Online monitoring and fault diagnosis method and system for laser cutting machine

Abstract

The invention relates to the technical field of laser cutting machine fault diagnosis, and discloses a laser cutting machine online monitoring and fault diagnosis method and system.The method comprises the steps that working state information of a laser cutting machine is collected, and a working state feature vector is constructed; constructing a fault prediction model of the laser cutting machine by using a Stacking ensemble learning method to predict a fault type and an occurrence probability of the fault type, and generating a 3D model of the laser cutting machine to carry out dynamic visual fault early warning; compared with the technical problems that complex data combination cannot be effectively processed and the prediction accuracy is poor especially under the condition of multi-dimensional sensor data and various fault types in the prior art, the machine learning model and the Stacking ensemble learning method are combined, so that the faults can be efficiently classified and predicted, and the fault prediction accuracy is improved. And whether the laser cutting machine has potential faults or not is accurately judged, so that the problem that accurate fault early warning cannot be carried out in real time in a traditional method is avoided.

Description

Technical Field

[0001] The present invention belongs to the technical field of laser cutting machine fault diagnosis, and particularly relates to an online monitoring and fault diagnosis method and system for a laser cutting machine. Background Art

[0002] At present, laser cutting machines are widely used in the manufacturing industry, especially in scenarios with high-precision and high-efficiency cutting requirements. However, various problems will inevitably occur during the long-term operation of laser cutting machines, resulting in equipment failures and shutdowns, which in turn affect production efficiency and product quality. Most of the existing laser cutting machine fault diagnosis methods rely on regular maintenance, manual inspections, or simple sensor monitoring. These methods have the following deficiencies: Fault warning is lagging. Most of the existing technologies can only alarm after a fault occurs and cannot detect potential fault risks in time before the fault occurs. For example, problems such as laser power fluctuations and manipulator overloads are often detected only after the equipment is severely damaged, resulting in low production efficiency and equipment damage; The accuracy of data analysis is low. The existing monitoring methods usually rely on single-sensor data and cannot effectively integrate the data of multiple sensors for comprehensive analysis. For example, it is difficult to accurately predict the overall health status of the equipment only through the data analysis of a laser power sensor or a temperature sensor, especially when the equipment is working under high load, and there are complex interactions between various parts of the equipment; Unable to process multi-dimensional data. Traditional methods usually have difficulty in dealing with the interaction effects of data from multiple sensors and equipment states, which affects the accuracy of fault diagnosis. In practical applications, the working state of a laser cutting machine is affected by multiple factors, including laser power, cutting area temperature, manipulator vibration, conveyor belt load, etc.; Strong dependence on manual intervention. Traditional fault diagnosis methods require operators to make decisions based on the results of manual inspections, and there are often human errors. Therefore, there is an urgent need for an online monitoring and fault diagnosis method and system for a laser cutting machine, which can, in a complex working environment, combine multi-dimensional sensor data and intelligent fault diagnosis methods to monitor the running state of the laser cutting machine in real time and accurately predict faults through intelligent analysis before the faults occur, so as to improve production efficiency and equipment reliability. Summary of the Invention

[0003] Aiming at the above-mentioned existing technical deficiencies, the purpose of the present invention is to propose an online monitoring and fault diagnosis method for a laser cutting machine, aiming to solve the technical problems in the prior art that, especially in the face of multi-dimensional sensor data and various fault types, complex data combinations cannot be effectively processed and the accuracy of fault prediction is poor.

[0004] To solve the above technical problems, the present invention adopts the following technical solutions: The present invention provides an online monitoring and fault diagnosis method for a laser cutting machine,

[0005] The online monitoring and fault diagnosis method for the laser cutting machine includes:

[0006] Step S10: The working state information of the laser cutting machine is collected in real time by sensors installed in the core part of the laser cutting machine, and the working state information is transmitted to a pre-constructed cloud database through the Ajax protocol;

[0007] Step S20: Obtain the working state information from the cloud database, and construct a normalization processing function, a Fourier transform function, and a moving window average function to process the working state information to obtain a working state feature vector;

[0008] Step S30: Construct a first laser cutting machine fault prediction model through the support vector machine algorithm SVM, construct a second laser cutting machine fault prediction model through the random forest algorithm RF, and use the Stacking ensemble learning method to use the outputs of SVM and RF as new features to construct a third laser cutting machine fault prediction model;

[0009] Step S40: Train the third laser cutting machine fault prediction model, and use the working state feature vector as the input of the trained model to obtain the fault type and its occurrence probability;

[0010] Step S50: Use PyThreeJS to generate a 3D model of the laser cutting machine, and perform visual warning in the 3D model according to the fault type and its occurrence probability.

[0011] Preferably, in step S30, the formula of the first laser cutting machine fault prediction model f(x) is:

[0012]

[0013] where x is the working state feature vector, N is the preset number of samples during model training, x i is the working state feature vector of sample i during preset model training, y i is the fault label during preset model training, K(x i , x) is the radial basis function, α i is the coefficient controlling the influence of each sample on classification, b is the preset bias term, and sgn is the activation function; the output of model f(x) is the fault type, and the probability P fault of the fault occurrence is obtained through Platt scaling, where A and B are parameters obtained through training.

[0014] Preferably, in step S30, the formula of the second laser cutting machine fault prediction model F RF is:

[0015]

[0016] Among them, T j (x) is the prediction result of the working state feature vector of the j-th pre-set decision tree; the model F RF outputs the fault type, and the probability of the fault occurrence is obtained through the probabilistic model Among them, σ is the Sigmoid function, which is used to convert the output of the tree into a probability value.

[0017] Preferably, in step S30, the third laser cutting machine fault prediction model F stacking has the formula:[[]]

[0018] F stacking =σ(W1·f(x)+W2·F RF +b)

[0019] Among them, W1 and W2 are the weights used to weight the outputs of SVM and RF in the third laser cutting machine fault prediction model F stacking

[0020] Preferably, in step S50, the fault types include abnormal laser power fluctuation, abnormal manipulator vibration, abnormal conveyor belt vibration, and abnormal temperature fluctuation in the cutting area; the steps of using PyThreeJS to generate the 3D model of the laser cutting machine specifically include: installing and importing the PyThreeJS library, creating a scene and a renderer, generating the laser source, manipulator, and conveyor belt data of the laser cutting machine, adding the laser source, manipulator, and conveyor belt data of the laser cutting machine to the scene, using the renderer to render the 3D scene to the browser, and using the animation loop function to update the 3D model in real time in combination with the working state information in the cloud database in step S10.

[0021] Preferably, in step S10, the working state information includes the output power of the laser source, the temperature of the cutting area, the manipulator vibration data, and the conveyor belt vibration data.

[0022] Preferably, in step S20, the working state feature vector includes the stability feature of the laser power, the vibration frequency feature of the manipulator, the vibration frequency feature of the conveyor belt, and the temperature fluctuation feature of the cutting area.

[0023] The present invention also provides a laser cutting machine online monitoring and fault diagnosis system, including:[[]]

[0024] A data acquisition and transmission module, which is used to collect the working state information of the laser cutting machine in real time through sensors installed in the core part of the laser cutting machine, and transmit the working state information to a pre-constructed cloud database through the Ajax protocol;

[0025] ​A data processing and feature construction module, which is used to obtain working status information from a cloud database, and construct a normalization processing function, a Fourier transform function, and a moving window average function to process the working status information to obtain a working status feature vector;

[0026] A fault prediction model construction module, which is used to construct a first laser cutting machine fault prediction model through the support vector machine algorithm SVM, construct a second laser cutting machine fault prediction model through the random forest algorithm RF, and use the Stacking ensemble learning method to use the outputs of SVM and RF as new features to construct a third laser cutting machine fault prediction model;

[0027] A model training and fault diagnosis module, which is used to train the third laser cutting machine fault prediction model, and use the working status feature vector as the input of the trained model to obtain the fault type and its occurrence probability;

[0028] A 3D visualization and fault warning module, which is used to generate a 3D model of the laser cutting machine using PyThreeJS, and perform visual warning in the 3D model according to the fault type and its occurrence probability.

[0029] The present invention also provides a computer program product, including an online monitoring and fault diagnosis program for a laser cutting machine. When the online monitoring and fault diagnosis program for the laser cutting machine is executed by a processor, the online monitoring and fault diagnosis method for the laser cutting machine is implemented.

[0030] The beneficial effects of the present invention are as follows: Compared with the prior art, especially in the face of multi-dimensional sensor data and diverse fault types, it is impossible to effectively process complex data combinations, and the accuracy of fault prediction is poor. Since the present invention combines machine learning models, it can accurately classify and predict multi-dimensional sensor data. Especially in the case of diverse fault types, the machine learning model can effectively learn the characteristics of each fault type, thereby improving the accuracy of prediction;

[0031] Since the Stacking method of the present invention combines multiple different machine learning models, it can make up for the deficiencies of a single model and improve the comprehensive accuracy of fault prediction; Description of the Drawings

[0032] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0033] Figure 1Schematic flow chart of the first embodiment of an on-line monitoring and fault diagnosis method for a laser cutting machine according to the present invention.

[0034] Figure 2 Schematic diagram of the equipment for an on-line monitoring and fault diagnosis method for a laser cutting machine according to the present invention. Detailed implementation manners

[0035] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0036] Embodiment 1: As Figure 1 shown, it is a schematic flow chart of the first embodiment of the on-line monitoring and fault diagnosis method for a laser cutting machine according to the present invention, and the first embodiment of the on-line monitoring and fault diagnosis method for a laser cutting machine according to the present invention is proposed.

[0037] In the first embodiment, the on-line monitoring and fault diagnosis method for the laser cutting machine includes:

[0038] Step S10: Real-time collect the working state information of the laser cutting machine through sensors installed in the core part of the laser cutting machine, and transmit the working state information to a pre-constructed cloud database through the Ajax protocol;

[0039] It should be noted that in step S10, the working state information includes the output power of the laser source, the temperature of the cutting area, the vibration data of the manipulator, and the vibration data of the conveyor belt. The Ajax protocol is an asynchronous data transmission method based on the HTTP protocol, which can realize real-time data exchange between the client and the cloud database, especially during the operation of the laser cutting machine, ensuring the immediacy and accuracy of the data.

[0040] It can be understood that the output power of the laser source reflects the cutting ability and stability of the laser cutting machine. Excessive temperature in the cutting area may affect the cutting quality. Excessive vibration may be a warning signal of manipulator jamming or overload. Abnormal vibration of the conveyor belt may cause waste not to be recycled in time or conveyor belt jamming. The data collected in real time by these sensors is transmitted to the cloud database through the Ajax protocol for subsequent data analysis and fault diagnosis.

[0041] Step S20: Obtain the working state information from the cloud database, and construct a normalization processing function, a Fourier transform function, and a moving window average function to process the working state information to obtain a working state feature vector;

[0042] It should be noted that in step S20, the working state feature vector includes the stability feature of the laser power, the vibration frequency feature of the manipulator, the vibration frequency feature of the conveyor belt, and the temperature fluctuation feature of the cutting area.

[0043] It can be understood that the working state feature vector is obtained by processing and analyzing the data collected by the sensors. The processed feature vector can accurately reflect the operating state of the laser cutting machine. The working state feature vector includes information in multiple dimensions, and through this information, it can help accurately predict whether the equipment has a fault and determine the type of fault and the probability of the occurrence of the fault.

[0044] It should be understood that the normalization processing function, the Fourier transform function, and the moving window average function are used to extract meaningful features from the original data to ensure that the subsequent fault diagnosis model can obtain reliable input data. Through the application of these functions, the noise interference can be effectively reduced, and the accuracy of fault prediction can be improved.

[0045] For example, during the operation of the laser cutting machine, the laser power may fluctuate due to different cutting materials or the aging of the laser source. At this time, the stability feature of the laser power can help detect whether there is an unstable power situation; the feature extracted from the vibration frequency of the manipulator through Fourier transform can reveal whether there are signs of jamming or overloading of the manipulator; the vibration frequency feature of the conveyor belt can help identify whether there is jamming or abnormal wear of the conveyor belt; the temperature fluctuation feature of the cutting area helps detect whether there is abnormal temperature, so as to judge whether it will affect the cutting quality or cause equipment damage.

[0046] Step S30: Construct the first laser cutting machine fault prediction model through the support vector machine algorithm SVM, construct the second laser cutting machine fault prediction model through the random forest algorithm RF, and use the Stacking ensemble learning method to construct the third laser cutting machine fault prediction model with the outputs of SVM and RF as new features;

[0047] It should be noted that in step S30, the formula for the first laser cutting machine fault prediction model f(x) is:

[0048]

[0049] where x is the working state feature vector, N is the number of samples during the preset model training, x i is the working state feature vector of sample i during the preset model training, y i is the fault label during the preset model training, K(x i , x) is the radial basis function, α iα is the coefficient to control the influence of each sample on classification, b is the preset bias term, and sgn is the activation function; the output of the model f(x) is the fault type, and the probability P of the fault occurrence is obtained through Platt scaling. fault , where A and B are parameters obtained through training.

[0050] In step S30, the formula of the second laser cutting machine fault prediction model F RF is:

[0051]

[0052] where T j (x) is the prediction result of the working state feature vector of the preset j-th decision tree; the output of the model F RF is the fault type, and the probability of the fault occurrence is obtained through a probability model where σ is the Sigmoid function, which is used to convert the output of the tree into a probability value.

[0053] In step S30, the formula of the third laser cutting machine fault prediction model F stacking is:

[0054] F stacking =σ(W1·f(x)+W2·F RF +b)

[0055] where W1 and W2 are the weights used to weight the outputs of the SVM and RF in the third laser cutting machine fault prediction model F stacking .

[0056] It can be understood that by adopting the Stacking ensemble learning method, the outputs of the SVM and RF models (i.e., their prediction results of the fault state) are used as new features and input into a meta-learning model (such as Logistic regression, neural network, etc.), further combining the advantages of the SVM and RF. The Stacking method can give full play to their advantages while ensuring the independence of each base learner, and improve the accuracy of the final prediction by combining multiple prediction results.

[0057] For example, assume that the working state feature vector of a laser cutting machine includes features such as laser power fluctuation, manipulator vibration frequency, conveyor belt vibration frequency, and cutting area temperature. When using SVM, the model may determine whether a fault occurs based on laser power fluctuation and manipulator vibration frequency; while when using RF, the model may rely on conveyor belt vibration frequency and temperature fluctuation for prediction. Through the Stacking method, the outputs of SVM and RF (such as the probability of predicting a fault) are used as new features and input into the meta-learning model, so as to obtain a more accurate fault prediction result. For example, if the SVM model predicts the probability of a laser power fluctuation fault to be 80%, and the RF model predicts the probability of a manipulator jamming fault to be 75%, the Stacking model will combine these two prediction values and obtain the final fault probability prediction through weighted average or other methods.

[0058] Step S40: Train the third laser cutting machine fault prediction model, and use the working state feature vector as the input of the trained model to obtain the fault type and its occurrence probability;

[0059] It can be understood that the training process involves using a historical dataset, including data in normal states and different fault types. These data will be used to train the Stacking model, that is, by combining the outputs of the SVM and RF models, to optimize the prediction accuracy. During the training process, methods such as cross-validation are used to ensure the stability and generalization ability of the model, so that the model can adapt to fault diagnosis under different future working states.

[0060] It should be understood that the prediction results of the fault type and its occurrence probability are obtained through the analysis and calculation of the model on the input feature vector. The output of the model includes the category of the fault and the probability of the occurrence of this fault, usually outputting a value ranging from 0 to 1, indicating the possibility of the occurrence of this fault.

[0061] For example, the working state feature vector obtained through step S20 includes the stability feature ΔP(t) of the laser power, the vibration frequency feature V hand (t) of the manipulator, the vibration frequency feature V belt (t) of the conveyor belt, and the temperature fluctuation feature ΔT(t) of the cutting area. Input these features into the Stacking ensemble learning model for training, and the model learns the relationships between different fault modes (such as laser power fluctuation and manipulator jamming) and these features based on historical data. The trained model can predict the fault type and its occurrence probability in real time. For example, when the newly input working state feature vector is: laser power fluctuation ΔP(t)=0.4, manipulator vibration frequency V hand (t)=1.2, conveyor belt vibration frequency V beltWhen P(t) = 1.1 and the temperature fluctuation ΔT(t) in the cutting area = 2.0, the model may predict the fault type of "laser power fluctuation" and output the probability of this fault occurring as 85%, prompting the operator that there may be a problem with the laser power and that an inspection is required.

[0062] Step S50: Use PyThreeJS to generate a 3D model of the laser cutting machine and perform visual warning in the 3D model according to the fault type and its occurrence probability.

[0063] It should be noted that this step uses the PyThreeJS library to generate a 3D visualization model of the laser cutting machine. Through PyThreeJS, interactive 3D graphics can be created in a Web environment, and each component of the laser cutting machine (such as the laser source, manipulator, conveyor belt, etc.) can be displayed in the model. The status of each component can be dynamically updated according to different fault types and occurrence probabilities through color changes, animation effects, etc., so as to provide intuitive warning information for the operator.

[0064] It should be understood that the generated 3D model is not only a static display tool, but a dynamic interaction interface closely integrated with the fault prediction system. The colors, animations, and labels in the 3D model will change according to the prediction results of the fault type, providing real-time feedback to help the operator understand the operating status of the cutting machine and its potential problems. Specifically, laser power fluctuations, abnormal vibrations of the manipulator, or conveyor belt load problems, etc. will all be manifested in the 3D model.

[0065] For example, during the operation of the laser cutting machine, the laser power sensor detects laser power fluctuations, and according to the prediction of the machine learning model, the probability of the fault occurring is 85%. In the 3D model, the part of the laser source will automatically turn red to indicate the high fault risk of this component. Similarly, if the probability of the fault of the manipulator vibration is 90%, the manipulator part will also turn red to remind the operator to check. If the probability of the fault occurring is low, the components of the model will continue to be displayed in green or blue, indicating their normal operating status. In addition, the operator can interact through the 3D interface, rotate and zoom the view, view the status of each component, and perform fault troubleshooting according to the report content.

[0066] Embodiment 2: In addition, an online monitoring and fault diagnosis system for a laser cutting machine provided by the present invention adopts an online monitoring and fault diagnosis method for a laser cutting machine in the above embodiment, and can solve the technical problem of online monitoring and fault diagnosis of a laser cutting machine. Compared with the prior art, the beneficial effects of the online monitoring and fault diagnosis system for a laser cutting machine provided by the present invention are the same as those of the online monitoring and fault diagnosis method for a laser cutting machine provided by the above embodiment, and other technical features in the online monitoring and fault diagnosis system for a laser cutting machine are the same as the features disclosed in the method of the above embodiment, and will not be elaborated herein.

[0067] Embodiment 3: The present invention provides an online monitoring and fault diagnosis device for a laser cutting machine. Please refer to Figure 2, An on-line monitoring and fault diagnosis device for a laser cutting machine includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein, the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute an on-line monitoring and fault diagnosis method for a laser cutting machine in the first embodiment above. The on-line monitoring and fault diagnosis device in the embodiment of the present invention may include, but is not limited to, mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (Personal Digital Assistant), PADs (Portable Application Description: tablet computers), PMPs (Portable Media Players), in-vehicle terminals (such as in-vehicle navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. An on-line monitoring and fault diagnosis device for a laser cutting machine is merely an example and should not impose any limitations on the functions and usage scope of the embodiments of the present invention. An on-line monitoring and fault diagnosis device for a laser cutting machine may include a processing device 1001 (such as a central processing unit, a graphics processing unit, etc.), which may perform various appropriate actions and processes according to a program stored in a read-only memory (ROM: Read Only Memory) 1002 or a program loaded from a storage device 1003 into a random access memory (RAM: Random Access Memory) 1004. In the RAM 1004, various programs and data required for the operation of an on-line monitoring and fault diagnosis device for a laser cutting machine are also stored. The processing device 1001, the ROM 1002, and the RAM 1004 are connected to each other through a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Generally, the following systems may be connected to the I / O interface 1006: an input device 1007 including, for example, a touch screen, a touchpad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, etc.; an output device 1008 including, for example, a liquid crystal display (LCD: Liquid Crystal Display), a speaker, a vibrator, etc.; a storage device 1003 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 1009. The communication device 1009 may allow an on-line monitoring and fault diagnosis device for a laser cutting machine to communicate with other devices wirelessly or wiredly to exchange data. Although an on-line monitoring and fault diagnosis device for a laser cutting machine with various systems is shown in the figure, it should be understood that it is not required to implement or have all the shown systems. More or fewer systems may be alternatively implemented or had.

[0068] Embodiment 4: The present invention further provides a computer program product, including a computer program which, when executed by a processor, implements the steps of an online monitoring and fault diagnosis method for a laser cutting machine as described above. The computer program product provided by the present invention can solve the technical problem of online monitoring and fault diagnosis of a laser cutting machine. Compared with the prior art, the beneficial effects of the computer program product provided by the present invention are the same as those of the online monitoring and fault diagnosis method for a laser cutting machine provided in the above embodiment, and will not be elaborated here.

[0069] Specifically, according to the embodiments disclosed by the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, the embodiments disclosed by the present invention include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program contains program codes for executing the methods shown in the flowcharts. In such an embodiment, the computer program can be downloaded and installed from the network through a communication device, or installed from a storage device 1003, or installed from a ROM 1002. When the computer program is executed by a processing device 1001, it executes the above functions defined in the methods of the embodiments disclosed by the present invention.

[0070] It should be understood that various parts of the present invention can be implemented by hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in a suitable manner in any one or more embodiments or examples.

[0071] Obviously, those skilled in the art can make various modifications and variations to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention is also intended to include these modifications and variations.

Claims

1. An online monitoring and fault diagnosis method for a laser cutting machine, characterized in that, The method includes: Step S10: Real-time collect the working state information of the laser cutting machine through the sensors installed in the core part of the laser cutting machine, and transmit the working state information to the pre-constructed cloud database through the Ajax protocol; Step S20: Obtain the working state information from the cloud database, and construct a normalization processing function, a Fourier transform function, and a moving window average function to process the working state information to obtain a working state feature vector; Step S30: Construct a first laser cutting machine fault prediction model through the support vector machine algorithm SVM, construct a second laser cutting machine fault prediction model through the random forest algorithm RF, and use the Stacking ensemble learning method to use the outputs of SVM and RF as new features to construct a third laser cutting machine fault prediction model; Step S40: Train the third laser cutting machine fault prediction model, and use the working state feature vector as the input of the trained model to obtain the fault type and its occurrence probability; Step S50: Use PyThreeJS to generate a 3D model of the laser cutting machine, and perform visual warning in the 3D model according to the fault type and its occurrence probability.

2. The online monitoring and fault diagnosis method of a laser cutting machine according to claim 1, characterized in that, In step S30, the formula of the first laser cutting machine fault prediction model f(x) is: Among them, x is the working state feature vector, N is the number of samples during the preset model training, and x i is the working state feature vector of sample i during the preset model training, and y i is the fault label during the preset model training, K(x i , x) is the radial basis function, and α i is the coefficient that controls the influence of each sample on classification, b is the preset bias term, and sgn is the activation function; the output of the model f(x) is the fault type, and the probability P of the fault occurrence is obtained through Platt scaling fault , where A and B are parameters obtained through training.

3. The online monitoring and fault diagnosis method of a laser cutting machine according to claim 2, characterized in that, In step S30, the formula of the second laser cutting machine fault prediction model F RF is as follows: Among them, T j (x) is the prediction result of the working state feature vector of the j-th pre-set decision tree; the model F RF outputs the fault type, and the probability of the fault occurrence is obtained through the probabilistic model where σ is the Sigmoid function, which is used to convert the output of the tree into a probability value.

4. The online monitoring and fault diagnosis method of a laser cutting machine according to claim 3, characterized in that, In step S30, the formula of the third laser cutting machine fault prediction model F stacking is as follows: F stacking = σ(W1·f(x)+W2·F RF +b) Among them, W1 and W2 are the weights used for weighting the outputs of SVM and RF in the third laser cutting machine fault prediction model F stacking ​ 5. The online monitoring and fault diagnosis method of a laser cutting machine according to claim 4, characterized in that, In step S50, the fault types include abnormal laser power fluctuation, abnormal manipulator vibration, abnormal conveyor belt vibration, and abnormal temperature fluctuation in the cutting area; The steps of using PyThreeJS to generate a 3D model of the laser cutting machine specifically include: installing and importing the PyThreeJS library, creating a scene and a renderer, generating the laser source, manipulator, and conveyor belt data of the laser cutting machine, adding the laser source, manipulator, and conveyor belt data of the laser cutting machine to the scene, using the renderer to render the 3D scene to the browser, and using the animation loop function to update the 3D model in real time in combination with the working state information in the cloud database in step S10.

6. The online monitoring and fault diagnosis method of a laser cutting machine according to claim 1, characterized in that, In step S10, the working state information includes the output power of the laser source, the temperature of the cutting area, the manipulator vibration data, and the conveyor belt vibration data.

7. The online monitoring and fault diagnosis method of a laser cutting machine according to claim 1, characterized in that, In step S20, the working state feature vector includes the stability feature of the laser power, the vibration frequency feature of the manipulator, the vibration frequency feature of the conveyor belt, and the temperature fluctuation feature of the cutting area.

8. An online monitoring and fault diagnosis system for a laser cutting machine, characterized in that, The online monitoring and fault diagnosis system of the laser cutting machine includes: A data acquisition and transmission module, which is used to real-time collect the working state information of the laser cutting machine through the sensors installed in the core part of the laser cutting machine, and transmit the working state information to the pre-constructed cloud database through the Ajax protocol; A data processing and feature construction module, which is used to obtain the working state information from the cloud database, and construct a normalization processing function, a Fourier transform function, and a moving window average function to process the working state information to obtain a working state feature vector; The fault prediction model construction module is used to construct the first laser cutting machine fault prediction model through the support vector machine algorithm SVM, construct the second laser cutting machine fault prediction model through the random forest algorithm RF, and use the Stacking ensemble learning method to construct the third laser cutting machine fault prediction model with the outputs of SVM and RF as new features; The model training and fault diagnosis module is used to train the third laser cutting machine fault prediction model, and use the working state feature vector as the input of the trained model to obtain the fault type and its occurrence probability; The 3D visualization and fault warning module is used to generate a 3D model of the laser cutting machine using PyThreeJS and perform visual warning in the 3D model according to the fault type and its occurrence probability.

9. An online monitoring and fault diagnosis device for a laser cutting machine, characterized in that, The laser cutting machine online monitoring and fault diagnosis device includes: a memory, a processor, and a laser cutting machine online monitoring and fault diagnosis program stored on the memory and executable on the processor. When the laser cutting machine online monitoring and fault diagnosis program is executed by the processor, it implements the laser cutting machine online monitoring and fault diagnosis method according to any one of claims 1 to 7.

10. A computer program product, characterized in that, The computer program product includes a laser cutting machine online monitoring and fault diagnosis program. When the laser cutting machine online monitoring and fault diagnosis program is executed by a processor, it implements the laser cutting machine online monitoring and fault diagnosis method according to any one of claims 1 to 7.

Images