits the order information to the intelligent board feeding system to assess the board grade and automatic feeding program according to the order requirements. Based on the parsed order information, the intelligent board feeding system selects an appropriate feeding program. This program involves determining how to arrange and stack the Grade A logs to meet the order requirements. The spider robotic arm uses a path planning algorithm to begin stacking the Grade A logs together according to the selected feeding program. This process is highly precise to ensure the stability and safety of stacking and splicing. The spider robotic arm makes real-time adjustments based on feedback from the vision positioning system to adapt to Grade A logs of different shapes and sizes. The spider robotic arm acquires information about the surrounding environment through sensors, including the position and movement of obstacles, to ensure operational safety and avoid collisions. The spider robotic arm with a vacuum suction cup mechanism selects an appropriate gripping posture based on the gripping strategy and motion planning algorithm, and precisely controls the vacuum suction cup mechanism to execute the gripping action. The spider robotic arm repeats the above steps, continuously gripping, stacking, feeding, and splicing according to the order quantity requirements until all materials are spliced.

[0067] As a specific example, in step 2, the PLC control system controls the material lifting and loading mechanism to prepare materials, as follows:

[0068] Step 2.1: Use a forklift to feed the Grade A logs into the material lifting and loading mechanism;

[0069] Step 2.2: The material lifting and loading mechanism is controlled by a servo motor to lift the Grade A logs to the work area to be grabbed.

[0070] Step 2.3: After arriving at the work area to be grabbed, the left and right alignment mechanisms perform a clamping action to press the Grade A wood board material firmly, so as to prevent the bottom material from being scattered due to vibration when grabbing the material.

[0071] As a specific example, in step 3, the visual inspection device detects scratches, dents, holes, and stains on the surface of the Grade A log board material, and obtains the inspection results and visual data, as follows:

[0072] Step 3.1: A visual inspection device is installed above the equipment. When the Grade A log board enters the inspection area, the visual inspection device is triggered to capture multiple frames of images in a short period of time and combine the images into one to reduce image noise.

[0073] Step 3.2: Image preprocessing, including image denoising, image enhancement, and image smoothing, to improve image quality and reduce the impact of noise on detection results;

[0074] Step 3.3: Feature extraction. Features in the image are extracted through edge detection, texture analysis, and color analysis.

[0075] Step 3.4: Use object detection algorithms to locate and identify different targets in the image. Object detection algorithms include region-based convolutional neural networks (R-CNN), fast region convolutional neural networks (Fast R-CNN), region convolutional neural networks (R-FCN), and single-stage detectors (such as YOLO and SSD). These algorithms can identify holes and blemishes in the image and provide their location information;

[0076] As a specific example, the visual recognition system employs a deep learning-based object detection algorithm: the YOLO algorithm, to detect and locate the product's position in the scene. The YOLO algorithm uses a single-stage object detection method, combined with anchor boxes to predict the target box, resulting in high detection speed and accuracy. Based on a pre-trained model, the algorithm achieves detection and recognition of different products through fine-tuning with a large amount of product data. Deep learning methods such as convolutional neural networks (CNN) can be used to learn image features, enabling the detection and determination of the number and size of holes and the number and size of stains on Grade A log boards to determine whether they meet the Grade A board conditions.

[0077] Step 3.5: Use semantic segmentation, instance segmentation or panorama segmentation algorithms to segment the image into different regions so that each region can be analyzed independently, separating the holes and stains of the Grade A log boards from the background and providing their size and shape information;

[0078] Step 3.6: Train the deep learning model using a large dataset of labeled samples; the dataset includes images of normal wooden boards and wooden boards with holes and stains of different sizes and numbers, and uses pixel-level annotations or bounding box annotations to mark the location and size of defects;

[0079] Step 3.7: Use data augmentation techniques to transform, rotate, and flip the original image to generate more samples, increasing the diversity and generalization ability of the dataset;

[0080] Step 3.8: Using the prepared dataset, select a suitable network structure, loss function and optimization algorithm, and make appropriate hyperparameter adjustments. Train the object detection or segmentation model to learn to identify the size and number of holes and blemishes.

[0081] Step 3.9: Apply the trained model to the real-time detection task. By inputting the image of the wooden board surface into the model, the prediction results are obtained, including the location, size, and number of holes and stains.

