Multi-layer multi-pass welding thermal response curve prediction method, system and equipment and storage medium
The prediction of multi-layer multi-pass welding thermal response curves through BP neural network solves the problem of difficult to predict multi-layer multi-pass welding thermal response curves in the prior art, and achieves rapid and economical acquisition of thermal response curves, improving the efficiency of welding process optimization and joint organization prediction.
Patent Information
- Application Number
- CN202311065447.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2023-08-23
- Publication Date
- 2025-02-25
AI Technical Summary
The prior art is difficult to effectively predict the thermal response curve of multi-layer multi-pass welding, resulting in limited welding process optimization and joint structure prediction.
By designing the BP neural network, using the thermal response curve test parameters as the training set, the BP neural network is trained to obtain the thermal response curve parameter prediction model. Enter the position coordinates and thermal input parameters of the welding point into this model to obtain the predicted thermal response curve parameters, thereby predicting the thermal response curve of the welding point.
It realizes rapid acquisition of thermal response curves under different process parameters, avoids multiple welding test measurements with thermocouples, reduces the workload and cost of tests, and improves the efficiency of welding process optimization and joint tissue prediction.
Smart Images
Figure CN119514246A_ABST
Abstract
Description
Technical Field
[0001] The embodiments of the present disclosure belong to the technical field, and specifically relate to a method, system, electronic device and storage medium for predicting a thermal response curve of a multi-layer multi-pass weld. Background Art
[0002] X80 pipeline steel is an ultra-low carbon, ultra-fine-grained low-alloy high-strength steel obtained through microalloying and controlled rolling and controlled cooling processes. It has good strength, toughness, corrosion resistance and fracture performance. It has been widely used at home and abroad. The China-Russia East Line Pipeline Project, the China-Myanmar Oil and Gas Pipeline and the medium- and long-term oil and gas pipeline network planning all use X80 pipeline steel as the main body for laying. The oil and gas pipeline has a large diameter and thick wall. Welding generally requires beveling the workpiece and using multi-layer and multi-pass welding. GMAW is a commonly used high-efficiency welding method for this type of pipeline steel. The local heating of its heat source will cause large residual stress and deformation near the weld, and make the weld structure coarse, causing the performance of this area to further decrease. Therefore, exploring the distribution of the temperature field of the weld joint is of great significance for optimizing the welding process and predicting the joint structure.
[0003] Microstructure prediction requires thermal response curves. Currently, the determination of thermal response curves is mainly achieved through thermocouples. However, the efficiency and cost of determining thermal response curves through experimental methods are low, and the use of GMAW multi-layer and multi-pass welding makes it more difficult to obtain thermal response curves. At present, there is no research on the prediction method of thermal response curves for multi-layer welding, which limits the further development of multi-layer welding optimization welding process and joint microstructure prediction. Summary of the invention
[0004] The embodiments of the present disclosure aim to solve at least one of the technical problems existing in the prior art, and provide a method, system, electronic device and storage medium for predicting thermal response curves of multi-layer and multi-pass welding.
[0005] One aspect of the present disclosure provides a method for predicting a multi-layer multi-pass welding thermal response curve, comprising:
[0006] Obtain the actual position coordinates of the welding point to be tested and the actual welding heat input parameters;
[0007] Inputting the actual position coordinates and the actual welding heat input parameters into a pre-trained thermal response curve parameter prediction model to obtain the thermal response curve prediction parameters of the welding point to be tested; wherein the thermal response curve parameter prediction model is obtained by pre-training a BP neural network using thermal response curve test parameters as a training set;
[0008] A predicted thermal response curve is obtained according to the thermal response curve prediction parameters.
[0009] Furthermore, the thermal response curve parameter prediction model is trained by the following steps:
[0010] According to the geometric characteristics, process characteristic parameters and high temperature performance parameters of the multi-layer and multi-pass welding joints, a temperature field finite element model is constructed;
[0011] Extracting a thermal response characteristic curve of a predetermined position according to the temperature field finite element model, analyzing characteristic parameters of the thermal response characteristic curve, and associating the analyzed characteristic parameters with position coordinates of the corresponding predetermined position and welding heat input parameters to form a training set;
[0012] The BP neural network is trained with position coordinates and welding heat input parameters as input and characteristic parameters as output, and a trained thermal response curve prediction model is obtained.
