CN116245216A - Workpiece welding quality prediction device and method based on OPC UA architecture - Google Patents
Workpiece welding quality prediction device and method based on OPC UA architecture Download PDFInfo
- Publication number
- CN116245216A CN116245216A CN202211741310.XA CN202211741310A CN116245216A CN 116245216 A CN116245216 A CN 116245216A CN 202211741310 A CN202211741310 A CN 202211741310A CN 116245216 A CN116245216 A CN 116245216A
- Authority
- CN
- China
- Prior art keywords
- workpiece
- quality prediction
- welding quality
- welding
- production data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000003466 welding Methods 0.000 title claims abstract description 142
- 238000000034 method Methods 0.000 title claims abstract description 57
- 238000004519 manufacturing process Methods 0.000 claims abstract description 89
- 230000008569 process Effects 0.000 claims abstract description 32
- 238000004891 communication Methods 0.000 claims abstract description 23
- 238000003860 storage Methods 0.000 claims abstract description 13
- 238000013527 convolutional neural network Methods 0.000 claims description 18
- 238000001514 detection method Methods 0.000 claims description 16
- 230000005540 biological transmission Effects 0.000 claims description 9
- 238000012549 training Methods 0.000 claims description 8
- 238000012545 processing Methods 0.000 claims description 6
- 238000011176 pooling Methods 0.000 description 12
- 238000012544 monitoring process Methods 0.000 description 9
- 230000006870 function Effects 0.000 description 7
- 238000004364 calculation method Methods 0.000 description 6
- 238000010586 diagram Methods 0.000 description 6
- 230000004913 activation Effects 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000007689 inspection Methods 0.000 description 2
- 238000007726 management method Methods 0.000 description 2
- 230000004044 response Effects 0.000 description 2
- 230000006978 adaptation Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000002950 deficient Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 238000003754 machining Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 210000002569 neuron Anatomy 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000033772 system development Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B23—MACHINE TOOLS; METAL-WORKING NOT OTHERWISE PROVIDED FOR
- B23K—SOLDERING OR UNSOLDERING; WELDING; CLADDING OR PLATING BY SOLDERING OR WELDING; CUTTING BY APPLYING HEAT LOCALLY, e.g. FLAME CUTTING; WORKING BY LASER BEAM
- B23K31/00—Processes relevant to this subclass, specially adapted for particular articles or purposes, but not covered by only one of the preceding main groups
- B23K31/12—Processes relevant to this subclass, specially adapted for particular articles or purposes, but not covered by only one of the preceding main groups relating to investigating the properties, e.g. the weldability, of materials
- B23K31/125—Weld quality monitoring
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
- G06Q10/06395—Quality analysis or management
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/04—Manufacturing
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- Economics (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Development Economics (AREA)
- Entrepreneurship & Innovation (AREA)
- Marketing (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Educational Administration (AREA)
- General Health & Medical Sciences (AREA)
- Game Theory and Decision Science (AREA)
- Operations Research (AREA)
- Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Manufacturing & Machinery (AREA)
- Primary Health Care (AREA)
- Evolutionary Computation (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Mechanical Engineering (AREA)
- General Factory Administration (AREA)
Abstract
The invention discloses a workpiece welding quality prediction device and method based on an OPC UA architecture, comprising a data acquisition and storage layer, an application layer and an equipment layer for acquiring production data of a workpiece welding process, wherein the data acquisition and storage layer comprises an OPC UA server which is in communication connection with the equipment layer; the application layer comprises an OPC UA client, a welding quality prediction system and an MES system, wherein the OPC UA client is in communication connection with an OPC UA server, displays production data and compares the production data with a set range; the welding quality prediction system is in communication connection with the data acquisition and storage layer, and production data are submitted to the welding quality prediction system to perform welding quality prediction; the MES system is used for marking workpieces with production data not within a set range or workpieces with unqualified welding quality prediction. The method can well solve the inconvenience caused by different equipment communication protocols, can rapidly and orderly collect production data of workpieces in the production process of a welding workshop, and provides a basis for later quality prediction and quality tracing.
Description
Technical Field
The invention relates to workpiece welding quality prediction, in particular to a workpiece welding quality prediction device and method based on an OPC UA architecture.
