An Online Fault Judgment Method and System for a Laser Welding Machine

Abstract

The present invention discloses an on-line fault determination method and system for a laser welding machine. The method includes: when the laser welding machine is running online, obtaining the real-time status data of the laser welding machine; calling the fault determination functions corresponding to several hyperspheres of the laser welding machine and respectively substituting the real-time status data of the laser welding machine into the fault determination functions D(x) corresponding to several hyperspheres for calculation to obtain calculation results; determining whether the calculation results are all greater than 0; if so, indicating that the laser welding machine has a fault.

Description

Technical Field

[0001] The present application relates to the technical field of cold rolling, and particularly to an online fault determination method and system for a laser welding machine. Background Art

[0002] A laser welding machine is a high-tech device integrating technologies such as optics, electricity, machinery, and automation. The laser welding machine belongs to deep penetration welding. It uses high voltage to generate laser, and through various mirrors and focusing mirrors in the laser transmission conduit, it accurately focuses on the butt joint part of the strip steel. Due to the high power of the laser beam, the tiny local part of the strip steel butt joint is quickly heated, melted, and evaporated to form a small hole. The melted material in the small hole then cools and solidifies to form a uniform and dense weld seam.

[0003] The laser welding machine is an important device in cold rolling. However, for the fault diagnosis of the laser welding machine, it currently remains in the offline diagnosis stage after an accident and lacks online control capabilities. Summary of the Invention

[0004] The present invention provides an online fault determination method and system for a laser welding machine to solve or partially solve the technical problem that the faults of the laser welding machine cannot be judged online, thereby lacking online control capabilities.

[0005] To solve the above technical problem, the present invention provides an online fault determination method for a laser welding machine, and the method includes:

[0006] When the laser welding machine is running online, obtaining real-time status data of the laser welding machine;

[0007] Invoking the fault determination functions corresponding to several hyperspheres of the laser welding machine Wherein, x is the real-time status data, x i , x j are samples in the same subset obtained by clustering based on the laser welding machine training set, n is the number of samples in the subset, is the global optimal solution corresponding to the minimum enclosing sphere optimization solution of the hypersphere, k(x, x i ), k(x i , x j ) are kernel functions respectively, k(x, x i ) represents the inner product of x, x i , k(x i , x j ) represents the inner product of x i , x j , and r is the radius of the hypersphere;

[0008] Substitute the real-time status data of the laser welding machine into the respective fault determination functions D(x) of the several hyperspheres for calculation to obtain the calculation results;

[0009] Determine whether all the calculation results are greater than 0;

[0010] If so, it means that the laser welding machine has a fault.

[0011] Preferably, the center C and radius r of the hypersphere are respectively:

[0012] where φ(x i ) represents the mapping relationship of x i ;

[0013]

[0014] Preferably, before obtaining the real-time status data of the laser welding machine, the method further includes:

[0015] Determine the center C and radius r of the hypersphere;

[0016] Use the center C and radius r of the hypersphere to obtain the fault determination function D(x); where, φ(x) represents the mapping relationship of x.

[0017] Preferably, the determination of the center C and radius r of the hypersphere includes:

[0018] Obtain a laser welding machine training set; where the samples in the laser welding machine training set refer to the status data samples when the laser welding machine is in a normal state;

[0019] Cluster the samples in the laser welding machine training set to obtain several non-overlapping subsets;

[0020] Based on the several non-overlapping subsets, determine the center C and radius r of the several hyperspheres; where one subset corresponds to one hypersphere, and each hypersphere has its own center C and radius r.

[0021] Preferably, the determination of the center C and radius r of the several hyperspheres based on the several non-overlapping subsets includes:

[0022] For each of the subsets, determine the relevant objective function and relevant constraint conditions:

[0023] where A is the penalty coefficient;

[0024] The improved sequential minimal optimization algorithm (SMO algorithm) is used to convert the related objective function into a bivariate objective function, and a penalty factor is added to the bivariate objective function for iterative optimization to determine the center and radius of the corresponding hypersphere; the penalty factor is used to prevent the related objective function from falling into the local optimal solution of the bivariate; the function obtained after adding the penalty factor to the bivariate objective function is: where is a bivariate, α t , α p , α2 are all Lagrange multipliers, x1, x2, x t , x p , x i , x j , x x are all samples in the same subset, k(x1,x2), k t (x t ,x1), k(x t ,x2), k(x p ) are all kernel functions, is the penalty factor.

