Zero-tailing cutting method, device and equipment of pipe cutting machine and storage medium

Abstract

The invention relates to the technical field of pipe cutting, in particular to a zero-tailing cutting method, device and equipment of a pipe cutting machine and a storage medium, and the method comprises the steps that parameter information of a to-be-machined pipe is obtained and input into a pre-constructed critical value prediction model, and a tailing critical value is obtained; cutting operation of the pipe to be machined is executed, and the remaining length is detected in real time; when the remaining length is smaller than or equal to the tailing critical value, real-time three-dimensional data of the to-be-machined pipe is obtained to generate a reverse cutting path used for guiding zero-tailing cutting operation; according to the method disclosed by the invention, dynamic prediction of the tailing critical value is realized through the critical value prediction model, so that extensive management of fixed tailing length in a traditional method is avoided, the trigger point can be dynamically adjusted according to the characteristics of different pipes, and the utilization rate of the tailings is remarkably improved; and when it is detected that the remaining length is smaller than or equal to the tailing critical value, reverse cutting logic is automatically switched to, the continuity of the cutting process is ensured, delay caused by manual intervention is reduced, and the production efficiency is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of pipe cutting, in particular to a pipe cutting machine zero tail cutting method, device, equipment and storage medium. BACKGROUND

[0002] In the fields of mechanical manufacturing, building materials, aerospace, etc., pipe cutting is an indispensable processing step. Traditional pipe cutting machines usually use fixed chucks to clamp the pipe and feed the cutting tool along the preset path to complete the cutting task. However, this traditional technology has exposed several significant limitations in practical application.

[0003] Firstly, due to the limitation of the fixed clamping position of the traditional pipe cutting machine, when the remaining length of the pipe is less than the minimum clamping distance of the chuck, the pipe cannot be clamped stably, resulting in the problem of fixed tail, i.e. the end part of the pipe cannot be effectively utilized. For materials with high cost, such as stainless steel and titanium alloy, this loss is particularly significant. Although some existing pipe cutting machines attempt to cut the tail by moving a single chuck, due to the lack of coordinated control of multiple chucks and precise path planning, it is difficult to eliminate the tail problem while ensuring cutting accuracy.

[0004] Secondly, the function of the traditional control system is limited to executing the preset program, and cannot respond to the changes in the remaining length of the pipe and the dynamic clamping requirements in real time, which makes it difficult to realize full-process automatic zero tail cutting. In addition, traditional pipe cutting machines are usually equipped with 2 to 3 fixed chucks, which are fixed in position and lack flexibility, and cannot dynamically adjust the posture of the pipe according to the changes in the cutting path. When cutting the end remaining short material, the single chuck clamping stability is insufficient, which can easily cause pipe vibration, resulting in a decrease in cutting quality or even cutting failure. This structural defect directly limits the feasibility of tail cutting and becomes one of the key factors that existing technology cannot break through the zero tail cutting bottleneck.

[0005] It can be seen that the existing technology needs to be improved and improved. SUMMARY

[0006] In order to overcome the shortcomings of the prior art, the purpose of the present application is to provide a pipe cutting machine zero tail cutting method, which aims to solve the problems of serious tail waste, insufficient cutting accuracy, poor adaptability, etc. of traditional pipe cutting machines, and realize the efficiency, accuracy and intelligence of pipe cutting.

[0007] A first aspect of the present invention provides a zero-tail cutting method for a pipe cutting machine, comprising: pre-constructing a critical value prediction model; obtaining parameter information of a pipe to be processed and inputting the pre-constructed critical value prediction model to obtain a tail critical value; performing a cutting operation on the pipe to be processed, and detecting the remaining length of the pipe to be processed in real time; when the remaining length of the pipe to be processed is ≤ the tail critical value, obtaining real-time three-dimensional data of the pipe to be processed; generating a reverse cutting path based on the real-time three-dimensional data of the pipe to be processed, and performing zero-tail cutting on the pipe to be processed based on the generated reverse cutting path.

[0008] Optionally, in a first implementation method of the first aspect of the present invention, the pre-constructed critical value prediction model includes: obtaining historical cutting data, the historical cutting data including historical cutting information of multiple pipes, the historical cutting information including basic pipe parameters, cutting process parameters and tail material related data; performing data cleaning and data enhancement processing on the historical cutting data to obtain pre-processed historical cutting data; pre-constructing a hybrid model, inputting the pre-processed historical cutting data into the pre-constructed hybrid model, and training the pre-constructed hybrid model in combination with the transfer learning method and the early stopping mechanism to obtain a critical value prediction model.