Background
The welding machine of various models is arranged in the automobile production welding workshop, and parameter data such as voltage, current, temperature and the like are required to be recorded during operation, and have great effects on quality prediction and quality problem tracing. In the past, data information was collected for accessing field devices, and each application developer was required to write a specific interface function. Because of the wide variety of field devices and the continual upgrading of products, great workload is often brought to users and software developers. This is often not enough to meet the practical needs of the job, and system integrators and developers are eagerly demanding a plug-and-play device driver with high efficiency, reliability, openness, interoperability. The problem of semantic interworking between different numerical control equipment is difficult to solve. At present, most research and application are focused on the aspects of numerical control equipment data acquisition, manufacturing workshop data acquisition, monitoring systems and the like, and network interconnection is basically realized. However, due to the fact that data isomerism exists among private data acquisition protocols of different numerical control equipment, semantic intercommunication cannot be achieved, and further development of workshop digitization is restricted. The difficulty of data processing further increases the difficulty of weld quality prediction and weld quality traceability.
In terms of welding quality prediction, the lack of a rapid acquisition and transmission method of real-time data makes it difficult to realize real-time prediction, and many manufacturers take this as a single plate, and after a prediction model is established, the model is used for a long time, and there is a delay in model update, so that the probability of prediction errors is increased. In terms of welding quality tracing, a vehicle body enters a quality inspection stage after welding, if quality inspection is unqualified, the vehicle body needs to be reprocessed, various data of the welding process are required to be acquired to determine distribution conditions of defective products and find out reasons for defects, and at present, after the production data of unqualified vehicle bodies are acquired, many manufacturers simply store the data or can store a certain number of data and then deliver the data to a quality prediction system for use, so that the quality prediction system is not updated timely.
Because the communication protocols of various welding machines and sensors are different, huge cost is required to be input for integration after data acquisition of a welding workshop. In the welding process, important parameters such as voltage, current and the like cannot be monitored and recorded in real time, so that the quality of a welded product cannot be guaranteed. Because of no unified communication protocol, when the workshop Manufacturing Execution System (MES) is designed, various different communication protocols and equipment adaptation can only be passively integrated, the difficulty of system development is improved, and the informatization construction of the whole workshop is hindered. The lack of a good data processing mode can bring great resistance to the real-time prediction of the welding quality and the quality tracing process of the subsequent quality prediction model.
The lack of a rapid acquisition and transmission method of real-time data causes difficulty in realizing real-time prediction, and under the condition that welding quality prediction and welding quality tracing are mutually divided, a quality prediction system is difficult to rapidly utilize the latest sample to update a structure, so that the method is unfavorable for actual production and has a certain progress space.
Disclosure of Invention
The invention aims to provide a workpiece welding quality prediction device and method based on an OPC UA architecture, which can well solve the inconvenience caused by different equipment communication protocols, can rapidly and orderly collect production data of workpieces in a production process of a welding workshop, and provides a basis for subsequent quality prediction and quality tracing.
In order to achieve the above purpose, the technical scheme adopted by the invention is as follows:
a workpiece welding quality prediction device based on an OPC UA architecture comprises an equipment layer, a data acquisition and storage layer and an application layer; the equipment layer is used for acquiring production data of a workpiece welding process; the data acquisition and storage layer comprises an OPC UA server, and the OPC UA server is in communication connection with the equipment layer and is used as a data acquisition end and a data transmission sending end; the application layer comprises an OPC UA client, a welding quality prediction system and an MES system, wherein the OPC UA client is in communication connection with an OPC UA server and is used as a receiving end of data transmission and used for displaying production data and comparing the production data with a set range; the welding quality prediction system is in communication connection with the data acquisition and storage layer, and production data of the workpiece welding process are submitted to the welding quality prediction system to perform welding quality prediction; the MES system is used for marking workpieces with production data not in a set range or workpieces with unqualified welding quality prediction.
Further, the data collection and storage layer also includes a database for storing production data.
Furthermore, a welding quality prediction model based on a convolutional neural network is integrated in the welding quality prediction system, and the welding quality prediction model takes production data of a workpiece welding process as input and takes a welding quality grade as output.
Further, if a certain welded workpiece is judged to be unqualified in the subsequent quality detection in the production process, the production data of the unqualified workpiece is extracted, and the welding quality prediction model is trained by utilizing the production data of the unqualified workpiece, so that the real-time updating and perfecting of the welding quality prediction model are realized.
