CN111231251A - Product detection method, equipment and system of injection molding machine - Google Patents
Product detection method, equipment and system of injection molding machine Download PDFInfo
- Publication number
- CN111231251A CN111231251A CN202010022321.7A CN202010022321A CN111231251A CN 111231251 A CN111231251 A CN 111231251A CN 202010022321 A CN202010022321 A CN 202010022321A CN 111231251 A CN111231251 A CN 111231251A
- Authority
- CN
- China
- Prior art keywords
- product
- model
- training
- injection molding
- 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.)
- Granted
Links
Images
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B29—WORKING OF PLASTICS; WORKING OF SUBSTANCES IN A PLASTIC STATE IN GENERAL
- B29C—SHAPING OR JOINING OF PLASTICS; SHAPING OF MATERIAL IN A PLASTIC STATE, NOT OTHERWISE PROVIDED FOR; AFTER-TREATMENT OF THE SHAPED PRODUCTS, e.g. REPAIRING
- B29C45/00—Injection moulding, i.e. forcing the required volume of moulding material through a nozzle into a closed mould; Apparatus therefor
- B29C45/17—Component parts, details or accessories; Auxiliary operations
- B29C45/76—Measuring, controlling or regulating
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B29—WORKING OF PLASTICS; WORKING OF SUBSTANCES IN A PLASTIC STATE IN GENERAL
- B29C—SHAPING OR JOINING OF PLASTICS; SHAPING OF MATERIAL IN A PLASTIC STATE, NOT OTHERWISE PROVIDED FOR; AFTER-TREATMENT OF THE SHAPED PRODUCTS, e.g. REPAIRING
- B29C45/00—Injection moulding, i.e. forcing the required volume of moulding material through a nozzle into a closed mould; Apparatus therefor
- B29C45/17—Component parts, details or accessories; Auxiliary operations
- B29C45/76—Measuring, controlling or regulating
- B29C45/7686—Measuring, controlling or regulating the ejected articles, e.g. weight control
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B29—WORKING OF PLASTICS; WORKING OF SUBSTANCES IN A PLASTIC STATE IN GENERAL
- B29C—SHAPING OR JOINING OF PLASTICS; SHAPING OF MATERIAL IN A PLASTIC STATE, NOT OTHERWISE PROVIDED FOR; AFTER-TREATMENT OF THE SHAPED PRODUCTS, e.g. REPAIRING
- B29C2945/00—Indexing scheme relating to injection moulding, i.e. forcing the required volume of moulding material through a nozzle into a closed mould
- B29C2945/76—Measuring, controlling or regulating
- B29C2945/76177—Location of measurement
- B29C2945/7629—Moulded articles
Landscapes
- Engineering & Computer Science (AREA)
- Manufacturing & Machinery (AREA)
- Mechanical Engineering (AREA)
- Injection Moulding Of Plastics Or The Like (AREA)
Abstract
The embodiment of the invention provides a product detection method, equipment and a system of an injection molding machine, wherein the method comprises the steps of obtaining production data generated by the injection molding machine in the production period of each product to be detected, the production data comprises production process data and/or mold cavity pressure data of the injection molding machine, inputting the production data into a pre-trained product detection model, and obtaining a detection result of the product to be detected, wherein the detection result is used for representing the class of the quality of the product to be detected in a preset quality standard, the product detection model is a model for determining the quality of the product to be detected based on neural network model training, and the quality of the plastic product produced by the injection molding machine is detected based on the pre-trained product detection model, so that the product detection method, equipment and system are not easily interfered by external environment, and the detection accuracy is improved.
Description
Technical Field
The embodiment of the invention relates to the field of intelligent manufacturing of injection molding machines, in particular to a product detection method, equipment and system of an injection molding machine.
Background
With the rapid development of the plastic machinery industry, plastic products produced by the plastic machinery industry are applied to various fields. With the increasing output of plastic products, there are increasing demands on production efficiency and product quality, and how to accurately detect and delete the defective products generated in the production process becomes an important problem.
In the prior art, the quality of products produced by an injection molding machine is mainly detected by a visual detection device, and plastic products are detected by an image recognition technology to identify defective products in the plastic products.
However, the visual detection device is additionally arranged for each injection molding machine, so that the cost of the injection molding machine is increased, and the plastic product is detected based on the image recognition technology, so that the detection accuracy is low due to the influence of the shooting effect, the illumination environment, the object shielding and the like.
Disclosure of Invention
The embodiment of the invention provides a product detection method, equipment and a system of an injection molding machine, which are used for improving the accuracy of plastic product detection and further improving the production efficiency and the product quality of injection molding equipment.
In a first aspect, an embodiment of the present invention provides a product inspection method for an injection molding machine, which is applied to an edge device, and the method includes:
acquiring production data generated by an injection molding machine in the production period of each product to be detected; the production data comprises production process data and/or mold cavity pressure data of the injection molding machine;
inputting the production data into a pre-trained product detection model to obtain a detection result of the product to be detected; the detection result is used for representing the category of the quality of the product to be detected in a preset quality standard;
the product detection model is a model for determining the quality of the product to be detected based on neural network model training.
In a specific implementation manner, before the inputting the production data into a pre-trained product detection model to obtain a detection result of the product to be detected, the method further includes:
and receiving the product detection model sent by the server.