[0082] Before the Grade A log boards enter the industrial camera inspection area, the surface is covered with wood chips and impurities, which will affect the clarity of the drawings acquired by the vision system. Therefore, an automated mechanism for removing wood chips and impurities was designed. Compressed air is used to blow away most of the wood chips from the surface of the board. Then, a soft brush mechanism is used to clean the surface wood chips, impurities and dust through repeated brushing.

[0083] The sawdust and dust adhering to the surface of the Grade A log boards affect the clarity of the images captured by the vision system. The vision system is also designed with filters to help remove sawdust, dust, or other noise from the images. For example, a median filter can be used to smooth the image and remove the effects of sawdust and dust. Image enhancement algorithms are also employed to enhance image contrast and clarity, making it easier to detect surface defects. For example, histogram equalization can improve the brightness distribution of the image. If sawdust and dust cause color distortion, color correction algorithms can be used to repair it. These algorithms can correct the color balance of the image, making it closer to the actual colors.

[0084] As a specific example, in step 4, products that fail to meet the Grade A standard for raw wood are placed into the waste disposal mechanism, as detailed below:

[0085] If the product fails the test, the spider robotic arm directly places it into the waste disposal area, awaiting further manual processing.

[0086] As a specific example, in step 5, the data information of the product rated as Grade A wood board in the test results is transmitted to the PC.

[0087] As a specific example, in step 6, based on visual data positioning, the spider robotic arm drives the vacuum adsorption mechanism to move to the middle position above the Grade A log board, as detailed below:

[0088] Step 6.1: Using object detection and localization algorithms, identify the location and center point of the Grade A log board through the machine learning model trained in Step 3;

[0089] Step 6.2: The PC calculates the path and distance the spider robotic arm needs to move and sends it to the PLC control system to move the vacuum adsorption mechanism to the center of the Grade A log board.

[0090] As a specific example, in step 7, the vacuum adsorption mechanism controls the vacuum pressure and vacuum area based on visual inspection data to pick up the Grade A wood boards, as detailed below:

[0091] Step 7.1: Separate the holes and stains of the Grade A wood boards detected by the vision system from the background, and divide the areas with dense holes according to the provided size and shape information;

[0092] Step 7.2: Transmit the data of densely porous areas to the PLC control system for calculation, control the vacuum adsorption mechanism to avoid densely porous areas, calculate the magnitude of vacuum pressure to avoid crushing the Grade A wood board due to excessive vacuum pressure, and use the calculated vacuum pressure to pick up the Grade A wood board.

[0093] As a specific example, in step 8, the spider robotic arm drives the vacuum adsorption mechanism to send the Grade A log into the splicing mechanism;

[0094] As a specific example, in step 9, the material lifting and loading mechanism automatically executes the lifting program to ensure that the material is within the grasping range of the spider robotic arm, as detailed below:

[0095] During the feeding process, when the number of Grade A logs decreases to the set threshold, the material lifting and loading mechanism automatically executes the lifting program to ensure that the material is within the grasping range of the spider robotic arm until there is no material in the material box.

[0096] The present invention will be further described below with reference to specific embodiments. The accompanying drawings are for illustrative purposes only, representing schematic diagrams rather than actual physical objects, and should not be construed as limiting the scope of this patent.

[0097] Example

[0098] In all the examples shown and discussed here, any specific values ​​should be interpreted as merely exemplary, not as limitations.

[0099] This embodiment provides an intelligent feeding device and method for a panel splicing machine. It uses a visual algorithm to assess whether the material meets the standard for Grade A wood panels and an automated feeding process. This solves the problems of high labor costs and low feeding efficiency caused by manual feeding, as well as the subjectivity of personnel in assessing Grade A wood panels, the lack of data support, and the high risk of work-related injuries.

[0100] This embodiment is achieved through the following technical solution:

[0101] S1: The MES order system receives customer order information. This includes order specifications, quantity, and other relevant information. The order system then transmits the order information to the intelligent panel feeding system for feeding according to the order requirements.

[0102] S2: The material lifting and loading mechanism prepares the materials, using a forklift to deliver three stacks of Grade A logs into the material lifting and loading mechanism. Then, the material lifting and loading mechanism, controlled by a servo motor, lifts the Grade A logs to the work area to be picked up.