Further, the production data includes welding process parameters and workpiece size data.
A workpiece welding quality prediction method based on OPC UA architecture comprises the following steps:
s1, starting welding processing of a workpiece, and collecting production data of a current workpiece welding process by a data collecting and storing layer through an equipment layer;
s2, sending the collected production data to an OPC UA client of an application layer for display, comparing the production data with a set range, if the production data of a certain workpiece is within the set range, carrying out S3, if the production data of the certain workpiece is not within the set range, marking the workpiece through an MES system, and carrying out key quality detection after the welding of the workpiece is completed;
and S3, submitting the collected production data to a welding quality prediction system for quality prediction, and judging whether the welding quality of the workpiece is qualified or not.
Further, if the predicted result is that the workpiece is unqualified in the S3, the workpiece is marked through the MES system, and key quality detection is performed after the welding of the workpiece is completed.
And S4, performing quality detection after the welding of the workpiece is finished, and if the quality detection of the workpiece is unqualified, extracting production data of the workpiece and updating a welding quality prediction model in a training welding quality prediction system.
The invention has the beneficial effects that:
1. the invention adopts OPC UA architecture to solve the inconvenience caused by different equipment communication protocols, can rapidly and orderly collect production data in the production process of a welding workshop, and provides a basis for the subsequent quality prediction and quality tracing.
2. According to the method, the welding quality of the workpiece is predicted in real time by using a welding quality prediction model based on a convolutional neural network, and if the predicted result is that the workpiece is unqualified, the workpiece is marked by an MES system, and key quality detection is performed after the welding of the workpiece is completed.
3. According to the invention, quality tracing is combined, and the production data of unqualified workpieces are extracted to update the welding quality prediction model in the training welding quality prediction system, so that the growth and perfection of the welding quality prediction model are maintained.
Drawings
FIG. 1 is a schematic structural diagram of a workpiece welding quality prediction device based on an OPC UA architecture;
FIG. 2 is a flowchart of a workpiece welding quality prediction method based on an OPC UA architecture;
FIG. 3 is a schematic diagram of an operational mode of the OPC UA architecture;
FIG. 4 is a schematic diagram of the architecture of OPC UA architecture;
FIG. 5 is a schematic diagram of a connection of an OPC UA client to an OPC UA server;
FIG. 6 is a schematic diagram of a convolutional neural network;
FIG. 7 is a schematic diagram of a convolutional neural network structure used to train a weld quality prediction model in accordance with the present invention.
Detailed Description
Further advantages and effects of the present invention will become readily apparent to those skilled in the art from the disclosure herein, by referring to the accompanying drawings and the preferred embodiments. The invention may be practiced or carried out in other embodiments that depart from the specific details, and the details of the present description may be modified or varied from the spirit and scope of the present invention. It should be understood that the preferred embodiments are presented by way of illustration only and not by way of limitation.
It should be noted that the illustrations provided in the following embodiments merely illustrate the basic concept of the present invention by way of illustration, and only the components related to the present invention are shown in the drawings and are not drawn according to the number, shape and size of the components in actual implementation, and the form, number and proportion of the components in actual implementation may be arbitrarily changed, and the layout of the components may be more complicated.
Referring to fig. 1, the workpiece welding quality prediction device based on OPC UA architecture includes an equipment layer, a data acquisition and storage layer and an application layer. The equipment layer comprises welding equipment and various sensors and is used for acquiring production data of a workpiece welding process, wherein the production data comprise welding process parameters and workpiece size data. The data acquisition and storage layer comprises an OPC UA server and a database for storing production data, and the OPC UA server is in communication connection with the equipment layer and is used as a data acquisition end and a data transmission sending end. The application layer comprises an OPC UA client, a welding quality prediction system and an MES system, wherein the OPC UA client is in communication connection with an OPC UA server and is used as a receiving end of data transmission and used for displaying production data and comparing the production data with a set range; the welding quality prediction system is in communication connection with the data acquisition and storage layer, and production data of the workpiece welding process are submitted to the welding quality prediction system to perform welding quality prediction; the MES system is used for marking workpieces with production data not in a set range or workpieces with unqualified welding quality prediction.