Optionally, the production process data includes: at least one of actual mold closing time, mold opening time, pressure maintaining end point, injection time, glue melting time, minimum residual quantity, glue melting end point, machine cycle time, pressure maintaining switching position, maximum injection pressure, maximum injection speed, glue melting starting point and switching pressure.
In a second aspect, an embodiment of the present invention provides a product detection method for an injection molding machine, which is applied to a server, and the method includes:
acquiring a plurality of training data uploaded by an auxiliary training machine, wherein each training data comprises production data and a corresponding quality label;
training by adopting a neural network model according to the plurality of training data to obtain a product detection model;
sending the product inspection model to at least one edge device; and each edge device is used for detecting the quality of the product to be monitored produced by the injection molding machine according to the product detection model.
In a specific implementation manner, the obtaining the product detection model by using neural network training according to the plurality of training data includes:
sequentially inputting the production data of the plurality of training data into a neural network model to obtain an output result of each production data;
and training the neural network model according to the loss function, the output result and the quality label of the training data, and ending the training process when a preset condition is met to obtain the product detection model.
Optionally, the preset conditions include: the training times are larger than the preset times, or the detection accuracy of the neural network model is larger than the preset accuracy.
In a specific implementation, the method further includes:
obtaining a plurality of model optimization data generated and uploaded by the auxiliary training machine in the production process, wherein each model optimization data comprises production data and a corresponding quality label;
optimizing the product inspection model according to the plurality of model optimization data;
and determining whether to update the product detection model of the at least one edge device according to the accuracy of the optimized product detection model and the accuracy of the product detection model before optimization.
Specifically, the comparing the accuracy of the product detection model with the accuracy of the product detection model before optimization, and updating the product detection model of the at least one edge device according to a preset condition includes:
respectively calculating the accuracy of the optimized product detection model and the accuracy of the product detection model before optimization according to the plurality of model optimization data;
and if the difference between the accuracy of the optimized product detection model and the accuracy of the product detection model before optimization is greater than a first preset value and/or the accuracy of the product detection model before optimization is lower than a second preset value, updating the product detection model of the at least one edge device.
In a third aspect, an embodiment of the present invention provides an edge device, including:
the acquisition module is used for acquiring production data generated by the injection molding machine in the production period of each product to be detected; the production data comprises production process data and a mold cavity pressure change curve of the injection molding machine.
The detection module is used for inputting the production data into a pre-trained product detection model to obtain a detection result of the product to be detected; the detection result is used for indicating whether the quality of the product to be detected meets a preset quality standard or not;
the product detection model is a model for determining the quality of the product to be detected based on neural network model training.
In a fourth aspect, an embodiment of the present invention provides a server, including:
the acquisition module is used for acquiring a plurality of training data uploaded by the auxiliary training machine, and each training data comprises production data and a corresponding quality label;
the processing module is used for obtaining a product detection model by adopting neural network training according to the plurality of training data;
a sending module for sending the product inspection model to at least one edge device; and each edge device is used for detecting the quality of the product to be monitored produced by the injection molding machine according to the product detection model.
In a fifth aspect, an embodiment of the present invention provides a product inspection system for an injection molding machine, including: the system comprises a cloud server, at least one edge device and an auxiliary training machine, wherein the edge device and the auxiliary training machine are connected with the cloud server; each edge device is connected with an injection molding machine;
the at least one edge device is used for executing the product inspection method of the injection molding machine of the first aspect;
the cloud server is used for executing the product detection method of the injection molding machine in the second aspect;
the auxiliary training machine is used for providing training data or model optimization data for the cloud server.
The embodiment of the invention provides a product detection method, equipment and a system of an injection molding machine.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly introduced below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and those skilled in the art can also obtain other drawings according to the drawings without inventive exercise.
FIG. 1 is a schematic diagram of an application of a product inspection system of an injection molding machine according to an embodiment of the present invention;
FIG. 2 is a flow chart of a first embodiment of a product inspection method of an injection molding machine according to an embodiment of the present invention;
FIG. 3 is a schematic flow chart of a second embodiment of a product inspection method of an injection molding machine according to an embodiment of the present invention;
FIG. 4 is a schematic flow chart of a third embodiment of a product inspection method of an injection molding machine according to an embodiment of the present invention;
FIG. 5 is a flowchart illustrating a fourth embodiment of a product inspection method for an injection molding machine according to an embodiment of the present invention;
FIG. 6 is a schematic flow chart of updating a product inspection model according to the present invention;
FIG. 7 is a diagram illustrating the results of a neural network model according to an embodiment of the present invention;
FIG. 8 is a schematic flow chart of a fifth embodiment of a product inspection method of an injection molding machine according to an embodiment of the present invention;
FIG. 9 is a schematic structural diagram of a first embodiment of an edge device according to an embodiment of the present invention;
fig. 10 is a schematic structural diagram of a first server according to an embodiment of the present invention;
fig. 11 is a schematic hardware structure diagram of an edge device according to an embodiment of the present invention;
fig. 12 is a schematic diagram of a hardware structure of a server according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. All other embodiments, which can be obtained by a person skilled in the art without any inventive step based on the embodiments of the present invention, are within the scope of the present invention.
The invention provides a product detection method, equipment and system of an injection molding machine, aiming at solving the problems in the prior art, a neural network model is pre-trained by adopting a large amount of data to obtain a product detection model, the quality of each product produced by the injection molding machine is detected through the continuously optimized product detection model, the detection result is more accurate, the injection molding machine is not required to be improved, and compared with the prior art that the product quality is detected through a visual detection device, the production cost is reduced.
Fig. 1 is a schematic application diagram of a product inspection system of an injection molding machine according to an embodiment of the present invention, and as shown in fig. 1, the product inspection system of the injection molding machine includes: the cloud server 01, at least one edge device 02 connected with the cloud server, also called edge node.
The edge devices 02 are any terminal device with a processor and used for providing entry points for a system, each edge device 02 is connected with an injection molding machine, or each edge device 02 belongs to a controller of the injection molding machine and collects production data from the injection molding machine, moreover, each edge device 02 is connected with the cloud server 01 in a wired or wireless mode and performs information interaction with the cloud server 01, for example, a pre-trained product detection model is downloaded from the cloud server 01, and a detection result or the production data of the injection molding machine is uploaded to the cloud server 01. Optionally, the edge device 02 may be connected to the router and then connected to the cloud server 01 in a wired or wireless manner.
The auxiliary training machine 03 is connected with the cloud server 01 and used for providing training data or model optimization data for the cloud server 01, the cloud server 01 trains to obtain a product detection model according to the training data uploaded by the auxiliary training machine 03, and the product detection model is optimized according to the model optimization data uploaded by the auxiliary training machine 03.
The cloud server 01 synchronizes the trained product detection model and the product detection model optimized in the production process to at least one edge device 02, so that each edge device detects each plastic product produced by the corresponding injection molding machine according to the latest product detection model.
Optionally, the training aid 03 can be directly connected with the cloud server 01, and also can be connected with the cloud server 01 through the edge device 02, for example, the edge device connected with the training aid 03 does not need to detect a product, and only provides training data and model optimization data of the training aid to the cloud server.
Exemplarily, the auxiliary training machine 03 is an injection molding machine or an injection molding device, and can periodically produce products identical to other injection molding machines, and produce production data in a production cycle of each product to be detected, optionally, the auxiliary training machine 03 is provided with a customized visual detection device for identifying the product quality of each product to be detected, and takes the identification result as a quality label of the production data corresponding to the product to be detected, and each production data and the corresponding quality label form training data or model optimization data to be uploaded to the cloud server, specifically, the training data or the model optimization data can be uploaded to the cloud database, so that the cloud server 01 trains a product detection model according to the training data, or optimizes the product detection model according to the model optimization data.
Optionally, the auxiliary training machine 03 is used as an injection molding machine or an injection molding device, production data are generated in the process of production or simulation production by the auxiliary training machine, corresponding quality labels are set for each production data, before the product detection model completes training, a plurality of training data are formed by a plurality of produced data and corresponding quality labels, the training data are sent to the cloud server 01, after the product detection model is trained, namely, in the production process of a product, a plurality of model optimization data are formed by a plurality of produced data and corresponding quality labels, and the training data are sent to the cloud server 01.
Optionally, the edge device 02 is further connected to a mold cavity pressure monitoring device, to obtain mold cavity pressure data in a production cycle of the product to be detected, for example, a variation curve of mold cavity pressure from the start of injection molding to the end of injection molding.
In a specific implementation manner, the injection molding machines and the training aids connected to at least one edge device 02 connected to the cloud server 01 belong to the same category of injection molding machines, where the same category refers to injection molding machines producing the same product and/or the same model, or may be injection molding machines classified into the same category through similarity calculation according to an artificial intelligence method, so that when the product detection model is applied to each edge device, the product detection model operates within a receivable range, for example, the accuracy of detecting the product quality by each edge device through the product detection model reaches more than 99%.
The product detection system of injection molding machine that this embodiment provided, including high in the clouds server 01 to and at least one edge device 02 and the training aiding machine 03 of being connected with high in the clouds server 01, an injection molding machine is connected to every edge device 02, has realized detecting the quality of product according to the production data and the product detection model of waiting to detect the product, and constantly optimizes the product detection model at the in-process of production, improves the accuracy of detecting.
On the basis of the above embodiments, the product inspection method of the injection molding machine provided in the following embodiments can be applied to any of the above system embodiments.
Fig. 2 is a schematic flowchart of a first embodiment of a product inspection method for an injection molding machine according to an embodiment of the present invention, an execution main body of the embodiment is an edge device 02 connected to the injection molding machine, or the edge device 02 integrated with the injection molding machine, as shown in fig. 2, the product inspection method for an injection molding machine includes:
s101: production data generated by the injection molding machine in the production period of each product to be detected is obtained.
It will be appreciated by those skilled in the art that the process of producing products by an injection molding machine is a cyclic process that produces one product to be tested per production cycle. In the step, the edge device acquires production data generated in each production period of the injection molding machine, so that the subsequent steps detect whether the corresponding product to be detected meets the quality requirement according to the production data of each period.