[0025] Preferably, for each of the subsets, determining the related objective function and related constraint conditions includes:

[0026] For each of the subsets, establishing a corresponding first objective function and first constraint condition;

[0027] Introducing Lagrange multipliers to combine and convert the first objective function and the first constraint condition into a second objective function, and determining the second constraint condition; wherein, the second constraint condition is the same as the related constraint condition;

[0028] Using Gaussian kernel nonlinear mapping to map the second objective function into the related objective function.

[0029] Preferably, the first objective function is: where x i is any sample data vector in the subset, A is the penalty coefficient, ξ i is the slack variable, ξ i ≥0, i = 1, 2,..., n;

[0030] The first constraint condition is:

[0031] The second objective function is: where α i , α j are Lagrange multipliers, k(x i ,x i ) is a kernel function.

[0032] The present invention also discloses an online fault determination system for a laser welding machine, including:

[0033] An acquisition module, configured to acquire real-time status data of the laser welding machine when the laser welding machine is running online;

[0034] An invocation module, configured to invoke the fault determination functions respectively corresponding to a plurality of hyperspheres of the laser welding machine Wherein, x is the real-time status data, x i , x j are samples in the same subset obtained by clustering based on a laser welding machine training set, n is the number of samples in the subset, is the global optimal solution corresponding to the minimum enclosing sphere optimization solution of the hypersphere, k(x, x i ), k(x i , x j ) are kernel functions respectively, k(x, x i ) represents the inner product of x, x i , k(x i , x j ) represents the inner product of x i , x j , and r is the radius of the hypersphere;

[0035] A calculation module, configured to respectively substitute the real-time status data of the laser welding machine into the fault determination function D(x) corresponding to each of the plurality of hyperspheres for calculation to obtain a calculation result;

[0036] A judgment module, configured to judge whether the calculation results are all greater than 0; if so, it indicates that the laser welding machine has a fault.

[0037] The present invention discloses a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, the steps of the above method are implemented.

[0038] The present invention discloses a computer device, including a memory, a processor, and a computer program stored on the memory and executable on the processor, and when the processor executes the program, the steps of the above method are implemented.

[0039] Through one or more technical solutions of the present invention, the present invention has the following beneficial effects or advantages:

[0040] The solution disclosed by the present invention acquires real-time status data of the laser welding machine when the laser welding machine is running online, and invokes the fault determination function Fault analysis is performed on the real-time status data of the laser welding machine, and online fault determination of the laser welding machine is carried out based on this, so as to make up for the deficiency of the online control ability of the laser welding machine and achieve the effect of quickly determining the faults of the laser welding machine online.

[0041] The above description is only an overview of the technical solution of the present invention. In order to be able to understand the technical means of the present invention more clearly, it can be implemented according to the content of the specification. And in order to make the above and other purposes, features and advantages of the present invention more obvious and understandable, the following specifically illustrates the specific embodiments of the present invention. Brief Description of the Drawings

[0042] By reading the following detailed description of the preferred embodiments, various other advantages and benefits will become clear to those of ordinary skill in the art. The drawings are only for the purpose of showing the preferred embodiments and are not considered to be a limitation of the present invention. And throughout the drawings, the same reference numerals are used to represent the same components. In the drawings:

[0043] Figure 1 Shows a flowchart of an online fault determination method for a laser welding machine according to an embodiment of the present invention;

[0044] Figure 2 Shows a schematic diagram of a hypersphere according to an embodiment of the present invention;

[0045] Figure 3 Shows a schematic diagram of an online fault determination system for a laser welding machine according to an embodiment of the present invention. Detailed Embodiments

[0046] The following will describe the exemplary embodiments of the present disclosure in more detail with reference to the drawings. Although the exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided so that the present disclosure can be more thoroughly understood and the scope of the present disclosure can be fully conveyed to those skilled in the art.

[0047] In order to solve the technical problem that the faults of the existing laser welding machine cannot be judged online, and thus lack the online control ability. The embodiments of the present invention provide an online fault determination method and system for a laser welding machine, which are used when the laser welding machine is running online to perform online fault determination on the laser welding machine.

[0048] In order to illustrate and explain the present invention, the following embodiments disclose an online fault determination method for a laser welding machine. This method can be applied to the laser welding machine or the control device for controlling the laser welding machine, depending on the actual situation. Refer to Figure 1 , the method includes:

[0049] Step 101: When the laser welding machine is running online, obtain the real-time status data of the laser welding machine.

[0050] Step 102: Call the fault determination function D(x) corresponding to each of several hyperspheres of the laser welding machine.

[0051] Step 103: Substitute the real-time status data of the laser welding machine into the fault determination function D(x) corresponding to each of the several hyperspheres for calculation to obtain the calculation results.

[0052] Step 104: Determine whether all the calculation results are greater than 0. If so, it indicates that the laser welding machine has a fault; otherwise, the laser welding machine has no fault.

[0053] In the above technical solution, when the laser welding machine is running online, the real-time status data of the laser welding machine is obtained, and the fault determination function is called to perform fault analysis on the real-time status data of the laser welding machine, and based on this, online fault determination of the laser welding machine is carried out to make up for the deficiency of the online control ability of the laser welding machine, achieving the effect of quickly determining the faults of the laser welding machine online.

[0054] In step 101, various sensors (collectively referred to as the sensor group in this embodiment) are provided on the laser welding machine to collect the real-time status data when the laser welding machine is running online. The real-time status data is generated by each moving part of the laser welding machine and includes: the gap width g w , the gap position g p , the welding speed ω s and the position p s , the laser power P, the welding wheel pressure T w , the leveling pressure T p . Therefore, the real-time status data x in this embodiment refers to a set of sensor data x = [g w , g p , ω s , p s , P, T w , T p . After the sensor group collects the real-time status data, it will be transmitted to the online fault determination system of the laser welding machine for fault determination.

[0055] In step 102, the number of hyperspheres is not limited in this embodiment. The fault determination function D(x) corresponding to each hypersphere is stored in the online fault determination system of the laser welding machine and is directly called when needed.

[0056] A hypersphere is the smallest closed sphere corresponding to the subset obtained by clustering the status data samples when the laser welding machine is in a normal state. The hypersphere contains the most status data samples while having the smallest possible radius. See Figure 2, is a schematic diagram of a hypersphere. The hypersphere contains state data samples of the laser welding machine when it is in a normal state ( Figure 2 Schematic for normal data), and the abnormal data is outside the hypersphere. Since the occurrence of faults in the laser welding machine has a slow characteristic and there is a lack of abnormal samples, it is difficult to construct a high-precision decision function using abnormal samples. Therefore, in this embodiment, several non-overlapping subsets are clustered from the state data samples of the laser welding machine when it is in a normal state, and then the corresponding hyperspheres are determined based on the several non-overlapping subsets. It can be seen that the hypersphere contains state data samples of the laser welding machine when it is in a normal state. Using such hyperspheres to represent the fault decision boundary can make up for the defect of insufficient abnormal samples of the laser welding machine.

[0057] Furthermore, the fault decision function D(x) corresponding to the hypersphere is constructed based on the center C and radius r of the hypersphere, and can be pre-constructed before obtaining the real-time state data of the laser welding machine. Specifically:

[0058] Among them, x is the real-time state data, φ(x) represents the mapping relationship of x, x i , x j are samples in the same subset obtained by clustering based on the laser welding machine training set, φ(x i ) represents the mapping relationship of x i , n is the number of samples in the subset, is the global optimal solution corresponding to the minimum enclosing sphere optimization solution of the hypersphere, k(x, x i ), k(x i , x j ) are kernel functions respectively, k(x, x i ) represents the inner product of x, x i , k(x i , x j ) represents the inner product of x i , x j , C and r are the center and radius of the hypersphere respectively,

[0059] From the construction process of the fault determination function D(x), it can be seen that the basic principle of its construction is as follows: for the real-time state data x of the laser welding machine, it is determined whether it is inside the hypersphere by its distance from the center C of the sphere, so as to determine whether there is a fault in the laser welding machine. If D(x)>0, it means that the distance from the center C of the hypersphere is outside the hypersphere. If all D(x) are greater than 0, it means that the laser welding machine has a fault. According to the basic principle of this construction, the center C of the sphere and the radius r are substituted into D(x) for calculation to obtain the final fault determination function D(x). It should be noted that since the center C of the sphere and the radius r are actually obtained by solving the relevant objective function, and the relevant objective function is the objective function obtained after Gaussian mapping, the center C of the sphere and the radius r obtained by solving are actually the mapped center of the sphere and the mapped radius. In this embodiment, the mapped center of the sphere and the mapped radius are directly used to construct the fault determination function, without having to find the exact preimage corresponding to the mapped center of the sphere to construct the fault determination function. In contrast, the amount of calculation can be reduced, and the construction method of the fault determination function is more simple and the accuracy meets the requirements.

[0060] Since before obtaining the real-time state data of the laser welding machine, the center C of the hypersphere and the radius r are first determined, and then the fault determination function D(x) is obtained by using the center C of the hypersphere and the radius r. Therefore, the center C of the hypersphere and the radius r of the hypersphere are the decisive factors for the accuracy of the fault determination of the laser welding machine. In order to obtain a high-precision center C of the sphere and radius r in this embodiment, the state data samples of the laser welding machine in the normal state are first collected for clustering to obtain several non-overlapping subsets to eliminate the coupling phenomenon existing between the moving parts of the laser welding machine, and then the center C of the hypersphere and the radius r corresponding to each subset are determined to accurately define the fault determination boundary of the laser welding machine, and the fault determination function is constructed accordingly to perform fault determination on the laser welding machine. Since each hypersphere in this embodiment is trained with several non-overlapping subsets, each hypersphere represents the data of each moving part of the laser welding machine that is not coupled in the normal state. The fault determination function constructed based on the center C of the hypersphere and the radius r can accurately perform fault determination on the laser welding machine and has the characteristic of high precision.

[0061] Specifically, the center C of the hypersphere and the radius r are obtained through offline training in the following manner:

[0062] Step 201, obtain the training set of the laser welding machine.

[0063] Specifically, the samples in the training set of the laser welding machine are the state data samples of the laser welding machine in the normal state, which are collected by the sensor group. The samples S in the training set of the laser welding machine = {x1, x2,... x m}, where x1, x2, x m respectively represent a set of state data samples. Taking x1 as an example, x1 = [g w (1), g p(1), ω s (1), p s (1), P(1), T w (1), T p (1)]。

[0064] Step 202: Cluster the samples in the laser welding machine training set to obtain several non - overlapping subsets.

[0065] In this embodiment, due to the coupling phenomenon existing in the moving parts of the laser welding machine, the samples in the laser welding machine training set are clustered into several non - overlapping subsets to eliminate the coupling relationship between the moving parts and make the fault determination accuracy higher.

[0066] Furthermore, various unsupervised clustering methods can be used to cluster the samples in the laser welding machine training set, such as the k - means clustering algorithm, the DBSCAN clustering algorithm, etc., to obtain several non - overlapping subsets. The clustering algorithm can be specifically selected according to the actual situation. Here, the k - means clustering algorithm is taken as an example for introduction, but it does not form a limitation. The specific steps of the k - means clustering algorithm are as follows: Randomly select several samples from the laser welding machine training set as the cluster centers; the cluster center is the mean vector of the cluster. Calculate the distance between each sample in the laser welding machine training set and each cluster center. Assign each sample to the cluster with the closest distance. Calculate the new cluster centers of each cluster. If the distance between the new cluster centers of each cluster and the cluster centers calculated last time for each cluster is less than the preset threshold, the clustering ends. Each subset after clustering is non - overlapping and contains its own state data samples. For example, x i 、x j are samples in the same subset, x i = [g w (i), g p (i), ω s (i), p s (i), P(i), T w (i), T p (i)], x j = [g w (j), g p (j), ω s (j), p s (j), P(j), T w (j), T p (j)].

[0067] Step 203: Determine the centers C and radii r of several hyperspheres based on the several non - overlapping subsets.

[0068] In this embodiment, a subset corresponds to a hypersphere, and each hypersphere has its own center and radius. For each subset, when determining the center and radius of its corresponding hypersphere, the following processing is performed:

[0069] First, determine the relevant objective function and relevant constraint conditions. Specifically, the relevant objective function is: The relevant constraint conditions are: A is the penalty coefficient. Secondly, use the improved sequential minimal optimization algorithm SMO algorithm to convert the relevant objective function into a bivariate objective function, and add a penalty factor to the bivariate objective function for iterative optimization to determine the center and radius of the corresponding hypersphere. Adding a penalty factor to the bivariate objective function is to avoid the relevant objective function falling into the local optimal solution of the bivariate. And the function obtained after adding the penalty factor to the bivariate objective function is:

[0070] Among them, is a bivariate, α t , α p , α2 are all Lagrange multipliers, x1, x2, x t , x p , x i , x j are all samples in the same subset, k(x1, x2), k x (x t , x1), k(x t , x2), k(x t , x p ) are all kernel functions, representing the inner product of relevant state data samples, is the penalty factor.

[0071] In this embodiment, by using the improved sequential minimal optimization algorithm SMO algorithm combined with the penalty factor to iteratively optimize the relevant objective function, the global optimal solution can be obtained, and then the center and radius of the hypersphere with high precision can be determined, providing data support for the fault determination function to accurately determine the faults of the laser welding machine.

[0072] The above process will be specifically introduced below. It should be noted that since the implementation processes of each subset are the same, in this embodiment, one subset T = {x1, x2,..., x n} is selected as an example for illustration, and other subsets are similar, where x i is a p-dimensional data vector, and x i ∈T.

[0073] Establish the first objective function and the first constraint condition. The radius of the smallest enclosing hypersphere obtained by training is r, and the center is C, and the following optimal solution should be satisfied:

[0074] The first objective function:

[0075] The first constraint condition:

[0076] where A is the penalty coefficient, and ξ i is the slack variable, ξ i ≥0, i = 1, 2,..., n.

[0077] Introduce Lagrange multipliers to combine and transform the first objective function and the first constraint condition into the second objective function, and determine the second constraint condition.

[0078] Specifically, introduce Lagrange multipliers α i , α j , α i ≥0, β i ≥0, and the corresponding Lagrangian function is:

[0079]

[0080] Take the partial derivatives of C, r, and ξ in Equation (3) respectively, and set their derivative values to 0.

[0081]

[0082] Thus, we obtain β i = A - α i ≥0, α i ≤A, substitute it into formula (3) to get:

[0083]

[0084] where k(x i , x i ), k(x i , x j ) are both kernel functions. Let Then the first objective function is transformed into the second objective function, and the second constraint condition is obtained:

[0085]

[0086] Use Gaussian kernel non - linear mapping to map the second objective function to a related objective function.

[0087] In this embodiment, in order to further eliminate the mutual coupling relationship between the moving parts in the laser welding machine, use Gaussian kernel non - linear mapping to map the second objective function to a related objective function. Specifically, map the data points in T to the high - dimensional space F. That is: use Gaussian kernel non - linear mapping φ: x → φ(x), at this time, Due to the constraint condition and the kernel function k(x i , x i ) = 1, then the second objective function and the second constraint condition in Equation (6) are mapped to the following related objective function and related constraint condition, where the second constraint condition and the related constraint condition are the same.

[0088]

[0089] Furthermore, the improved SMO algorithm is used to process the related objective function, and the radius r and the center C of the corresponding hypersphere are obtained. Among them,

[0090] In the specific implementation process, the improved SMO algorithm can be used to obtain the optimal solution α of formula (7) * , and the improved SMO algorithm uses one-time update of two variables to solve such optimization problems, and writes formula (7) as a bivariate function about and . Among them, the multipliers other than α1 and α2 are regarded as constants. To avoid falling into the local optimal solution of the bivariate and and accelerate the global optimization process, a penalty term is added to the bivariate function as follows:

[0091]

[0092] Denote where γ is a constant. Then where L = max(0, γ - A), H = min(A, γ) is substituted into Equation (8) to obtain:

[0093]

[0094]

[0095] Let Solve to get:

[0096]

[0097] Denote It can be obtained that:

[0098]

[0099] Denote Get

[0100] From this, the optimized α1 and α2 can be obtained and According to the variable selection principle, find α that does not satisfy the condition 0 ≤ α i ≤ A i , and find the α that makes the largest j , thereby accelerating the iteration speed until all α i satisfy the constraint conditions, and the iteration terminates to obtain the optimal solution of formula (7) Furthermore, the radius r of the hypersphere and the center C of the sphere are obtained.

[0101] The above is the specific implementation process of obtaining the radius r of the hypersphere and the center C of the sphere by using the SMO algorithm.

[0102] Refer to Figure 2 , which is a schematic diagram of the hypersphere. The hypersphere contains the state data samples of the laser welding machine when it is working normally ( Figure 2 shown as normal data). Therefore, for the real-time state data x, substitute it into the fault determination function D(x) corresponding to each hypersphere for calculation and obtain the calculation result. If the calculation results are all greater than 0, it means that it is outside all hyperspheres, then the laser welding machine has a fault, otherwise the laser welding machine has no fault. If the laser welding machine has a fault, subsequent alarm, shutdown and other processing can be carried out.

[0103] Based on the same inventive concept as in the foregoing embodiments, refer to Figure 3 , this embodiment of the present invention also provides an online fault determination system for a laser welding machine, including:

[0104] An acquisition module 301, configured to acquire the real-time state data of the laser welding machine when the laser welding machine is running online;

[0105] A call module 302, configured to call the fault determination functions corresponding to several hyperspheres of the laser welding machine Wherein, x is the real-time state data, x i , x j are samples in the same subset obtained by clustering based on the laser welding machine training set, n is the number of samples in the subset, is the global optimal solution corresponding to the minimum enclosing sphere optimization solution of the hypersphere, k(x, x i ), k(x i , x j ) are kernel functions respectively, k(x, x i ) represents the inner product of x, x i , k(x i , x j ) represents x i , x jThe inner product, where r is the radius of the hypersphere;

[0106] A calculation module 303, configured to respectively input the real-time state data of the laser welding machine into the fault determination function D(x) corresponding to each of the several hyperspheres for calculation to obtain a calculation result;

[0107] A judgment module 304, configured to judge whether the calculation results are all greater than 0; if so, it indicates that the laser welding machine has a fault.

[0108] Based on the same inventive concept as in the foregoing embodiments, an embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, the steps of any of the foregoing methods are implemented.

[0109] Based on the same inventive concept as in the foregoing embodiments, an embodiment of the present invention further provides a computer device, including a memory, a processor, and a computer program stored on the memory and executable on the processor, and when the processor executes the program, the steps of any of the foregoing methods are implemented.

[0110] The algorithms and displays provided herein are not inherently related to any particular computer, virtual system, or other device. Various general-purpose systems can also be used in conjunction with the teachings herein. The structure required to construct such systems will be apparent from the above description. In addition, the present invention is not directed to any particular programming language. It should be understood that the content of the present invention described herein can be implemented using various programming languages, and the description of a particular language above is for the purpose of disclosing the best mode of the present invention.

[0111] In the specification provided herein, a large number of specific details are set forth. However, it can be understood that the embodiments of the present invention can be practiced without these specific details. In some instances, well-known methods, structures, and techniques have not been shown in detail so as not to obscure the understanding of this specification.

[0112] Similarly, it should be understood that, in order to streamline this disclosure and assist in understanding one or more of the various inventive aspects, in the foregoing description of the exemplary embodiments of the present invention, the various features of the present invention are sometimes grouped together in a single embodiment, figure, or description thereof. However, the disclosed method should not be construed as reflecting the intention that the claimed invention requires more features than are expressly recited in each claim. Rather, as reflected in the following claims, the inventive aspects lie in less than all of the features of the single foregoing disclosed embodiment. Thus, the claims following the detailed description are hereby expressly incorporated into this detailed description, with each claim standing on its own as a separate embodiment of the present invention.

[0113] Those skilled in the art can understand that the modules in the devices in the embodiments can be adaptively changed and arranged in one or more devices different from the embodiments. The modules or units or components in the embodiments can be combined into one module or unit or component, and in addition, they can be divided into multiple sub-modules or sub-units or sub-components. Except that at least some of such features and / or processes or units are mutually exclusive, any combination can be adopted to combine all the features disclosed in this specification (including the accompanying claims, abstract and drawings) and all the processes or units of any method or device so disclosed. Unless otherwise explicitly stated, each feature disclosed in this specification (including the accompanying claims, abstract and drawings) can be replaced by an alternative feature that provides the same, equivalent or similar purpose.

[0114] In addition, those skilled in the art can understand that although some of the embodiments herein include certain features included in other embodiments rather than other features, the combination of the features of different embodiments means that it is within the scope of the present invention and forms different embodiments. For example, in the following claims, any one of the claimed embodiments can be used in any combination.

[0115] Each component embodiment of the present invention can be implemented in hardware, or in software modules running on one or more processors, or in a combination thereof. Those skilled in the art should understand that a microprocessor or a digital signal processor (DSP) can be used in practice to implement some or all of the functions of some or all of the components of the gateway, proxy server, and system according to the embodiments of the present invention. The present invention can also be implemented as a device or apparatus program (such as a computer program and a computer program product) for executing part or all of the methods described herein. Such a program for implementing the present invention can be stored on a computer-readable medium, or can be in the form of one or more signals. Such signals can be downloaded from an Internet website, or provided on a carrier signal, or provided in any other form.

[0116] It should be noted that the above embodiments illustrate the present invention rather than limit the present invention, and those skilled in the art can design alternative embodiments without departing from the scope of the appended claims. In the claims, any reference signs placed between parentheses shall not be construed as limiting the claim. The word "comprising" does not exclude the presence of elements or steps not listed in the claim. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. The present invention can be implemented by means of hardware including several different elements and by means of a suitably programmed computer. In the unit claims listing several devices, several of these devices may be embodied by the same item of hardware. The use of the words first, second, and third, etc. does not denote any order. These words may be interpreted as names.

Claims

1. An online fault determination method for a laser welding machine, characterized in that, The method comprises: When the laser welder is running online, obtaining real-time status data of the laser welder; Invoke the fault determination functions corresponding to several hyperspheres of the laser welding machine wherein x is the real-time state data, x i , x j are samples in the same subset obtained by clustering based on the laser welding machine training set, n is the number of samples in the subset, is the global optimal solution corresponding to the optimization solution of the minimum enclosing sphere for the hypersphere, k(x, x i ), k(x i , x j ) are kernel functions respectively, k(x, x i ) represents the inner product of x, x i , k(x i , x j ) represents the inner product of x i , x j ; r is the radius of the hypersphere; the center C and radius r of the hypersphere are obtained according to the following steps: Obtain the laser welding machine training set; wherein, the samples in the laser welding machine training set refer to the state data samples when the laser welding machine is in a normal state; Cluster the samples in the laser welding machine training set to obtain several non-intersecting subsets; Determine the center C and radius r of the several hyperspheres based on the several non-intersecting subsets, including: For each subset, determine the relevant objective function and relevant constraint conditions: wherein A is the penalty coefficient, α i and α j are Lagrange multipliers; Use the improved sequential minimal optimization algorithm SMO algorithm to transform the relevant objective function into a bivariate objective function, and add a penalty factor to the bivariate objective function for iterative optimization to determine the center and radius of the hypersphere; The penalty factor is used to prevent the relevant objective function from falling into the local optimal solution of the bivariate; The function obtained after adding the penalty factor to the bivariate objective function is: wherein is a bivariate, α t , α p , α2 are all Lagrange multipliers, x1, x2, x t , x p , x i , x j are all samples in the same subset, k(x1, x2), k(x t , x1), k(x t , x2), k(x t , x p ) are all kernel functions, is the penalty factor; One subset corresponds to one hypersphere, and each hypersphere has its own center C and radius r; The real-time status data of the laser welder are respectively brought into the fault judgment functions D(x) corresponding to the respective hyperspheres to perform calculations to obtain calculation results; Determine whether the calculation results are all greater than 0; If so, it indicates that the laser welder has a fault.

2. The method according to claim 1, characterized in that The center C and radius r of the hypersphere are: in, Indicates x i The mapping relationship; 3. The method according to claim 2, characterized in that Before obtaining the real-time status data of the laser welder, the method further includes: Determine the center C and radius r of the hypersphere; Obtain the fault determination function D(x) by using the center C and radius r of the hypersphere; where, represents the mapping relationship of x.

4. The method according to claim 1, characterized in that, Determining relevant objective functions and relevant constraints for each subset includes: For each of the subsets, establishing a corresponding first objective function and a first constraint condition; Introducing a Lagrange multiplier to combine the first objective function and the first constraint condition into a second objective function, and determining a second constraint condition; wherein the second constraint condition is the same as the related constraint condition; The second objective function is mapped into the related objective function using Gaussian kernel nonlinear mapping.

5. The method according to claim 4, characterized in that The first objective function is as follows: where x i is any sample data vector in the subset, A is the penalty coefficient, and ξ i is the slack variable, ξ i ≥0, i = 1, 2,..., n; The first constraint condition is as follows: The second objective function is: Among them, α i , α j is the Langrange multiplier, k(x i ,x i ) is the kernel function.

6. An online fault judgment system for a laser welder, characterized in that: include: An acquisition module, used for obtaining real-time status data of the laser welder when the laser welder is running online; A calling module is used to call the fault judgment functions corresponding to the respective hyperspheres of the laser welding machine in, x is the real-time status data, x i 、x j are samples in the same subset obtained by clustering based on the laser welder training set, n is the number of samples in the subset, is the global optimal solution corresponding to the minimum closed sphere optimization solution of the hypersphere, k(x,x i )、k(x i ,x j ) are kernel functions, k(x,x i ) represents x,x i The inner product of k(x i ,x j ) represents x i ,x j The inner product of r is the radius of the hypersphere; the center C and the radius r of the hypersphere are obtained according to the following steps: obtaining a laser welder training set; wherein the samples in the laser welder training set refer to state data samples when the laser welder is in a normal state; clustering the samples in the laser welder training set to obtain a number of non-overlapping subsets; determining the centers C and radii r of the hyperspheres based on the non-overlapping subsets, including: determining, for each of the subsets, a relevant objective function and relevant constraints: Wherein, A is the penalty coefficient; the improved sequential minimum optimization algorithm (SMO) is used to convert the relevant objective function into a two-variable objective function, and the penalty factor is added to the two-variable objective function to perform iterative optimization to determine the center and radius corresponding to the hypersphere; the penalty factor is used to prevent the relevant objective function from falling into the local optimal solution of the two variables; the function obtained after adding the penalty factor to the two-variable objective function is: in, is a bivariate, α t , α p , α2 are all Lagrange multipliers, x1, x2, x t 、x p 、x i 、x j are all samples from the same subset, k(x1,x2), k(x t ,x1),k(x t ,x2),k(x t ,x p ) are kernel functions, is the penalty factor; one subset corresponds to a hypersphere, and each hypersphere has its own center C and radius r; a calculation module, configured to bring the real-time status data of the laser welder into the fault judgment functions D(x) corresponding to the respective hyperspheres for calculation to obtain calculation results; The judgment module is used to judge whether the calculation results are all greater than 0; if so, it indicates that there is a fault in the laser welder.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by a processor, the steps of the method according to any one of claims 1 to 5 are implemented.

8. A computer device, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, When the processor executes the program, the steps of the method according to any one of claims 1 to 5 are implemented.

Images