The OPC UA architecture consists of an OPC UA client and an OPC UA server, wherein the OPC UA server is responsible for the data acquisition work of the underlying device and responding to the request of the client. The OPC UA client is responsible for making a request to the OPC UA server, and the OPC UA server responds to send the received information to the user. The working modes of the OPC UA customer service end and the OPC UA server are shown in figure 3.
Referring to fig. 4, the specific steps for the opc UA client to interact with the server are as follows:
(1) an OPC UA client sends a service request;
(2) the service request is sent to the server after passing through the OPC UA communication stack;
(3) after receiving the request, the OPC UA server calls a corresponding service set to execute tasks on nodes in an address space or in modules such as monitoring subscription and the like;
(4) the OPC UA server returns a corresponding request response to the OPC UA client;
(5) the response is sent to the OPC UA client via the OPC UA communications stack;
(6) the OPC UA client processes the information returned by the server;
(7) the OPC UA client returns node information and the like acquired from the address space of the OPC UA server to the user.
The principle of applying the OPC UA protocol to the workshop is that an OPC UA server is embedded into various bottom production devices of the workshop, data acquired by various numerical control systems are mapped into an OPC UA server address space correspondingly, and data communication is started after a client of the OPC UA is connected with a server.
Data transmission is performed by using an OPC UA architecture, see fig. 5, and the specific steps include:
(1) search server
The OPC UA client first needs to establish a connection with a discovery server, and then obtains a registered list of OPC UA servers through the discovery server, thereby realizing data communication.
The OPC UA server registers by calling a register Server () method, and after the OPC UA client establishes connection with the discovery server, the OPC UA client searches all registered OPC UA servers by calling a FindServers () method provided by the discovery server, thereby acquiring the required relevant description information of the OPC UA servers, including all information in the connection process and the like. Finally, the terminal information of the OPC UA server is obtained by calling the GetEndpoints () method, and this terminal information contains information such as IP address and security settings required for establishing a connection between the OPC UA client and the OPC UA server.
(2) Browsing address space
The address space is a collection of information with various meanings in the OPC UA server. The OPC UA unifies address space, service and security models, and unified access of information is realized by providing a unified data interface, so that the communication rate of data is improved.
The address space of OPC UA is composed of some nodes inside the OPC UA server, these nodes are also called as basic units of address space, data in the address space is represented by these nodes, and the nodes of OPC UA have various types, and different types of nodes have some type-determined attributes in addition to some general attributes. The more important node types are three, namely: object nodes, variable nodes and method nodes.
Nodes of object types in the address space manage the objects themselves and their included variables, methods, and bound events. The OPC UA client can obtain the variable values of the object nodes in the address space of the OPC UA server by means of reading and subscribing. The variable type node is used for storing the bottom layer information of the workshop equipment acquired by the OPC UA server. The values of the variable nodes in the address space can be read or modified by the OPC UA client, and the data changes of the variable nodes can be subscribed to for monitoring. The method node is configured by an OPC UA server, and comprises the modes of setting input parameters of an OPC UA client and returning output results, and is generally simple in data processing.
The data generated by the workshop equipment or the sensor are bound in the address space of the server in the form of nodes, and each node object represents one data item. The OPC UA client accesses the address space of the workshop equipment, the sensor and the like to acquire data in a browsing, reading and subscribing mode.
And the OPC UA client submits the NodeId and the filtering condition of the initial node to the OPC UA server by using a Browse service, and searches all node lists in an address space of the server.
After the client acquires the node list, the attribute value of the node to be accessed is Read from the address space by using the Read service with nodeToread as an input parameter, and the Read result is displayed through an updateAttributeList () function.
(3) Subscribing to data
The implementation of data real-time acquisition relies on OPC UA periodically acquiring data through a publish/subscribe mode. First, the OPC UA client initializes a subscription service to create a monitoring item. And then the OPC UA server samples the monitoring items at equal intervals and returns the data to the OPC UA client.
Subscription implementation relies primarily on CreateSubscription, publish, deleteSubscription, createMonitoredItems, deleteMonitoredItems, modifyMonitoredItems methods, as shown in table 1:
table 1 subscription implementation function and function meaning
| Function of | Functional meaning |
| CreateSubscription() | New subscription item |
| Publish() | Publishing subscription items |
| DeleteSubscription() | Deleting subscription items |
| CreateMonitoredItems() | New monitoring item |
| ModifyMonitoredItems() | Modifying a monitoring item |
| DeleteMonitoredItems() | Deleting monitoring items |
When creating a subscription item using createsubdescription (), a publish-subscribe interval, a modify-publish interval, a request lifetime number, a modify lifetime number, and the like can be set. When creating a monitoring item using createMonitoredItems (), it is necessary to specify a list of returned results.
When detecting the welding quality of a workpiece, an automobile manufacturer has a classification mode for the welding quality of the workpiece, and trains a welding quality prediction model based on a convolutional neural network model according to different production data and corresponding welding quality grades.
The welding quality prediction system is integrated with a welding quality prediction model based on a convolutional neural network, wherein the welding quality prediction model takes production data of a workpiece welding process as input and takes welding quality grade as output.
Referring to fig. 6, a Convolutional Neural Network (CNN) is a deep neural network with a convolutional structure, including an input layer, a convolutional layer, a pooling layer, a full-connection layer, and an output layer. Meanwhile, the convolutional neural network CNN adopts a local connection and weight sharing mode, so that network complexity is reduced, and the learning rate of the network is improved.
The convolution layers are mainly used for learning features, input data of the convolution layers are output data of the input layers or the pooling layers, each convolution kernel carries out convolution calculation with one layer of feature map, and after an activation function, the output feature map can be obtained. The calculation formula of the convolution process is as follows:
wherein: l is the layer number of the convolutional neural network structure, j is the jth channel, and +.>
The j-th channel output of the convolution layer l is output, and f is an activation function; n (N) j Is a feature map subset;
is a convolution kernel; * For convolution operation, ++>
For additional biasing.
The pooling layer is also called a downsampling layer, and dimension reduction is carried out on the feature map extracted by the convolution layer, so that the network calculation amount is effectively reduced. Common pooling methods include average pooling and maximum pooling. Maximum pooling computes the feature map entered in the pooling window by computing the maximum value. The calculation formula of the pooling layer is as follows:
wherein: l represents the layer number of the convolutional neural network structure, down is the pooling operation process, w and +.>
Respectively representing the pooling weights and the additional bias.
In the fully connected network, the feature map of the upper layer is spliced into one-dimensional features serving as the input of the fully connected network, and fully connected calculation is performed. The calculation formula is as follows:
wherein: w (w) j Is the weight of the fully connected network.
The training process of the convolutional neural network is similar to that of the traditional neural network and is divided into two stages of forward propagation and backward propagation. In the forward propagation stage, sample data of a training set is input into a convolutional neural network, further, a predicted value of the sample is calculated, meanwhile, an error between the predicted value and an actual value is obtained, and if the error exceeds a set value, the error enters into an error counter propagation stage. The back propagation process is to propagate the error between the predicted value and the true value layer by layer through the hidden layer to the input layer in a specific form. As the network counter propagates, the error signal acts on the layers of neurons, continually correcting weights and biases until the error of the network falls below a threshold.
The design is put into a convolutional neural network and consists of 1 input layer, 2 convolutional layers, 2 pooling layers, 1 full-connection layer and 1 output layer.
The production data such as welding process voltage, current, temperature and the like are used as input, the welding quality grade is used as output, and the model is trained according to the actual production data and can be used as a welding quality prediction model after being trained well.
Referring to fig. 7, the convolutional neural network structure for training the welding quality prediction model of the present invention is shown, and after the welding quality prediction model is trained, the convolutional neural network structure is applied to actual production to realize real-time welding quality prediction. In the welding process of the current workpiece, the current workpiece is input into a welding quality prediction model according to the acquired production data to obtain a predicted welding quality grade, and if the predicted quality is unqualified, the current workpiece is marked so as to facilitate the subsequent key detection.
If a certain welded workpiece is judged to be unqualified in the subsequent quality detection in the production process, the production data of the unqualified workpiece is extracted, and the welding quality prediction model is trained by utilizing the production data of the unqualified workpiece, so that the real-time updating and perfection of the welding quality prediction model are realized, the shelf time of the unqualified product data is reduced, and more accurate service can be provided for subsequent production.
Referring to fig. 2, the workpiece welding quality prediction method based on OPC UA architecture includes the following steps:
s1, starting production, starting welding processing of a workpiece, and collecting production data of the current workpiece welding process by a data collecting and storing layer through an equipment layer, wherein the production data comprise welding technological parameters and workpiece size data obtained by welding equipment and sensors of the equipment layer. The OPC UA server collects production data of the current workpiece welding process and stores the production data in a database.