[0009] Optionally, in a second implementation of the first aspect of the present invention, the pre-constructed hybrid model includes a feature engineering layer, a fully connected layer, a material classification sub-model, a geometric parameter sub-model, a feature fusion layer and a multi-layer perceptron, the feature engineering layer is used to perform feature conversion processing on the pre-processed historical cutting data, the output end of the feature engineering layer is connected to the input end of the fully connected layer, and the output end of the fully connected layer is respectively connected to the input end of the material classification sub-model and the input end of the geometric parameter sub-model; the material classification sub-model is used to output the material feature vector to the feature fusion layer, the geometric parameter sub-model is used to output the geometric feature vector to the feature fusion layer, the output end of the feature fusion layer is connected to the multi-layer perceptron, and the multi-layer perceptron is used to output the tail material critical value.

[0010] Optionally, in a third implementation method of the first aspect of the present invention, the method of obtaining parameter information of the pipe to be processed and inputting it into a pre-built critical value prediction model to obtain the tailing critical value includes: obtaining parameter information of the pipe to be processed, the parameter information including material parameters and specification parameters of the pipe to be processed; inputting the obtained parameter information into the pre-built critical value prediction model to obtain the tailing critical value corresponding to the pipe to be processed.

[0011] Optionally, in a fourth implementation of the first aspect of the present invention, the cutting operation is performed on the pipe to be processed and the remaining length of the pipe to be processed is detected in real time, including: obtaining cutting requirements and obtaining a three-dimensional model of the pipe to be processed, the cutting requirements including the incision position, incision shape and cutting angle; taking the front end of the pipe to be processed as the starting point, and using the A* algorithm to generate a forward cutting path based on the parameter information, cutting requirements and three-dimensional model of the pipe to be processed; cutting the pipe to be processed based on the generated forward cutting path, and detecting the remaining length of the pipe to be processed in real time.

[0012] Optionally, in a fifth implementation of the first aspect of the present invention, generating a reverse cutting path based on the real-time three-dimensional data of the pipe to be processed includes: processing the real-time three-dimensional data of the pipe to be processed, and performing three-dimensional modeling based on the processed real-time three-dimensional data to obtain a real-time three-dimensional model of the pipe to be processed; confirming the shape of the pipe to be processed based on the real-time three-dimensional model, wherein the shape includes a cylindrical shape and a straight tube shape; when the pipe to be processed is cylindrical, taking the end origin of the pipe to be processed as the starting point, generating a spiral basic trajectory along the X-axis direction of the pipe according to cutting requirements; when the pipe to be processed is a straight tube shape, taking the end of the pipe to be processed as the starting point, generating a linear interpolation trajectory using a cubic spline interpolation algorithm;

[0013] The helical basic trajectory or linear interpolation trajectory is optimized to obtain the reverse cutting path.

[0014] Optionally, in a sixth implementation of the first aspect of the present invention, the spiral basic trajectory or the linear interpolation trajectory is optimized to obtain a reverse cutting path, including: pre-constructing a weighted multi-objective function including material utilization, cutting time and incision roughness, wherein the weight of material utilization, the weight of cutting time and the weight of incision roughness are determined by a hierarchical analysis method; obtaining the performance parameters of the pipe cutting machine, and setting constraints based on the performance parameters of the pipe cutting machine; and optimizing the spiral basic trajectory or the linear interpolation trajectory using a genetic algorithm based on the weighted multi-objective function and the constraints to obtain a reverse cutting path.

[0015] The second aspect of the present invention provides a zero-tail cutting device for a pipe cutting machine, comprising: a construction module for pre-constructing a critical value prediction model; a prediction module for acquiring parameter information of a pipe to be processed and inputting the pre-constructed critical value prediction model to obtain a tail critical value; a detection module for performing a cutting operation on the pipe to be processed and detecting the remaining length of the pipe to be processed in real time; an acquisition module for acquiring real-time three-dimensional data of the pipe to be processed when the remaining length of the pipe to be processed is ≤ the tail critical value; and a generation module for generating a reverse cutting path based on the real-time three-dimensional data of the pipe to be processed, and performing zero-tail cutting on the pipe to be processed based on the generated reverse cutting path.

[0016] The third aspect of the present invention provides a zero-tail cutting device for a pipe cutting machine, which includes: a memory and at least one processor, wherein instructions are stored in the memory; at least one processor calls the instructions in the memory to enable the zero-tail cutting device for the pipe cutting machine to perform each step of the zero-tail cutting method for the pipe cutting machine described in any one of the above items.

