Anode guide rod assembly identification method and system for aluminum electrolysis anode scrap pressure stripping

Abstract

The invention relates to the technical field of industrial automation and intelligent identification. The invention provides an anode guide rod assembly identification method and system for aluminum electrolysis anode scrap pressure stripping, and the method comprises the steps: obtaining related historical data, and obtaining a historical preprocessing data set; constructing an initial recognition model, and training the initial recognition model to obtain a recognition model; real-time data of the anode guide rod assembly are collected and preprocessed, and a preprocessed data set is obtained; performing feature extraction on the contour information in the preprocessed data set to obtain a feature data set; the contour information in the feature data set is identified, and the pressing-disengaging type of the anode guide rod assembly is judged according to a preset standard; and according to the recognition result, a catenary system is controlled to adjust the action of the corresponding turnout, and the anode guide rod assembly is arranged to the pressing-disengaging equipment of the corresponding pressing-disengaging type. The problems that in the prior art, in the aluminum electrolysis anode scrap pressure stripping process, the identification of the anode guide rod assembly is inaccurate, the efficiency is low, and pressure stripping equipment and technological parameters are adjusted depending on manual judgment are solved.

Description

Technical Field

[0001] The present invention relates to the field of industrial automation and intelligent identification technology, and in particular to a method and system for identifying an anode guide rod assembly for aluminum electrolysis residual electrode press-off. Background Art

[0002] In the aluminum electrolysis production process, the anode guide rod assembly is an important part of the electrolytic cell. It carries the anode carbon block and conducts electrolysis reaction through electric current. As the electrolysis process proceeds, the anode carbon block will gradually be consumed to form a residual anode. In order to ensure the continuity and efficiency of the electrolysis process, the residual anode needs to be regularly pressed off the anode guide rod assembly and replaced with a new anode carbon block. The traditional anode guide rod assembly residual anode pressing process mainly relies on manual operation, which is not only inefficient but also poses a safety hazard. With the development of automation and intelligent technology, the aluminum electrolysis industry has begun to explore the use of automated equipment to press off the residual anode guide rod assemblies. However, to achieve automated pressing, it is first necessary to accurately identify the type and state of the anode guide rod assembly in order to select appropriate pressing equipment and process parameters.

[0003] In addition, the existing identification methods can only identify the basic type of anode guide rod assembly, but cannot accurately determine its press-off type, that is, it is impossible to determine whether the anode guide rod assembly needs to be pressed off, the difficulty of pressing off, and the applicable pressing off equipment, etc. As a result, in actual applications, automated equipment often needs to adjust the pressing off equipment and process parameters based on manual judgment, thus limiting the efficiency and accuracy of automated pressing off. Summary of the invention

[0004] The purpose of the present invention is to provide a method and system for identifying an anode guide rod assembly for pressing and stripping of aluminum electrolysis scraps, aiming to solve the problems of inaccurate identification of anode guide rod assemblies and low efficiency in the prior art during pressing and stripping of aluminum electrolysis scraps, as well as reliance on manual judgment to adjust pressing and stripping equipment and process parameters.

[0005] The present invention is achieved through the following technical solutions:

[0006] A method for identifying an anode guide rod assembly for aluminum electrolysis residual electrode pressure release, comprising the steps of:

[0007] Acquire relevant historical data of the anode guide rod assembly, construct a historical data set, and preprocess the historical data set to obtain a historical preprocessed data set;

[0008] An initial recognition model is constructed based on a machine learning algorithm, and the initial recognition model is trained using a historical preprocessed data set to obtain a recognition model;

[0009] Collecting real-time image data of the anode guide rod assembly to be pressed off, and preprocessing the collected real-time image data to obtain a preprocessed data set;

[0010] Extracting features from the contour information of the anode guide rod assembly in the preprocessed data set to obtain a feature data set;

[0011] The contour information in the feature data set is identified through the recognition model, and the pressure-release type of the anode guide rod assembly to be pressed off is determined according to the preset pressure-release standard of the anode butts, so as to obtain the recognition result;

[0012] According to the identification results, the catenary system is controlled to adjust the corresponding turnout action, and the anode guide rod assembly to be pressed off the residual pole is discharged to the pressing and stripping equipment of the corresponding pressing and stripping type for residual pole pressing and stripping treatment.

[0013] Optionally, the specific process of obtaining relevant historical data of the anode guide rod assembly and constructing the historical data set is:

[0014] Several sample data of anode guide rod assemblies in aluminum electrolysis production lines are collected, including image data and corresponding dimensional data of anode guide rod assemblies of different specifications, different degrees of wear and different production batches.

[0015] Optionally, the specific process of preprocessing the historical data set to obtain the historical preprocessed data set is:

[0016] The image data in the historical data set is denoised; the denoised image data is grayscaled to convert the color image into a grayscale image; the grayscale image is binarized; the binarized image is morphologically processed to eliminate small noise points or fill small holes in the image; the contour of the morphologically processed image is extracted to obtain the contour information of the anode guide rod assembly, and the contour information is associated with the corresponding size data to form a historical preprocessing data set.

[0017] Optionally, the specific process of constructing the initial recognition model based on the machine learning algorithm and training the initial recognition model through the historical preprocessed data set to obtain the recognition model is:

[0018] Based on the machine learning algorithm, an initial recognition model is constructed according to the recognition requirements of the anode guide rod assembly;

[0019] The contour information of the anode guide rod assembly in the historical preprocessing data set is used as input features, and the corresponding size data is used as labels to train the initial recognition model;

[0020] Through several iterative trainings, until the recognition accuracy of the initial recognition model reaches a preset threshold or no longer improves significantly, the training is stopped to obtain the recognition model.

[0021] Optionally, the specific process of collecting real-time image data of the anode guide rod assembly to be pressed off by the residual anode and preprocessing the collected real-time image data to obtain the preprocessed data set is:

[0022] The image data of the anode guide rod assembly to be pressed off by the image acquisition equipment installed on the aluminum electrolysis production line are captured in real time; the captured real-time image data are denoised; the denoised real-time image data are grayed; the grayscale image is binarized to divide the pixels in the image into foreground and background; the binarized image is morphologically processed; the contour of the morphologically processed image is extracted, and the contour information of the anode guide rod assembly is extracted by using an edge detection algorithm, and the extracted contour information is used as a preprocessing data set.

[0023] Optionally, the specific process of extracting features from the contour information of the anode guide rod assembly in the preprocessed data set to obtain the feature data set is:

[0024] The contour information of the anode guide rod assembly in the preprocessed data set is subjected to geometric feature extraction, texture feature extraction and shape feature extraction. A comprehensive feature data set is constructed through the comprehensive extraction of geometric features, texture features and shape features. The contour geometric parameters of the anode guide rod assembly are obtained through geometric feature extraction. The texture information of the contour edge is analyzed through texture feature extraction. The shape features of the anode guide rod assembly are identified through shape feature extraction.

[0025] Optionally, the specific process of identifying the contour information in the feature data set by using the recognition model and judging the pressure-release type of the anode guide rod assembly to be pressure-released according to the preset pressure-release standard of the anode butts is as follows:

[0026] The feature data set is input into the trained recognition model; the recognition model uses the learned feature mapping relationship to parse and match the input feature data set; the specific specifications of the anode guide rod assembly are identified by comparing the feature data set with the known features stored in the recognition model; based on the preset residual pole press-off standard, the recognition model outputs the corresponding recognition results according to the wear degree or scrap status of the anode guide rod assembly, including the specification type of the anode guide rod assembly, whether it meets the scrap standard, and the recommended type of press-off equipment.

[0027] Optionally, according to the identification result, the catenary system is controlled to adjust the corresponding turnout action, and the anode guide rod assembly to be pressed off is discharged to the pressing off equipment of the corresponding pressing off type, and the specific process of the remaining pole pressing off treatment is:

[0028] The identification result is converted into a control signal, and the control signal is sent to the catenary system; the catenary system automatically adjusts the action of the catenary turnout according to the received control signal, and transports the anode guide rod assembly to the corresponding pressing and stripping equipment along a preset path; the pressing and stripping equipment performs residual pole pressing and stripping treatment on the anode guide rod assembly according to the preset pressing and stripping parameters and procedures.

[0029] The present invention also provides an anode guide rod assembly identification system for aluminum electrolysis residual electrode pressure-off, which is used to implement the above-mentioned anode guide rod assembly identification method for aluminum electrolysis residual electrode pressure-off, comprising:

[0030] An image acquisition module is provided on the aluminum electrolysis production line to capture the image data of the anode guide rod assembly to be pressed off in real time;

[0031] A preprocessing module, used to preprocess the collected real-time image data and extract the contour information of the anode guide rod assembly;

[0032] The feature extraction module is used to extract the geometric features, texture features and shape features of the anode guide rod assembly contour information and construct a comprehensive feature data set;

[0033] The recognition model module is used to build a recognition model based on a machine learning algorithm, and to determine the pressure-off type and scrap status of the anode guide rod assembly according to a preset residual anode pressure-off standard to obtain a recognition result;

[0034] The control module is used to generate a control signal according to the recognition result of the recognition model, and send the signal to the catenary system to control the action of the catenary turnout and transport the anode guide rod assembly to the corresponding pressing and stripping equipment for residual pole pressing and stripping treatment;

[0035] Display module: used to display the measurement results, identification results and the control status of the entire measurement and identification process.

[0036] Optionally, it also includes:

[0037] Calibration and maintenance module, used to regularly calibrate and maintain the performance of each component of the system;

[0038] A light source adjustment module is used to automatically adjust the intensity and color temperature of the light source according to the brightness of the image acquisition environment;

[0039] Safety monitoring module, used to monitor the operating status of the catenary system and the pressure-release equipment in real time;

[0040] The remote control module is used for remote users to view the system's operating status, measurement results, and identification results in real time, and to set parameters and diagnose faults, and to remotely manage and maintain the system.

[0041] The technical solution of the present invention has at least the following advantages and beneficial effects:

[0042] Improve construction efficiency: By accurately obtaining the geographic data information of the construction site and the actual parameters of the construction machinery and vehicles, a driving route that is more in line with the actual situation can be formulated, which not only reduces unnecessary driving time, but also avoids construction delays caused by unreasonable routes, thereby significantly improving the overall construction efficiency.

[0043] Optimize resource utilization: By optimizing the initial driving path, the optimized driving path is obtained, which fully considers the actual conditions of the construction site and the performance characteristics of the construction machinery and vehicles, ensures the feasibility and efficiency of the path, and helps to reasonably allocate construction resources such as fuel, manpower and time, reducing resource waste.

[0044] Enhanced safety: Key nodes are set on the driving optimization path, and corresponding operation guidelines are formulated for each key node, providing drivers with clear driving instructions and operation suggestions. This not only helps drivers better grasp the key points in the driving process, but also effectively avoids safety accidents caused by improper operation, thereby improving safety during the construction process.

[0045] Improve the level of automation: Inputting the optimized driving path and operation guide into the navigation system of the construction machinery vehicle can realize the execution of the automatic driving system or provide reference for the driver. It not only improves the automation level of the construction machinery vehicle, reduces the complexity and difficulty of manual operation, but also provides the driver with more convenient and accurate navigation support.

[0046] Strong adaptability: It fully considers the complexity and variability of the construction site. By acquiring the geographic data information of the construction site and the actual parameters of the construction machinery and vehicles in real time, it can flexibly adjust the driving path and operation guidelines. It has strong adaptability and flexibility and can cope with various complex construction environments and task requirements. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 This is a schematic flow chart of a method for identifying an anode guide rod assembly for aluminum electrolysis residual electrode pressing and falling off according to Example 1 of the present invention;

[0048] Figure 2 This is a schematic diagram of the working principle of the anode guide rod assembly identification system for aluminum electrolysis residual electrode pressure release according to Example 2 of the present invention. DETAILED DESCRIPTION

[0049] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations.

[0050] Example 1

[0051] Reference Figure 1 , a method for identifying an anode guide rod assembly for aluminum electrolysis residual electrode pressure release, comprising the steps of:

[0052] Step 1: Obtain relevant historical data of the anode guide rod assembly, construct a historical data set, and preprocess the historical data set to obtain a historical preprocessed data set.

[0053] In this embodiment, the specific process of obtaining relevant historical data of the anode guide rod assembly and constructing the historical data set is as follows:

[0054] Several sample data of anode guide rod assemblies in aluminum electrolysis production lines are collected, including image data and corresponding dimensional data of anode guide rod assemblies of different specifications, different wear degrees and different production batches. Representative samples of anode guide rod assemblies are selected from aluminum electrolysis production lines. Anode guide rod assemblies include guide rods, explosive blocks and steel claws. The samples of anode guide rod assemblies cover different sample specifications, such as 240KA series guide rods: 130mm×130mm×2278mm; 300KA series guide rods: 150mm×150mm×2278mm; 240KA explosive blocks: 165mm×165mm×50mm; 300KA explosive blocks: 200mm×220mm×52mm; 240KA steel claws: 1040mm×140mm×430mm; 300KA steel claws: 1040mm×160mm×430mm; wear degrees range from slight wear to severe wear; and different production batches to reflect the variability in the production process. Use high-resolution image acquisition equipment (such as industrial cameras) to take pictures of the selected anode guide rod assemblies to ensure that the images are clear and complete, and can accurately reflect the contours and detailed features of the anode guide rod, explosive block and steel claw. For each anode guide rod assembly, use precise measuring tools (such as vernier calipers or laser rangefinders) to measure the initial anode guide rod cross-sectional diameter, explosive block size and steel claw diameter, as well as the corresponding dimensional data after use, and record the corresponding dimensional data. The collected image data and corresponding dimensional data are organized into a structured data set, and each data point contains an image and the corresponding anode guide rod cross-sectional diameter, explosive block size and steel claw diameter electric dimensional information. The image data can be annotated to mark the contours or key points of the anode guide rod, explosive block and steel claw so that the algorithm can learn these features during the training process.

[0055] In this embodiment, the specific process of preprocessing the historical data set to obtain the historical preprocessed data set is as follows:

[0056] The image data in the historical data set is denoised, and image denoising algorithms (such as Gaussian filtering, median filtering, etc.) are used to remove noise in the image and improve image quality; the denoised image data is grayed to convert the color image into a gray image, simplify the image information, and reduce the amount of calculation; the gray image is binarized and threshold segmented to divide the pixels in the image into two categories: foreground (anode guide rod and steel claw) and background to form a binary image; the binarized image is morphologically processed, such as dilation, corrosion, opening operation, closing operation, etc., to eliminate small noise points or fill small holes in the image; the image after morphological processing is contour extracted, and the contour information of the anode guide rod and the steel claw is extracted using an edge detection algorithm (such as Canny edge detection, Sobel edge detection, etc.), and the contour information is associated with the corresponding size data to form a historical preprocessing data set.

[0057] Step 2: Build an initial recognition model based on the machine learning algorithm, and train the initial recognition model through the historical preprocessing data set to obtain the recognition model.

[0058] In this embodiment, an initial recognition model is constructed based on a machine learning algorithm, and the initial recognition model is trained using a historical preprocessed data set. The specific process of obtaining the recognition model is as follows:

[0059] Based on the machine learning algorithm, an initial recognition model is constructed according to the recognition requirements of the anode guide rod assembly; the contour information of the anode guide rod assembly in the historical preprocessing data set is used as the input feature, and the corresponding size data is used as the label to train the initial recognition model; after several iterations of training, until the recognition accuracy of the initial recognition model reaches a preset threshold or no longer significantly improves, the training is stopped to obtain the recognition model. A machine learning algorithm based on image processing can be selected, such as a convolutional neural network (CNN) or a support vector machine (SVM). For example, a convolutional neural network is used to construct an initial recognition model. The structure of the model includes an input layer, a convolutional layer, a pooling layer, a fully connected layer, and an output layer. The input layer receives the preprocessed image data, that is, the contour information of the anode guide rod assembly, the convolutional layer and the pooling layer are used to extract features in the image, the fully connected layer is used to map the extracted features to the output layer, and the output layer outputs the predicted size data or classification results. The training data is extracted from the historical preprocessing data set. The training data includes the contour image and the corresponding size data of the anode guide rod assembly. The contour image is used as an input feature and the size data is used as a label to train the recognition model. The training data is input into the initial recognition model, the output of the model is calculated by forward propagation, and the parameters of the model are adjusted by the back propagation algorithm to minimize the error between the output and the label. The training process requires multiple iterations, and each iteration uses a portion of the training data for training and evaluates the performance of the model. If the performance of the model reaches the preset threshold or no longer improves significantly, the training is stopped. During the training process, a validation data set can be used to evaluate the performance of the model. The validation data set should be independent of the training data set and contain the same type of anode guide rod assembly images and size data. By comparing the prediction results of the model on the validation data set with the true label, the accuracy, robustness and generalization ability of the model can be evaluated. According to the evaluation results, the structure of the model can be optimized, such as increasing the number of convolutional layers, adjusting the size of the convolution kernel, changing the pooling method, etc. The optimized model can be retrained and its performance evaluated again until a satisfactory recognition effect is achieved. After multiple iterations of training and optimization, the final recognition model is obtained. The recognition model can accurately identify the contour information of the anode guide rod assembly and predict its size data or classification results.

[0060] Step 3: collecting real-time image data of the anode guide rod assembly to be pressed off, and preprocessing the collected real-time image data to obtain a preprocessed data set.

[0061] In this embodiment, real-time image data of the anode guide rod assembly to be pressed off is collected, and the collected real-time image data is preprocessed to obtain the specific process of the preprocessed data set:

[0062] The image data of the anode guide rod assembly to be pressed off by the image acquisition equipment installed on the aluminum electrolysis production line are captured in real time; the captured real-time image data are denoised; the denoised real-time image data are grayed; the grayscale image is binarized to divide the pixels in the image into foreground and background; the binarized image is morphologically processed; the contour of the morphologically processed image is extracted, and the contour information of the anode guide rod assembly is extracted by using an edge detection algorithm, and the extracted contour information is used as a preprocessing data set.

[0063] In this embodiment, a high-resolution ultra-clear camera can be installed on the aluminum electrolysis production line, especially in the anode assembly workshop and the residual anode pressing and stripping area, to ensure that the camera can cover all the anode guide rod assemblies to be tested and can capture clear image data. The camera is started to capture the image data of the anode guide rod assembly to be pressed and stripped in real time. The camera should have high sensitivity and fast response capabilities to capture images on the dynamic production line. The collected real-time image data is denoised using image denoising algorithms (such as Gaussian filtering, median filtering, etc.) to remove noise and interference in the image and improve image quality; the color image is converted into a grayscale image to simplify image information and reduce the amount of calculation; the grayscale image is threshold segmented to divide the pixels in the image into two categories: foreground (anode guide rod and steel claw) and background to form a binary image; the binary image is morphologically processed, such as expansion, corrosion, opening operation, closing operation, etc., to eliminate small noise points or fill small holes in the image to make the image clearer and more accurate; the edge detection algorithm (such as Canny edge detection, Sobel edge detection, etc.) is used to extract the contour of the image after morphological processing to extract the contour information of the anode guide rod assembly. The extracted contour information of the anode guide rod assembly is stored as a preprocessing data set for subsequent feature extraction and recognition processing; the preprocessing data set is verified to ensure the accuracy and integrity of the data.

[0064] Step 4: Extract features of the contour information of the anode guide rod assembly in the preprocessed data set to obtain a feature data set.

[0065] In this embodiment, the specific process of extracting features from the contour information of the anode guide rod assembly in the preprocessed data set to obtain the feature data set is as follows:

[0066] The contour information of the anode guide rod assembly in the preprocessed data set is subjected to geometric feature extraction, texture feature extraction and shape feature extraction. A comprehensive feature data set is constructed through the comprehensive extraction of geometric features, texture features and shape features. The contour geometric parameters of the anode guide rod assembly are obtained through geometric feature extraction. The texture information of the contour edge is analyzed through texture feature extraction. The shape features of the anode guide rod assembly are identified through shape feature extraction.

[0067] In this embodiment, the geometric feature extraction can calculate the contour circumference of the anode guide rod assembly and the area enclosed by the contour of the anode guide rod assembly through the collected image data information; the diameter of the anode guide rod and the steel claw can be estimated by fitting a circular or elliptical contour, for example, by mathematical methods such as the least square method; for non-circular parts, such as the non-circular parts of the explosion block and the steel claw, the ratio of its length to width can be calculated to distinguish components of different shapes. Texture feature extraction can analyze the grayscale changes of the contour edge of the anode guide rod assembly through the grayscale generation matrix (GLCM) to extract texture features such as contrast, correlation, energy and homogeneity; by calculating the local binary pattern (LBP) of the contour edge, the texture details are captured; by performing wavelet transform on the contour image, the texture features at different scales are extracted. Shape feature extraction includes calculating the Hu moment of the contour of the anode guide rod assembly; converting the contour information into a frequency domain representation and extracting its Fourier descriptor to capture the global and local features of the shape; using the template matching method, the pre-processed contour is matched with the preset template to identify its shape type. The geometric features, texture features and shape features extracted above are integrated into a feature data set; the feature data set is normalized to ensure that different features have the same weight in the subsequent recognition model; if necessary, the feature data set is reduced in dimension to reduce the amount of calculation and improve the recognition efficiency.

[0068] Step 5: The contour information in the feature data set is identified through the recognition model, and the pressure-release type of the anode guide rod assembly to be pressure-released is determined according to the preset pressure-release standard of the anode butts to obtain the recognition result.

[0069] In this embodiment, the contour information in the feature data set is identified by using the recognition model, and the pressure-release type of the anode guide rod assembly to be pressed off is determined according to the preset butt pressure-release standard. The specific process of obtaining the recognition result is as follows:

[0070] The feature data set is input into the trained recognition model; the recognition model uses the learned feature mapping relationship to parse and match the input feature data set, including but not limited to the analysis of geometric features, texture features and shape features; by comparing the feature data set with the known features stored in the recognition model, the specific specifications of the anode guide rod assembly are identified, including the cross-sectional diameter of the anode guide rod, the size of the explosion block and the diameter of the steel claw; based on the preset residual pole pressing and stripping standard, the recognition model compares the identified specification information with the threshold in the standard according to the wear degree or scrap state of the anode guide rod assembly. If the wear degree of the anode guide rod assembly exceeds the preset threshold or reaches the scrap standard, the recognition model will mark it as a component that needs to be pressed and stripped, and output the corresponding recognition result, including the specification type of the anode guide rod assembly, whether it meets the scrap standard and the recommended type of pressing and stripping equipment. The recognition result will be used to guide the subsequent residual pole pressing and stripping process. Before outputting the result, the recognition result can be verified to ensure its accuracy and reliability, which can be achieved by comparing with known standard data, using cross-validation and other methods; if there is an error or uncertainty in the recognition result, it can be fed back to the recognition model for further optimization and adjustment.

[0071] Step 6: According to the identification result, the catenary system is controlled to adjust the corresponding turnout action, and the anode guide rod assembly to be pressed off the residual pole is discharged to the pressing and stripping equipment of the corresponding pressing and stripping type for residual pole pressing and stripping treatment.

[0072] In this embodiment, according to the recognition result, the catenary system is controlled to adjust the corresponding turnout action, and the anode guide rod assembly to be pressed off is discharged to the pressing off equipment of the corresponding pressing off type. The specific process of the remaining pole pressing off treatment is as follows:

[0073] The identification result is converted into a control signal, and the control signal is sent to the catenary system; the catenary system automatically adjusts the action of the catenary turnout according to the received control signal, and transports the anode guide rod assembly to the corresponding pressing and stripping equipment along a preset path; the pressing and stripping equipment performs residual pole pressing and stripping treatment on the anode guide rod assembly according to the preset pressing and stripping parameters and procedures.

[0074] In this embodiment, the system receives the recognition result from the recognition model, which includes the specification type of the anode guide rod assembly, whether it reaches the scrap standard, and the recommended type of pressing and stripping equipment; according to the recognition result, the system automatically generates a corresponding control signal, which needs to include the identification information of the anode guide rod assembly (such as number or position information), the identification of the target pressing and stripping equipment, and any necessary operation instructions, such as conveying speed, turnout steering, etc. The control signal is sent to the control unit of the catenary system through a predefined communication interface (such as Ethernet, wireless LAN, etc.). After receiving the control signal, the control unit of the catenary system verifies the signal integrity and correctness to ensure that the control signal is accurate. The catenary system monitors the current state of each turnout in real time to ensure that the turnout is in an operable state. According to the instructions in the control signal, the catenary system adjusts the corresponding turnout and changes the conveying path of the anode guide rod assembly, including the steering and locking of the turnout, so as to ensure that the anode guide rod assembly can accurately enter the conveying line of the target pressing and stripping equipment. When the anode guide rod assembly approaches the target pressing and stripping equipment, the system confirms its position through sensors or visual recognition technology to ensure that the anode guide rod assembly can accurately park at the designated position of the pressing and stripping equipment. The pressing and stripping equipment automatically sets the corresponding pressing and stripping parameters, such as pressure, time, etc., according to the pressing and stripping equipment type and the specification information of the anode guide rod assembly in the identification result. The pressing and stripping equipment performs residual pole pressing and stripping treatment on the anode guide rod assembly according to the preset pressing and stripping parameters and procedures, including the steps of clamping the anode guide rod assembly, applying pressure, and separating the residual poles. After the pressing and stripping equipment completes the pressing and stripping operation, it will feedback the processing results (such as whether the pressing and stripping is successful, the state of the anode guide rod assembly after pressing and stripping, etc.) to the system. During the whole process, the system monitors the conveying status of the anode guide rod assembly, the turnout status, and the operating status of the pressing and stripping equipment in real time through sensors and monitoring equipment. Once an abnormal situation is found (such as turnout failure, conveyor belt failure, pressing and stripping equipment failure, etc.), the system immediately starts the abnormal handling process, including stopping transportation, alarm prompts, starting backup equipment, etc.

[0075] Example 2

[0076] Based on Example 1, refer to Figure 2 In this embodiment, a system for identifying an anode guide rod assembly for aluminum electrolysis waste anode pressure-off is provided, which is used to implement the anode guide rod assembly identification method for aluminum electrolysis waste anode pressure-off, including:

[0077] An image acquisition module is provided on the aluminum electrolysis production line to capture the image data of the anode guide rod assembly to be pressed off in real time;

[0078] A preprocessing module, used to preprocess the collected real-time image data and extract the contour information of the anode guide rod assembly;

[0079] The feature extraction module is used to extract the geometric features, texture features and shape features of the anode guide rod assembly contour information and construct a comprehensive feature data set;

[0080] The recognition model module is used to build a recognition model based on a machine learning algorithm, and to determine the pressure-off type and scrap status of the anode guide rod assembly according to a preset residual anode pressure-off standard to obtain a recognition result;

[0081] The control module is used to generate a control signal according to the recognition result of the recognition model, and send the signal to the catenary system to control the action of the catenary turnout and transport the anode guide rod assembly to the corresponding pressing and stripping equipment for residual pole pressing and stripping treatment;

[0082] Display module: used to display the measurement results, identification results and the control status of the entire measurement and identification process.

[0083] In this embodiment, it also includes:

[0084] Calibration and maintenance module, used to regularly calibrate and maintain the performance of each component of the system;

[0085] A light source adjustment module is used to automatically adjust the intensity and color temperature of the light source according to the brightness of the image acquisition environment;

[0086] Safety monitoring module, used to monitor the operating status of the catenary system and the pressure-release equipment in real time;

[0087] The remote control module is used for remote users to view the system's operating status, measurement results, and identification results in real time, and to set parameters and diagnose faults, and to remotely manage and maintain the system.

[0088] In this embodiment, the image acquisition module is arranged on the aluminum electrolysis production line, especially in the anode assembly workshop and the residual pole pressing and stripping area, for real-time capturing the image data of the anode guide rod assembly to be pressed and stripped by the residual pole. The preprocessing module receives the real-time image data transmitted by the image acquisition module, and performs preprocessing steps such as denoising, graying, binarization, morphological processing and contour extraction to extract the contour information of the anode guide rod assembly. The feature extraction module extracts the geometric features, texture features and shape features of the contour information of the anode guide rod assembly extracted by the preprocessing module; the geometric features include the contour perimeter, area, etc.; the texture features are extracted by grayscale generation matrix, local binary pattern and other methods; the shape features are obtained by calculating Hu moment, Fourier descriptor and other methods. The recognition model module constructs a recognition model based on the machine learning algorithm and uses the historical preprocessing data set for training; the trained recognition model receives the feature data set transmitted by the feature extraction module, performs parsing and matching, and identifies the specific specifications of the anode guide rod assembly; according to the preset residual pole pressing and stripping standard, the recognition model determines the pressing and stripping type and scrapping status of the anode guide rod assembly, and outputs the recognition result. The control module receives the recognition result of the recognition model module and generates a corresponding control signal; the control signal is sent to the catenary system, and the catenary system adjusts the turnout action according to the signal, such as controlling the turnout to go straight or bend, that is, controlling the transportation path of the turnout, and transporting the anode guide rod assembly to the corresponding pressing device along the preset path; the pressing device performs residual pole pressing and stripping treatment on the anode guide rod assembly according to the preset pressing and stripping parameters and procedures. The display module is used to display the measurement results, recognition results and the control status of the entire measurement and recognition process, so that the operator can understand the system status in real time; the safety monitoring module monitors the operating status of the catenary system and the pressing and stripping equipment in real time to ensure the safe operation of the system; once an abnormal situation is found, the system immediately starts the abnormal handling process, including stopping transportation, alarm prompts, starting backup equipment, etc. The calibration and maintenance module regularly calibrates and maintains the performance of each component of the system to ensure the long-term stable operation of the system; the light source adjustment module automatically adjusts the intensity and color temperature of the light source according to the brightness of the image acquisition environment to optimize the image acquisition effect. The remote control module allows remote users to view the system's operating status, measurement results, and identification results in real time, and to set parameters and diagnose faults; remote users can remotely manage and maintain the system to improve the flexibility and reliability of the system.

Claims

1. A method for identifying an anode guide rod assembly for aluminum electrolysis residual electrode pressure release, characterized in that: Includes steps: Acquire relevant historical data of the anode guide rod assembly, construct a historical data set, and preprocess the historical data set to obtain a historical preprocessed data set; An initial recognition model is constructed based on a machine learning algorithm, and the initial recognition model is trained using a historical preprocessed data set to obtain a recognition model; Collecting real-time image data of the anode guide rod assembly to be pressed off, and preprocessing the collected real-time image data to obtain a preprocessed data set; Extracting features from the contour information of the anode guide rod assembly in the preprocessed data set to obtain a feature data set; The contour information in the feature data set is identified by the recognition model, and the pressure-release type of the anode guide rod assembly to be pressed off is determined according to the preset pressure-release standard of the anode butts, so as to obtain the recognition result; According to the identification results, the catenary system is controlled to adjust the corresponding turnout action, and the anode guide rod assembly to be pressed off is discharged to the pressing equipment of the corresponding pressing type for the residual pole pressing off treatment.

2. The method for identifying anode guide rod components for aluminum electrolysis residual electrode pressure-off according to claim 1, characterized in that: The specific process of obtaining the relevant historical data of the anode guide rod assembly and constructing the historical data set is as follows: Several sample data of anode guide rod assemblies in aluminum electrolysis production lines are collected, including image data and corresponding dimensional data of anode guide rod assemblies of different specifications, different degrees of wear and different production batches.

3. The method for identifying anode guide rod components for aluminum electrolysis residual electrode pressure-off according to claim 2, characterized in that: The specific process of preprocessing the historical data set to obtain the historical preprocessed data set is as follows: The image data in the historical data set is denoised; the denoised image data is grayscaled to convert the color image into a grayscale image; the grayscale image is binarized; the binarized image is morphologically processed to eliminate small noise points or fill small holes in the image; the contour of the morphologically processed image is extracted to obtain the contour information of the anode guide rod assembly, and the contour information is associated with the corresponding size data to form a historical preprocessing data set.

4. The method for identifying anode guide rod components for aluminum electrolysis residual electrode pressure-off according to claim 1, characterized in that: The specific process of constructing the initial recognition model based on the machine learning algorithm and training the initial recognition model through the historical preprocessing data set to obtain the recognition model is as follows: Based on the machine learning algorithm, an initial recognition model is constructed according to the recognition requirements of the anode guide rod assembly; The contour information of the anode guide rod assembly in the historical preprocessing data set is used as input features, and the corresponding size data is used as labels to train the initial recognition model; Through several iterative trainings, until the recognition accuracy of the initial recognition model reaches a preset threshold or no longer improves significantly, the training is stopped to obtain the recognition model.

5. The method for identifying anode guide rod components for aluminum electrolysis residual electrode pressure-off according to claim 1, characterized in that: The specific process of collecting the real-time image data of the anode guide rod assembly to be pressed off and preprocessing the collected real-time image data to obtain the preprocessed data set is as follows: The image data of the anode guide rod assembly to be pressed off by the image acquisition equipment installed on the aluminum electrolysis production line are captured in real time; the captured real-time image data is de-noised; Perform grayscale processing on the real-time image data after denoising; perform binarization processing on the grayscale image to divide the pixels in the image into foreground and background; Perform morphological processing on the binarized image; The contour of the image after morphological processing is extracted, and the edge detection algorithm is used to extract the contour information of the anode guide rod assembly, and the extracted contour information is used as the preprocessing data set.

6. The method for identifying anode guide rod components for aluminum electrolysis residual electrode pressure-off according to claim 1, characterized in that: The specific process of extracting features from the contour information of the anode guide rod assembly in the preprocessed data set to obtain the feature data set is as follows: The contour information of the anode guide rod assembly in the preprocessed data set is subjected to geometric feature extraction, texture feature extraction and shape feature extraction. A comprehensive feature data set is constructed through the comprehensive extraction of geometric features, texture features and shape features. The contour geometric parameters of the anode guide rod assembly are obtained through geometric feature extraction. The texture information of the contour edge is analyzed through texture feature extraction. The shape features of the anode guide rod assembly are identified through shape feature extraction.

7. The method for identifying anode guide rod components for aluminum electrolysis residual electrode pressure release according to claim 1, characterized in that: The specific process of identifying the contour information in the feature data set by using the recognition model and judging the pressure-release type of the anode guide rod assembly to be pressure-released according to the preset pressure-release standard of the anode butts is as follows: The feature data set is input into the trained recognition model; the recognition model uses the learned feature mapping relationship to parse and match the input feature data set; the specific specifications of the anode guide rod assembly are identified by comparing the feature data set with the known features stored in the recognition model; based on the preset residual pole press-off standard, the recognition model outputs the corresponding recognition results according to the wear degree or scrap status of the anode guide rod assembly, including the specification type of the anode guide rod assembly, whether it meets the scrap standard, and the recommended type of press-off equipment.

8. The method for identifying anode guide rod components for aluminum electrolysis residual electrode pressure release according to claim 1, characterized in that: According to the identification result, the catenary system is controlled to adjust the corresponding turnout action, and the anode guide rod assembly to be pressed off is discharged to the pressing off equipment of the corresponding pressing off type. The specific process of the residual pole pressing off treatment is as follows: The identification result is converted into a control signal, and the control signal is sent to the catenary system; the catenary system automatically adjusts the action of the catenary turnout according to the received control signal, and transports the anode guide rod assembly to the corresponding pressing and stripping equipment along a preset path; the pressing and stripping equipment performs residual pole pressing and stripping treatment on the anode guide rod assembly according to the preset pressing and stripping parameters and procedures.

9. An anode guide rod assembly identification system for aluminum electrolysis waste anode pressure-off, used to implement the anode guide rod assembly identification method for aluminum electrolysis waste anode pressure-off as claimed in any one of claims 1 to 8, characterized in that: include: An image acquisition module is provided on the aluminum electrolysis production line to capture the image data of the anode guide rod assembly to be pressed off in real time; A preprocessing module, used to preprocess the collected real-time image data and extract the contour information of the anode guide rod assembly; The feature extraction module is used to extract the geometric features, texture features and shape features of the anode guide rod assembly contour information and construct a comprehensive feature data set; The recognition model module is used to build a recognition model based on a machine learning algorithm, and to determine the pressure-off type and scrap status of the anode guide rod assembly according to a preset residual anode pressure-off standard to obtain a recognition result; The control module is used to generate a control signal according to the recognition result of the recognition model, and send the signal to the catenary system to control the action of the catenary turnout and transport the anode guide rod assembly to the corresponding pressing and stripping equipment for residual pole pressing and stripping treatment; Display module: used to display the measurement results, identification results and the control status of the entire measurement and identification process.

10. The anode guide rod assembly identification system for aluminum electrolysis residual electrode pressure release according to claim 9, characterized in that: Also includes: Calibration and maintenance module, used to regularly calibrate and maintain the performance of each component of the system; A light source adjustment module is used to automatically adjust the intensity and color temperature of the light source according to the brightness of the image acquisition environment; Safety monitoring module, used to monitor the operating status of the catenary system and the pressure-release equipment in real time; The remote control module is used for remote users to view the system's operating status, measurement results, and identification results in real time, and to set parameters and diagnose faults, and to remotely manage and maintain the system.

Images