Wherein the production data comprises production process data and/or mold cavity pressure data of the injection molding machine; optionally, the production process data of the injection molding machine includes: at least one of actual mold closing time, mold opening time, pressure maintaining end point, injection time, glue melting time, minimum residual quantity, glue melting end point, machine cycle time, pressure maintaining switching position, maximum injection pressure, maximum injection velocity, glue melting starting point and switching pressure, for example, as shown in table 1, each column of data in table 1 is production process data obtained in one production cycle.
| Cycle counting | 1 | 2 | … |
| Actual mold clamping time | 0.81 | 0.81 | … |
| Time of opening mould | 1.1 | 1.1 | … |
| Pressure maintaining end point | 34 | 34 | … |
| Time of injection | 0.18 | 0.18 | … |
| Time of melting | 2.46 | 2.41 | … |
| Minimum residual | 34 | 34 | … |
| Melt glue terminal | 126.1 | 126 | … |
| Cycle time of machine | 5.62 | 5.62 | … |
| Pressure maintaining switch position | 82.1 | 82.1 | … |
| Maximum injection pressure | 2094 | 2113 | … |
| Maximum shooting velocity | 392.6 | 392.4 | … |
| Starting point of melt adhesive | 32.3 | 32.4 | … |
| Switching pressure | 1310 | 1311 | … |
TABLE 1
S102: and inputting the production data into a pre-trained product detection model to obtain a detection result of the product to be detected.
The production data obtained in step S101 is input into a pre-trained product detection model, and the product detection model outputs a detection result of the product to be detected, where the detection result is used to indicate a category of the quality of the product to be detected in a preset quality standard, and illustratively, the detection result may be a product that meets the quality standard and does not meet a quality label, such as a defective product or a genuine product, and further, the product that does not meet the quality standard is classified, such as a defective product having a flash or a defective product having a defect, or the detection result may be classified according to the quality standard, such as a primary genuine product, a secondary genuine product, a primary defective product, a secondary defective product, a waste report, and the like.
The product detection model is a model for determining the quality of the product to be detected based on neural network model training.
According to the product detection method of the injection molding machine, the obtained production data generated in the production period of each product to be detected are input into the pre-trained product detection model, so that the detection result of each product to be detected is obtained, and compared with the prior art that the product quality of each product is identified through an image identification technology, the product detection method is not easily interfered by the external environment, and the obtained detection result is more accurate.
Fig. 3 is a schematic flowchart of a second flowchart of an embodiment of a product detection method for an injection molding machine according to an embodiment of the present invention, where an execution main body of the embodiment is a server, and may specifically be a cloud server 01, as shown in fig. 3, the product detection method for an injection molding machine includes:
s201: a plurality of training data uploaded by an auxiliary training machine is obtained.
The auxiliary training machine can be an injection molding machine provided with a visual detection device, products to be detected are produced through controlling the auxiliary training machine to obtain production data of the products to be detected, corresponding quality labels are set for the production data according to the quality of each product to be detected, exemplarily, the quality labels can be set according to the detection result of the visual detection device, the quality of the products to be detected can also be judged in a manual detection mode, the corresponding quality labels are set for the production data, and each production data and the corresponding quality labels form training data.
The auxiliary training machine can upload training data after each production cycle is finished, or upload a plurality of training data at one time, and the scheme does not require the auxiliary training machine.
S202: and training by adopting a neural network model according to a plurality of training data to obtain a product detection model.
And training the neural network model by adopting a plurality of training data, so that the accuracy of the neural network model meets the requirement of product quality detection, and obtaining the product detection model.
In a specific implementation manner, the step includes a process as shown in fig. 4, and fig. 4 is a schematic flow chart of a third embodiment of the product detection method of the injection molding machine according to the embodiment of the present invention.
S2021: and sequentially inputting the production data of the plurality of training data into the neural network model to obtain an output result of each production data.
S2022: and training the neural network model according to the loss function, the output result and the quality label of the training data, and ending the training process when the preset condition is met to obtain the product detection model.
In steps S2021 to S2022, the production data of the plurality of training data are sequentially input to the neural network model to obtain an output result of each production data, the output result is a quality classification of the product to be detected and corresponds to the quality label in the training data, and the neural network model is continuously trained by using a loss function according to the quality label corresponding to each production data and the output result, so as to improve the accuracy of the neural network model, and the training process is ended until the neural network model satisfies a preset condition, so as to obtain the product detection model.
Optionally, the preset conditions include that the training process is ended when the training times are greater than a preset number, namely the training times of the neural network model reach the preset number, the detection accuracy of the neural network model is greater than a preset accuracy, for example, the accuracy obtained by detecting N groups of production data by the neural network model is calculated to be greater than the preset accuracy α, and the loss function value is smaller than a preset threshold value, namely the loss function value is continuously converged in the training process of the neural network model until the loss function value is converged to be smaller than the preset threshold value, so that the training process is ended.
S203: the product inspection model is sent to at least one edge device.
In this step, the cloud server sends the trained product detection model to at least one edge device, that is, each edge device receives the product detection model sent by the server, so that each edge device performs quality detection on the product to be detected produced by the injection molding machine according to the product detection model.
In this embodiment, training the neural network model according to the training data that the training machine provided through the high in the clouds server to obtain the higher product detection model of the degree of accuracy, and with product detection model synchronization to every edge device with high in the clouds server connection, make every edge device can carry out accurate quality inspection to the product of waiting of injection molding machine production according to product detection model, in order to discern wherein unsatisfied quality requirement's substandard product.
Before the embodiments shown in fig. 3 to 4 are applied to the product quality detection using the product detection model, the product detection model is continuously optimized in the process of producing products by the injection molding machine according to the embodiment shown in fig. 5, so as to improve the accuracy of the product detection model.
Fig. 5 is a schematic flowchart of a fourth embodiment of a product detection method of an injection molding machine according to an embodiment of the present invention, where an execution main body of the embodiment is a server, and may specifically be a cloud server 01, as shown in fig. 5, the product detection method of the injection molding machine includes:
s301: and acquiring a plurality of model optimization data generated and uploaded by the auxiliary training machine in the production process.
Similar to step S201, the training aid and the other injection molding machines for manufacturing products cooperate, for example, the same production steps are used to produce the same products, the training aid obtains the production data of the products to be detected during the production of the products to be detected, and sets a corresponding quality label for each production data according to the quality of each product to be detected, illustratively, the quality label may be set according to the result detected by the visual detection device, or the quality of the products to be detected may be determined by a manual detection method, and sets a corresponding quality label for each production data, and each production data and the corresponding quality label constitute a model optimization data.
S302: optimizing the product inspection model based on the plurality of model optimization data.
The cloud server continues to train the product detection model according to the model optimization data uploaded by the auxiliary training machine so as to optimize the product detection model and improve the accuracy of the product detection model.
S303: and determining whether to update the product detection model of the at least one edge device according to the accuracy of the optimized product detection model and the accuracy of the product detection model before optimization.
Referring to fig. 6, for a schematic flow chart of updating a product detection model provided by the present invention, a product detection model before optimization and a product detection model after optimization are stored in a cloud server at the same time, production data of a plurality of model optimization data are respectively input into the product detection model before optimization and the product detection model after optimization, respective accuracies are calculated, and whether to update the product detection model after optimization to at least one edge device is determined according to the accuracy of the product detection model before optimization and the accuracy of the product detection model after optimization.
For example, if the accuracy of the product inspection model before the optimization is lower than a second preset value, the product inspection model in the at least one edge device is updated to the optimized product inspection model, or if the accuracy of the product inspection model before the optimization is lower than the second preset value, the accuracy of the product inspection model before the optimization is updated to the optimized product inspection model.
In this embodiment, the product detection model is continuously optimized in the production process of the injection molding machine to improve the accuracy of the product detection model, so that the detection of the product detection model is more consistent with the actually produced product, and the accuracy of quality detection on the currently produced product is improved.
In a specific implementation manner, the neural network model in the cloud server may be a network structure in any shape. Optionally, the neural network model adopts a fully-connected network structure, as shown in fig. 7, which is a schematic diagram of a result of the neural network model provided in the embodiment of the present invention, and includes an input layer, a hidden layer, and an output layer. The output of the output layer is the same as the dimension of the label, for example, the dimension can be 2 dimensions, namely, the label is a good product or a defective product, and also can be multidimensional, namely, the good product has fins and the defective product is short of materials.
Optionally, the loss function is a mean square error. The neural network is optimized by back propagation according to a loss function.
On the basis of the above embodiments, taking an edge device as an example of a controller integrated in an injection molding machine, and fig. 8 is a schematic flow chart of a fifth embodiment of a product detection method for an injection molding machine according to an embodiment of the present invention, as shown in fig. 8, a plurality of injection molding machines connected to a cloud server belong to the same class of injection molding machines, including an injection molding machine 1, an injection molding machine 2, and an injection molding machine 3, where the injection molding machine 1 is an auxiliary training machine, the controller of the auxiliary training machine collects production process data of the auxiliary training machine and/or mold cavity pressure data of a mold cavity pressure monitoring device to form production data, and sets a corresponding quality label for each production data, for example, a detection result obtained by a visual detection device or a manual detection is used as a quality label for the production data to obtain a plurality of training data, and uploads the plurality of training data to the cloud server in batch, and training the neural network model by the cloud server according to the training data to obtain a product detection model. The auxiliary training machine obtains a plurality of model optimization data in the product production process, the specific process is similar to the process for obtaining the training data, the process is not repeated, the cloud server optimizes the product detection model according to the model optimization data, judges whether the optimized product detection model meets the preset condition according to a preset decision mechanism, and updates the optimized product detection model to the injection molding machine connected with the optimized product detection model when the preset condition is met. The injection molding machines 2 and 3 are application machines, and the injection molding machines 2 and 3 perform quality detection on the to-be-detected product produced by the injection molding machines according to the product detection model sent by the cloud server to obtain a detection result, namely a quality product or a defective product.
Fig. 9 is a schematic structural diagram of a first embodiment of an edge device according to the present invention, and as shown in fig. 9, the edge device 10 includes:
the acquisition module 11 is used for acquiring production data generated by the injection molding machine in the production cycle of each product to be detected; the production data comprises production process data and a mold cavity pressure change curve of the injection molding machine.
The detection module 12 is configured to input the production data into a pre-trained product detection model to obtain a detection result of the product to be detected; the detection result is used for indicating whether the quality of the product to be detected meets a preset quality standard or not;
the product detection model is a model for determining the quality of the product to be detected based on neural network model training.
The device provided in this embodiment may be used to implement the technical solution of the above method embodiment, and the implementation principle and technical effect are similar, which are not described herein again.
Fig. 10 is a schematic structural diagram of a first embodiment of a server according to the present invention, and as shown in fig. 10, the server 20 includes:
the acquisition module 21 is configured to acquire a plurality of training data uploaded by an auxiliary training machine, where each training data includes production data and a corresponding quality label;
the processing module 22 is configured to obtain a product detection model by using neural network training according to the plurality of training data;
a sending module 23, configured to send the product detection model to at least one edge device; and each edge device is used for detecting the quality of the product to be monitored produced by the injection molding machine according to the product detection model.
The device provided in this embodiment may be used to implement the technical solution of the above method embodiment, and the implementation principle and technical effect are similar, which are not described herein again.
Fig. 11 is a schematic hardware structure diagram of an edge device according to an embodiment of the present invention. As shown in fig. 11, the control device 60 of the present embodiment includes: a processor 601 and a memory 602; wherein
A memory 602 for storing computer-executable instructions;
the processor 601 is configured to execute the computer executable instructions stored in the memory to implement the steps performed by the edge device in the above embodiments. Reference may be made in particular to the relevant description in the preceding method embodiments.
Alternatively, the memory 602 may be separate or integrated with the processor 601.
When the memory 602 is provided separately, the control device further includes a bus 603 for connecting the memory 602 and the processor 601.
Fig. 12 is a schematic diagram of a hardware structure of a server according to an embodiment of the present invention. As shown in fig. 12, the server 70 of the present embodiment includes: a processor 701 and a memory 702; wherein
A memory 702 for storing computer-executable instructions;
the processor 701 is configured to execute the computer execution instructions stored in the memory, so as to implement the steps performed by the server or the cloud server in the foregoing embodiments. Reference may be made in particular to the description in relation to the method embodiments described above.
Alternatively, the memory 702 may be separate or integrated with the processor 701.
When the memory 702 is provided separately, the control device further includes a bus 703 for connecting the memory 702 and the processor 701.
The embodiment of the invention also provides a computer-readable storage medium, wherein a computer executing instruction is stored in the computer-readable storage medium, and when a processor executes the computer executing instruction, the product detection method of the injection molding machine is realized.
Those of ordinary skill in the art will understand that: all or a portion of the steps for implementing the above-described method embodiments may be performed by hardware associated with program instructions. The foregoing program may be stored in a computer-readable storage medium. When executed, the program performs steps comprising the method embodiments described above; and the aforementioned storage medium includes: various media that can store program codes, such as ROM, RAM, magnetic or optical disks.
Finally, it should be noted that: the above embodiments are only used to illustrate the technical solution of the present invention, and not to limit the same; while the invention has been described in detail and with reference to the foregoing embodiments, it will be understood by those skilled in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some or all of the technical features may be equivalently replaced; and the modifications or the substitutions do not make the essence of the corresponding technical solutions depart from the scope of the technical solutions of the embodiments of the present invention.
Claims (10)
1. A product inspection method for an injection molding machine, applied to an edge device, the method comprising:
acquiring production data generated by an injection molding machine in the production period of each product to be detected; the production data comprises production process data and/or mold cavity pressure data of the injection molding machine;
inputting the production data into a pre-trained product detection model to obtain a detection result of the product to be detected; the detection result is used for representing the category of the quality of the product to be detected in a preset quality standard;
the product detection model is a model for determining the quality of the product to be detected based on neural network model training.
2. The method of claim 1, wherein before the inputting the production data into a pre-trained product inspection model to obtain the inspection result of the product to be inspected, the method further comprises:
and receiving the product detection model sent by the server.
3. The method of claim 1 or 2, wherein the production process data comprises: at least one of actual mold closing time, mold opening time, pressure maintaining end point, injection time, glue melting time, minimum residual quantity, glue melting end point, machine cycle time, pressure maintaining switching position, maximum injection pressure, maximum injection speed, glue melting starting point and switching pressure.
4. A product inspection method for an injection molding machine, applied to a server, the method comprising:
acquiring a plurality of training data uploaded by an auxiliary training machine, wherein each training data comprises production data and a corresponding quality label;
training by adopting a neural network model according to the plurality of training data to obtain a product detection model;
sending the product inspection model to at least one edge device; and each edge device is used for detecting the quality of the product to be monitored produced by the injection molding machine according to the product detection model.
5. The method of claim 4, wherein the training using a neural network to obtain the product detection model according to the plurality of training data comprises:
sequentially inputting the production data of the plurality of training data into a neural network model to obtain an output result of each production data;
and training the neural network model according to the loss function, the output result and the quality label of the training data, and ending the training process when a preset condition is met to obtain the product detection model.
6. The method according to claim 5, wherein the preset conditions include: the training times are larger than the preset times, or the detection accuracy of the neural network model is larger than the preset accuracy.
7. The method of claim 4, further comprising:
obtaining a plurality of model optimization data generated and uploaded by the auxiliary training machine in the production process, wherein each model optimization data comprises production data and a corresponding quality label;
respectively calculating the accuracy of the optimized product detection model and the accuracy of the product detection model before optimization according to the plurality of model optimization data;
and if the difference between the accuracy of the optimized product detection model and the accuracy of the product detection model before optimization is greater than a first preset value and/or the accuracy of the product detection model before optimization is lower than a second preset value, updating the product detection model of the at least one edge device.
8. An edge device, comprising:
the acquisition module is used for acquiring production data generated by the injection molding machine in the production period of each product to be detected; the production data comprises production process data and a mold cavity pressure change curve of the injection molding machine.
The detection module is used for inputting the production data into a pre-trained product detection model to obtain a detection result of the product to be detected; the detection result is used for indicating whether the quality of the product to be detected meets a preset quality standard or not;
the product detection model is a model for determining the quality of the product to be detected based on neural network model training.
9. A server, comprising:
the acquisition module is used for acquiring a plurality of training data uploaded by the auxiliary training machine, and each training data comprises production data and a corresponding quality label;
the processing module is used for obtaining a product detection model by adopting neural network training according to the plurality of training data;
a sending module for sending the product inspection model to at least one edge device; and each edge device is used for detecting the quality of the product to be monitored produced by the injection molding machine according to the product detection model.
10. A product inspection system for an injection molding machine, comprising: the system comprises a cloud server, at least one edge device and an auxiliary training machine, wherein the edge device and the auxiliary training machine are connected with the cloud server; each edge device is connected with an injection molding machine;
the at least one edge device is used for executing the product inspection method of the injection molding machine according to any one of claims 1 to 3;
the cloud server is used for executing the product detection method of the injection molding machine of any one of claims 4 to 7;
the auxiliary training machine is used for providing training data or model optimization data for the cloud server.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010022321.7A CN111231251B (en) | 2020-01-09 | 2020-01-09 | Product detection method, equipment and system of injection molding machine |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010022321.7A CN111231251B (en) | 2020-01-09 | 2020-01-09 | Product detection method, equipment and system of injection molding machine |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111231251A true CN111231251A (en) | 2020-06-05 |
| CN111231251B CN111231251B (en) | 2022-02-01 |
Family
ID=70868407
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010022321.7A Active CN111231251B (en) | 2020-01-09 | 2020-01-09 | Product detection method, equipment and system of injection molding machine |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111231251B (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112330233A (en) * | 2021-01-05 | 2021-02-05 | 广州中和互联网技术有限公司 | Injection molding product quality detection method based on data model |
| CN112677438A (en) * | 2020-12-08 | 2021-04-20 | 上海微亿智造科技有限公司 | Segmented construction and screening method and system based on injection molding and machine debugging data characteristics |
| CN112749623A (en) * | 2020-12-08 | 2021-05-04 | 上海微亿智造科技有限公司 | Processing and feature extraction method and system for high-frequency sensor data of injection molding process |
| CN113112040A (en) * | 2021-04-16 | 2021-07-13 | 广东美的厨房电器制造有限公司 | Detection method, terminal, server, detection system and readable storage medium |
| CN116787698A (en) * | 2023-07-27 | 2023-09-22 | 九河精微塑胶工业(深圳)有限公司 | Mold for realizing double-runner injection molding and process for matching injection molding equipment |
| CN117245874A (en) * | 2023-11-14 | 2023-12-19 | 广东美的制冷设备有限公司 | Injection molding quality detection method, device, injection molding machine, equipment and storage medium |
| CN117455316A (en) * | 2023-12-20 | 2024-01-26 | 中山市东润智能装备有限公司 | Method for acquiring data of injection molding factory equipment |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20010051858A1 (en) * | 2000-06-08 | 2001-12-13 | Jui-Ming Liang | Method of setting parameters for injection molding machines |
| CN1353039A (en) * | 2000-11-03 | 2002-06-12 | 盟立自动化股份有限公司 | Intelligent control method of ejection former |
| CN101168282A (en) * | 2006-10-27 | 2008-04-30 | 日精树脂工业株式会社 | Support apparatus of injection-molding machine |
| CN108257121A (en) * | 2018-01-09 | 2018-07-06 | 北京百度网讯科技有限公司 | The newer method, apparatus of product defects detection model, storage medium and terminal device |
| CN108828409A (en) * | 2018-08-03 | 2018-11-16 | 南方电网科学研究院有限责任公司 | Fault detection system based on edge calculation |
| CN109142547A (en) * | 2018-08-08 | 2019-01-04 | 广东省智能制造研究所 | A kind of online lossless detection method of acoustics based on convolutional neural networks |
| JP2019014187A (en) * | 2017-07-10 | 2019-01-31 | 株式会社日本製鋼所 | Injection molding machine system including computer and a plurality of injection molding machines |
| CN110120044A (en) * | 2019-05-20 | 2019-08-13 | 北京百度网讯科技有限公司 | Article defects detection method and device, edge device, server |
| CN110598761A (en) * | 2019-08-26 | 2019-12-20 | 深圳大学 | Dispensing detection method and device and computer readable storage medium |
-
2020
- 2020-01-09 CN CN202010022321.7A patent/CN111231251B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20010051858A1 (en) * | 2000-06-08 | 2001-12-13 | Jui-Ming Liang | Method of setting parameters for injection molding machines |
| CN1353039A (en) * | 2000-11-03 | 2002-06-12 | 盟立自动化股份有限公司 | Intelligent control method of ejection former |
| CN101168282A (en) * | 2006-10-27 | 2008-04-30 | 日精树脂工业株式会社 | Support apparatus of injection-molding machine |
| JP2019014187A (en) * | 2017-07-10 | 2019-01-31 | 株式会社日本製鋼所 | Injection molding machine system including computer and a plurality of injection molding machines |
| CN108257121A (en) * | 2018-01-09 | 2018-07-06 | 北京百度网讯科技有限公司 | The newer method, apparatus of product defects detection model, storage medium and terminal device |
| CN108828409A (en) * | 2018-08-03 | 2018-11-16 | 南方电网科学研究院有限责任公司 | Fault detection system based on edge calculation |
| CN109142547A (en) * | 2018-08-08 | 2019-01-04 | 广东省智能制造研究所 | A kind of online lossless detection method of acoustics based on convolutional neural networks |
| CN110120044A (en) * | 2019-05-20 | 2019-08-13 | 北京百度网讯科技有限公司 | Article defects detection method and device, edge device, server |
| CN110598761A (en) * | 2019-08-26 | 2019-12-20 | 深圳大学 | Dispensing detection method and device and computer readable storage medium |
Non-Patent Citations (1)
| Title |
|---|
| 李琎: "注塑制品表面缺陷识别及工艺映射方法研究", 《中国优秀硕士学位论文全文数据库 工程科技I辑》 * |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112677438A (en) * | 2020-12-08 | 2021-04-20 | 上海微亿智造科技有限公司 | Segmented construction and screening method and system based on injection molding and machine debugging data characteristics |
| CN112749623A (en) * | 2020-12-08 | 2021-05-04 | 上海微亿智造科技有限公司 | Processing and feature extraction method and system for high-frequency sensor data of injection molding process |
| CN112677438B (en) * | 2020-12-08 | 2021-11-16 | 上海微亿智造科技有限公司 | Segmented construction and screening method and system based on injection molding and machine debugging data characteristics |
| CN112749623B (en) * | 2020-12-08 | 2023-04-07 | 上海微亿智造科技有限公司 | Processing and feature extraction method and system for high-frequency sensor data of injection molding process |
| CN112330233A (en) * | 2021-01-05 | 2021-02-05 | 广州中和互联网技术有限公司 | Injection molding product quality detection method based on data model |
| CN112330233B (en) * | 2021-01-05 | 2021-06-01 | 广州中和互联网技术有限公司 | Injection molding product quality detection method based on data model |
| CN113112040A (en) * | 2021-04-16 | 2021-07-13 | 广东美的厨房电器制造有限公司 | Detection method, terminal, server, detection system and readable storage medium |
| CN116787698A (en) * | 2023-07-27 | 2023-09-22 | 九河精微塑胶工业(深圳)有限公司 | Mold for realizing double-runner injection molding and process for matching injection molding equipment |
| CN117245874A (en) * | 2023-11-14 | 2023-12-19 | 广东美的制冷设备有限公司 | Injection molding quality detection method, device, injection molding machine, equipment and storage medium |
| CN117245874B (en) * | 2023-11-14 | 2024-04-02 | 广东美的制冷设备有限公司 | Injection molding quality detection method, device, injection molding machine, equipment and storage medium |
| CN117455316A (en) * | 2023-12-20 | 2024-01-26 | 中山市东润智能装备有限公司 | Method for acquiring data of injection molding factory equipment |
| CN117455316B (en) * | 2023-12-20 | 2024-04-19 | 中山市东润智能装备有限公司 | Method for acquiring data of injection molding factory equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111231251B (en) | 2022-02-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111231251B (en) | Product detection method, equipment and system of injection molding machine | |
| CN109754089B (en) | Model training system and method | |
| CN110531722B (en) | Process parameter recommendation system and method based on data acquisition | |
| CN111428329B (en) | Model-based machine learning system | |
| CN112906155B (en) | Virtual measurement method for injection molding information | |
| CN114757286A (en) | Multi-class fault data generation method based on conditional countermeasure generation network | |
| CN104392244A (en) | Synthetic aperture radar image classifying method based on stacked automatic coding machines | |
| CN110851920B (en) | Automatic generation method of main reinforcement line of die material pressing device | |
| CN113821637A (en) | Long text classification method and device, computer equipment and readable storage medium | |
| CN117312879A (en) | Injection molding machine production data supervision and early warning method, system and medium | |
| CN111310611B (en) | Method for detecting cell view map and storage medium | |
| CN116834244A (en) | Image monitoring alarm system and method for injection mold | |
| KR102546356B1 (en) | Optimization process algorithm automatic judgment system and method using machine learning model | |
| CN111695631A (en) | Method, device, equipment and medium for extracting verification fault features based on SAE | |
| CN116223529A (en) | Intelligent detection method and system for production of film-pressed gloves | |
| CN110936574A (en) | Method for judging working state of injection molding machine | |
| CN114530163B (en) | Method and system for adopting life cycle of voice recognition equipment based on density clustering | |
| Ma et al. | Intelligent recommendation system of the injection molding process parameters based on CAE simulation, process window, and machine learning | |
| CN110287520B (en) | Automatic recognition method for punching elements | |
| US20220297362A1 (en) | Artificial intelligence-based injection molding system and method for generating molding condition in injection molding system | |
| CN113988493A (en) | Method and device for constructing terminal evaluation model and storage medium | |
| CN116541714B (en) | Power grid regulation model training method, system, computer equipment and storage medium | |
| CN115496131B (en) | Equipment health state classification method based on multiple pre-training neural networks | |
| CN112529038B (en) | Method and device for identifying main board material and storage medium | |
| CN116611850B (en) | System for detecting and tracing engine assembly quality curve |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |