CN117798498A - Method and system for automatically adjusting welding abnormality of intelligent laser welding machine - Google Patents

Abstract

The invention relates to the field of remote monitoring, and discloses a method and a system for automatically adjusting welding abnormality of an intelligent laser welding machine, wherein the method comprises the following steps: collecting historical operation data and current operation data of the laser welding machine, identifying historical operation abnormality corresponding to the historical operation data, identifying an operation state of the laser welding machine, identifying abnormal influence parameters of the laser welding machine, and analyzing abnormal association relation between the abnormal influence parameters and the laser welding machine; analyzing the continuous operation state of the laser welding machine, identifying the abnormal state of the continuous operation state, and constructing a parameter particle swarm of the laser welding machine; calculating the state fitness of the particle swarm corresponding to the particle swarm, determining the optimal particle swarm of the laser welding machine, adjusting the welding parameters of the laser welding machine to obtain adjustment parameters, and realizing abnormal automatic adjustment of the laser welding machine based on the adjustment parameters. The invention can improve the intellectualization of automatic adjustment of welding abnormality of the intelligent laser welding machine.

Description

Method and system for automatically adjusting welding abnormality of intelligent laser welding machine

Technical Field

The invention relates to the field of remote monitoring, in particular to a method and a system for automatically adjusting welding abnormality of an intelligent laser welding machine.

Background

A laser welder is a modern welding device that heats a material with a high energy laser beam to locally melt the material and form a weld, thereby completing a welding process of a metal or other material. Automatic adjustment of welding abnormality of the laser welding machine realizes automation and intellectualization of the welding process, improves welding quality and production efficiency, and reduces labor cost.

At present, the automatic adjustment of the welding abnormality of the laser welder mainly comprises the steps of collecting welding parameters of the welding process of the intelligent laser welder through a sensor, identifying the abnormality and automatically adjusting the welding parameters by identifying whether the welding parameters are larger than a preset standard threshold value, wherein the method can only identify and adjust the abnormality which occurs, and can not prevent the abnormality which possibly occurs, so that the automatic adjustment of the welding abnormality of the intelligent laser welder is not intelligent.

Disclosure of Invention

The invention provides a method and a system for automatically adjusting welding abnormality of an intelligent laser welding machine, which mainly aim to improve the intellectualization of automatically adjusting the welding abnormality of the intelligent laser welding machine.

In order to achieve the above object, the present invention provides a method for automatically adjusting welding abnormality of an intelligent laser welding machine, comprising:

Collecting historical operation data and current operation data of a laser welding machine, identifying historical operation abnormality corresponding to the historical operation data, and performing abnormality marking on the historical operation data based on the historical operation abnormality to obtain marked operation data;

constructing an anomaly identification network architecture of the laser welding machine, optimizing network architecture parameters of the anomaly identification network architecture based on the marking operation data and a preset loss function to obtain an optimized anomaly identification model, and calculating model anomaly identification performance of the optimized anomaly identification model;

when the model abnormality recognition performance accords with a preset standard performance, based on the current operation data, recognizing the operation state of the laser welding machine by using the optimized abnormality recognition model, recognizing abnormality influence parameters of the laser welding machine, and analyzing abnormality association relation between the abnormality influence parameters and the laser welding machine;

analyzing the continuous operation state of the laser welding machine based on the abnormal association relation and the operation state, identifying the abnormal state of the continuous operation state, and constructing a parameter particle swarm of the laser welding machine based on the abnormal state;

Calculating the state fitness of the particle group corresponding to the parameter particle group by using a preset welding state evaluation function, determining the optimal particle group of the laser welding machine based on the state fitness, adjusting the welding parameters of the laser welding machine based on the optimal particle group, obtaining adjustment parameters, and realizing abnormal automatic adjustment of the laser welding machine based on the adjustment parameters.

Optionally, the performing exception marking on the historical operation data based on the historical operation exception to obtain marked operation data includes:

constructing a unified time axis of the historical operation abnormality and the historical operation data;

serializing the historical operation data based on the unified time axis to obtain serial historical operation data;

identifying a time node of the historical operational anomaly;

extracting historical operation data of a target sequence corresponding to the time node;

and marking the sequence historical operation data to obtain the marked operation data.

Optionally, the constructing an anomaly identification network architecture of the laser welder includes:

identifying a monitoring requirement of the laser welding machine;

calculating a demand complexity coefficient of the monitoring demand;

Determining an architecture component of the laser welder based on the demand complexity factor;

and constructing an anomaly identification network architecture of the laser welding machine based on the architecture component.

Optionally, the optimizing the network architecture parameters of the anomaly identification network architecture based on the tag operation data and a preset loss function to obtain an optimized anomaly identification model includes:

calculating an architecture parameter gradient of an architecture parameter corresponding to the abnormal identification network architecture by using the marking operation data and the loss function;

calculating target architecture parameters of the anomaly identification network architecture based on the architecture parameter gradient;

and optimizing network architecture parameters of the anomaly identification network architecture based on the architecture parameter gradient to obtain an optimized anomaly identification model.

Optionally, calculating an architecture parameter gradient of the architecture parameter corresponding to the anomaly identification network architecture by using the tag operation data and the loss function includes:

extracting welding parameters from the marking operation data;

outputting a predicted value of the framework by utilizing the abnormal recognition network framework based on the welding parameters;

based on the architecture predicted value and the loss function, calculating an architecture parameter gradient of an architecture parameter corresponding to the anomaly identification network architecture by using the following formula:

；

Wherein,architecture parameter gradient representing weights in corresponding architecture parameters of network architecture, +.>Representing a loss function->Architecture predictor +.>Gradient of->Representing anomaly identificationActivating function in network architecture to recognize neuron output value corresponding to network architecture abnormality +.>Derivative of>Representing neuron output value +.>Weight in architecture parameters corresponding to network architecture>Is used for the gradient of (a),architecture parameter gradient representing bias in corresponding architecture parameters of the network architecture, +.>Representing neuron output value +.>Bias in corresponding architecture parameters of network architecture>Is a gradient of (a).