And S2, sending the collected production data to an OPC UA client of an application layer for display, wherein the OPC UA client compares the production data with a set range, if the production data of a certain workpiece is within the set range, S3 is carried out, if the production data of the certain workpiece is not within the set range, the workpiece is marked by an MES system, and after the welding of the workpiece is completed, the important quality of the workpiece is detected.
S3, submitting the collected production data to a welding quality prediction system for quality prediction, inputting the production data to a welding quality prediction model in the welding quality prediction system, outputting a welding quality grade, and judging whether the welding quality of the workpiece is qualified. If the predicted result is that the workpiece is unqualified, marking the workpiece through an MES system, and performing key quality detection on the workpiece after the welding of the workpiece is finished.
And S4, performing quality detection after the welding of the workpiece is finished, and if the quality detection of the workpiece is unqualified, extracting production data of the workpiece and updating a welding quality prediction model in a training welding quality prediction system.
S1 to S4 are repeated until the production is completely finished.
When a welding workshop starts to process, production data of a workpiece welding process can be conveniently collected through an OPC UA server based on an OPC UA architecture, and intuitively displayed through an OPC UA client, and the method is convenient for using subsequent data due to the uniformity and standardization of the OPC UA architecture, and meanwhile, data from different welding machines, different devices and different sensors are efficiently collected and converted into a uniform format, and the collected production data is stored in a database. When the OPC UA client finds that the machining data of the current car body has obvious errors, an error result is submitted to the MES management system, the workpiece is marked, and the marked problem workpiece is detected in a key way. And submitting the current workpiece production data to a welding quality prediction model for quality prediction, if the prediction result is unqualified, submitting the unqualified result to an MES management system, marking the workpiece, and detecting the marked problem workpiece with emphasis. If a problem workpiece appears in the vehicle body quality detection process, the production data of the workpiece are extracted from the database and used for updating the training welding quality prediction model, and the previous process is repeated until the production is finished.
The above embodiments are merely preferred embodiments for fully explaining the present invention, and the scope of the present invention is not limited thereto. Equivalent substitutions and modifications will occur to those skilled in the art based on the present invention, and are intended to be within the scope of the present invention.
Claims (8)
1. Work piece welding quality prediction device based on OPCUA framework, its characterized in that: the system comprises an equipment layer, a data acquisition and storage layer and an application layer;
the equipment layer is used for acquiring production data of a workpiece welding process;
the data acquisition and storage layer comprises an OPCUA server, and the OPCUA server is in communication connection with the equipment layer and is used as a data acquisition end and a data transmission sending end;
the application layer comprises an OPCUA client, a welding quality prediction system and an MES system, wherein the OPCUA client is in communication connection with an OPCUA server and is used as a receiving end of data transmission and used for displaying production data and comparing the production data with a set range; the welding quality prediction system is in communication connection with the data acquisition and storage layer, and production data of the workpiece welding process are submitted to the welding quality prediction system to perform welding quality prediction; the MES system is used for marking workpieces with production data not in a set range or workpieces with unqualified welding quality prediction.
2. The OPCUA architecture-based workpiece weld quality prediction apparatus according to claim 1, wherein: the data acquisition and storage layer also includes a database for storing production data.
3. The OPCUA architecture-based workpiece welding quality prediction apparatus according to claim 1 or 2, wherein: the welding quality prediction system is integrated with a welding quality prediction model based on a convolutional neural network, wherein the welding quality prediction model takes production data of a workpiece welding process as input and takes welding quality grade as output.
4. The OPCUA architecture-based workpiece weld quality prediction apparatus according to claim 3, wherein: if a certain welded workpiece is judged to be unqualified in the subsequent quality detection in the production process, the production data of the unqualified workpiece is extracted, and the welding quality prediction model is trained by utilizing the production data of the unqualified workpiece, so that the real-time updating and perfecting of the welding quality prediction model are realized.
5. The OPCUA architecture-based workpiece welding quality prediction apparatus according to claim 1 or 2, wherein: the production data includes welding process parameters and workpiece size data.
