Visual inspection, marking and processing integrated system of aluminum punching machine
By integrating the integrated visual inspection mark processing system on the aluminum stamping machine, and using deep learning and reinforcement learning algorithms for automatic scheduling and correction, the problems of errors in traditional detection methods and poor accuracy of automation solutions are solved, and efficient and accurate aluminum processing is achieved and detection.
Patent Information
- Application Number
- CN202411204471.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-30
- Publication Date
- 2024-12-03
AI Technical Summary
Traditional manual detection methods are prone to errors, especially in large-scale production, and the existing automation solutions have poor detection accuracy, which cannot effectively ensure the quality control of aluminum processing.
A integrated visual inspection mark processing system for aluminum stamping machine is designed, using an integrated detection system and an online monitoring and correction system. Through automatic scheduling and real-time monitoring of laser cutting and precision detection equipment, high-precision cutting of aluminum is achieved. and detection, combined with deep learning and reinforcement learning algorithms for automatic scheduling and correction.
Automatic quality control is realized, eliminating the impact of human and environmental factors on detection accuracy, significantly improving production efficiency, ensuring the accuracy and consistency of the detection results, and reducing defective rates.
Smart Images

Figure CN119057211A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of aluminum material processing, and in particular to a visual detection, marking and processing integrated system for an aluminum material punching machine. Background Art
[0002] As an advanced processing method, laser cutting technology has the characteristics of high efficiency, precision and strong adaptability. It is particularly suitable for cutting metal materials such as aluminum. Through the high energy density of the laser beam, laser cutting can achieve fast and contactless cutting of aluminum, and the cutting surface quality is high, and no subsequent processing is required. At the same time, laser cutting technology can also be controlled by computer to achieve complex cutting shapes and patterns, which improves the flexibility and diversity of processing.
[0003] Precision machining technology is a processing process that requires high precision and high quality. In the field of aluminum processing, precision machining technology can achieve high-precision grinding, polishing and cutting of aluminum surfaces to meet the requirements of specific applications for material properties, surface quality, etc. Through the application of precision machining technology, the processing accuracy and surface quality of aluminum can be further improved, thereby enhancing product performance and competitiveness.
[0004] The prior art has the following problems:
[0005] 1. Traditional manual inspection methods are prone to errors, especially in large-scale production. As working hours increase, operator fatigue and decreased attention will make quality control difficult to ensure. In the current automation solution, the conventional machine vision of the packaged CCD camera is used for inspection, which is seriously affected by the environment and has poor inspection accuracy. Summary of the invention
[0006] Technical issues solved
[0007] In view of the shortcomings of the prior art, the present invention provides an integrated visual inspection and marking processing system for an aluminum stamping machine, which solves the problems of low integration of aluminum processing and low qualified rate of finished products caused by the inability to monitor during the processing.
[0008] Technical Solution
[0009] To achieve the above objectives, the present invention is implemented through the following technical solutions: a visual inspection and marking processing integrated system for an aluminum stamping machine, the visual inspection and marking integrated machine is composed of an integrated inspection system and an online monitoring and correction system, and the inspection process flow in the visual inspection and marking integrated machine is as follows:
[0010] Sp1: Task setting: the operator sets the parameters and requirements of the inspection task on the control platform according to the customer order or production plan, including the size, shape and quantity of aluminum materials;
[0011] Sp2: Automatic scheduling, operating the control platform of the integrated detection system, automatically scheduling the working order and time of laser cutting equipment and precision detection equipment according to the parameters and priorities set by the task to maximize production efficiency and resource utilization;
[0012] Sp3: Aluminum material loading: the operator places the aluminum material to be tested on the workbench and confirms the testing position and direction through the control platform;
[0013] Sp4: Laser cutting detection. The laser cutting equipment of the integrated detection system accurately cuts the aluminum material into the required size and shape according to the parameters set by the task. At the same time, the online monitoring system and correction system monitor the size deviation and surface quality in the cutting process in real time and make real-time corrections.
[0014] Sp5: Precision inspection and processing. After cutting is completed, the precision inspection equipment conducts further precision inspection of the aluminum material according to the parameters set by the task, including drilling, turning, milling, etc. The online monitoring system and correction system continuously monitor the size and surface quality during the inspection process, and provide real-time feedback to the control platform for adjustment;
[0015] Sp6: Automatic correction. When the online monitoring and correction system detects deviations or defects in the detection process, the control platform will automatically adjust the parameters of the laser cutting equipment and precision detection equipment to ensure the accuracy and consistency of the detection results.
[0016] Sp7: Inspection completion and unloading. After the inspection is completed, the operator confirms the inspection quality and quantity on the control platform, and then unloads the finished aluminum from the workbench for packaging and storage.