Optionally, when the model anomaly identification performance meets a preset standard performance, identifying the operation state of the laser welding machine by using the optimized anomaly identification model based on the current operation data, including:

when the model anomaly identification performance accords with a preset standard performance, performing feature extraction on the current operation data by utilizing a convolution layer in the optimized anomaly identification model to obtain a data feature map;

extracting a feature map from the data feature map by using a pooling layer in the optimization anomaly identification model to obtain a target feature map;

introducing nonlinear properties into the target feature map by using an activation function in the optimized anomaly identification model to obtain a nonlinear property feature map;

And connecting the nonlinear property characteristic graphs by using a full connection layer in the optimized abnormal recognition model to obtain the running state of the laser welding machine.

Optionally, the identifying the abnormal influence parameter of the laser welder includes:

constructing a parameter test group of welding parameters corresponding to the laser welding machine;

identifying a test running state of the laser welder corresponding to the test welding parameters in the parameter test group;

based on the parameter test group and the test running state, constructing an association relation curve between corresponding welding parameters and running states of the laser welding machine;

and identifying abnormal influence parameters of the laser welding machine based on the association relation curve.

Optionally, the identifying, based on the association relationship, an abnormal influence parameter of the laser welding machine includes:

identifying a fluctuation area of the association relation curve;

calculating the fluctuation coefficient of the association relation curve based on the fluctuation area by using the following formula:

；

wherein,fluctuation coefficient representing association curve, +.>Represents the relation of->Fluctuation amplitude of individual fluctuation areas, +.>Representing the association curve->Represents the relation of- >The angular frequency of the highest point corresponding to the wave zone, < >>Represents the relation of->The individual fluctuation zones correspond to the highest values,/->Represents the relation of->The individual fluctuation zones correspond to the lowest value, +.>Represents the relation of->Duration of the individual fluctuation zones, +.>Representing the number of fluctuation areas corresponding to the association relation curve;

and determining an abnormal influence parameter of the laser welding machine based on the fluctuation coefficient.

Optionally, the analyzing the continuous operation state of the laser welder based on the abnormal association relationship and the operation state includes:

identifying a parameter predicted value of a welding parameter corresponding to the laser welding machine;

analyzing the predicted running state of the laser welding machine based on the parameter predicted value and the abnormal association relation;

serializing the predicted running state to obtain a sequence predicted running state;

and analyzing the continuous operation state of the laser welding machine based on the sequence prediction operation state.

In order to solve the above problems, the present invention further provides a system for automatically adjusting welding abnormality of an intelligent laser welding machine, the system comprising:

the operation data marking module is used for collecting historical operation data and current operation data of the laser welding machine, identifying historical operation abnormality corresponding to the historical operation data, and marking the historical operation data abnormally based on the historical operation abnormality to obtain marked operation data;

The anomaly identification model construction module is used for constructing an anomaly identification network architecture of the laser welding machine, optimizing network architecture parameters of the anomaly identification network architecture based on the marking operation data and a preset loss function to obtain an optimized anomaly identification model, and calculating model anomaly identification performance of the optimized anomaly identification model;

the abnormal influence parameter identification module is used for identifying the running state of the laser welding machine by utilizing the optimized abnormal recognition model, identifying the abnormal influence parameters of the laser welding machine and analyzing the abnormal association relation between the abnormal influence parameters and the laser welding machine when the abnormal recognition performance of the model accords with the preset standard performance;

the parameter particle swarm construction module is used for analyzing the continuous operation state of the laser welding machine based on the abnormal association relation and the operation state, identifying the abnormal state of the continuous operation state and constructing the parameter particle swarm of the laser welding machine based on the abnormal state;

the abnormal automatic adjustment module is used for calculating the state fitness of the particle group corresponding to the parameter particle group by using a preset welding state evaluation function, determining the optimal particle group of the laser welding machine based on the state fitness, adjusting the welding parameters of the laser welding machine based on the optimal particle group to obtain adjustment parameters, and realizing abnormal automatic adjustment of the laser welding machine based on the adjustment parameters.

According to the embodiment of the invention, a data basis is provided for constructing the association relationship between the operation data and the operation abnormality in the later period by identifying the history operation abnormality corresponding to the history operation data; further, the embodiment of the invention carries out abnormality marking on the historical operation data based on the historical operation abnormality to obtain marked operation data which can be used as data for model training in the later period; according to the embodiment of the invention, the basis for improving the later-stage abnormality recognition model of the laser welding machine can be constructed by constructing the abnormality recognition network architecture of the laser welding machine; optionally, the embodiment of the invention optimizes the network architecture parameters of the anomaly identification network architecture based on the marking operation data and a preset loss function to obtain an optimized anomaly identification model, which can optimize the performance of the anomaly identification network architecture to train out a target anomaly identification model, further, the embodiment of the invention can evaluate whether the model can accurately identify the operation anomaly of the laser welder by calculating the model anomaly identification performance of the optimized anomaly identification model, thereby improving the anomaly monitoring effect on the operation of the laser welder, optionally, when the model anomaly identification performance accords with the preset standard performance, based on the current operation data, the operation state of the laser welder can be identified by using the optimized anomaly identification model to provide basis for the operation anomaly evaluation of the laser welder in the later period, and further, the embodiment of the invention can predict the operation state of the laser welder based on the anomaly association relation and the operation state, thereby improving the intellectualization of the operation process of the laser welder, finally, and finally, the invention can realize the repair of the optimum particle swarm welding by using the optimal particle swarm welding parameter by using the preset parameter, and can realize the repair of the optimum particle swarm welding by using the optimal particle swarm welding parameter. Therefore, the method and the system for automatically adjusting the welding abnormality of the intelligent laser welding machine can improve the intellectualization of automatically adjusting the welding abnormality of the intelligent laser welding machine.

Drawings

FIG. 1 is a flow chart of a method for automatically adjusting welding anomalies in an intelligent laser welder according to an embodiment of the present invention;

FIG. 2 is a functional block diagram of a system for automatically adjusting welding anomalies in an intelligent laser welder according to one embodiment of the present invention;

FIG. 3 is a schematic diagram of an electronic device of a system for automatically adjusting welding abnormality of an intelligent laser welding machine according to an embodiment of the present invention;

the achievement of the objects, functional features and advantages of the present invention will be further described with reference to the accompanying drawings, in conjunction with the embodiments.

Detailed Description

It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.

The embodiment of the application provides a method for automatically adjusting welding abnormality of an intelligent laser welding machine. The execution main body of the method for automatically adjusting welding abnormality of the intelligent laser welding machine comprises, but is not limited to, at least one of a server side, a terminal and the like which can be configured to execute the method provided by the embodiment of the application. In other words, the method for automatically adjusting welding abnormality of the intelligent laser welding machine may be performed by software or hardware installed in a terminal device or a server device, and the software may be a blockchain platform. The service end includes but is not limited to: a single server, a server cluster, a cloud server or a cloud server cluster, and the like. The server may be an independent server, or may be a cloud server that provides cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (Content Delivery Network, CDN), and basic cloud computing services such as big data and artificial intelligence platforms.

Referring to fig. 1, a flow chart of a method for automatically adjusting welding anomalies of an intelligent laser welding machine according to an embodiment of the invention is shown. In this embodiment, the method for automatically adjusting welding abnormality of the intelligent laser welding machine includes:

s1, collecting historical operation data and current operation data of a laser welding machine, identifying historical operation abnormality corresponding to the historical operation data, and carrying out abnormality marking on the historical operation data based on the historical operation abnormality to obtain marked operation data.

In the embodiment of the invention, the historical operation data and the current operation data refer to operation data of the historical operation record of the laser welding machine and operation data of the current operation record, such as welding temperature, welding speed, welding seam shape and the like.

Optionally, the embodiment of the invention provides a data basis for constructing the association relationship between the operation data and the operation abnormality in the later period by identifying the history operation abnormality corresponding to the history operation data. The historical operation abnormality refers to an abnormal state generated in the historical operation process of the laser welding machine, such as an abnormal state of power failure, overhigh temperature and the like. In detail, the historical operation abnormality may be identified by script recording an operation log of the laser welding machine.

Further, the embodiment of the invention carries out abnormality marking on the historical operation data based on the historical operation abnormality to obtain the marked operation data which can be used as data for model training in the later period. The marking operation data refers to a data set obtained by marking operation data corresponding to an operation abnormality in the historical operation data. In detail, the anomaly tagging of the historical operating data may be data tagging by a data tagging tool.

Optionally, as an embodiment of the present invention, the performing, based on the historical operation anomaly, anomaly marking on the historical operation data to obtain marked operation data includes: constructing a unified time axis of the historical operation abnormality and the historical operation data; serializing the historical operation data based on the unified time axis to obtain serial historical operation data; identifying a time node of the historical operational anomaly; extracting historical operation data of a target sequence corresponding to the time node; and marking the sequence historical operation data to obtain the marked operation data.

The unified time axis is a time axis constructed by the historical operation abnormality and the historical operation data according to the same time reference point, the sequence historical operation data is a data set obtained by arranging the historical operation data according to the recorded time point, the time node is a node on the unified time axis, in which the historical operation abnormality occurs, and the target sequence historical operation data is a set corresponding to the historical operation data on the time node.

S2, constructing an anomaly identification network architecture of the laser welding machine, optimizing network architecture parameters of the anomaly identification network architecture based on the marking operation data and a preset loss function to obtain an optimized anomaly identification model, and calculating model anomaly identification performance of the optimized anomaly identification model.

According to the embodiment of the invention, the basis for improving the later-stage abnormality recognition model of the laser welding machine can be constructed by constructing the abnormality recognition network architecture of the laser welding machine. The anomaly identification network architecture refers to an initial network architecture of an anomaly identification model of the laser welding machine, such as a hierarchical structure, an activation function, a network element, and the like.

As one embodiment of the present invention, the constructing an anomaly identification network architecture of the laser welder includes: identifying a monitoring requirement of the laser welding machine; calculating a demand complexity coefficient of the monitoring demand; determining an architecture component of the laser welder based on the demand complexity factor; and constructing an anomaly identification network architecture of the laser welding machine based on the architecture component.

The monitoring requirement refers to a requirement for identifying monitoring of the laser welding machine, such as abnormal monitoring requirements of overheat temperature, overload load and the like in the operation process of the laser welding machine, the requirement complexity coefficient refers to the complexity degree of the monitoring requirement, and the architecture component refers to a network architecture component for monitoring the abnormality of the laser welding machine, such as a convolution layer, a circulation unit, a nonlinear activation function and the like.

Optionally, the embodiment of the invention optimizes the network architecture parameters of the anomaly identification network architecture based on the marking operation data and a preset loss function, and obtains an optimized anomaly identification model to optimize the performance of the anomaly identification network architecture to train a target anomaly identification model. The loss function is an index for measuring the difference between the model predicted value and the actual value, and for classification problems, a cross entropy loss function is commonly used.

Optionally, as an embodiment of the present invention, the optimizing the network architecture parameters of the anomaly identification network architecture based on the tag operation data and a preset loss function to obtain an optimized anomaly identification model includes: calculating an architecture parameter gradient of an architecture parameter corresponding to the abnormal identification network architecture by using the marking operation data and the loss function; calculating target architecture parameters of the anomaly identification network architecture based on the architecture parameter gradient; and optimizing network architecture parameters of the anomaly identification network architecture based on the architecture parameter gradient to obtain an optimized anomaly identification model.

The architecture parameter gradient refers to a partial derivative of a loss function on architecture parameters corresponding to the anomaly identification network architecture, and the target architecture parameter refers to a new architecture parameter of the anomaly identification network architecture calculated through an optimization algorithm. The optimization algorithm may be Gradient Descent (Gradient Descent), random Gradient Descent (Stochastic Gradient Descent, SGD), adam, RMSprop, etc.

Optionally, as an optional embodiment of the present invention, the calculating, using the tag operation data and the loss function, an architecture parameter gradient of an architecture parameter corresponding to the anomaly identification network architecture includes: extracting welding parameters from the marking operation data; outputting a predicted value of the framework by utilizing the abnormal recognition network framework based on the welding parameters; based on the architecture predicted value and the loss function, calculating an architecture parameter gradient of an architecture parameter corresponding to the anomaly identification network architecture by using the following formula:

；

wherein,architecture parameter gradient representing weights in corresponding architecture parameters of network architecture, +.>Representing a loss function->Architecture predictor +.>Gradient of->Output value of activation function corresponding to abnormal recognition network architecture in abnormal recognition network architecture>Derivative of>Representing neuron output value +.>Weight in architecture parameters corresponding to network architecture>Is used for the gradient of (a),architecture parameter gradient representing bias in corresponding architecture parameters of the network architecture, +.>Representing neuron output value +.>Bias in corresponding architecture parameters of network architecture>Is a gradient of (a).

Wherein the neuron output valueRefers to the signal value that is transmitted from the neuron after processing by the activation function. This value is the result of the internal calculation of the input signal received by the neuron. The Weight (Weight) refers to the strength of the connection between neurons or the strength of the information transfer. In neural networks, each connection can be considered as a weight that determines the importance of information transfer from one individual to another. The Bias (Bias) is then a fixed value applied to the output of the neuron, which allows the output of the activation function of the neuron to shift when no input is received. The presence of the bias term allows greater flexibility in the neural network, and by adjusting the bias values, the network can more easily fit non-linear relationships in the data.

Further, according to the embodiment of the invention, whether the model can accurately identify the operation abnormality of the laser welding machine can be evaluated by calculating the model abnormality identification performance of the optimized abnormality identification model, so that the abnormality monitoring effect on the operation of the laser welding machine is improved. The model abnormality recognition performance refers to the accuracy of the model of the abnormality recognition model in recognizing the operation abnormality of the laser welding machine. In detail, the model anomaly recognition performance of the calculation of the optimized anomaly recognition model may be evaluated by calculating a recall ratio of the optimized anomaly recognition model.

And S3, when the model abnormality identification performance accords with preset standard performance, based on the current operation data, identifying the operation state of the laser welding machine by using the optimized abnormality identification model, identifying abnormality influence parameters of the laser welding machine, and analyzing abnormality association relation between the abnormality influence parameters and the laser welding machine.

Optionally, when the model abnormality recognition performance meets a preset standard performance, based on the current operation data, the operation state of the laser welding machine is recognized by using the optimized abnormality recognition model, so that a basis can be provided for the operation abnormality evaluation of the laser welding machine in a later stage. The operation state refers to a state of the laser welding machine in current operation, such as stable operation, overheat temperature, overload and the like.

Optionally, as an embodiment of the present invention, when the model anomaly identification performance meets a preset standard performance, identifying the operation state of the laser welding machine by using the optimized anomaly identification model based on the current operation data includes: when the model anomaly identification performance accords with a preset standard performance, performing feature extraction on the current operation data by utilizing a convolution layer in the optimized anomaly identification model to obtain a data feature map; extracting a feature map from the data feature map by using a pooling layer in the optimization anomaly identification model to obtain a target feature map; introducing nonlinear properties into the target feature map by using an activation function in the optimized anomaly identification model to obtain a nonlinear property feature map; and connecting the nonlinear property characteristic graphs by using a full connection layer in the optimized abnormal recognition model to obtain the running state of the laser welding machine.

The method comprises the steps of extracting features in input data, capturing features with different scales by setting different convolution kernel sizes and numbers, wherein a pooling layer is used for reducing the dimension of a feature map, reducing the calculation amount and retaining important feature information, an activation function is used for introducing nonlinear properties, a common activation function comprises ReLU, sigmoid, tanh and other functions, a fully connected layer is used for connecting outputs of the previous layers and obtaining a final prediction result through matrix multiplication and offset addition, a data feature map is a feature map extracted through the convolution layer, a target feature map is a feature map obtained by extracting useful information of the feature map, and a nonlinear property feature map is a feature map obtained by introducing nonlinear properties to the downsampled features.

Furthermore, the embodiment of the invention can provide basis for predicting the running state trend of the laser welding machine in the later period by identifying the abnormal influence parameters of the laser welding machine. The abnormal influence parameter refers to a parameter that influences the operation state of the laser welding machine, such as the operation power, the operation time and the like of the laser welding machine.

Further, as an embodiment of the present invention, the identifying the abnormal influence parameter of the laser welder includes: constructing a parameter test group of welding parameters corresponding to the laser welding machine; identifying a test running state of the laser welder corresponding to the test welding parameters in the parameter test group; based on the parameter test group and the test running state, constructing an association relation curve between corresponding welding parameters and running states of the laser welding machine; and identifying abnormal influence parameters of the laser welding machine based on the association relation curve.

The parameter test set is used for testing the welding parameters and the running state of the laser welding machine, the parameter test set changes one welding parameter value, in addition, the welding parameter value is kept unchanged, the test running state is the corresponding running state of the laser welding machine in the parameter test set, the association relation curve is the relation curve between the welding parameters and the running state, for example, the abscissa is the running power of the laser welding machine, and the ordinate is the test running state.

Optionally, as an optional embodiment of the present invention, the identifying, based on the association relationship curve, an abnormal influence parameter of the laser welding machine includes: identifying a fluctuation area of the association relation curve; calculating the fluctuation coefficient of the association relation curve based on the fluctuation area by using the following formula:

；

wherein,fluctuation coefficient representing association curve, +.>Represents the relation of->Fluctuation amplitude of individual fluctuation areas, +.>Representing the association curve->Represents the relation of->The angular frequency of the highest point corresponding to the wave zone, < >>Represents the relation of->The individual fluctuation zones correspond to the highest values,/->Represents the relation of->The individual fluctuation zones correspond to the lowest value, +.>Represents the relation of->Duration of the individual fluctuation zones, +.>Representing the number of fluctuation areas corresponding to the association relation curve;

and determining an abnormal influence parameter of the laser welding machine based on the fluctuation coefficient.

According to the embodiment of the invention, the accuracy of analyzing the trend of the running state of the laser welding machine can be further improved by analyzing the abnormal incidence relation between the abnormal influence parameters and the laser welding machine. The abnormal association relationship refers to an influence relationship of the abnormal influence parameter on the operation state of the laser welding machine, for example, the higher the power is, the higher the operation temperature of the laser welding machine is. In detail, the abnormal association relationship can be analyzed through the abnormal influence parameter corresponding association relationship curve.

S4, analyzing the continuous operation state of the laser welding machine based on the abnormal association relation and the operation state, identifying the abnormal state of the continuous operation state, and constructing the parameter particle swarm of the laser welding machine based on the abnormal state.

Optionally, the embodiment of the invention analyzes the continuous operation state of the laser welding machine based on the abnormal association relationship and the operation state, so that the operation state of the laser welding machine can be predicted, thereby improving the intellectualization of identifying the abnormal operation process of the laser welding machine. The continuous operation state refers to a continuous state in which future operation of the laser welding machine is predicted, for example, the laser welding machine continuously and stably operates from the current beginning, and the laser welding machine continuously increases from the current beginning operation temperature.

Optionally, as an embodiment of the present invention, the analyzing the continuous operation state of the laser welding machine based on the abnormal association relationship and the operation state includes: identifying a parameter predicted value of a welding parameter corresponding to the laser welding machine; analyzing the predicted running state of the laser welding machine based on the parameter predicted value and the abnormal association relation; serializing the predicted running state to obtain a sequence predicted running state; and analyzing the continuous operation state of the laser welding machine based on the sequence prediction operation state.

The parameter predicted value refers to a value of a welding parameter corresponding to the laser welding machine in a next operation process, the predicted operation state refers to a predicted operation state calculated by a certain group of parameter predicted values, and the sequence predicted operation state refers to an operation state set obtained by arranging the predicted operation states according to a time sequence.

In the embodiment of the invention, the abnormal state refers to a state of abnormal operation of the laser welding machine, such as abnormal states of overheat temperature, overload and the like. In detail, the abnormal state may be identified by comparing the continuous operation state with a preset abnormal threshold. Wherein the abnormal threshold is a standard threshold for evaluating whether the laser welding machine is abnormal in operation.

Optionally, according to the embodiment of the invention, by constructing the parameter particle swarm of the laser welding machine, the welding parameters for repairing the abnormal operation of the laser welding machine can be obtained through continuous particle swarm optimization, so that the effect of repairing the abnormal operation of the laser welding machine is improved. Wherein, the parameter particle group refers to randomly initializing a group of particles in a defined domain of welding parameters, each particle representing a potential welding parameter setting.

S5, calculating the state fitness of the particle group corresponding to the parameter particle group by using a preset welding state evaluation function, determining an optimal particle group of the laser welding machine based on the state fitness, adjusting welding parameters of the laser welding machine based on the optimal particle group to obtain adjustment parameters, and realizing abnormal automatic adjustment of the laser welding machine based on the adjustment parameters.

According to the embodiment of the invention, the condition fitness of the particle group corresponding to the parameter particle group is calculated by utilizing the preset welding condition evaluation function, so that the parameter particle group can be screened as the basis. The state fitness refers to the fitness of the parameter particle swarm in the current parameter environment. The welding state evaluation function refers to a function for calculating the fitness of the parameter particle swarm, and in detail, the welding state evaluation function can be written by python codes.

In the embodiment of the present invention, the optimal particle group refers to a particle group with the highest fitness among the parameter particle groups, and in detail, the determining the optimal particle group of the laser welding machine may select the particle group with the highest fitness.

Optionally, in the embodiment of the present invention, the welding parameters of the laser welding machine are adjusted based on the optimal particle set, so that the adjustment parameters can be obtained to repair the abnormal operation state of the laser welding machine in time, thereby improving the intellectualization of repairing the abnormal operation state of the laser welding machine. The adjustment parameters refer to parameters obtained by updating welding parameters of the laser welding machine through the optimal particle set.

According to the embodiment of the invention, a data basis is provided for constructing the association relationship between the operation data and the operation abnormality in the later period by identifying the history operation abnormality corresponding to the history operation data; further, the embodiment of the invention carries out abnormality marking on the historical operation data based on the historical operation abnormality to obtain marked operation data which can be used as data for model training in the later period; according to the embodiment of the invention, the basis for improving the later-stage abnormality recognition model of the laser welding machine can be constructed by constructing the abnormality recognition network architecture of the laser welding machine; optionally, the embodiment of the invention optimizes the network architecture parameters of the anomaly identification network architecture based on the marking operation data and a preset loss function to obtain an optimized anomaly identification model, which can optimize the performance of the anomaly identification network architecture to train out a target anomaly identification model, further, the embodiment of the invention can evaluate whether the model can accurately identify the operation anomaly of the laser welder by calculating the model anomaly identification performance of the optimized anomaly identification model, thereby improving the anomaly monitoring effect on the operation of the laser welder, optionally, when the model anomaly identification performance accords with the preset standard performance, based on the current operation data, the operation state of the laser welder can be identified by using the optimized anomaly identification model to provide basis for the operation anomaly evaluation of the laser welder in the later period, and further, the embodiment of the invention can predict the operation state of the laser welder based on the anomaly association relation and the operation state, thereby improving the intellectualization of the operation process of the laser welder, finally, and finally, the invention can realize the repair of the optimum particle swarm welding by using the optimal particle swarm welding parameter by using the preset parameter, and can realize the repair of the optimum particle swarm welding by using the optimal particle swarm welding parameter. Therefore, the method for automatically adjusting the welding abnormality of the intelligent laser welding machine can improve the intellectualization of automatically adjusting the welding abnormality of the intelligent laser welding machine.

Fig. 2 is a functional block diagram of a system for automatically adjusting welding abnormality of an intelligent laser welding machine according to an embodiment of the present invention.

The system 200 for automatically adjusting welding abnormality of an intelligent laser welding machine can be installed in electronic equipment. Depending on the functions implemented, the system 200 for automatically adjusting welding anomalies of the intelligent laser welding machine may include an operation data marking module 201, an anomaly identification model construction module 202, an anomaly impact parameter identification module 203, a parameter particle swarm construction module 204, and an anomaly automatic adjustment module 205. The module of the invention, which may also be referred to as a unit, refers to a series of computer program segments, which are stored in the memory of the electronic device, capable of being executed by the processor of the electronic device and of performing a fixed function.

In the present embodiment, the functions concerning the respective modules/units are as follows:

the operation data marking module 201 is configured to collect historical operation data and current operation data of a laser welding machine, identify a historical operation abnormality corresponding to the historical operation data, and perform an abnormality marking on the historical operation data based on the historical operation abnormality to obtain marked operation data;

The anomaly identification model construction module 202 is configured to construct an anomaly identification network architecture of the laser welding machine, optimize network architecture parameters of the anomaly identification network architecture based on the marking operation data and a preset loss function, obtain an optimized anomaly identification model, and calculate a model anomaly identification performance of the optimized anomaly identification model;

the abnormality influencing parameter identification module 203 is configured to identify an operation state of the laser welder, identify an abnormality influencing parameter of the laser welder, and analyze an abnormality association relationship between the abnormality influencing parameter and the laser welder based on the current operation data when the model abnormality identification performance meets a preset standard performance;

the parameter particle swarm construction module 204 is configured to analyze a continuous operation state of the laser welding machine based on the abnormal association relationship and the operation state, identify an abnormal state of the continuous operation state, and construct a parameter particle swarm of the laser welding machine based on the abnormal state;

the abnormality automatic adjustment module 205 is configured to calculate a state fitness of a particle set corresponding to the parameter particle set by using a preset welding state evaluation function, determine an optimal particle set of the laser welder based on the state fitness, adjust welding parameters of the laser welder based on the optimal particle set, obtain adjustment parameters, and implement abnormality automatic adjustment of the laser welder based on the adjustment parameters.

In detail, each module in the system 200 for automatically adjusting welding abnormality of an intelligent laser welder in the embodiment of the present invention adopts the same technical means as the method for automatically adjusting welding abnormality of an intelligent laser welder in the drawings when in use, and can produce the same technical effects, which are not described herein.

The embodiment of the invention provides electronic equipment for realizing a method for automatically adjusting welding abnormality of an intelligent laser welding machine.

Referring to fig. 3, the electronic device may include a processor 30, a memory 31, a communication bus 32, and a communication interface 33, and may further include a computer program stored in the memory 31 and executable on the processor 30, such as a method program for automatically adjusting welding anomalies of an intelligent laser welding machine.

The processor may be formed by an integrated circuit in some embodiments, for example, a single packaged integrated circuit, or may be formed by a plurality of integrated circuits packaged with the same function or different functions, including one or more central processing units (Central Processing Unit, CPU), a microprocessor, a digital processing chip, a graphics processor, a combination of various control chips, and the like. The processor is a Control Unit (Control Unit) of the electronic device, connects various parts of the entire electronic device using various interfaces and lines, executes or executes programs or modules stored in the memory (for example, a program for performing automatic adjustment of welding abnormality of the intelligent laser welder, etc.), and invokes data stored in the memory to perform various functions of the electronic device and process the data.

The memory includes at least one type of readable storage medium including flash memory, removable hard disk, multimedia card, card memory (e.g., SD or DX memory, etc.), magnetic memory, magnetic disk, optical disk, etc. The memory may in some embodiments be an internal storage unit of the electronic device, such as a mobile hard disk of the electronic device. The memory may in other embodiments also be an external storage device of the electronic device, such as a plug-in mobile hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card) or the like, which are provided on the electronic device. Further, the memory may also include both internal storage units and external storage devices of the electronic device. The memory can be used for storing application software installed in the electronic equipment and various data, such as codes of a program for automatically adjusting welding abnormality based on the intelligent laser welding machine, and the like, and can be used for temporarily storing data which is already output or is to be output.

The communication bus may be a peripheral component interconnect standard (Peripheral Component Interconnect, PCI) bus, or an extended industry standard architecture (Extended Industry Standard Architecture, EISA) bus, among others. The bus may be classified as an address bus, a data bus, a control bus, etc. The bus is arranged to enable a connection communication between the memory and at least one processor or the like.

The communication interface is used for communication between the electronic equipment and other equipment, and comprises a network interface and a user interface. Optionally, the network interface may include a wired interface and/or a wireless interface (e.g., WI-FI interface, bluetooth interface, etc.), typically used to establish a communication connection between the electronic device and other electronic devices. The user interface may be a Display (Display), an input unit such as a Keyboard (Keyboard), or alternatively a standard wired interface, a wireless interface. Alternatively, in some embodiments, the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, an OLED (Organic Light-Emitting Diode) touch, or the like. The display may also be referred to as a display screen or display unit, as appropriate, for displaying information processed in the electronic device and for displaying a visual user interface.

For example, although not shown, the electronic device may further include a power source (such as a battery) for powering the respective components, and preferably, the power source may be logically connected to the at least one processor through a power management system, so as to perform functions of charge management, discharge management, and power consumption management through the power management system. The power supply may also include one or more of any of a direct current or alternating current power supply, a recharging system, a power failure detection circuit, a power converter or inverter, a power status indicator, and the like. The electronic device may further include various sensors, bluetooth modules, wi-Fi modules, etc., which are not described herein.

It should be understood that the embodiments described are for illustrative purposes only and are not limited to this configuration in the scope of the patent application.

The program stored in the memory of the electronic device for automatically adjusting welding abnormality of the intelligent laser welding machine is a combination of a plurality of instructions, and when the program runs in the processor, the program can realize:

collecting historical operation data and current operation data of a laser welding machine, identifying historical operation abnormality corresponding to the historical operation data, and performing abnormality marking on the historical operation data based on the historical operation abnormality to obtain marked operation data;

constructing an anomaly identification network architecture of the laser welding machine, optimizing network architecture parameters of the anomaly identification network architecture based on the marking operation data and a preset loss function to obtain an optimized anomaly identification model, and calculating model anomaly identification performance of the optimized anomaly identification model;

when the model abnormality recognition performance accords with a preset standard performance, based on the current operation data, recognizing the operation state of the laser welding machine by using the optimized abnormality recognition model, recognizing abnormality influence parameters of the laser welding machine, and analyzing abnormality association relation between the abnormality influence parameters and the laser welding machine;

Analyzing the continuous operation state of the laser welding machine based on the abnormal association relation and the operation state, identifying the abnormal state of the continuous operation state, and constructing a parameter particle swarm of the laser welding machine based on the abnormal state;

calculating the state fitness of the particle group corresponding to the parameter particle group by using a preset welding state evaluation function, determining the optimal particle group of the laser welding machine based on the state fitness, adjusting the welding parameters of the laser welding machine based on the optimal particle group, obtaining adjustment parameters, and realizing abnormal automatic adjustment of the laser welding machine based on the adjustment parameters.

Specifically, the specific implementation method of the above instruction by the processor may refer to descriptions of related steps in the corresponding embodiment of the drawings, which are not repeated herein.

Further, the electronic device integrated modules/units, if implemented in the form of software functional units and sold or used as stand-alone products, may be stored in a computer readable storage medium. The computer readable storage medium may be volatile or nonvolatile. For example, the computer readable medium may include: any entity or system capable of carrying the computer program code, a recording medium, a U disk, a removable hard disk, a magnetic disk, an optical disk, a computer Memory, a Read-Only Memory (ROM).

The present invention also provides a computer readable storage medium storing a computer program which, when executed by a processor of an electronic device, can implement:

collecting historical operation data and current operation data of a laser welding machine, identifying historical operation abnormality corresponding to the historical operation data, and performing abnormality marking on the historical operation data based on the historical operation abnormality to obtain marked operation data;

constructing an anomaly identification network architecture of the laser welding machine, optimizing network architecture parameters of the anomaly identification network architecture based on the marking operation data and a preset loss function to obtain an optimized anomaly identification model, and calculating model anomaly identification performance of the optimized anomaly identification model;

when the model abnormality recognition performance accords with a preset standard performance, based on the current operation data, recognizing the operation state of the laser welding machine by using the optimized abnormality recognition model, recognizing abnormality influence parameters of the laser welding machine, and analyzing abnormality association relation between the abnormality influence parameters and the laser welding machine;

analyzing the continuous operation state of the laser welding machine based on the abnormal association relation and the operation state, identifying the abnormal state of the continuous operation state, and constructing a parameter particle swarm of the laser welding machine based on the abnormal state;

Calculating the state fitness of the particle group corresponding to the parameter particle group by using a preset welding state evaluation function, determining the optimal particle group of the laser welding machine based on the state fitness, adjusting the welding parameters of the laser welding machine based on the optimal particle group, obtaining adjustment parameters, and realizing abnormal automatic adjustment of the laser welding machine based on the adjustment parameters.

In the several embodiments provided by the present invention, it should be understood that the disclosed apparatus, system and method may be implemented in other manners. For example, the system embodiments described above are merely illustrative, e.g., the division of the modules is merely a logical function division, and other manners of division may be implemented in practice.

The modules described as separate components may or may not be physically separate, and components shown as modules may or may not be physical units, may be located in one place, or may be distributed over multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

In addition, each functional module in the embodiments of the present invention may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit. The integrated units can be realized in a form of hardware or a form of hardware and a form of software functional modules.

It will be evident to those skilled in the art that the invention is not limited to the details of the foregoing illustrative embodiments, and that the present invention may be embodied in other specific forms without departing from the spirit or essential characteristics thereof.

The present embodiments are, therefore, to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein. Any reference signs in the claims shall not be construed as limiting the claim concerned.

The embodiment of the application can acquire and process the related data based on the artificial intelligence technology. Among these, artificial intelligence (Artificial Intelligence, AI) is the theory, method, technique and application system that uses a digital computer or a digital computer-controlled machine to simulate, extend and extend human intelligence, sense the environment, acquire knowledge and use knowledge to obtain optimal results.

Furthermore, it is evident that the word "comprising" does not exclude other elements or steps, and that the singular does not exclude a plurality. Multiple units or systems as set forth in the system claims may also be implemented by means of one unit or system in software or hardware. The terms first, second, etc. are used to denote a name, but not any particular order.

Finally, it should be noted that the above-mentioned embodiments are merely for illustrating the technical solution of the present invention and not for limiting the same, and although the present invention has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that modifications and equivalents may be made to the technical solution of the present invention without departing from the spirit and scope of the technical solution of the present invention.

Claims (10)

1. A method for automatically adjusting welding anomalies of an intelligent laser welding machine, the method comprising:

collecting historical operation data and current operation data of a laser welding machine, identifying historical operation abnormality corresponding to the historical operation data, and performing abnormality marking on the historical operation data based on the historical operation abnormality to obtain marked operation data;

constructing an anomaly identification network architecture of the laser welding machine, optimizing network architecture parameters of the anomaly identification network architecture based on the marking operation data and a preset loss function to obtain an optimized anomaly identification model, and calculating model anomaly identification performance of the optimized anomaly identification model;

when the model abnormality recognition performance accords with a preset standard performance, based on the current operation data, recognizing the operation state of the laser welding machine by using the optimized abnormality recognition model, recognizing abnormality influence parameters of the laser welding machine, and analyzing abnormality association relation between the abnormality influence parameters and the laser welding machine;

Analyzing the continuous operation state of the laser welding machine based on the abnormal association relation and the operation state, identifying the abnormal state of the continuous operation state, and constructing a parameter particle swarm of the laser welding machine based on the abnormal state;

calculating the state fitness of the particle group corresponding to the parameter particle group by using a preset welding state evaluation function, determining the optimal particle group of the laser welding machine based on the state fitness, adjusting the welding parameters of the laser welding machine based on the optimal particle group, obtaining adjustment parameters, and realizing abnormal automatic adjustment of the laser welding machine based on the adjustment parameters.

2. The method for automatically adjusting welding anomalies of an intelligent laser welder according to claim 1, wherein the anomaly marking the historical operating data based on the historical operating anomalies results in marked operating data, comprising:

constructing a unified time axis of the historical operation abnormality and the historical operation data;

serializing the historical operation data based on the unified time axis to obtain serial historical operation data;

identifying a time node of the historical operational anomaly;

Extracting historical operation data of a target sequence corresponding to the time node;

and marking the sequence historical operation data to obtain the marked operation data.

3. The method for automatically adjusting welding anomalies of a smart laser welder according to claim 1, wherein the constructing an anomaly identification network architecture of the laser welder comprises:

identifying a monitoring requirement of the laser welding machine;

calculating a demand complexity coefficient of the monitoring demand;

determining an architecture component of the laser welder based on the demand complexity factor;

and constructing an anomaly identification network architecture of the laser welding machine based on the architecture component.

4. The method for automatically adjusting welding anomalies of an intelligent laser welder according to claim 1, wherein optimizing network architecture parameters of the anomaly identification network architecture based on the tag operation data and a preset loss function to obtain an optimized anomaly identification model comprises:

calculating an architecture parameter gradient of an architecture parameter corresponding to the abnormal identification network architecture by using the marking operation data and the loss function;

calculating target architecture parameters of the anomaly identification network architecture based on the architecture parameter gradient;

And optimizing network architecture parameters of the anomaly identification network architecture based on the architecture parameter gradient to obtain an optimized anomaly identification model.

5. The method for automatically adjusting welding anomalies in a smart laser welder according to claim 4, wherein calculating an architecture parameter gradient for the corresponding architecture parameter of the anomaly identification network using the tag operational data and the loss function comprises:

extracting welding parameters from the marking operation data;

outputting a predicted value of the framework by utilizing the abnormal recognition network framework based on the welding parameters;

based on the architecture predicted value and the loss function, calculating an architecture parameter gradient of an architecture parameter corresponding to the anomaly identification network architecture by using the following formula:

；

wherein,architecture parameter gradient representing weights in corresponding architecture parameters of network architecture, +.>Representing a loss function->Architecture predictor +.>Gradient of->Output value of activation function corresponding to abnormal recognition network architecture in abnormal recognition network architecture>Derivative of>Representing neuron output value +.>Weight in architecture parameters corresponding to network architecture>Gradient of->Architecture parameter gradient representing bias in corresponding architecture parameters of the network architecture, +. >Representing neuron output value +.>Bias in corresponding architecture parameters of network architecture>Is a gradient of (a).

6. The method for automatically adjusting welding anomalies of an intelligent laser welder according to claim 1, wherein when the model anomaly identification performance meets a preset standard performance, identifying an operating state of the laser welder using the optimized anomaly identification model based on the current operating data, comprising:

when the model anomaly identification performance accords with a preset standard performance, performing feature extraction on the current operation data by utilizing a convolution layer in the optimized anomaly identification model to obtain a data feature map;

extracting a feature map from the data feature map by using a pooling layer in the optimization anomaly identification model to obtain a target feature map;

introducing nonlinear properties into the target feature map by using an activation function in the optimized anomaly identification model to obtain a nonlinear property feature map;

and connecting the nonlinear property characteristic graphs by using a full connection layer in the optimized abnormal recognition model to obtain the running state of the laser welding machine.

7. The method for automatically adjusting welding anomalies of a smart laser welder according to claim 1, wherein the identifying anomaly influencing parameters of the laser welder includes:

Constructing a parameter test group of welding parameters corresponding to the laser welding machine;

identifying a test running state of the laser welder corresponding to the test welding parameters in the parameter test group;

based on the parameter test group and the test running state, constructing an association relation curve between corresponding welding parameters and running states of the laser welding machine;

and identifying abnormal influence parameters of the laser welding machine based on the association relation curve.

8. The method for automatically adjusting welding anomalies of a smart laser welder according to claim 7, wherein the identifying anomaly influencing parameters of the laser welder based on the correlation curve includes:

identifying a fluctuation area of the association relation curve;

calculating the fluctuation coefficient of the association relation curve based on the fluctuation area by using the following formula:

；

wherein,fluctuation coefficient representing association curve, +.>Represents the relation of->Fluctuation amplitude of individual fluctuation areas, +.>Representing the association curve->Represents the relation of->The wave regions correspond to the angular frequencies of the highest points,represents the relation of->The individual fluctuation zones correspond to the highest values,/- >Represents the relation of->The individual fluctuation zones correspond to the lowest value, +.>Represents the relation of->Duration of the individual fluctuation zones, +.>Representing the number of fluctuation areas corresponding to the association relation curve;

and determining an abnormal influence parameter of the laser welding machine based on the fluctuation coefficient.

9. The method for automatically adjusting welding anomalies of an intelligent laser welder according to claim 1, wherein analyzing a continuous operating state of the laser welder based on the anomaly correlations and the operating state includes:

identifying a parameter predicted value of a welding parameter corresponding to the laser welding machine;

analyzing the predicted running state of the laser welding machine based on the parameter predicted value and the abnormal association relation;

serializing the predicted running state to obtain a sequence predicted running state;

and analyzing the continuous operation state of the laser welding machine based on the sequence prediction operation state.

10. A system for automatic adjustment of welding anomalies in an intelligent laser welder, characterized in that it comprises:

The operation data marking module is used for collecting historical operation data and current operation data of the laser welding machine, identifying historical operation abnormality corresponding to the historical operation data, and marking the historical operation data abnormally based on the historical operation abnormality to obtain marked operation data;

the anomaly identification model construction module is used for constructing an anomaly identification network architecture of the laser welding machine, optimizing network architecture parameters of the anomaly identification network architecture based on the marking operation data and a preset loss function to obtain an optimized anomaly identification model, and calculating model anomaly identification performance of the optimized anomaly identification model;

the abnormal influence parameter identification module is used for identifying the running state of the laser welding machine by utilizing the optimized abnormal recognition model, identifying the abnormal influence parameters of the laser welding machine and analyzing the abnormal association relation between the abnormal influence parameters and the laser welding machine when the abnormal recognition performance of the model accords with the preset standard performance;

the parameter particle swarm construction module is used for analyzing the continuous operation state of the laser welding machine based on the abnormal association relation and the operation state, identifying the abnormal state of the continuous operation state and constructing the parameter particle swarm of the laser welding machine based on the abnormal state;

The abnormal automatic adjustment module is used for calculating the state fitness of the particle group corresponding to the parameter particle group by using a preset welding state evaluation function, determining the optimal particle group of the laser welding machine based on the state fitness, adjusting the welding parameters of the laser welding machine based on the optimal particle group to obtain adjustment parameters, and realizing abnormal automatic adjustment of the laser welding machine based on the adjustment parameters.