[0017] A fourth aspect of the present invention provides a computer-readable storage medium having instructions stored thereon, and when the instructions are executed by a processor, the various steps of the zero-tail cutting method for a pipe cutting machine described above are implemented.

[0018] In the technical solution of the present invention, a critical value prediction model is used to dynamically predict the critical value of tailings, thereby avoiding the extensive management of fixed tailing lengths in traditional methods. The trigger point can be dynamically adjusted according to the characteristics of different pipes, significantly improving the utilization rate of tailings. When it is detected that the remaining length is ≤ the critical value of tailings, it automatically switches to the reverse cutting logic to ensure the continuity of the cutting process and reduce delays caused by manual intervention. In addition, this automated process reduces the risk of manual misjudgment, thereby improving overall production efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 A logic flow chart of a zero-tail cutting method for a pipe cutting machine provided in an embodiment of the present invention;

[0020] Figure 2 A schematic structural diagram of a zero-tail cutting device for a pipe cutting machine provided in an embodiment of the present invention;

[0021] Figure 3 A schematic structural diagram of the zero-tail cutting device of the pipe cutting machine provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0022] The present invention provides a zero-tail cutting method, device, equipment and storage medium for a pipe cutting machine. In the present invention, the terms "first", "second", "third", "fourth", etc. (if any) in the description and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "including" or "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or are inherent to these processes, methods, products or devices.

[0023] The present application discloses a zero-tail cutting method. The pipe cutting machine that can realize the cutting method adopts a modular design, including four core modules: a pipe clamping system, a cutting execution system, an intelligent detection system and a motion control system. Each module system realizes data interaction through industrial Ethernet; wherein, the pipe clamping system includes a four-chuck free mechanism and a floating support mechanism. The four-chuck free mechanism includes four chucks, each chuck realizes free movement along the axial X-axis of the pipe through a linear guide and a servo electric cylinder, and is equipped with a rotary servo motor to realize 360° circumferential rotation. The four chucks are divided into two groups, the front chuck and the first middle chuck are the two chucks at the feed end, and the second middle chuck and the rear chuck are the two chucks at the discharge end, which move in coordination through a master-slave control mode: the chuck at the feed end fixes the initial position of the pipe, and the chuck at the discharge end is dynamically adjusted with the cutting process to ensure that the pipe is always in a tensioned state; the floating support mechanism is an elastic support bracket arranged between the two chucks, which adopts air bearing technology to automatically adapt to the bending deformation of the pipe and reduce vibration interference during cutting; the cutting execution system includes a cutting head assembly, the cutting The head assembly includes a cutting unit and a five-axis motion platform; the cutting unit integrates a fiber laser cutting head and a high-speed mechanical cutting tool holder, and adapts to the cutting requirements of different materials through an electric switching mechanism; the five-axis motion platform adopts an XYZ linear axis combined with an AB rotary axis structure, with an X-axis travel of 2000mm, a Y / Z axis travel of 300mm, and an A / B axis rotation range of ±90°, which can support bevel cutting at any angle; the intelligent detection system includes a laser length gauge, a three-dimensional line laser scanner and an inertial measurement unit. The laser length gauge is installed on the outside of the front chuck and uses the triangulation principle to monitor the total length and remaining length changes of the pipe in real time; the three-dimensional line laser scanner is arranged at the front end of the cutting head to generate point cloud data and reconstruct the three-dimensional model of the pipe (including parameters such as diameter, curvature, surface defects, etc.) in real time through a point cloud fitting algorithm; the inertial measurement unit is installed on the rear chuck to detect the pipe attitude angle in real time and compensate for position deviation caused by pipe vibration or deformation; the motion control system includes a main controller and a servo drive. The model of the main controller can be Siemens S7-1500T PLC, the servo drive can be Panasonic MINAS A6 series.

[0024] Furthermore, the present application discloses a zero tail material cutting method of a pipe cutting machine. For ease of understanding, the specific process of the embodiment of the present invention is described below. Figure 1 In one embodiment of the present invention, a method for cutting a pipe with zero tailings by a pipe cutting machine includes:

[0025] 101. Pre-build critical value prediction model;

[0026] In this embodiment, when the pre-built critical value prediction model is deployed, the critical value prediction model can be updated online; for example, when a certain amount of new data is accumulated, such as 500 groups, incremental training is automatically triggered to improve the adaptability of the critical value prediction model to new materials or new processes; in addition, the knowledge distillation technology can be used to compress the critical value prediction model into a lightweight version, which can shorten the inference delay time to meet the real-time production demand.

[0027] 102. Obtain the parameter information of the pipe to be processed and input it into the pre-built critical value prediction model to obtain the tailing critical value;

[0028] 103. Perform cutting operation on the pipe to be processed and real-time detect the remaining length of the pipe to be processed;

[0029] 104. When the remaining length of the pipe to be processed is less than or equal to the tailing critical value, obtain the real-time three-dimensional data of the pipe to be processed;

[0030] In this embodiment, a laser length measuring instrument installed on the outside of the front chuck can be used to real-time monitor the total length and the change of the remaining length of the pipe to be processed, and the detection frequency of the laser length measuring instrument is greater than or equal to 100 Hz, which can ensure the real-time data.

[0031] 105. Generate a reverse cutting path based on the real-time three-dimensional data of the pipe to be processed, and perform zero tailing cutting on the pipe to be processed based on the generated reverse cutting path.

[0032] The application discloses a zero tailing cutting method of a pipe cutting machine, which realizes dynamic prediction of the tailing critical value through a critical value prediction model, thereby avoiding the extensive management of fixed tailing length in the traditional method, dynamically adjusting the trigger point according to the characteristics of different pipes, and significantly improving the utilization rate of tailing; when the remaining length is detected to be less than or equal to the tailing critical value, the reverse cutting logic is automatically switched to, ensuring the continuity of the cutting process and reducing the delay caused by manual intervention; in addition, the automatic process reduces the risk of manual misjudgment, thereby improving the overall production efficiency.

[0033] Further, in the embodiment of the application, the pre-built critical value prediction model comprises:

[0034] 201. Obtain historical cutting data, wherein the historical cutting data comprises historical cutting information of a plurality of pipes, and the historical cutting information comprises pipe basic parameters, cutting process parameters and tailing related data;

[0035] In the embodiment, the historical cutting data covers pipe base parameters (material, diameter, wall thickness, length, hardness, etc.), cutting process parameters (power, speed, cutting angle, cutting type, etc.), tail-related data (actual tail length, tail availability, cut quality grade), and environmental parameters (temperature, humidity); the historical cutting data sources include factory historical cutting logs, real-time sensor collected data, and manually labeled samples, with a single material sample amount of ≥5000 groups to ensure the sufficiency of the critical value prediction model training.

[0036] 202. The historical cutting data is subjected to data cleaning and data enhancement processing to obtain preprocessed historical cutting data;

[0037] In the embodiment, the data cleaning processing includes outlier processing and missing value filling; specifically, the IQR quartile method is used to remove tail length abnormal data, and 95% of normal samples are retained; the KNN interpolation method is used to fill numerical value type parameters, and the Mode is used to fill category type parameters.

[0038] In the embodiment, for small sample materials such as titanium alloy, the SMOTE oversampling method is used to generate synthetic data, and for data distribution imbalance problems such as stainless steel samples accounting for 70% and aluminum alloy accounting for 20%, the data enhancement processing is realized through category weight adjustment and optimization.

[0039] 203. A hybrid model is pre-built, the preprocessed historical cutting data is input into the pre-built hybrid model, the pre-built hybrid model is trained in combination with a transfer learning method and an early stopping mechanism, and a critical value prediction model is obtained;

[0040] In the embodiment, through transfer learning, only 5-20 groups of new material samples are needed to complete model fine-tuning, that is, 200 epochs of training on a general pipe data set and 50 epochs of fine-tuning for specific factory data are needed to reduce the dependence on single factory data; for the early stopping mechanism, the training is terminated when the validation set loss does not decrease for 10 consecutive epochs, avoiding overfitting and improving the model generalization ability.

[0041] In the embodiment, by constructing the critical value prediction model, the tail critical value generated in the production process of various different materials and different specifications of pipes can be accurately predicted, overcoming the limitations of traditional empirical formulas in dealing with complex materials and the difficulty of adapting to changes; and the critical value prediction model is a hybrid model including a material classification sub-model and a geometric parameter sub-model, which can separate the influence of material characteristics and geometric parameters, avoid the fitting deviation of a single model for complex characteristics, and improve the prediction accuracy.

[0042] In this embodiment, the pre-built hybrid model includes a feature engineering layer, a fully connected layer, a material classification sub-model, a geometric parameter sub-model, a feature fusion layer and a multi-layer perceptron;

[0043] The feature engineering layer is used to perform feature conversion processing on the pre-processed historical cutting data. Specifically, the feature engineering layer is used to standardize continuous parameters (such as diameter and wall thickness), perform One-Hot encoding on categorical parameters (such as material), and construct combined features (such as diameter-to-thickness ratio D / t and length-to-diameter ratio L / D) to mine hidden associations between parameters; the output end of the feature engineering layer is connected to the input end of the fully connected layer, and the output end of the fully connected layer is respectively connected to the input end of the material classification sub-model and the input end of the geometric parameter sub-model;

[0044] The material classification sub-model is used to output material feature vectors to the feature fusion layer. The material classification sub-model uses an 8-layer convolutional neural network, each layer containing a 3×3 convolution kernel and a MaxPooling layer. It can extract hierarchical features of materials (such as "metal / non-metal", "stainless steel / carbon steel", "hardness grade") and output a 128-dimensional material feature vector;

[0045] The geometric parameter sub-model is used to output a geometric feature vector to the feature fusion layer. The geometric parameter sub-model captures the long-distance dependency between geometric parameters (such as "the longer the length and the smaller the diameter, the longer the tail material needs to be") through the multi-head attention mechanism of the Transformer encoder, and outputs a 128-dimensional geometric feature vector containing parameter associations;

[0046] The feature fusion layer concatenates the material feature vector and the geometric feature vector into a 256-dimensional vector and outputs it to the multilayer perceptron. The concatenated vector undergoes a three-layer MLP nonlinear transformation in the multilayer perceptron and outputs a tail critical value. In the multilayer perceptron, the first layer includes 512 neurons for coarse screening of key features, the second layer includes 256 neurons for refined analysis of features, and the third layer includes 128 neurons for final integration and outputs a specific numerical value, namely the tail critical value. In the multilayer perceptron, the activation function uses LeakyReLU to retain negative value features (such as negative tolerance compensation).

[0047] Furthermore, in an embodiment of the present invention, obtaining parameter information of the pipe to be processed and inputting it into a pre-built critical value prediction model to obtain the tailing critical value includes:

[0048] 301. Obtain parameter information of the pipe to be processed, wherein the parameter information includes material parameters and specification parameters of the pipe to be processed;

[0049] In this embodiment, the material information can be input by scanning a QR code on a pipe label, or can be manually selected by an operator; the specification parameters can be obtained from the design drawings of the pipe to be processed.

[0050] 302. Input the acquired parameter information into a pre-built critical value prediction model to obtain a tailing critical value corresponding to the pipe to be processed.

[0051] Furthermore, in an embodiment of the present invention, the cutting operation on the pipe to be processed and the real-time detection of the remaining length of the pipe to be processed include:

[0052] 401. Obtain cutting requirements and a three-dimensional model of the pipe to be processed, wherein the cutting requirements include a cutting position, a cutting shape, and a cutting angle;

[0053] In this embodiment, the three-dimensional model of the pipe to be processed can be obtained through the design drawings, or point cloud data can be generated by a three-dimensional line laser scanner arranged at the front end of the cutting head. Based on the point cloud data, a point cloud fitting algorithm is used to reconstruct the three-dimensional model of the pipe to be processed; by obtaining the three-dimensional model of the pipe to be processed, defect parameters such as the curvature and ovality of the pipe can be identified, so that the forward cutting path can avoid defects on the surface of the pipe, such as pits, cracks, etc., thereby avoiding quality abnormalities during the cutting process.

[0054] 402. Taking the front end of the pipe to be processed as the starting point, based on the parameter information, cutting requirements and three-dimensional model of the pipe to be processed, an A* algorithm is used to generate a forward cutting path;

[0055] In this embodiment, the front end of the pipe to be processed is used as the starting point, and the optimization goal is to pursue the shortest cutting path and the smallest chuck interference, while taking into account the kinematic constraints of the cutting head, such as the maximum acceleration limit of 10m / s. 2 , and then generate a smooth forward cutting trajectory; using the A* algorithm, it can efficiently calculate the optimal forward cutting path within 100 milliseconds, significantly improving cutting efficiency and enhancing the smoothness of the path, thereby effectively reducing the wear of the cutting head during the cutting process.

[0056] 403. Cut the pipe to be processed based on the generated forward cutting path, and detect the remaining length of the pipe to be processed in real time.

[0057] Furthermore, in an embodiment of the present invention, the generating of the reverse cutting path based on the real-time three-dimensional data of the pipe to be processed includes:

[0058] 501. Processing the real-time three-dimensional data of the pipe to be processed, and performing three-dimensional modeling based on the processed real-time three-dimensional data to obtain a real-time three-dimensional model of the pipe to be processed;

[0059] In this embodiment, the real-time three-dimensional data is real-time point cloud data fed back by a three-dimensional line laser scanner arranged at the front end of the cutting head. Point cloud registration technology is used to align the real-time point cloud data with the three-dimensional model of the pipe to be processed to update the end position and shape changes of the pipe to be processed, such as the slight deformation generated during the cutting process, to obtain a real-time three-dimensional model. This real-time three-dimensional model can compensate for the end position deviation caused by thermal deformation during cutting, thereby ensuring the accuracy of reverse cutting.

[0060] 502. Confirming the shape of the pipe to be processed based on the real-time three-dimensional model, wherein the shape includes a cylindrical shape and a straight pipe shape;

[0061] 503. When the pipe to be processed is cylindrical, the end origin of the pipe to be processed is used as the starting point, and a spiral basic trajectory is generated along the X-axis direction of the pipe according to the cutting requirements;

[0062] In this embodiment, the helix angle is calculated according to the incision angle, and a helical cutting trajectory with adjustable pitch is generated based on the calculated helix angle.

[0063] 504. When the pipe to be processed is a straight pipe, a linear interpolation trajectory is generated using a cubic spline interpolation algorithm with the end of the pipe to be processed as the starting point;

[0064] In this embodiment, a cubic spline interpolation algorithm is used to generate a linear interpolation trajectory starting from the end of the pipe to be processed and toward the current chuck clamping direction. The smoothness of the generated linear interpolation trajectory is ensured by adjusting the interpolation node density.

[0065] In this embodiment, exclusive trajectories are generated for different tube shapes, which can achieve precise control and improve process adaptability; the spiral trajectory evenly distributes cutting force and heat through spiral movement, achieving uniform cutting in all directions, and can adapt to round tubes of different diameters and lengths; the linear trajectory achieves rapid cutting through linear movement, and is suitable for straight tubes that do not require high cutting accuracy.

[0066] 505. Optimize the spiral basic trajectory or the linear interpolation trajectory to obtain a reverse cutting path.

[0067] Furthermore, in an embodiment of the present invention, the step of optimizing the spiral basic trajectory or the linear interpolation trajectory to obtain a reverse cutting path includes:

[0068] 601. Pre-constructing a weighted multi-objective function including material utilization, cutting time and cut roughness. The weights of material utilization, cutting time and cut roughness are determined by the hierarchical analysis method.

[0069] In the embodiment, a weighted multi-objective function including material utilization (tail length ≤5mm), cutting time (minimize empty cutting path) and kerf roughness (Ra≤1.6μm) is established; wherein the weights of the weighted multi-objective function are determined by the analytic hierarchy process, and under typical working conditions, the weight of the material utilization > the weight of the cutting time > the weight of the kerf roughness, and the weights can be dynamically adjusted according to production requirements; for example, the weight of the material utilization can be 0.6, the weight of the cutting time can be 0.3, and the weight of the kerf roughness can be 0.1.

[0070] 602, acquire the performance parameters of the pipe cutting machine, and set the constraint conditions based on the performance parameters of the pipe cutting machine;

[0071] In the embodiment, the constraint conditions include a chuck motion range constraint and a cutting head acceleration limit, the chuck motion range constraint is dynamically calculated according to the mechanical stroke of the chuck, and the cutting head acceleration limit includes an X-axis acceleration limit and a Y-axis acceleration limit, specifically, the X-axis and Y-axis accelerations ≤10m / s 2 , to ensure motion stability.

[0072] 603, based on the weighted multi-objective function and the constraint conditions, the spiral line basic trajectory or the linear interpolation trajectory is optimized by using a genetic algorithm to obtain a reverse cutting path;

[0073] In the embodiment, the genetic algorithm is used for effective processing when optimizing the spiral line basic trajectory or the linear interpolation trajectory; first, through real number coding, each gene corresponds to the coordinates of the key points of the trajectory and related cutting parameters such as cutting speed and cutting power P, and the gene length of each individual is equal to the product of the number of trajectory points and 5; in the selection operation stage, the roulette selection method is used to select individuals with higher fitness; then, the trajectory is processed in sections, and single-point crossover operation is performed to increase the diversity of the population; in addition, Gaussian mutation is performed on the trajectory point coordinates with a probability of 0.05, and the cutting parameters are adjusted to ensure the quality of the solution; through multiple iterations, the optimal solution, i.e. the reverse cutting path, is finally obtained; using the genetic algorithm for solving not only improves the accuracy of the generated reverse cutting path, but also optimizes the cutting efficiency and cost.

[0074] The zero tail cutting method of the pipe cutting machine in the embodiment of the application is described above, and the zero tail cutting device of the pipe cutting machine in the embodiment of the application is described below, please refer to Figure 2 , one embodiment of the zero tail cutting device of the pipe cutting machine in the embodiment of the application includes:

[0075] The construction module 701 is configured to pre-construct a critical value prediction model.

[0076] The prediction module 702 is configured to acquire parameter information of the pipe to be processed, and input the pre-constructed critical value prediction model to obtain the tailing critical value.

[0077] The detection module 703 is configured to perform a cutting operation on the pipe to be processed, and detect the remaining length of the pipe to be processed in real time.

[0078] The acquisition module 704 is configured to acquire real-time three-dimensional data of the pipe to be processed when the remaining length of the pipe to be processed is less than or equal to the tailing critical value.

[0079] The generation module 705 is configured to generate a reverse cutting path based on the real-time three-dimensional data of the pipe to be processed, and perform zero-tailing cutting on the pipe to be processed based on the generated reverse cutting path.

[0080] Based on the same idea as the method in the above embodiment, the device provided by the present application can implement the method of the above embodiment.

[0081] The above Figure 2 The zero-tailing cutting device of the pipe cutting machine in the embodiment of the present application is described in detail from the perspective of a modular functional entity, and the zero-tailing cutting device of the pipe cutting machine in the embodiment of the present application is described in detail from the perspective of hardware processing.

[0082] Figure 3 is a structural schematic diagram of a zero-tailing cutting device of a pipe cutting machine provided by the embodiment of the present application. The zero-tailing cutting device 800 of the pipe cutting machine can have relatively large differences due to different configurations or performances, and can include one or more than one processor (central processing unit, CPU) 810 and a memory 820, and one or more than one storage medium 830 (for example, one or more than one mass storage device) storing an application program 833 or data 832. The memory 820 and the storage medium 830 can be temporary storage or persistent storage. The program stored in the storage medium 830 can include one or more than one module (not shown in the figure), and each module can include a series of instruction operations in the zero-tailing cutting device 800 of the pipe cutting machine. Furthermore, the processor 810 can be configured to communicate with the storage medium 830, and execute a series of instruction operations in the storage medium 830 on the zero-tailing cutting device 800 of the pipe cutting machine, so as to implement the steps of the zero-tailing cutting method of the pipe cutting machine provided by each method embodiment.

[0083] The zero tail material cutting device 800 of the pipe cutting machine may further include one or more power supplies 840, one or more wired or wireless network interfaces 850, one or more input and output interfaces 860, and / or one or more operating systems 831, such as Windows Server, Mac OS X, Unix, Linux, FreeBSD, etc. It will be understood by those skilled in the art that Figure 3 The illustrated zero-tail cutting device structure of the pipe cutting machine does not constitute a limitation to the zero-tail cutting device of the pipe cutting machine, and may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently.

[0084] The present invention also provides a computer-readable storage medium, which can be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium. The computer-readable storage medium stores instructions, which, when executed on a computer, enable the computer to execute the steps of the zero-tail cutting method of a pipe cutting machine.

[0085] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described systems, devices, and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0086] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), magnetic disk or optical disk, etc., various media that can store program code.

[0087] Finally, it should be noted that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art will be able to modify the technical solutions described in the aforementioned embodiments or substitute equivalents for some of the technical features. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

Claims

1. A zero tail material cutting method for a pipe cutting machine, characterized in that: include: Pre-built critical value prediction model; Obtain parameter information of the pipe to be processed and input it into a pre-built critical value prediction model to obtain the tailing critical value; Perform cutting operations on the pipe to be processed and detect the remaining length of the pipe to be processed in real time; When the remaining length of the pipe to be processed is less than or equal to the tailing critical value, the real-time three-dimensional data of the pipe to be processed is obtained; A reverse cutting path is generated based on real-time three-dimensional data of the pipe to be processed, and the pipe to be processed is cut with zero tailing based on the generated reverse cutting path.

2. The zero-tail cutting method of the pipe cutting machine according to claim 1, characterized in that: The pre-built critical value prediction model includes: Acquire historical cutting data, wherein the historical cutting data includes historical cutting information of a plurality of pipes, wherein the historical cutting information includes basic parameters of the pipes, cutting process parameters, and tailings-related data; Perform data cleaning and data enhancement processing on the historical cutting data to obtain preprocessed historical cutting data; A pre-built hybrid model is input into the pre-built hybrid model, and the pre-processed historical cutting data is trained on the pre-built hybrid model by combining the transfer learning method and the early stopping mechanism to obtain a critical value prediction model.

3. The zero-tail cutting method of the pipe cutting machine according to claim 2, characterized in that: The pre-built hybrid model includes a feature engineering layer, a fully connected layer, a material classification sub-model, a geometric parameter sub-model, a feature fusion layer and a multi-layer perceptron. The feature engineering layer is used to perform feature conversion processing on the pre-processed historical cutting data. The output end of the feature engineering layer is connected to the input end of the fully connected layer, and the output end of the fully connected layer is respectively connected to the input end of the material classification sub-model and the input end of the geometric parameter sub-model; the material classification sub-model is used to output the material feature vector to the feature fusion layer, and the geometric parameter sub-model is used to output the geometric feature vector to the feature fusion layer. The output end of the feature fusion layer is connected to the multi-layer perceptron, and the multi-layer perceptron is used to output the tail material critical value.

4. The zero-tail cutting method of the pipe cutting machine according to claim 1, characterized in that: The method of obtaining parameter information of the pipe to be processed and inputting it into a pre-built critical value prediction model to obtain the tailing critical value includes: Obtaining parameter information of the pipe to be processed, wherein the parameter information includes material parameters and specification parameters of the pipe to be processed; The acquired parameter information is input into a pre-built critical value prediction model to obtain the tailing critical value corresponding to the pipe to be processed.

5. The zero-tail cutting method of the pipe cutting machine according to claim 1, characterized in that: The method of performing a cutting operation on the pipe to be processed and detecting the remaining length of the pipe to be processed in real time includes: Obtaining cutting requirements and a three-dimensional model of the pipe to be processed, wherein the cutting requirements include a cut position, a cut shape, and a cut angle; Starting from the front end of the pipe to be processed, the A* algorithm is used to generate a forward cutting path based on the parameter information, cutting requirements and 3D model of the pipe to be processed; The pipe to be processed is cut based on the generated forward cutting path, and the remaining length of the pipe to be processed is detected in real time.

6. The zero-tail cutting method of the pipe cutting machine according to claim 5, characterized in that: The method of generating a reverse cutting path based on the real-time three-dimensional data of the pipe to be processed includes: Processing the real-time 3D data of the pipe to be processed, and performing 3D modeling based on the processed real-time 3D data to obtain a real-time 3D model of the pipe to be processed; Confirming the shape of the pipe to be processed based on the real-time three-dimensional model, wherein the shape includes a cylindrical shape and a straight pipe shape; When the pipe to be processed is cylindrical, the origin of the end of the pipe to be processed is used as the starting point, and a spiral basic trajectory is generated along the X-axis direction of the pipe according to the cutting requirements; When the pipe to be processed is a straight pipe, the end of the pipe to be processed is used as the starting point and the cubic spline interpolation algorithm is used to generate a linear interpolation trajectory; The helical basic trajectory or linear interpolation trajectory is optimized to obtain the reverse cutting path.

7. The zero-tail cutting method of a pipe cutting machine according to claim 6, characterized in that: The step of optimizing the helical basic trajectory or the linear interpolation trajectory to obtain a reverse cutting path includes: A weighted multi-objective function including material utilization, cutting time and cut roughness is pre-built. The weights of material utilization, cutting time and cut roughness are determined by the analytic hierarchy process. Obtain the performance parameters of the pipe cutting machine and set constraints based on the performance parameters of the pipe cutting machine; Based on the weighted multi-objective function and constraints, the genetic algorithm is used to optimize the spiral basic trajectory or the linear interpolation trajectory to obtain the reverse cutting path.

8. A zero tail material cutting device for a pipe cutting machine, characterized in that: include: A building module for pre-building a critical value prediction model; The prediction module is used to obtain parameter information of the pipe to be processed and input it into a pre-built critical value prediction model to obtain the tailing critical value; The detection module is used to perform cutting operations on the pipe to be processed and detect the remaining length of the pipe to be processed in real time; An acquisition module is used to acquire real-time three-dimensional data of the pipe to be processed when the remaining length of the pipe to be processed is less than or equal to the tailing critical value; The generation module is used to generate a reverse cutting path based on the real-time three-dimensional data of the pipe to be processed, and to perform zero-tail material cutting on the pipe to be processed based on the generated reverse cutting path.

9. A zero-tail cutting device for a pipe cutting machine, characterized in that: The zero tailing cutting device of the pipe cutting machine includes: a memory and at least one processor, wherein the memory stores instructions; At least one of the processors calls the instructions in the memory to enable the zero-tail cutting device of the pipe cutting machine to perform each step of the zero-tail cutting method of the pipe cutting machine according to any one of claims 1-7.

10. A computer-readable storage medium having instructions stored thereon, characterized in that: When the instructions are executed by the processor, the various steps of the zero-tail cutting method of the pipe cutting machine as described in any one of claims 1 to 7 are implemented.

Images