[0017] Preferably, the integrated detection system further includes the following contents:
[0018] Equipment integration module: including laser cutting equipment and precision testing equipment, and the laser cutting equipment and precision testing equipment are located at different positions on the same workbench surface and are uniformly controlled by the control platform;
[0019] Control platform: controls and monitors the laser cutting equipment and precision testing equipment inside the equipment integration module. The operator sets the testing parameters, monitors the testing process, and records and stores the testing data for subsequent analysis through the control platform;
[0020] Intelligent production scheduling module: automatically schedules the working order of laser cutting and precision inspection equipment according to different inspection tasks and priorities, and realizes automatic collaboration and coordination between equipment to maximize production efficiency and resource utilization.
[0021] Preferably, the online monitoring and correction system further includes the following contents:
[0022] Real-time monitoring module: including laser measurement, image recognition, sensor monitoring, etc., which can monitor the parameters such as size, shape and surface quality of aluminum materials in real time during the inspection process;
[0023] Automatic correction module: used for automatic correction monitoring during the detection process. Once deviations or defects are detected during the detection process, the system can automatically adjust the parameters of the laser cutting or precision detection equipment to achieve accuracy and consistency of the detection results;
[0024] Data analysis and feedback module: Analyze and process the detection data monitored in real time, and provide timely feedback to the operator.
[0025] Preferably, the integrated detection system includes the following algorithm:
[0026] Automatic Scheduling Algorithm: A scheduling algorithm based on deep learning automatically optimizes the working order of laser cutting and precision inspection equipment according to the parameters and priorities of the inspection tasks;
[0027] Detection parameter optimization algorithm: Combined with deep neural network technology, it automatically adjusts the detection parameters according to process requirements;
[0028] Fault monitoring and early warning algorithm: Adopt a fault prediction model based on machine learning to monitor the operating status of the equipment and abnormal conditions during the detection process in real time, issue alarms in time and provide corresponding processing suggestions.
[0029] Preferably, the online monitoring and correction system includes the following algorithm:
[0030] Real-time data acquisition algorithm: using multi-sensor fusion technology to obtain various parameters of the detection process from equipment, sensors, and image recognition systems in real time;
[0031] Data analysis and processing algorithms: Use deep learning algorithms to analyze and process data collected in real time;
[0032] Automatic correction feedback algorithm: Based on reinforcement learning technology, the parameters of the equipment are automatically adjusted according to the results of data analysis to maintain the accuracy of the test results consistent;
[0033] Learning optimization algorithm: Compare and analyze historical detection data with real-time collected data, and use adaptive learning methods to continuously optimize the performance and accuracy of the online monitoring and correction system.
[0034] Preferably, the automatic scheduling algorithm further includes an adaptive optimization module based on reinforcement learning, which can autonomously learn and continuously optimize the scheduling strategies of laser cutting equipment and precision testing equipment according to the execution results and resource consumption of different tasks in the production process, thereby improving equipment utilization and reducing production time.
[0035] Preferably, the online monitoring and correction system also includes a virtual detection module based on digital twin technology. The virtual detection module synchronizes the data of the physical system and the virtual system in real time, performs virtual simulation and prediction on each step in the production process, and dynamically adjusts the actual operating parameters according to the simulation results, thereby further improving the detection accuracy and reducing the defect rate.
[0036] Preferably, the visual detection and marking integrated machine also includes a visual detection and marking unit, further comprising:
[0037] Image recognition module: Based on the convolutional neural network algorithm, it analyzes the image data during the aluminum inspection process and identifies surface defects and unqualified areas in real time;
[0038] Visual marking module: Mark the defective area based on the image recognition results to facilitate subsequent processing and quality inspection.
[0039] Preferably, the image recognition module further comprises:
[0040] Deep learning image enhancement algorithm: used to improve the accuracy and stability of image recognition, especially the recognition effect under complex lighting conditions;
[0041] Real-time image processing algorithm: Through parallel computing technology, large-scale image data can be quickly processed and analyzed to ensure real-time identification of defects during inspection.
[0042] Preferably, the visual marking module further comprises:
[0043] Marking feedback algorithm: Based on the identification results of the defect area, the marking signal is automatically generated and marked through the visual marking device;
[0044] Intelligent labeling optimization algorithm: Combined with deep learning algorithms, it optimizes labeling accuracy and efficiency based on historical detection data and real-time detection results, and reduces mislabeling and missed labels.
[0045] Beneficial Effects
[0046] The present invention provides a visual inspection and marking processing integrated system for an aluminum stamping machine.
[0047] Beneficial effects:
[0048] 1. The present invention realizes automated quality control through high-precision visual inspection and online monitoring, eliminates the influence of human and environmental factors on the inspection accuracy, and the integrated inspection and scheduling system optimizes the working sequence and time of the equipment through the automatic scheduling algorithm, significantly improving the overall production efficiency.
[0049] 2. The real-time monitoring module of the system of the present invention can accurately collect various parameters in the detection process through multi-sensor fusion technology, and analyze and process the data in real time through deep learning algorithms. Any deviation will be corrected immediately, ensuring the stability and reliability of the production process.
[0050] 3. The system of the present invention records and stores all detection and marking data in real time, which is not only used for quality traceability, but also provides a reliable basis for subsequent production optimization. This data-driven production model helps to continuously improve production processes and product quality. BRIEF DESCRIPTION OF THE DRAWINGS
[0051] Figure 1 is a system structure diagram of the present invention;
[0052] Figure 2 It is a process flow chart of the present invention. DETAILED DESCRIPTION
[0053] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. Specific embodiment one:
[0055] like Figure 1-2 As shown, a visual inspection and marking processing integrated system for an aluminum stamping machine is shown. The visual inspection and marking integrated machine is composed of an integrated inspection system and an online monitoring and correction system. The inspection process flow in the visual inspection and marking integrated machine is as follows:
[0056] Sp1: Task setting: the operator sets the parameters and requirements of the inspection task on the control platform according to the customer order or production plan, including the size, shape and quantity of aluminum materials;
[0057] Sp2: Automatic scheduling, operating the control platform of the integrated detection system, automatically scheduling the working order and time of laser cutting equipment and precision detection equipment according to the parameters and priorities set by the task to maximize production efficiency and resource utilization;
[0058] Sp3: Aluminum material loading: the operator places the aluminum material to be tested on the workbench and confirms the testing position and direction through the control platform;
[0059] Sp4: Laser cutting detection. The laser cutting equipment of the integrated detection system accurately cuts the aluminum material into the required size and shape according to the parameters set by the task. At the same time, the online monitoring system and correction system monitor the size deviation and surface quality in the cutting process in real time and make real-time corrections.
[0060] Sp5: Precision inspection and processing. After cutting is completed, the precision inspection equipment conducts further precision inspections on the aluminum material according to the parameters set by the task, such as drilling, turning, milling, etc. The online monitoring system and correction system continuously monitor the size and surface quality during the inspection process, and provide real-time feedback to the control platform for adjustment;
[0061] Sp6: Automatic correction. When the online monitoring and correction system detects deviations or defects in the detection process, the control platform will automatically adjust the parameters of the laser cutting equipment and precision detection equipment to ensure the accuracy and consistency of the detection results.
[0062] Sp7: Inspection completion and unloading. After the inspection is completed, the operator confirms the inspection quality and quantity on the control platform, and then unloads the finished aluminum from the workbench for packaging and storage.
[0063] In actual applications, operators set the parameters of various inspection tasks on the control platform according to the different requirements of customer orders or production plans. This includes but is not limited to the size (including length, width, thickness), shape (including rectangular, circular or other complex geometric shapes), material properties (including alloy type, hardness requirements), etc. of aluminum. The setting of parameters directly affects the subsequent processing process, including the generation of cutting paths, the formulation of inspection standards, and the benchmark for quality control.
[0064] The integrated inspection system uses advanced scheduling algorithms to automatically schedule the working order and time of each device according to the set priority when multiple tasks are carried out simultaneously. This scheduling process not only maximizes production efficiency, but also reasonably allocates resources to avoid equipment idling or task accumulation. If some tasks require a long cutting time, the system will prioritize the inspection of other short-time tasks to ensure the efficient operation of the entire production line. In addition, the system will also consider the maintenance time of the equipment and the frequency of consumables replacement to further optimize the scheduling process.
[0065] In this step, the operator not only needs to accurately place the aluminum material to be inspected on the workbench, but also needs to confirm the inspection position and direction according to the characteristics of the material. The visual inspection system can automatically identify the placement status of the aluminum material, including whether it is skewed or covered by debris, and prompt the operator to make adjustments in real time. This process ensures the accuracy of subsequent cutting and inspection operations. For factories with a higher degree of automation, robotic arms can also be introduced to automatically load and position aluminum materials to reduce errors caused by human operations.
[0066] The laser cutting equipment is the core of the entire process. It uses high-precision laser to cut aluminum according to the parameters set by the task. The visual inspection system monitors the quality of the cutting edge in real time during the cutting process, including the finish of the cutting surface, the straightness of the edge, etc. At the same time, the online monitoring system feedbacks the dimensional deviation information through the sensor. If the deviation during the cutting process is detected to be beyond the set range, the system will immediately adjust the laser cutting parameters (including laser power, cutting speed, and focus position) to ensure the accuracy and consistency of the cutting. This closed-loop control system ensures that the quality of each batch of products can meet strict industrial standards.
[0067] After cutting, the aluminum will be transferred to the precision testing equipment for further processing. At this time, the equipment will perform drilling, turning, milling and other operations according to different processing requirements. The online monitoring and correction system will continuously monitor the size and surface quality of each step to ensure that the final product meets the predetermined specifications. During the drilling process, the system will monitor the position and rotation speed of the drill bit in real time to detect whether the size and depth of the hole meet the requirements. If any deviation is found, the system will immediately adjust the processing parameters of the equipment to avoid the production of defective products.
[0068] When the online monitoring system detects deviations or quality problems during processing, the control platform will automatically make corrections. For example, if the surface roughness of the aluminum material is detected to exceed the set standard, the system will appropriately reduce the laser cutting speed or increase the laser power to improve the finish of the cut surface. This automatic correction process not only improves production efficiency, but also reduces the frequency of human intervention, thereby improving the overall intelligence level of the production line.
[0069] After all the inspection and correction steps are completed, the operator needs to confirm the quality and quantity of the inspection results on the control platform. At this time, the system will generate a detailed inspection report, including the execution of each process step, inspection parameters and result data. This report will become an important basis for product quality traceability. Finally, the confirmed qualified aluminum will be automatically or manually unloaded from the workbench, and packaged and stored according to the production plan for further processing or shipment.
[0070] This vision inspection and marking all-in-one solution not only greatly improves the precision and consistency of aluminum processing, but also significantly improves production efficiency and reduces production costs through the application of automation and intelligent technology. The system's online monitoring and automatic correction functions ensure real-time optimization of the entire processing process, greatly reducing the defective rate, thereby meeting customers' demand for high-quality aluminum products.
[0071] The integrated detection and marking machine is the core component of the entire aluminum stamping machine system, responsible for realizing high-precision visual inspection and automatic marking functions. The overall composition and working form are as follows:
[0072] Industrial cameras (high-resolution CCD cameras): used to capture surface images of aluminum materials and provide high-precision visual information. These cameras usually have high frame rates and high-definition resolutions, and can capture details such as tiny surface defects and dimensional deviations of aluminum materials.
[0073] Laser marking machine: used to mark the surface of aluminum, usually used to mark serial numbers, batch numbers, QR codes or customized logos. Laser marking machines can accurately control the marking position, depth and shape to ensure the clarity and durability of the mark.
[0074] Lighting system (LED): Provides stable and uniform lighting conditions to ensure that the images captured by the camera are clear and free of shadows. The lighting system is usually adjusted according to the reflective characteristics of aluminum to avoid excessive reflections or glare.
[0075] Image processing unit (GPU or FPGA, GPU is preferred): responsible for real-time processing of image data collected by industrial cameras, extracting the geometric dimensions, surface quality and other features of aluminum. The image processing unit uses algorithm analysis to determine whether the aluminum meets the set quality standards and feeds the results back to the control system.
[0076] Mechanical motion system (XYZ axis motion platform): used to accurately control the position of the camera and the marking machine, so that it can flexibly cover various parts of the aluminum. The XYZ axis motion platform is usually driven by a servo motor, with a precision ball screw or linear guide rail to ensure high-precision positioning.
[0077] Online monitoring and correction module: monitors various parameters during the inspection process (such as size, position, surface quality, etc.) in real time, and automatically adjusts system settings when deviations occur to ensure the accuracy of inspection and marking.
[0078] Control system (PLC or industrial computer): The core control unit of the entire all-in-one machine, responsible for task scheduling, equipment coordination, and data processing and transmission. The control system communicates with each hardware component through industrial Ethernet or other bus systems to achieve synchronous operation of the entire all-in-one machine.
[0079] The working principle of the integrated detection and marking machine is based on the combination of visual inspection technology and laser marking technology. Through the collaborative work of hardware and software, efficient and accurate aluminum material detection and marking can be achieved.
[0080] Image acquisition and processing: The industrial camera will first capture the surface image of the aluminum material. The image processing unit processes the image through the built-in algorithm to extract key features such as the geometric dimensions, surface defects, and material information of the aluminum material. When the image processing unit detects any defects or deviations that exceed the preset standards, it will immediately issue correction instructions through the control system.
[0081] Automatic marking: After the inspection is completed and the quality of the aluminum is determined to be qualified, the laser marking machine starts working. According to the marking information provided by the control system (such as serial number, QR code, etc.), the laser marking machine will perform precise laser marking at the specified position. The XYZ axis motion platform ensures that the marking machine can cover any position of the aluminum material to meet the marking requirements of different products.
[0082] Online monitoring and correction: During the entire inspection and marking process, the online monitoring module continuously collects various real-time data and compares them with preset standards. When the marking position deviation or the inspection quality does not meet the requirements is detected, the system will automatically adjust the position of the marking machine or re-execute the inspection procedure to ensure the accuracy and consistency of the final product.
[0083] Initialization and task setting: The operator enters the task parameters in the control system, including the inspection standard, marking content and quantity. After the system is initialized, each hardware component (such as camera, marking machine, light source, motion platform, etc.) starts self-checking to ensure normal working status.
[0084] Automatic inspection and marking process: After the aluminum is positioned on the workbench, the inspection and marking machine automatically starts. First, the camera system scans the entire aluminum surface, and the image processing unit analyzes the inspection results. After confirmation, the laser marking machine marks according to the preset content. If a problem is found during the inspection, the system will automatically pause and prompt the operator to check or automatically correct it.
[0085] Data storage and quality traceability: All inspection and marking data will be stored in the control system in real time, including the inspection results, marking content, processing time, etc. of each piece of aluminum. These data are not only used for quality traceability, but also provide data support for subsequent production optimization.
[0086] Equipment maintenance and calibration: The inspection and marking machine needs to be regularly maintained and calibrated, especially the accuracy calibration of the camera and laser marking machine. The control system will record the use status of the equipment and remind the operator to perform maintenance when maintenance is required. Specific embodiment 2:
[0088] like Figure 1-2 As shown, an integrated detection system of a visual inspection and marking processing system for an aluminum stamping machine further includes the following contents:
[0089] Equipment integration module: including laser cutting equipment and precision testing equipment, and the laser cutting equipment and precision testing equipment are located at different positions on the same workbench surface and are uniformly controlled by the control platform;
[0090] Control platform: controls and monitors the laser cutting equipment and precision testing equipment inside the equipment integration module. The operator sets the testing parameters, monitors the testing process, and records and stores the testing data for subsequent analysis through the control platform;
[0091] Intelligent production scheduling module: automatically schedules the working order of laser cutting and precision inspection equipment according to different inspection tasks and priorities, and realizes automatic collaboration and coordination between equipment to maximize production efficiency and resource utilization.
[0092] The integrated detection system includes the following algorithms:
[0093] Automatic Scheduling Algorithm: The scheduling algorithm based on deep learning automatically optimizes the working order of laser cutting and precision testing equipment according to the parameters and priorities of the detection tasks; The scheduling algorithm based on deep learning is responsible for optimizing the working order of laser cutting and precision testing equipment according to the parameters and priorities of the detection tasks. The core is as follows:
[0094]
[0095] S * : Optimal scheduling order, P(T i |S): Task T i The success probability under the scheduling order S is predicted by the deep learning model, U(T i ): Task T i The priority or utility value of a
[0096] Task parameters and priorities are used as inputs, and the success probability of each task under different scheduling orders is predicted through a deep neural network model; the expected value of the total utility under all possible scheduling orders is calculated using the above formula; the optimal scheduling order is output, and tasks are executed in this order, automatically optimizing the working order of equipment, maximizing production efficiency, reducing task delays, and improving resource utilization.
[0097] Detection parameter optimization algorithm: Combined with deep neural network technology, the detection parameters are automatically adjusted according to the process requirements; The detection parameter optimization algorithm combines deep neural network technology to automatically adjust the detection parameters according to the process requirements, and the following formula is obtained:
[0098]
[0099] P * : Optimal detection parameter set.
[0100] The loss function measures the gap between the current detection parameters P and the process requirements X.
[0101] Initialize the detection parameter P; input the process requirement X and the current detection parameter P into the deep neural network to calculate the detection effect and the loss of the target The detection parameters are updated through the back-propagation algorithm to minimize the loss function; the update is iterated until the loss function converges, and the optimized detection parameters are output. The detection parameters are automatically adjusted to adapt to different process requirements, ensuring detection accuracy and reducing manual adjustment errors.
[0102] Fault monitoring and early warning algorithm: adopts a fault prediction model based on machine learning to monitor the operating status of the equipment and abnormal conditions during the detection process in real time, issue alarms in time and provide corresponding processing suggestions. The fault monitoring and early warning algorithm adopts a fault prediction model based on machine learning to monitor the operating status of the equipment and abnormal conditions during the detection process in real time, and issue early warnings. The core is as follows:
[0103]
[0104] Predicted probability of failure.
[0105] σ: Activation function, usually the Sigmoid function, mapped to the failure probability between (0,1).
[0106] W: Model weight matrix.
[0107] F: Feature vector, including equipment operating status data such as temperature, vibration, and power consumption.
[0108] b: Bias term.
[0109] Obtain feature data F from device sensors in real time, input the feature data into the trained machine learning model, and output the failure probability when If a certain threshold is exceeded (the threshold is adjusted according to the actual working conditions), a fault warning is triggered and recommended treatment measures are given. Equipment failures can be detected in advance, downtime and production losses can be reduced, and maintenance plans can be optimized.
[0110] The online monitoring and correction system further includes the following:
[0111] Real-time monitoring module: including laser measurement, image recognition, sensor monitoring, etc., which can monitor the parameters such as size, shape and surface quality of aluminum materials in real time during the inspection process;
[0112] Automatic correction module: used for automatic correction monitoring during the detection process. Once deviations or defects are detected during the detection process, the system can automatically adjust the parameters of the laser cutting or precision detection equipment to achieve accuracy and consistency of the detection results;
[0113] Data analysis and feedback module: Analyze and process the detection data monitored in real time, and provide timely feedback to the operator.
[0114] The online monitoring and correction system includes the following algorithms:
[0115] Real-time data acquisition algorithm: Utilize multi-sensor fusion technology to obtain various parameters in the detection process from equipment, sensors, and image recognition systems in real time; use the following formula:
[0116]
[0117] D: fused data vector.
[0118] D i : Data collected by the (i)th sensor.
[0119] α i : The weight of the (i)th sensor, usually determined by Kalman filtering or Bayesian update algorithm.
[0120] Collect data from multiple sensors simultaneously and get D i , through the weighted fusion algorithm, the sensor data D i The integrated data is fused into a comprehensive data vector D, which is transmitted to the subsequent processing module in real time for real-time monitoring and adjustment, providing high-precision, multi-dimensional detection data to ensure the comprehensiveness and accuracy of parameters during the detection process.
[0121] Data analysis and processing algorithms: Use deep learning algorithms to analyze and process data collected in real time and identify abnormal situations;
[0122] A=f(D;θ)
[0123] A: Analysis results, including normal or abnormal labels and their confidence.
[0124] f(D; θ): deep learning model, which processes data D based on parameter θ and outputs analysis results.
[0125] The fused real-time data D is input into the deep learning model. The model processes the data through a multi-layer neural network, extracts features and analyzes the normality of the data, outputs the analysis result A, and transmits it to the automatic correction module for feedback. It analyzes the detection data in real time, quickly identifies potential problems and gives prompts, reducing missed detections and false detections.
[0126] Automatic correction feedback algorithm: Based on reinforcement learning technology, the parameters of the equipment are automatically adjusted according to the results of data analysis to maintain the accuracy of the test results consistent;
[0127]
[0128] P t : Current device parameters.
[0129] R(P t ,A t ): The detection result feedback under the current parameters is used as the reward function.
[0130] α: learning rate, controls the parameter adjustment step size.
[0131] The gradient of the reward function indicates the direction of parameter adjustment.
[0132] Get the current test result feedback from the data analysis module A t , adjust the equipment parameters according to the feedback information, so that the test results gradually tend to the optimal, gradually optimize the parameter settings through multiple iterations until the test results meet the accuracy requirements, and automatically adjust the test equipment parameters in real time to maintain the consistency of the test results and reduce manual intervention.
[0133] Learning optimization algorithm: compare and analyze historical detection data with real-time collected data, and use adaptive learning methods to continuously optimize the performance and accuracy of the online monitoring and correction system;
[0134]
[0135] W t : The current weight of the model.
[0136] Error function between historical data and real-time data.
[0137] β: learning rate, which controls the step size of model weight adjustment.
[0138] The historical detection data D hist With real-time data real Compare, calculate the error, and update the weight W of the deep learning model according to the error t ,With the passage of time and the accumulation of data, the model performance is gradually optimized, ,which improves the detection accuracy of the system.,Through the guidance of historical data, the monitoring and ,correction system is continuously optimized, and the system accuracy and reliability are ,gradually improved.
[0139] The online monitoring and correction system also includes a virtual detection module based on digital twin technology. The virtual detection module synchronizes the data of the physical system and the virtual system in real time, performs virtual simulation and prediction on each step in the production process, and dynamically adjusts the actual operating parameters according to the simulation results, thereby further improving the detection accuracy and reducing the defect rate.
[0140] The automatic scheduling algorithm further includes an adaptive optimization module based on reinforcement learning, which can autonomously learn and continuously optimize the scheduling strategies of laser cutting equipment and precision testing equipment according to the execution results and resource consumption of different tasks in the production process, thereby improving equipment utilization and reducing production time.
[0141] The visual detection and marking integrated machine also includes a visual detection and marking unit, and the visual detection and marking unit further includes:
[0142] Image recognition module: Based on the convolutional neural network algorithm, the image data in the aluminum material inspection process is analyzed to identify surface defects and unqualified areas in real time; the image recognition module further includes:
[0143] Deep learning image enhancement algorithm: used to improve the accuracy and stability of image recognition, especially the recognition effect under complex lighting conditions. It uses deep learning technology to enhance the input image data and reduce interference caused by factors such as lighting and noise.
[0144] I textenh =G(I textinput ,theta)
[0145] I textenh :Enhanced image,I textinput : The original input image, G(I textinput ,θ): Enhancement function, mapping function based on the deep learning model parameter θ.
[0146] Input image I textinput In the deep learning network model, the model extracts the features of the image through multiple convolutional layers, pooling layers and other structures, and adjusts factors such as illumination, contrast, and noise in the fully connected layer. After network processing, the enhanced image θ is output, which is used for subsequent image recognition processing to improve the recognition accuracy. It improves the stability and accuracy of image recognition under complex lighting conditions and reduces the impact of image noise and insufficient contrast.
[0147] Real-time image processing algorithm: Through parallel computing technology, large-scale image data can be quickly processed and analyzed to ensure real-time identification of defects during detection. The real-time image processing algorithm uses parallel computing technology to quickly process and analyze large-scale image data to ensure real-time identification of defects during detection. The formula is as follows:
[0148] D processed =H(D input ,P)
[0149] D processed : Processed image data. D input : Input large-scale image data. H(D input ,P): parallel processing function, where P represents the parallel parameters in the processing process (including the number of cores and threads for parallel computing, etc.).
[0150] The input large-scale image data D input Block processing, each block of image data is processed through parallel computing, using multi-core processors or GPU and other hardware acceleration for fast calculation, integrating the processed data of each block to form a complete processing result D processed This data is then used to detect defective areas in real time, ensuring efficient and accurate identification. Parallel computing greatly improves the speed of image data processing to ensure real-time performance, especially in inspection scenarios where large-scale image data needs to be processed, significantly reducing latency.
[0151] Visual marking module: Based on the image recognition results, the defective area is marked to facilitate subsequent processing and quality inspection. The visual marking module further includes:
[0152] Marking feedback algorithm: Based on the identification results of the defect area, the marking signal is automatically generated and marked through the visual marking device;
[0153] M=g(R)
[0154] M: Mark signal.
[0155] g(R): A function that generates a label signal based on the detection result g(R).
[0156] According to the detection results, the defective area g(R) is identified, and the corresponding marking signal M is generated according to the identification results. The marking signal is sent to the visual marking device, and a mark is automatically generated in the corresponding area to accurately mark the detected defective area, which is convenient for subsequent processing and maintenance and improves the detection efficiency.
[0157] Intelligent labeling optimization algorithm: Combined with deep learning algorithms, it optimizes labeling accuracy and efficiency based on historical detection data and real-time detection results, and reduces mislabeling and missed labels.
[0158] The algorithm formula is as follows:
[0159]
[0160] M opt : Optimal labeling scheme, R i : The i-th detection result, f(Ri ,M):Marking function, which evaluates the marking effect of the i-th detection result under the marking scheme M.
[0161] Collect and input historical detection data and real-time detection results, analyze the data through deep learning models, generate multiple possible marking schemes M, evaluate each scheme, and select the optimal scheme M that can maximize marking accuracy and efficiency opt Automatically apply the optimal marking solution to mark defective areas, reduce mis-marking and missed marking, improve marking accuracy and efficiency, reduce subsequent problems caused by inaccurate marking, and gradually reduce mis-marking and missed marking through optimization of historical data.
[0162] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the statement "comprising a reference structure" do not exclude the existence of other identical elements in the process, method, article or device including the elements.
[0163] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.
Claims
1. A visual inspection and marking processing integrated system for an aluminum stamping machine, characterized in that: The visual inspection and marking integrated machine is composed of an integrated inspection system and an online monitoring and correction system. The inspection process flow in the visual inspection and marking integrated machine is as follows: Sp1: Task setting: the operator sets the parameters and requirements of the inspection task on the control platform according to the customer order or production plan, including the size, shape and quantity of aluminum materials; Sp2: Automatic scheduling, operating the control platform of the integrated detection system, automatically scheduling the working order and time of laser cutting equipment and precision detection equipment according to the parameters and priorities set by the task to maximize production efficiency and resource utilization; Sp3: Aluminum material loading: the operator places the aluminum material to be tested on the workbench and confirms the testing position and direction through the control platform; Sp4: Laser cutting detection. The laser cutting equipment of the integrated detection system accurately cuts the aluminum material into the required size and shape according to the parameters set by the task. At the same time, the online monitoring system and correction system monitor the size deviation and surface quality in the cutting process in real time and make real-time corrections. Sp5: Precision inspection and processing. After cutting is completed, the precision inspection equipment conducts further precision inspection of the aluminum material according to the parameters set by the task, including drilling, turning, and milling. The online monitoring system and correction system continuously monitor the size and surface quality during the inspection process, and provide real-time feedback to the control platform for adjustment; Sp6: Automatic correction. When the online monitoring and correction system detects deviations or defects in the detection process, the control platform will automatically adjust the parameters of the laser cutting equipment and precision detection equipment to ensure the accuracy and consistency of the detection results. Sp7: Inspection completion and unloading. After the inspection is completed, the operator confirms the inspection quality and quantity on the control platform, and then unloads the finished aluminum from the workbench for packaging and storage.
2. The visual inspection and marking processing integrated system for an aluminum stamping machine according to claim 1, characterized in that: The integrated detection system further includes the following contents: Equipment integration module: including laser cutting equipment and precision testing equipment, and the laser cutting equipment and precision testing equipment are located at different positions on the same workbench surface and are uniformly controlled by the control platform; Control platform: controls and monitors the laser cutting equipment and precision testing equipment inside the equipment integration module. The operator sets the testing parameters, monitors the testing process, and records and stores the testing data for subsequent analysis through the control platform; Intelligent production scheduling module: automatically schedules the working order of laser cutting and precision inspection equipment according to different inspection tasks and priorities, and realizes automatic collaboration and coordination between equipment to maximize production efficiency and resource utilization.
3. The visual inspection and marking processing integrated system for an aluminum stamping machine according to claim 1, characterized in that: The online monitoring and correction system further includes the following contents: Real-time monitoring module: including laser measurement, image recognition, and sensor monitoring, which can monitor the size, shape, and surface quality parameters of aluminum materials in real time during the inspection process; Automatic correction module: used for automatic correction monitoring during the detection process. Once deviations or defects are detected during the detection process, the system can automatically adjust the parameters of the laser cutting or precision detection equipment to achieve accuracy and consistency of the detection results. Data analysis and feedback module: Analyze and process the detection data monitored in real time, and provide timely feedback to the operator.
4. The visual inspection and marking processing integrated system for an aluminum stamping machine according to claim 1, characterized in that: The integrated detection system includes the following algorithms: Automatic Scheduling Algorithm: A scheduling algorithm based on deep learning automatically optimizes the working order of laser cutting and precision inspection equipment according to the parameters and priorities of the inspection tasks; Detection parameter optimization algorithm: Combined with deep neural network technology, it automatically adjusts the detection parameters according to process requirements; Fault monitoring and early warning algorithm: Adopt a fault prediction model based on machine learning to monitor the operating status of the equipment and abnormal conditions during the detection process in real time, issue alarms in time and provide corresponding processing suggestions.
5. The visual inspection and marking processing integrated system for an aluminum stamping machine according to claim 1, characterized in that: The online monitoring and correction system includes the following algorithms: Real-time data acquisition algorithm: using multi-sensor fusion technology to obtain various parameters of the detection process from equipment, sensors, and image recognition systems in real time; Data analysis and processing algorithms: Use deep learning algorithms to analyze and process data collected in real time; Automatic correction feedback algorithm: Based on reinforcement learning technology, the parameters of the equipment are automatically adjusted according to the results of data analysis to maintain the accuracy of the test results consistent; Learning optimization algorithm: Compare and analyze historical detection data with real-time collected data, and use adaptive learning methods to continuously optimize the performance and accuracy of the online monitoring and correction system.
6. The visual inspection and marking processing integrated system for an aluminum stamping machine according to claim 4, characterized in that: The automatic scheduling algorithm further includes an adaptive optimization module based on reinforcement learning, which can autonomously learn and continuously optimize the scheduling strategies of laser cutting equipment and precision testing equipment according to the execution effects and resource consumption of different tasks in the production process, thereby improving equipment utilization and reducing production time.
7. The visual inspection and marking processing integrated system for an aluminum stamping machine according to claim 1, characterized in that: The online monitoring and correction system also includes a virtual detection module based on digital twin technology. The virtual detection module synchronizes the data of the physical system and the virtual system in real time, performs virtual simulation and prediction on each step in the production process, and dynamically adjusts the actual operating parameters according to the simulation results, thereby further improving the detection accuracy and reducing the defect rate.
8. The visual inspection and marking processing integrated system for an aluminum stamping machine according to claim 1, characterized in that: The visual detection and marking integrated machine also includes a visual detection and marking unit, and the visual detection and marking unit further includes: Image recognition module: Based on the convolutional neural network algorithm, it analyzes the image data during the aluminum inspection process and identifies surface defects and unqualified areas in real time; Visual marking module: Mark the defective area based on the image recognition results to facilitate subsequent processing and quality inspection.
9. The visual inspection and marking processing integrated system for an aluminum stamping machine according to claim 8, characterized in that: The image recognition module further comprises: Deep learning image enhancement algorithm: used to improve the accuracy and stability of image recognition, especially the recognition effect under complex lighting conditions; Real-time image processing algorithm: Through parallel computing technology, large-scale image data can be quickly processed and analyzed to ensure real-time identification of defects during inspection.
10. The visual inspection and marking processing integrated system for an aluminum stamping machine according to claim 8, characterized in that: The visual marking module further comprises: Marking feedback algorithm: Based on the identification results of the defect area, the marking signal is automatically generated and marked through the visual marking device; Intelligent labeling optimization algorithm: Combined with deep learning algorithms, it optimizes labeling accuracy and efficiency based on historical detection data and real-time detection results, and reduces mislabeling and missed labels.
Citation Information
Cited By
Rubber product rubber raw material pretreatment method
CN119658762A
Method for rapidly correcting perpendicularity of laser cutting machine
CN119703459A
Machining process of automobile motor shell bearing steel bushing based on numerical control technology
CN120287008A
Improved manufacturing method for head of core rod of PQF continuous rolling unit
CN120347489A
Aluminum material laser cutting and precision machining integrated process
CN120438980A