[0013] Furthermore, the analyzing the characteristic parameters of the thermal response characteristic curve includes:
[0014] The thermal response characteristic curve is expressed using the following formula (1):
[0015] T=T H exp [-b×(tc)]+T 0 (1)
[0016] Among them, t is the time variable, T is the corresponding temperature at time t, T H is the initial temperature adjustment parameter, T 0 is the ambient temperature adjustment parameter, b is the curve curvature adjustment parameter, and c is the multi-pass welding fitting adjustment parameter;
[0017] The initial temperature adjustment parameter T H , Ambient temperature adjustment parameter T 0 , curve curvature adjustment parameter b and multi-pass welding fitting adjustment parameter c are used as characteristic parameters.
[0018] Furthermore, before constructing the temperature field finite element model, the method further includes: obtaining a thermal response curve of the predetermined position;
[0019] After constructing the temperature field finite element model, the method further comprises: verifying the temperature field finite element model using the thermal response curve.
[0020] Another aspect of the present disclosure provides a multi-layer multi-pass welding thermal response curve prediction system, comprising:
[0021] An acquisition module is used to obtain the actual position coordinates of the welding point to be tested and the actual welding heat input parameters;
[0022] A prediction module, used for inputting the actual position coordinates and the actual welding heat input parameters into a pre-trained thermal response curve parameter prediction model to obtain the thermal response curve prediction parameters of the welding point to be tested; wherein the thermal response curve parameter prediction model is obtained by training a BP neural network in advance using the thermal response curve test parameters as a training set;
[0023] The curve module is used to predict the parameters according to the thermal response curve to obtain a predicted thermal response curve.
[0024] Furthermore, the system also includes a training module, which is used to:
[0025] According to the geometric characteristics, process characteristic parameters and high temperature performance parameters of the multi-layer and multi-pass welding joints, a temperature field finite element model is constructed;
[0026] Extracting a thermal response characteristic curve of a predetermined position according to the temperature field finite element model, analyzing characteristic parameters of the thermal response characteristic curve, and associating the analyzed characteristic parameters with position coordinates of the corresponding predetermined position and welding heat input parameters to form a training set;
[0027] The BP neural network is trained with position coordinates and welding heat input parameters as input and characteristic parameters as output, and a trained thermal response curve prediction model is obtained.
[0028] Furthermore, the training module also includes an analysis module, which is used to:
[0029] The thermal response characteristic curve is expressed using the following formula (1):
[0030] T=T H exp [-b×(tc)]+T 0 (1)
[0031] Among them, t is the time variable, T is the corresponding temperature at time t, T H is the initial temperature adjustment parameter, T 0 is the ambient temperature adjustment parameter, b is the curve curvature adjustment parameter, and c is the multi-pass welding fitting adjustment parameter;
[0032] The initial temperature adjustment parameter T H , Ambient temperature adjustment parameter T 0 , curve curvature adjustment parameter b and multi-pass welding fitting adjustment parameter c are used as characteristic parameters.
[0033] Furthermore, the training module also includes a verification module for obtaining a thermal response curve of the predetermined position before constructing the temperature field finite element model; and
[0034] After the temperature field finite element model is constructed, the temperature field finite element model is calibrated using the thermal response curve.
[0035] Another aspect of the present disclosure provides an electronic device, comprising:
[0036] at least one processor; and,
[0037] The memory communicatively connected to the at least one processor is used to store one or more programs, and when the one or more programs are executed by the at least one processor, the at least one processor can implement the multi-layer multi-pass welding thermal response curve prediction method as described above.
[0038] Another aspect of the present disclosure provides a computer-readable storage medium storing a computer program, wherein the computer program, when executed by a processor, implements the method for predicting thermal response curves of multi-layer and multi-pass welding as described above.
[0039] The disclosed embodiments provide a method, system, electronic device, and storage medium for predicting a thermal response curve for a multi-layer, multi-pass weld. By designing a BP neural network, the BP neural network is trained using a training set obtained from a thermal response curve test to obtain a thermal response curve parameter prediction model. The position coordinates and thermal input parameters of the welding point of the multi-layer, multi-pass welding process are input into the model to obtain the predicted thermal response curve parameters, so that the thermal response curve of the welding point can be predicted based on these parameters. The present invention combines the thermal response curve test with a neural network, and can achieve rapid acquisition of a thermal response curve under different process parameters, avoiding the use of thermocouples for multiple welding test measurements, reducing the workload of the test, and reducing the cost of obtaining a thermal response curve for a multi-layer, multi-pass weld. BRIEF DESCRIPTION OF THE DRAWINGS
[0040] Figure 1 A schematic diagram of a process of predicting a thermal response curve of multi-layer multi-pass welding according to an embodiment of the present disclosure;
[0041] Figure 2 A schematic diagram of a training process of a thermal response curve parameter prediction model according to another embodiment of the present disclosure;
[0042] Figure 3 It is a schematic diagram of a temperature field finite element model of another embodiment of the present disclosure;
[0043] Figure 4 A schematic diagram of a predicted thermal response curve and a thermocouple measured thermal response curve according to another embodiment of the present disclosure;
[0044] Figure 5 It is a structural schematic diagram of a multi-layer multi-pass welding thermal response curve prediction system according to another embodiment of the present disclosure;
[0045] Figure 6The figure is a schematic diagram of the structure of an electronic device according to another embodiment of the present disclosure. DETAILED DESCRIPTION
[0046] The following will be combined with the drawings in the embodiments of the present disclosure to clearly and completely describe the technical solutions in the embodiments of the present disclosure. Obviously, the described embodiments are only part of the embodiments of the present disclosure, not all of the embodiments. Based on the embodiments in the present disclosure, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present disclosure.
[0047] In addition, the described features, structures or characteristics may be combined in one or more embodiments in any suitable manner. In the following description, many specific details are provided to provide a full understanding of the embodiments of the present disclosure. However, those skilled in the art will appreciate that the technical solutions of the present disclosure may be practiced without one or more of the specific details, or other methods, components, devices, steps, etc. may be adopted. In other cases, known methods, devices, implementations or operations are not shown or described in detail to avoid blurring the various aspects of the present disclosure.
[0048] The flowcharts shown in the accompanying drawings are only exemplary and do not necessarily include all the contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps can be decomposed, and some operations / steps can be combined or partially combined, so the actual execution order may change according to actual conditions.
[0049] It should be understood that although the terms first, second, third, etc. may be used in this disclosure to describe various components, these components should not be limited by these terms. These terms are used to distinguish one component from another component. Therefore, the first component discussed below can be referred to as the second component without departing from the teachings of the concepts of the present disclosure. As used in this disclosure, the term "and / or" includes any one of the associated listed items and all combinations of one or more.
[0050] Those skilled in the art will appreciate that the drawings are merely schematic diagrams of example embodiments, and the modules or processes in the drawings are not necessarily necessary for implementing the present disclosure, and therefore cannot be used to limit the protection scope of the present disclosure.
[0051] like Figure 1 As shown, an embodiment of the present disclosure provides a method for predicting a multi-layer multi-pass welding thermal response curve, comprising the following steps:
[0052] Step S11, obtaining the actual position coordinates of the welding point to be tested and the actual welding heat input parameters.
[0053] Specifically, the welding point to be tested is the position on the welding base material such as the X80 pipeline steel weld joint where the thermal response curve needs to be predicted, and the position coordinates of the position corresponding to the finite element model and its heat input parameters are obtained. The heat input parameters are the line energy calculated from the process characteristic parameters, namely welding current, welding voltage and welding speed.
[0054] Step S12: inputting the actual position coordinates and the actual welding heat input parameters into a pre-trained thermal response curve parameter prediction model to obtain the thermal response curve prediction parameters of the welding point to be tested.
[0055] Specifically, the position coordinates of the welding point and the heat input parameters obtained in step S11 are input into the pre-trained thermal response curve parameter prediction model, and the model outputs the thermal response curve prediction parameters, including the initial temperature adjustment parameter T H , Ambient temperature adjustment parameter T 0 , curve curvature adjustment parameter b, multi-pass welding fitting adjustment parameter c.
[0056] The thermal response curve parameter prediction model is obtained by training the BP neural network using the thermal response curve test parameters as a training set in advance. Figure 2 As shown, the specific training steps are as follows:
[0057] Step S21, obtaining a thermal response characteristic curve.
[0058] Specifically, small holes are punched at predetermined positions of the weld joints before the welding test. Three small holes can be punched in each layer of the multi-layer multi-pass welding of X80 pipeline steel. A thermocouple is placed in each small hole, and the thermocouple is used to record the thermal response characteristic curve during the welding process. The thermal response characteristic curves of the characteristic positions of the weld and the heat-affected zone of the X80 steel grade D1219×22mm steel pipe are tested separately. Since the thermocouple may be damaged during the welding process, the thermal response characteristic curve recorded by the damaged thermocouple is removed, and the normal thermal response characteristic curve is retained.
[0059] Step S22: construct a temperature field finite element model according to the geometric characteristics, process characteristic parameters and high temperature performance parameters of the multi-layer and multi-pass welding joints.
[0060] Among them, the geometric characteristics of the welding joint include groove characteristics and groove dimensions, which are used to construct the geometric model of the butt joint; the process characteristic parameters include welding current, welding voltage and welding speed; the high-temperature performance parameters are the characteristic parameters of the parent material, which are obtained by conducting high-temperature performance tests on the welding parent material such as X80 steel grade D1219×22mm specification steel pipe and the formed weld in advance.
[0061] Step S23, verifying the temperature field finite element model.
[0062] Specifically, the adjustable parameters of the heat source model in the temperature field finite element model are adjusted according to the thermal response characteristic curve obtained in step S21 to obtain a temperature field finite element model suitable for the X80 pipeline steel welding process, and the verification is completed. The verified temperature field finite element model is as follows: Figure 3 shown.
[0063] Step S24: extracting a thermal response characteristic curve of a predetermined position according to the temperature field finite element model, analyzing characteristic parameters of the thermal response characteristic curve, and associating the analyzed characteristic parameters with the position coordinates of the corresponding predetermined position and welding heat input parameters to form a training set.
[0064] Specifically, a large number of thermal response characteristic curves are extracted from the temperature field finite element model. It can be understood that each thermal response characteristic curve corresponds to a position coordinate and a set of welding heat input parameters. These thermal response characteristic curves are analyzed and mathematical functions that can uniformly express these thermal response characteristic curves are selected. The specific analysis and selection process is as follows:
[0065] For the selection of mathematical functions of the temperature-time curve of the welding cooling process, it can be simplified as the cooling process of a high-temperature metal block for calculation. The conclusion obtained by the simplified calculation is generalized to be the temperature change of the welding process. According to the principle of heat transfer, the rate at which a hot object dissipates heat to the surrounding environment is proportional to the temperature difference between the object and the surrounding environment, as described by Newton's law of cooling. Therefore, from the perspective of convective heat transfer, it can be expressed mathematically as:
[0066]
[0067] Where: dQ / dt is the heat loss rate of the metal block, H is the heat transfer coefficient, A is the surface area of the metal block, T is the current temperature of the metal block, T s is the ambient temperature.
[0068] Assuming that the metal block is isotropic and uniform, and all parts except the surface are insulated, the lumped parameter method can be used to describe the cooling process of the high-temperature metal block. This method assumes that the entire block can be regarded as a single homogeneous entity with a uniform temperature. Using the lumped capacitance method and the above equation, the function expression of the temperature of the metal block during the cooling process can be derived, and this function is in the form of exponential decay, and the expression is:
[0069]
[0070] Where: T 0 is the initial temperature of the metal block, m is the mass of the metal block, C p is the specific heat capacity of the metal block.
[0071] The temperature-time curve obtained by the above expression is a direct input function. It is inaccurate to directly apply this function to express the welding process, because in the high temperature range, radiation heat transfer cannot be ignored. However, the main research temperature range of this invention is the cooling process of 800-500℃. After calculation, radiation heat transfer will have a significant effect above 1200℃, so radiation heat transfer is not considered. Therefore, the expression used in this paper to fit the cooling curve of multi-layer multi-pass welding of X80 pipeline steel is:
[0072] T=T H exp[-b×(tc)]+T 0
[0073] Where: T is the temperature corresponding to time t, T H Adjust the parameters for a mathematical function based on the initial temperature, T 0 is the adjustment parameter of the mathematical function based on the ambient temperature, b is the adjustment parameter for adjusting the curvature of the temperature drop curve, T is the temperature-time variable to be calculated, and c is the adjustment parameter for the translation of the adjustment function along the x-axis, which is used to fit the multi-pass welding situation.
[0074] It should be noted that T H With T 0 They are not the actual peak temperature and ambient temperature, but are parameters used to fit the function. They are numerically close to the peak temperature or ambient temperature, but have no practical significance.
[0075] Subsequently, multiple sets of training data are extracted to form the training set of the BP neural network: for the thermal response curve of multi-layer and multi-pass welding of X80 pipeline steel to be fitted, it is grouped and simplified. It is assumed that its cooling curve is represented by y = f(x), and the known node data are the temperatures corresponding to the continuous time points (t x ,T x ), where the time step of each two data is 0.4s. It is hoped that different functions y = φ (x) can be constructed for multiple cooling of each weld to approximate y = f (x). In the interpolation process, the fitting generally meets more stringent requirements, that is, Since the fitting function used in the present invention does not consider the influence of radiation heat transfer, it is allowed that some points do not strictly conform to the above formula, but the errors of the two should be minimized during the calculation process. At present, the typical method to reduce the deviation and facilitate calculation is to calculate the sum of squares of the deviations. The sum of squares of the deviations η can be calculated using the following formula:
[0076]
[0077] For the function studied in this paper It can be written as follows:
[0078] T=T Hexp[-b×(tc)]+T 0
[0079] Controlling η to a set value below The relevant function parameters of Expand it into this mathematical function, the parameters of which are the target parameters sought in this article.
[0080] By calling the curve_fit library in scipy, a common data processing library in Python, the curve can be fitted according to the simplification of the thermal response curve of multi-layer and multi-pass welding mentioned above, and the characteristic parameter T can be extracted. H , b, c, T 0 The extracted large number of characteristic parameters are matched one by one with the corresponding position coordinates and welding heat input parameters to form the training set of the BP neural network.
[0081] Step S25, respectively using the position coordinates and welding heat input parameters as input and the characteristic parameters as output, training the BP neural network to obtain a trained thermal response curve prediction model.
[0082] Specifically, the training set obtained in step S24 is used to train the BP neural network. During the training, the learning rate is constantly checked to obtain the optimal solution. The loss function adopts the mean square error loss function (MSE).
[0083] The thermal response curve parameter prediction model with trained weights and thresholds is called, and the actual position coordinates of the welding point to be tested and the actual welding heat input parameters obtained in step S11 are input, so that the characteristic parameters of the thermal response curve of the welding point to be tested can be output.
[0084] Step S13: obtaining a predicted thermal response curve according to the thermal response curve prediction parameters.
[0085] Specifically, the actual characteristic parameter T predicted in the above step S12 is H , b, c, T 0 , using the formula T = T H exp[-b×(tc)]+T 0 , and the actual predicted thermal response curve of the welding point to be tested is obtained by using the mapping software. As a verification, the comparison between the predicted thermal response curve and the actual measured thermal response curve of the thermocouple is shown in Figure 2. Figure 4 As shown, the lines in the figure represent the predicted thermal response curves, and the points represent the measured thermal response curves of the thermocouples.
[0086] A method for predicting the thermal response curve of multi-layer and multi-pass welding in an embodiment of the present disclosure constructs a thermal response prediction model based on a BP neural network, and verifies the GMAW finite element model of X80 pipeline steel through the measured data of the GMAW welding test of X80 pipeline steel. On this basis, the thermal response curves at different welding process parameters and different positions are studied, and the BP network prediction model is constructed and trained based on this as a sample to identify the complex nonlinear relationship between the welding process parameters, the position of the point and the thermal response curve, so that the thermal response curve can be quickly obtained under different process parameters, avoiding the use of thermocouples for multiple welding test measurements, which can save a lot of test costs and time, and is of great significance for optimizing the X80 pipeline steel welding process, predicting the organization of the X80 pipeline steel welded joint, and then optimizing the performance of the X80 pipeline steel welded joint.
[0087] like Figure 5 As shown, another embodiment of the present disclosure provides a multi-layer multi-pass welding thermal response curve prediction system, comprising:
[0088] An acquisition module 510 is used to acquire the actual position coordinates of the welding point to be tested and the actual welding heat input parameters;
[0089] The prediction module 520 is used to input the actual position coordinates and the actual welding heat input parameters into a pre-trained thermal response curve parameter prediction model to obtain the thermal response curve prediction parameters of the welding point to be tested; wherein the thermal response curve parameter prediction model is obtained by training the BP neural network in advance using the thermal response curve test parameters as a training set;
[0090] The curve module 530 is used to obtain a predicted thermal response curve according to the thermal response curve prediction parameters.
[0091] Specifically, the acquisition module 510 is used to implement the function of step S11 in the previous embodiment, the prediction module 520 is used to implement the function of step S12 in the previous embodiment, and the curve module 530 is used to implement the function of step S13 in the previous embodiment.
[0092] For example, Figure 5 As shown, the multi-layer multi-pass welding thermal response curve prediction system of the embodiment of the present disclosure further includes a training module 540 for implementing the functions of steps S21 to S25 in the previous embodiment.
[0093] For example, Figure 5 As shown, the training module 540 also includes an analysis module 541 for implementing the analysis process in step S24 in the previous embodiment.
[0094] For example, Figure 5As shown, the training module 540 also includes a verification module 542 for implementing the functions of steps S21 and S23 in the previous embodiment.
[0095] A multi-layer multi-pass welding thermal response curve prediction system of the disclosed embodiment builds a thermal response prediction model based on a BP neural network, and verifies the GMAW finite element model of X80 pipeline steel through the measured data of the GMAW welding test of X80 pipeline steel. On this basis, the thermal response curves at different welding process parameters and different positions are studied, and the BP network prediction model is constructed and trained based on this as a sample to identify the complex nonlinear relationship between the welding process parameters, the position of the point and the thermal response curve, so that the thermal response curve can be quickly obtained under different process parameters, avoiding the use of thermocouples for multiple welding test measurements, which can save a lot of test costs and time, and is of great significance for optimizing the X80 pipeline steel welding process, predicting the organization of the X80 pipeline steel welded joint, and then optimizing the performance of the X80 pipeline steel welded joint.
[0096] like Figure 6 As shown, another embodiment of the present disclosure provides an electronic device, including:
[0097] At least one processor 601; and a memory 602 in communication with the at least one processor 601, for storing one or more programs, which, when executed by the at least one processor 601, enables the at least one processor 601 to implement the multi-layer multi-pass welding thermal response curve prediction method described above.
[0098] Among them, the memory and the processor are connected in a bus manner, and the bus may include any number of interconnected buses and bridges, and the bus connects various circuits of one or more processors and memories together. The bus can also connect various other circuits such as peripherals, voltage regulators, and power management circuits, which are well known in the art and are therefore not further described herein. The bus interface provides an interface between the bus and the transceiver. The transceiver can be one element or multiple elements, such as multiple receivers and transmitters, providing a unit for communicating with various other devices on a transmission medium. The data processed by the processor is transmitted on a wireless medium via an antenna, and further, the antenna also receives data and transmits the data to the processor.
[0099] The processor is responsible for managing the bus and general processing, and can also provide various functions, including timing, peripheral interfaces, voltage regulation, power management, and other control functions. Memory can be used to store data used by the processor when performing operations.
[0100] Yet another embodiment of the present disclosure provides a computer-readable storage medium, and when the computer program is executed by a processor, the above-mentioned method for predicting the thermal response curve of multi-layer and multi-pass welding can be implemented.
[0101] The computer-readable medium may be included in the apparatus, device, or system of the present disclosure, or may exist independently.
[0102] Computer-readable storage media may be any tangible media that contains or stores a program, which may be an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device. More specific examples include, but are not limited to, an electrical connection with one or more wires, a portable computer disk, a hard disk, optical fiber, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof.
[0103] The computer-readable storage medium may also include a data signal propagated in baseband or as part of a carrier wave, in which the computer-readable program code is carried. Specific examples include but are not limited to electromagnetic signals, optical signals, or any suitable combination thereof.
[0104] It is to be understood that the above embodiments are merely exemplary embodiments used to illustrate the principles of the present disclosure, but the present disclosure is not limited thereto. For those of ordinary skill in the art, various modifications and improvements can be made without departing from the spirit and substance of the present disclosure, and these modifications and improvements are also considered to be within the scope of protection of the present disclosure.
Claims
1. A method for predicting thermal response curves of multi-layer and multi-pass welding, characterized in that: include: Obtain the actual position coordinates of the welding point to be tested and the actual welding heat input parameters; Inputting the actual position coordinates and the actual welding heat input parameters into a pre-trained thermal response curve parameter prediction model to obtain the thermal response curve prediction parameters of the welding point to be tested; wherein the thermal response curve parameter prediction model is obtained by pre-training a BP neural network using thermal response curve test parameters as a training set; A predicted thermal response curve is obtained according to the thermal response curve prediction parameters.
2. The method according to claim 1, characterized in that The thermal response curve parameter prediction model is trained by the following steps: According to the geometric characteristics, process characteristic parameters and high temperature performance parameters of the multi-layer and multi-pass welding joints, a temperature field finite element model is constructed; Extracting a thermal response characteristic curve of a predetermined position according to the temperature field finite element model, analyzing characteristic parameters of the thermal response characteristic curve, and associating the analyzed characteristic parameters with position coordinates of the corresponding predetermined position and welding heat input parameters to form a training set; The BP neural network is trained with position coordinates and welding heat input parameters as input and characteristic parameters as output, and a trained thermal response curve prediction model is obtained.
3. The method according to claim 2, characterized in that The analyzing the characteristic parameters of the thermal response characteristic curve includes: The thermal response characteristic curve is expressed using the following formula (1): T=T H exp[-b×(t-c)]+T0 (1) Among them, t is the time variable, T is the corresponding temperature at time t, T H is the initial temperature adjustment parameter, T0 is the ambient temperature adjustment parameter, b is the curve curvature adjustment parameter, and c is the multi-pass welding fitting adjustment parameter; The initial temperature adjustment parameter T H , ambient temperature adjustment parameter T0, curve curvature adjustment parameter b and multi-pass welding fitting adjustment parameter c are used as characteristic parameters.
4. The method according to claim 2 or 3, characterized in that: Before constructing the temperature field finite element model, the method further includes: obtaining a thermal response curve of the predetermined position; After constructing the temperature field finite element model, the method further comprises: verifying the temperature field finite element model using the thermal response curve.
5. A multi-layer multi-pass welding thermal response curve prediction system, characterized in that: The system comprises: An acquisition module is used to obtain the actual position coordinates of the welding point to be tested and the actual welding heat input parameters; A prediction module, used for inputting the actual position coordinates and the actual welding heat input parameters into a pre-trained thermal response curve parameter prediction model to obtain the thermal response curve prediction parameters of the welding point to be tested; wherein the thermal response curve parameter prediction model is obtained by training a BP neural network in advance using the thermal response curve test parameters as a training set; The curve module is used to predict the parameters according to the thermal response curve to obtain a predicted thermal response curve.
6. The system according to claim 5, characterized in that The system also includes a training module for: According to the geometric characteristics, process characteristic parameters and high temperature performance parameters of the multi-layer and multi-pass welding joints, a temperature field finite element model is constructed; Extracting a thermal response characteristic curve of a predetermined position according to the temperature field finite element model, analyzing characteristic parameters of the thermal response characteristic curve, and associating the analyzed characteristic parameters with position coordinates of the corresponding predetermined position and welding heat input parameters to form a training set; The BP neural network is trained with position coordinates and welding heat input parameters as input and characteristic parameters as output, and a trained thermal response curve prediction model is obtained.
7. The system according to claim 6, characterized in that The training module also includes an analysis module, which is used to: The thermal response characteristic curve is expressed using the following formula (1): T=T H exp[-b×(t-c)]+T0 (1) Among them, t is the time variable, T is the corresponding temperature at time t, T H is the initial temperature adjustment parameter, T0 is the ambient temperature adjustment parameter, b is the curve curvature adjustment parameter, and c is the multi-pass welding fitting adjustment parameter; The initial temperature adjustment parameter T H , ambient temperature adjustment parameter T0, curve curvature adjustment parameter b and multi-pass welding fitting adjustment parameter c are used as characteristic parameters.
8. The system according to claim 6 or 7, characterized in that: The training module further includes a verification module for obtaining a thermal response curve of the predetermined position before constructing a temperature field finite element model; and After the temperature field finite element model is constructed, the temperature field finite element model is calibrated using the thermal response curve.
9. An electronic device, characterized in that: include: at least one processor; as well as, A memory communicatively connected to the at least one processor is used to store one or more programs, and when the one or more programs are executed by the at least one processor, the at least one processor can implement the multi-layer multi-pass welding thermal response curve prediction method described in any one of claims 1 to 4.
10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method for predicting a thermal response curve of a multi-layer multi-pass welding according to any one of claims 1 to 4 is implemented.