[0103] Upon reaching the work area to be grasped, the left and right alignment mechanisms perform a clamping action to press the Grade A wood board material firmly, preventing the bottom material from scattering due to vibration during material grasping.

[0104] S3: Trigger the vision camera to acquire an image of the wood surface. This involves using a high-resolution camera system to detect scratches, dents, holes, stains, and other defects on the surface. The vision inspection system then transmits the detection results to the control system for further processing.

[0105] The MES order system receives customer order information, including specifications, quantity, and other relevant details. The system then transmits this information to the intelligent feeding system of the splicing machine to determine the grade of the wood planks and the appropriate automatic feeding program based on the order requirements. The intelligent feeding system selects the appropriate feeding program based on the parsed order information. This program determines how to arrange and stack the Grade A logs to meet the order requirements. The spider robotic arm uses a path planning algorithm to begin stacking the Grade A logs according to the selected feeding program. This process is highly precise to ensure the stability and safety of the stacking and splicing. The spider robotic arm makes real-time adjustments based on feedback from the vision positioning system to accommodate Grade A logs of different shapes and sizes. The spider robotic arm uses sensors to acquire information about the surrounding environment, including the position and movement of obstacles, to ensure operational safety and avoid collisions. The spider robotic arm, equipped with a vacuum suction cup mechanism, selects the appropriate gripping posture based on the gripping strategy and motion planning algorithm, and precisely controls the vacuum suction cup mechanism to execute the gripping action. The spider-like robotic arm repeats the above steps, continuously grabbing, stacking, feeding, and assembling materials according to the order quantity requirements until all materials are assembled.

[0106] An industrial camera is installed above the device. When the Grade A log board enters the detection area, the industrial camera is triggered to capture multiple frames of images in a short period of time and combine the images into one, which can effectively reduce the noise of the image.

[0107] The visual recognition system employs the YOLO algorithm, a deep learning-based object detection algorithm, to detect and locate products within a scene. The YOLO algorithm uses a single-stage object detection method, combined with anchor boxes for bounding box prediction, resulting in high detection speed and accuracy. Based on a pre-trained model, the algorithm is fine-tuned using a large amount of product data to achieve the detection and recognition of different products. Deep learning methods such as convolutional neural networks (CNNs) can be used to learn image features, enabling the detection and determination of the number and size of holes and the number and size of stains on Grade A log boards to determine whether they meet the Grade A board requirements.

[0108] Before the Grade A logs enter the industrial camera inspection area, their surface is covered with sawdust and other impurities. This can affect the clarity of the drawings acquired by the vision system.

[0109] The sawdust and impurities on the surface of the Grade A log board affect the clarity of the drawings acquired by the vision system. In this system, an automated mechanism for removing sawdust and impurities is designed first. Compressed air is used to blow away most of the sawdust from the surface of the board. Then, a soft brush mechanism is used to clean the sawdust, impurities and dust on the surface through repeated brushing.

[0110] The sawdust and dust adhering to the surface of the Grade A log boards affect the clarity of the images captured by the vision system. The vision system is also designed with filters to help remove sawdust, dust, or other noise from the images. For example, a median filter can be used to smooth the image and remove the effects of sawdust and dust. Image enhancement algorithms are also employed to enhance image contrast and clarity, making it easier to detect surface defects. For example, histogram equalization can improve the brightness distribution of the image. Color correction is used to repair color distortion caused by sawdust and dust. These algorithms can correct the color balance of the image, making it closer to the actual colors.

[0111] S4: Products that do not meet the Grade A standard for raw wood boards are removed. Products that fail the visual inspection are placed directly into the waste disposal area by the spider robotic arm, awaiting further manual processing.

[0112] S5: Products rated as Grade A solid wood boards will have their size information transmitted to the system through visual inspection, and the system's internal algorithm will determine the stacking and splicing procedures for the Grade A solid wood boards.

[0113] S6: Using visual data, the robotic arm with a vacuum adsorption mechanism moves to the middle position above the Grade A log board;

[0114] S7: The vacuum adsorption mechanism controls the vacuum pressure and vacuum area based on visual inspection information. Holes and stains on the Grade A wood boards detected by the vision system are separated from the background, providing information such as their size and shape, and defining areas with dense pores. After acquiring data on these densely pore-filled areas, the data is transmitted to the PLC for calculation, controlling the vacuum adsorption mechanism to avoid creating a vacuum in these areas. The vacuum pressure is calculated to prevent excessive pressure from crushing the Grade A wood boards (2mm thick).

[0115] S8: The robotic arm with a vacuum adsorption mechanism picks up the Grade A logs via step S7 and then places them into the splicing machine at a speed of 3 seconds per board. This achieves automatic splicing and meets production efficiency requirements.

[0116] S9: When the equipment is performing stacking and splicing feeding operations, the number of Grade A log boards will decrease continuously. The material lifting and loading mechanism will automatically execute the lifting program to ensure that the material is within the grasping range of the spider robotic arm until there is no material in the material frame.

[0117] In the description of this invention patent, it should be understood that the structures, proportions, sizes, etc., illustrated in the accompanying drawings are merely for illustrative purposes to aid those skilled in the art and to facilitate understanding and reading. They are not intended to limit the implementation conditions of this invention and therefore have no substantial technical significance. Any modifications to the structure, changes in proportions, or adjustments to size, without affecting the effectiveness and purpose of this invention, should still fall within the scope of the technical content disclosed in this invention. Furthermore, the terms such as "upper," "lower," "left," "right," "middle," "first," and "second" used in this specification are merely for clarity and not intended to limit the scope of implementation of this invention. Changes or adjustments to their relative relationships, without substantially altering the technical content, should also be considered within the scope of implementation of this invention.

[0118] The terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, unless otherwise stated, "a plurality of" means two or more.

[0119] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation" and "setting" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal communication between two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0120] For those skilled in the art, various variations or modifications can be made based on the above description. It is neither necessary nor possible to exhaustively list all possible implementations. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of the claims of this invention.

Claims

1. An intelligent feeding device for a panel assembly machine, characterized in that, Includes PC terminal, PLC control system, material lifting and loading mechanism, spider robotic arm, vacuum adsorption mechanism, and vision inspection device; The PC terminal is used to receive product order information, process the data, and send the processed data to the PLC control system. The PLC control system is the control center of the intelligent feeding system, controlling the operation of each component; The material lifting and loading mechanism is used to lift the Grade A logs to the working area where the spider robotic arm will grasp them. The spider robotic arm is equipped with a vacuum adsorption mechanism at its end, which uses vacuum adsorption to transport the Grade A log board from the work area to be grasped to the board splicing mechanism. The visual inspection device is used to inspect the Grade A log boards, obtain the inspection results and visual positioning data of the Grade A log boards, determine whether the Grade A log boards are qualified, and provide data support for the operation of the spider robotic arm and vacuum adsorption mechanism. The PLC control system controls the material lifting and loading mechanism to prepare materials, as follows: A forklift is used to feed the Grade A logs into the material lifting and loading mechanism. The material lifting and loading mechanism is controlled by a servo motor to lift the Grade A logs to the work area to be grabbed. Upon reaching the work area to be grabbed, the left and right alignment mechanisms perform a clamping action to press the Grade A wood board material firmly, preventing the bottom material from scattering due to vibration during the grabbing process.

2. A method for intelligent feeding of a panel assembly machine, characterized in that, Includes the following steps: Step 1: The MES order system obtains product order information and transmits it to the PC. Step 2: The PLC control system controls the material lifting and loading mechanism to prepare materials, as detailed below: Step 2.1: Use a forklift to feed the Grade A logs into the material lifting and loading mechanism; Step 2.2: The material lifting and loading mechanism is controlled by a servo motor to lift the Grade A logs to the work area to be grabbed. Step 2.3: After arriving at the work area to be grabbed, the left and right alignment mechanisms perform a clamping action to press the Grade A wood board material firmly, so as to prevent the bottom material from being scattered due to vibration when grabbing the material. Step 3: The visual inspection device detects scratches, dents, holes, stains and other defects on the surface of the Grade A wood board material to obtain the inspection results and visual data. Step 4: Place the products that fail to meet the Grade A standard for raw wood into the waste disposal area. Step 5: Transmit the data information of products rated as Grade A log boards from the test results to the PC. Step 6: Based on visual data positioning, the spider robotic arm drives the vacuum adsorption mechanism to move to the middle position above the Grade A log board; Step 7: The vacuum adsorption mechanism controls the vacuum pressure and vacuum area based on the visual inspection data to pick up the Grade A wood boards. Step 8: The spider-like robotic arm drives the vacuum adsorption mechanism to feed the Grade A log into the splicing mechanism; Step 9: The material lifting and loading mechanism automatically executes the lifting program to ensure that the material is within the grasping range of the spider robotic arm.

3. The intelligent feeding method for a panel assembly machine according to claim 2, characterized in that, The product order information in step 1 includes product model, product quantity, customer information, and machine type information.

4. The intelligent feeding method for the panel assembly machine according to claim 2, characterized in that, In step 3, the visual inspection device detects scratches, dents, holes, and stains on the surface of the Grade A log board material, and obtains the inspection results and visual data, as follows: Step 3.1: Use a visual inspection device to capture multiple frames of images of Grade A log boards, and combine the images into one to reduce image noise; Step 3.2: Image preprocessing, including image denoising, image enhancement, and image smoothing, to improve image quality and reduce the impact of noise on detection results; Step 3.3: Feature extraction. Features in the image are extracted through edge detection, texture analysis, and color analysis. Step 3.4: Use region-based convolutional neural networks, fast region convolutional neural networks, region convolutional neural networks, or single-stage detectors to identify holes and blemishes in the image and their location information. Step 3.5: Use semantic segmentation, instance segmentation or panorama segmentation algorithms to segment the image into different regions so that each region can be analyzed independently, separating the holes and stains of the Grade A log boards from the background and providing their size and shape information; Step 3.6: Train the deep learning model using a large dataset of labeled samples; The sample dataset includes images of normal wooden boards and wooden boards with holes and stains of different sizes and numbers, with pixel-level annotations or bounding box annotations used to mark the location and size of defects; Step 3.7: Use data augmentation techniques to transform, rotate, and flip the original image to generate more samples, increasing the diversity and generalization ability of the dataset; Step 3.8: Using the prepared dataset, select a suitable network structure, loss function and optimization algorithm, and make appropriate hyperparameter adjustments. Train the object detection or segmentation model to learn to identify the size and number of holes and blemishes. Step 3.9: Apply the trained model to the real-time detection task. By inputting the image of the wooden board surface into the model, the prediction results are obtained, including the location, size, and number of holes and stains.

5. The intelligent feeding method for a panel assembly machine according to claim 2, characterized in that, In step 6, based on visual data positioning, the spider robotic arm drives the vacuum adsorption mechanism to move to the center position above the Grade A log board, as detailed below: Step 6.1: Using object detection and localization algorithms, identify the location and center point of the Grade A log board through the machine learning model trained in Step 3; Step 6.2: The PC calculates the path and distance the spider robotic arm needs to move and sends it to the PLC control system to move the vacuum adsorption mechanism to the center of the Grade A log board.

6. The intelligent feeding method for a panel assembly machine according to claim 2, characterized in that, In step 7, the vacuum adsorption mechanism controls the vacuum pressure and vacuum area based on visual inspection data, as follows: Step 7.1: Separate the holes and stains of the Grade A wood boards detected by the vision system from the background, and divide the areas with dense holes according to the provided size and shape information; Step 7.2: Transmit the data of densely porous areas to the PLC control system for calculation, control the vacuum adsorption mechanism to avoid densely porous areas, calculate the magnitude of vacuum pressure to avoid crushing the Grade A wood board due to excessive vacuum pressure, and use the calculated vacuum pressure to pick up the Grade A wood board.

7. The intelligent feeding method for a panel assembly machine according to claim 2, characterized in that, In step 9, the material lifting and loading mechanism automatically executes the lifting program to ensure that the material is within the grasping range of the spider robotic arm, as detailed below: During the feeding process, when the number of Grade A logs decreases to the set threshold, the material lifting and loading mechanism automatically executes the lifting program to ensure that the material is within the grasping range of the spider robotic arm until there is no material in the material box.

Images