6. The workpiece welding quality prediction method based on the OPCUA framework is characterized by comprising the following steps of:
s1, starting welding processing of a workpiece, and collecting production data of a current workpiece welding process by a data collecting and storing layer through an equipment layer;
s2, sending the collected production data to an OPCUA client of an application layer for display, comparing the production data with a set range, if the production data of a certain workpiece is within the set range, carrying out S3, if the production data of the certain workpiece is not within the set range, marking the workpiece through an MES system, and carrying out key quality detection after the welding of the workpiece is completed;
and S3, submitting the collected production data to a welding quality prediction system for quality prediction, and judging whether the welding quality of the workpiece is qualified or not.
7. The method for predicting welding quality of a workpiece based on an OPCUA framework according to claim 6, wherein if the predicted result is that the workpiece is not qualified in S3, the workpiece is marked by an MES system, and the key quality is detected after the welding of the workpiece is completed.
8. The method for predicting welding quality of a workpiece based on OPCUA architecture according to claim 6 or 7, further comprising S4, performing quality detection after welding of the workpiece, and if the quality detection of the workpiece is not qualified, extracting production data of the workpiece and updating a welding quality prediction model in a training welding quality prediction system.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211741310.XA CN116245216A (en) | 2022-12-31 | 2022-12-31 | Workpiece welding quality prediction device and method based on OPC UA architecture |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211741310.XA CN116245216A (en) | 2022-12-31 | 2022-12-31 | Workpiece welding quality prediction device and method based on OPC UA architecture |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116245216A true CN116245216A (en) | 2023-06-09 |
Family
ID=86625297
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211741310.XA Pending CN116245216A (en) | 2022-12-31 | 2022-12-31 | Workpiece welding quality prediction device and method based on OPC UA architecture |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116245216A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117436769A (en) * | 2023-12-20 | 2024-01-23 | 山东方垠智能制造有限公司 | Structural part welding quality monitoring method, system, storage medium and equipment |
-
2022
- 2022-12-31 CN CN202211741310.XA patent/CN116245216A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117436769A (en) * | 2023-12-20 | 2024-01-23 | 山东方垠智能制造有限公司 | Structural part welding quality monitoring method, system, storage medium and equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| KR101917006B1 (en) | Semiconductor Manufacturing Yield Prediction System and Method based on Machine Learning | |
| US20150269495A1 (en) | Methods and system for providing simultaneous multi-task ensemble learning | |
| CN111414457A (en) | Intelligent question-answering method, device, equipment and storage medium based on federal learning | |
| CN111796957B (en) | Transaction abnormal root cause analysis method and system based on application log | |
| US20090118842A1 (en) | Manufacturing prediction server | |
| CN116245216A (en) | Workpiece welding quality prediction device and method based on OPC UA architecture | |
| CN112149352B (en) | Prediction method for marketing activity clicking by combining GBDT automatic characteristic engineering | |
| KR20200010624A (en) | Big Data Integrated Diagnosis Prediction System Using Machine Learning | |
| CN111123223A (en) | General development platform, management system and method for radar health management | |
| CN112541729A (en) | Big data based visual intelligent management and control method and system for production whole process | |
| CN116128254A (en) | Embedded intelligent manufacturing resource configuration method and terminal system based on edge calculation | |
| CN111159559A (en) | Method for constructing recommendation engine according to user requirements and user behaviors | |
| US20240161068A1 (en) | Machine learning based vehicle service recommendation system | |
| US20090119077A1 (en) | Use of simulation to generate predictions pertaining to a manufacturing facility | |
| CN117689256A (en) | Aluminum alloy casting product quality tracing method | |
| CN113111139A (en) | Alarm detection method and device based on Internet of things sensor | |
| CN112799949A (en) | Model optimization method, device, system and equipment | |
| CN110458664B (en) | User travel information prediction method, device, equipment and storage medium | |
| CN111447193A (en) | Method and device for anomaly detection of real-time data stream | |
| CN116628381A (en) | AI prediction-based web page data design module fault lossless recovery method | |
| CN110855519A (en) | Network flow prediction method | |
| CN115619475A (en) | Commodity recommendation method, commodity recommendation system and related devices | |
| CN114595842A (en) | Advertisement playing equipment group event studying and judging method and system based on real-time calculation | |
| CN113361811A (en) | Method, system, device and computer readable storage medium for predicting operation state | |
| CN112783977A (en) | Mass data search implementation method based on big data |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |