CN115239645A - 3D printing abnormity detection method, device, equipment and storage medium - Google Patents

3D printing abnormity detection method, device, equipment and storage medium Download PDF

Info

Publication number
CN115239645A
CN115239645A CN202210784999.8A CN202210784999A CN115239645A CN 115239645 A CN115239645 A CN 115239645A CN 202210784999 A CN202210784999 A CN 202210784999A CN 115239645 A CN115239645 A CN 115239645A
Authority
CN
China
Prior art keywords
printing
data
printer
image data
machine learning
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
Application number
CN202210784999.8A
Other languages
Chinese (zh)
Inventor
廖继盛
郑顺昌
卢松柏
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shenzhen Shengma Youchuang Technology Co ltd
Original Assignee
Shenzhen Shengma Youchuang Technology Co ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Shenzhen Shengma Youchuang Technology Co ltd filed Critical Shenzhen Shengma Youchuang Technology Co ltd
Priority to CN202210784999.8A priority Critical patent/CN115239645A/en
Publication of CN115239645A publication Critical patent/CN115239645A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00Image analysis
    • G06T7/0002Inspection of images, e.g. flaw detection
    • G06T7/0004Industrial image inspection
    • BPERFORMING OPERATIONS; TRANSPORTING
    • B29WORKING OF PLASTICS; WORKING OF SUBSTANCES IN A PLASTIC STATE IN GENERAL
    • B29CSHAPING OR JOINING OF PLASTICS; SHAPING OF MATERIAL IN A PLASTIC STATE, NOT OTHERWISE PROVIDED FOR; AFTER-TREATMENT OF THE SHAPED PRODUCTS, e.g. REPAIRING
    • B29C64/00Additive manufacturing, i.e. manufacturing of three-dimensional [3D] objects by additive deposition, additive agglomeration or additive layering, e.g. by 3D printing, stereolithography or selective laser sintering
    • B29C64/30Auxiliary operations or equipment
    • B29C64/386Data acquisition or data processing for additive manufacturing
    • BPERFORMING OPERATIONS; TRANSPORTING
    • B33ADDITIVE MANUFACTURING TECHNOLOGY
    • B33YADDITIVE MANUFACTURING, i.e. MANUFACTURING OF THREE-DIMENSIONAL [3-D] OBJECTS BY ADDITIVE DEPOSITION, ADDITIVE AGGLOMERATION OR ADDITIVE LAYERING, e.g. BY 3-D PRINTING, STEREOLITHOGRAPHY OR SELECTIVE LASER SINTERING
    • B33Y50/00Data acquisition or data processing for additive manufacturing
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06NCOMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00Machine learning
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T3/00Geometric image transformations in the plane of the image
    • G06T3/40Scaling of whole images or parts thereof, e.g. expanding or contracting
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00Indexing scheme for image analysis or image enhancement
    • G06T2207/20Special algorithmic details
    • G06T2207/20081Training; Learning
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00Indexing scheme for image analysis or image enhancement
    • G06T2207/30Subject of image; Context of image processing
    • G06T2207/30108Industrial image inspection
    • G06T2207/30144Printing quality

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Chemical & Material Sciences (AREA)
  • Materials Engineering (AREA)
  • General Physics & Mathematics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Manufacturing & Machinery (AREA)
  • Software Systems (AREA)
  • Artificial Intelligence (AREA)
  • Data Mining & Analysis (AREA)
  • Evolutionary Computation (AREA)
  • Medical Informatics (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Mathematical Physics (AREA)
  • Mechanical Engineering (AREA)
  • Optics & Photonics (AREA)
  • Quality & Reliability (AREA)

Abstract

The invention discloses a 3D printing abnormity detection method, a device, equipment and a storage medium, wherein the method comprises the following steps: the method comprises the steps of obtaining 3D printing image data in a 3D printing process through an image sensor arranged at a preset position of a 3D printer, obtaining model information of the 3D printer, searching a target machine learning model according to the model information of the 3D printer, inputting the 3D printing image data into the target machine learning model for analysis, obtaining an analysis result, and performing abnormal detection based on the analysis result to obtain a detection result. According to the 3D printing abnormity detection method, the abnormity detection is carried out on the 3D printing image data in the 3D printing process through the target machine learning model corresponding to the model information of the 3D printer, and compared with a mode that the abnormity in the 3D printing process is detected through manual observation in the prior art, the 3D printing abnormity detection method effectively reduces the labor cost.

Description

3D printing abnormity detection method, device, equipment and storage medium
Technical Field
The invention relates to the technical field of printing, in particular to a 3D printing abnormity detection method, a device, equipment and a storage medium.
Background
In an FDM (Fused Deposition Modeling) 3D printer, a material is melted at a high temperature and flows into a fine nozzle, and then a desired model is printed layer by layer from bottom to top according to an imported three-dimensional drawing. The photocuring 3D printer is used for curing the resin material layer by layer through a specific light source and transmitting a projection screen, and printing a model to be printed from top to bottom.
At present, when the FMD or the photocuring 3D printer is abnormal in the printing process, if the model is printed in a staggered mode, the model is printed in a position offset mode, the spray head does not discharge materials, the discharge materials are not adhered to the model, manual observation and judgment are needed, if the abnormal condition occurs, unattended operation is carried out, time and materials of a user can be wasted, and therefore the manual cost is greatly increased through the manual observation detection mode.
The above is only for the purpose of assisting understanding of the technical aspects of the present invention, and does not represent an admission that the above is prior art.
Disclosure of Invention
The invention mainly aims to provide a 3D printing abnormity detection method, a device, equipment and a storage medium, and aims to solve the technical problem that in the prior art, the labor cost is increased due to the fact that abnormity appearing in the 3D printing process is detected in a manual observation mode.
In order to achieve the above object, the present invention provides a 3D printing anomaly detection method, including the steps of:
acquiring 3D printing image data in a 3D printing process through an image sensor arranged at a preset position of a 3D printer;
acquiring model information of a 3D printer, and searching a target machine learning model according to the model information of the 3D printer;
inputting the 3D printing image data into the target machine learning model for analysis to obtain an analysis result;
and carrying out anomaly detection based on the analysis result to obtain a detection result.
Optionally, the target machine learning model includes a data cleaning module, a feature engineering module, a feature scaling module, and a state prediction module, and the step of inputting the 3D printing image data into the target machine learning model for analysis to obtain an analysis result includes:
inputting the 3D printing image data into the data cleaning module for data cleaning to obtain data to be predicted;
inputting the data to be predicted into the feature engineering module for vectorization to obtain a feature vector;
inputting the feature vector into the feature scaling module for preprocessing to obtain a target test vector;
and inputting the target test vector to the state prediction module for performing abnormity prediction, and taking a prediction result as an analysis result.
Optionally, the step of inputting the 3D printing image data into the data cleaning module to perform data cleaning to obtain data to be predicted includes:
and inputting the 3D printing image data into the data cleaning module so that the data cleaning module deletes data which does not accord with the standard of the standard data in the 3D printing image data based on the standard of the standard data to obtain data to be predicted.
Optionally, the preprocessing includes normalization, and the step of inputting the feature vector to the feature scaling module for preprocessing to obtain a target test vector includes:
and when the number of the feature vectors is lower than a preset number threshold, inputting the feature vectors into the feature scaling module for normalization processing to obtain target test vectors under preset dimensions.
Optionally, the preprocessing includes a normalization process, and the step of inputting the feature vector to the feature scaling module for preprocessing to obtain a target test vector includes:
and when the number of the feature vectors reaches a preset number threshold, inputting the feature vectors into the feature scaling module for standardization, and obtaining a target test vector in a preset limited interval.
Optionally, the step of performing anomaly detection based on the analysis result to obtain a detection result includes:
reading abnormal data proportion in the analysis result;
and when the abnormal data ratio reaches a preset ratio, judging that abnormality exists in the 3D printing process, and taking the judgment result as a detection result.
Optionally, before the step of acquiring 3D printing image data in a 3D printing process by an image sensor disposed at a preset position of the 3D printer, the method further includes:
acquiring abnormal image data and standard image data of 3D printers of different models;
constructing a model training sample according to the abnormal image data and the standard image data;
performing iterative training on an initial machine learning model according to the model training sample to obtain a preset machine learning model, wherein the initial machine learning model comprises a neural network algorithm model, a guided aggregation algorithm model or a random forest algorithm model;
and associating the preset machine learning model with the corresponding 3D printer model information.
Further, to achieve the above object, the present invention also proposes a 3D printing abnormality detection apparatus, the apparatus including:
the image acquisition unit is used for acquiring 3D printing image data in a 3D printing process through an image sensor arranged at a preset position of the 3D printer;
the machine learning model searching unit is used for acquiring the model information of the 3D printer and searching a target machine learning model according to the model information of the 3D printer;
the analysis unit is used for inputting the 3D printing image data into the target machine learning model for analysis to obtain an analysis result;
and the abnormality detection unit is used for carrying out abnormality detection on the basis of the analysis result to obtain a detection result.
Further, to achieve the above object, the present invention also proposes a 3D printing abnormality detection apparatus, the apparatus including: a memory, a processor and a 3D printing anomaly detection program stored on the memory and executable on the processor, the 3D printing anomaly detection program being configured to implement the steps of the 3D printing anomaly detection method as described above.
Furthermore, in order to achieve the above object, the present invention further provides a storage medium, wherein the storage medium stores a 3D printing abnormality detection program, and when the 3D printing abnormality detection program is executed by a processor, the 3D printing abnormality detection program implements the steps of the 3D printing abnormality detection method as described above.
According to the method, 3D printing image data in the 3D printing process are obtained through an image sensor arranged at a preset position of a 3D printer, then model information of the 3D printer is obtained, a target machine learning model is searched according to the model information of the 3D printer, the 3D printing image data are input into the target machine learning model to be analyzed, an analysis result is obtained, and finally anomaly detection is carried out based on the analysis result to obtain a detection result. According to the 3D printing abnormity detection method, the abnormity detection is carried out on the 3D printing image data in the 3D printing process through the target machine learning model corresponding to the model information of the 3D printer, and compared with a mode that the abnormity in the 3D printing process is detected through manual observation in the prior art, the 3D printing abnormity detection method effectively reduces the labor cost.
Drawings
Fig. 1 is a schematic structural diagram of a 3D printing anomaly detection device in a hardware operating environment according to an embodiment of the present invention;
FIG. 2 is a flowchart illustrating a 3D printing anomaly detection method according to a first embodiment of the present invention;
FIG. 3 is a schematic diagram of the main components of a 3D printer with a machine learning model running in an anomaly detection device;
FIG. 4 is a schematic diagram of the main components of a 3D printer with a machine learning model running in a cloud server;
FIG. 5 is a flowchart illustrating a 3D printing anomaly detection method according to a second embodiment of the present invention;
FIG. 6 is a flowchart illustrating a 3D printing anomaly detection method according to a third embodiment of the present invention;
fig. 7 is a block diagram illustrating a configuration of a 3D printing abnormality detection apparatus according to a first embodiment of the present invention.
The implementation, functional features and advantages of the objects of the present invention will be further explained with reference to the accompanying drawings.
Detailed Description
It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
Referring to fig. 1, fig. 1 is a schematic structural diagram of a 3D printing anomaly detection device in a hardware operating environment according to an embodiment of the present invention.
As shown in fig. 1, the 3D printing abnormality detection apparatus may include: a processor 1001, such as a Central Processing Unit (CPU), a communication bus 1002, a user interface 1003, a network interface 1004, and a memory 1005. Wherein a communication bus 1002 is used to enable connective communication between these components. The user interface 1003 may include a Display screen (Display), an input unit such as a Keyboard (Keyboard), and the optional user interface 1003 may also include a standard wired interface, a wireless interface. The network interface 1004 may optionally include a standard wired interface, a Wireless interface (e.g., a Wireless-Fidelity (Wi-Fi) interface). The Memory 1005 may be a high-speed Random Access Memory (RAM) or a Non-Volatile Memory (NVM), such as a disk Memory. The memory 1005 may alternatively be a storage device separate from the processor 1001.
Those skilled in the art will appreciate that the configuration shown in fig. 1 does not constitute a limitation of the 3D printing anomaly detection apparatus, and may include more or fewer components than those shown, or some components in combination, or a different arrangement of components.
As shown in fig. 1, a memory 1005, which is a storage medium, may include therein an operating system, a network communication module, a user interface module, and a 3D printing abnormality detection program.
In the 3D printing abnormality detection apparatus shown in fig. 1, the network interface 1004 is mainly used for data communication with a network server; the user interface 1003 is mainly used for data interaction with a user; the processor 1001 and the memory 1005 of the 3D printing abnormality detection apparatus according to the present invention may be provided in the 3D printing abnormality detection apparatus, and the 3D printing abnormality detection apparatus calls the 3D printing abnormality detection program stored in the memory 1005 through the processor 1001 and executes the 3D printing abnormality detection method provided by the embodiment of the present invention.
An embodiment of the present invention provides a 3D printing anomaly detection method, and referring to fig. 2, fig. 2 is a flowchart illustrating a first embodiment of the 3D printing anomaly detection method according to the present invention.
In this embodiment, the 3D printing anomaly detection method includes the following steps:
step S10: and acquiring 3D printing image data in the 3D printing process through an image sensor arranged at a preset position of the 3D printer.
It should be noted that the execution subject of the method of this embodiment may be a computing service device with data processing, network communication, and program running functions, such as a mobile phone, a tablet computer, a personal computer, and the like, or may be other electronic devices capable of implementing the same or similar functions. The 3D printing abnormality detection method provided in the present embodiment and each of the following embodiments is specifically described here with the above-described 3D printing abnormality detection apparatus (simply referred to as detection apparatus).
It can be understood that the 3D printer related to this embodiment may be a 3D printer, and may include an FMD3D printer, an LCD (Liquid Crystal Display) photocuring 3D printer, an SLA (Stereo photocuring application) photocuring 3D printer, a DLP (Digital Light Processing, digital Light Processing projector) photocuring 3D printer, and the like.
It can be understood that, the image sensor can be used to obtain the image or graphic data of the model change in the 3D printing process, such as a camera or a video camera, and the number of the image sensors can be 1 or more, and the corresponding number can be set according to the actual requirement. The image sensor and the detection device may be connected by means of MIPI (Mobile Industry Processor Interface), SPI (Serial peripheral Interface), UART (Universal Asynchronous Receiver/Transmitter), and the like. The image sensor can also be assisted by a module with an image processing function so as to improve the quality of image data acquisition.
It should be noted that the preset position may be a corresponding position for ensuring that the image sensor can accurately monitor the 3D printing model in the 3D printer.
It is understood that the 3D printing image data may be image data of a 3D printing model during 3D printing, which may be detected by an image sensor.
In a specific implementation, the detection device can acquire images or graphic data of model changes in a 3D printing process detected by an image sensor at a preset position of the 3D printer at regular time or in real time.
Step S20: and acquiring the model information of the 3D printer, and searching a target machine learning model according to the model information of the 3D printer.
It should be noted that the 3D printer model information may be identification information for identifying the current 3D printer, such as a hardware identification number or a version number, and each 3D printer has model information corresponding thereto for a user or a machine to identify.
It will be appreciated that the target machine learning model may be a model for analyzing anomalies in the 3D printed image as described above. A large number of abnormal images and normal images stored in the history 3D printing process can be selected as samples and obtained through training. Because the model printed by each 3D printer model changes, the machine learning model corresponding to each 3D printer model is different. During training, the 3D printer model information is determined, then the machine learning models corresponding to the 3D printer model information are determined, all the machine learning models are stored in a main control of the detection equipment in a unified mode, and all the machine learning models can also be stored in a cloud server associated with the detection equipment in a unified mode.
In a specific implementation, the detection device may obtain the model information of the 3D printer, and search, according to the model information of the 3D printer, a current machine learning model corresponding to the model information of the 3D printer from the machine learning model.
Referring to fig. 3, fig. 3 is a schematic diagram of main components of a 3D printer in which a machine learning model operates in an abnormality detection device.
As shown in fig. 3, a machine learning model is run in the detection device. Correspondingly, the 3D printer may include the above-mentioned machine learning model, detection device, camera, and 3D printer basic component, wherein the machine learning model may be embedded in the detection device (specifically may be embedded in the main control in the detection device), and the detection device is connected with the camera and the 3D printer basic component, respectively.
It should be noted that the basic components of the 3D printer may be the basic components required by the 3D printer in the 3D printing process of the model.
It will be appreciated that in an FDM 3D printer, the basic components of the 3D printer may be a spray head, a thermal bed, a temperature sensor, a motor, an extruder, a memory.
In a specific implementation, the spray head can be used for melting materials into a fluid state, the thermal bed can be used for heating the 3D printing bottom plate, so that a 3D printing object can be fixed on the bottom, and the temperature sensor can be used for detecting the temperature of the spray head; motors may be used to control X, Y, Z axis movement of the 3D printer, extruders may be used to feed materials into the nozzle assemblies, and memories may be used to store software firmware, slice files, image and graphics files, etc., with memory types including memory cards, EMMC (Embedded multimedia Card), non-volatile memory, etc.
It can be appreciated that in a photocuring 3D printer, the basic components of the 3D printer can be a light source, a motor, an image projection device, a memory, a limit switch, a chute.
In a specific implementation, the light source can be formed by a light curing material, the light source comprises a UV light source, a laser light source and the like, the motor can be used for controlling Z-axis movement, the image projection device can be used for projecting images of a 3D printing model, the device comprises an LCD (liquid crystal display), a DLP (digital light processing) device and the like, the storage can be used for storing software firmware, slice files, images, graphic files and the like, the types of the storage comprise a storage card, an EMMC (embedded multi-media card), a nonvolatile storage and the like, the limit switch can be used for limiting the maximum limit and the zero point of the Z-axis movement, and the trough can be used for placing materials of the light curing 3D printer.
It should be noted that, based on the 3D printer in fig. 3, a schematic diagram of main components of the 3D printer in which the machine learning model shown in fig. 4 runs in the cloud server is proposed.
As shown in fig. 4, the machine learning model runs in a cloud server. Correspondingly, the 3D printer can comprise the machine learning model, the detection device, the camera, the 3D printer basic component and the network communication module, and the cloud server can comprise the machine learning model and the network communication module. The machine learning model storage and cloud server are connected with the camera and the 3D printer basic assembly respectively, the cloud server is connected with the detection device through the network communication module, and the cloud server is further connected with the client.
In specific implementation, the functions of the basic components of the 3D printer are consistent with those of the above components, and are not described, the network communication module can be in other modes such as WIFI, ethernet, 4G and 5G, and the client can be used for controlling the 3D printer and receiving the state of the 3D printer, and the client comprises a mobile terminal or a PC.
It should be understood that the principle of the FDM 3D printer and the photocuring 3D printer with the machine learning model running in the cloud server is similar to that described above.
Step S30: and inputting the 3D printing image data into the target machine learning model for analysis to obtain an analysis result.
In a specific implementation, the detection device acquires images or graphic data detected by the image sensor at regular time or in real time, inputs the images or graphic data into the target machine learning model, so that the target machine learning model analyzes the detected graphics or graphic data based on normal graphic data, and when the detected image data is inconsistent with the image data of the normal 3D printing model in the target machine learning model, the detection device can determine that the image data is abnormal, thereby acquiring the ratio of the abnormal data to the normal data, and taking the ratio as an analysis result.
Step S40: and carrying out anomaly detection based on the analysis result to obtain a detection result.
In a specific implementation, the detection device may perform a ratio of abnormal data to normal data in the analysis result, and when the ratio of the abnormal data is greater than that of the normal data, it may be determined that the 3D printing process is abnormal, and otherwise, it may be determined that the 3D printing process is normal.
Further, in order to improve the efficiency of detection, in this embodiment, step S40 may include:
step S401: and reading abnormal data proportion in the analysis result.
It should be noted that the abnormal data ratio may be a percentage of the abnormal data to the total data, or a ratio of the abnormal data to the normal data.
Step S402: and when the proportion of the abnormal data reaches a preset proportion value, judging that abnormality exists in the 3D printing process, and taking the judgment result as a detection result.
The abnormality may be a 3D printing model layer dislocation, a 3D printing model position shift, a non-discharge of the head, and a discharge but non-adhesion model.
It is understood that the preset percentage value may be a ratio of abnormal data to normal data, and if the ratio is a ratio of abnormal data to total data, the ratio is 50% and the abnormal data is considered to be equal to the normal data, and if the ratio is higher than 50%, the ratio of abnormal data is considered to be more.
In specific implementation, the preset ratio value can be set to be a numerical value higher than 50%, when the abnormal data ratio is higher than the preset ratio value, it is determined that the 3D printing process is abnormal, the detection device can perform preset operations such as pausing, stopping or 3D printing again on the 3D printer, and the determination result is used as a detection result and fed back to the client and the 3D printer.
It should be understood that, the detection device can quickly generate a detection result by reading the abnormal data proportion in the analysis result and performing abnormality judgment on the abnormal data proportion and the preset proportion value, thereby effectively improving the detection efficiency.
This embodiment acquires 3D through the image sensor who sets up in 3D printer preset position department and prints the 3D image data that prints in-process, then acquires 3D printer model information to look for target machine learning model according to 3D printer model information, print image data input to target machine learning model with 3D again and carry out the analysis, obtain the analysis result, carry out abnormal detection based on the analysis result at last, obtain the testing result. Because the anomaly detection is performed on the 3D printing image data in the 3D printing process through the target machine learning model corresponding to the model information of the 3D printer, compared with a mode of detecting the anomaly in the 3D printing process through manual observation in the prior art, the 3D printing anomaly detection method effectively reduces labor cost.
Referring to fig. 5, fig. 5 is a flowchart illustrating a 3D printing anomaly detection method according to a second embodiment of the present invention.
Based on the first embodiment described above, in order to improve the quality of the 3D printing anomaly detection analysis result, in the present embodiment, the step S30 includes:
step S301: and inputting the 3D printing image data into the data cleaning module for data cleaning to obtain data to be predicted.
It is noted that the target machine learning model includes a data cleaning module, a feature engineering module, a feature scaling module, and a state prediction module.
It will be appreciated that the data cleansing module may be used to re-audit and verify the received 3D printed image data to delete duplicate information, correct existing errors, and provide data consistency.
In specific implementation, the 3D printing image data is input to the data cleaning module, so that the data cleaning module can remove data with obvious errors such as data formats, repeated data, unreasonable data or contradictory contents, and the like, and the quality of the image is improved.
It should be understood that the data to be predicted may be high-quality 3D printed image data that has undergone the data cleansing process described above.
Further, in order to improve the effect of data cleansing, in this embodiment, the step S301 may include:
step S3011: and inputting the 3D printing image data into the data cleaning module so that the data cleaning module deletes data which does not accord with the standard of the standard data in the 3D printing image data based on the standard of the standard data to obtain data to be predicted.
The reference data standard may be a standard corresponding to 3D print image data without significant errors, and may be determined in advance from normal image data.
In a specific implementation, after the detection device inputs the 3D printing image data into the data cleaning module in the target machine learning model, the data cleaning module may process the 3D printing image data according to the standard of the reference data, and delete data that does not conform to the standard of the reference data to obtain image data with higher quality as the prediction data. As the detection equipment improves the quality of 3D printing image data through the standard of the reference data, the data cleaning effect can be effectively improved.
Step S302: and inputting the data to be predicted to the feature engineering module for vectorization to obtain a feature vector.
It should be noted that the feature engineering module may be configured to exchange data to be predicted, so that the data to be predicted can be identified by the feature engineering module.
In a specific implementation, the detection device inputs the data to be predicted to the feature engineering module, so that the feature engineering module performs data exchange on the data to be predicted to generate vectorized data, and feeds the vectorized data serving as a feature vector back to the detection device.
Step S303: and inputting the feature vector into the feature scaling module for preprocessing to obtain a target test vector.
It should be noted that the feature scaling module may be configured to scale the feature vector to avoid a problem that the feature vector cannot be directly used by the target machine learning model due to incompleteness or inconsistency of the feature vector, and accordingly, the pre-processing may be the above scaling processing.
In a specific implementation, the detection device inputs the feature vector to a feature scaling module in the target machine learning model, so that the feature scaling module scales the feature vector to obtain a target test vector which can be directly used by the target machine learning model.
Further, in order to improve the zooming effect, in this embodiment, the step S303 may include:
step S3031: and when the number of the feature vectors is lower than a preset number threshold, inputting the feature vectors into the feature scaling module for normalization processing to obtain target test vectors under preset dimensions.
It should be noted that the preset number threshold may be used as a criterion for determining the number of feature vectors, that is, the number of feature vectors is greater above the threshold, and the number of feature vectors is lower below the threshold. Correspondingly, when the number of the feature vectors is low, the feature scaling can be performed in a normalization manner, so as to improve the scaling effect.
It is understood that the normalization can be a process of scaling the feature vector to [0,1] to convert the data with units into data without units, and the normalization includes Min Max Scaler, function conversion, inverse cotangent conversion, and the like.
It should be noted that the preset dimension may be a dimension corresponding to the unified data metric, and may be preset in the feature scaling module.
It can be understood that when the feature vectors are few, the influence of the units is eliminated by unifying the measuring standards of the data, so that the feature scaling module can process the feature vectors more quickly and quickly.
It should be understood that the target test vector may be the normalized feature vector described above for further processing.
Step S3032: and when the number of the feature vectors reaches a preset number threshold, inputting the feature vectors into the feature scaling module for standardization, and obtaining target test vectors in a preset limited interval.
When the number of feature vectors is high, the feature scaling may be performed in a standardized manner to improve the scaling effect.
It is understood that normalization can be a process of scaling the feature vectors to make the feature vectors fall into a defined interval without changing the distribution of the original feature vectors, and the normalization includes z-score normalization or StandardScaler normalization.
It should be noted that the preset limited interval may be the above-mentioned limited interval, and may be preset in the feature scaling module based on the standard image data.
It can be understood that when the number of feature vectors is large, the normalization method is complicated, the normalization can play a role in unifying dimensions, and meanwhile, the normalized feature vectors are comparable and convenient for the next data processing.
It should be understood that the target test vector may be the normalized feature vector described above for further processing.
It should be noted that the feature scaling module reasonably selects a normalization and standardization processing mode according to the number of the feature vectors, thereby effectively improving the scaling effect.
Step S304: and inputting the target test vector to the state prediction module for performing abnormity prediction, and taking a prediction result as an analysis result.
It should be noted that the state prediction module may be a module for performing exception prediction on the target test vector.
In a specific implementation, the detection device inputs the target test vector to a state prediction module of a target machine learning model, so that the state prediction module judges the target test vector based on a standard vector corresponding to normal image data, when data in the target test vector is inconsistent with the standard vector, the data can be judged to be abnormal data and marked, otherwise, the normal data is not marked, the ratio of the marked abnormal data to total data is generated, and the ratio is used as an analysis result.
In this embodiment, the 3D print image data is input to the data cleaning module to perform data cleaning, to obtain data to be predicted, then the data to be predicted is input to the feature engineering module to perform vectorization, to obtain a feature vector, then the feature vector is input to the feature scaling module to perform preprocessing, to obtain a target test vector, and finally the target test vector is input to the state prediction module to perform anomaly prediction, and a prediction result is used as an analysis result. According to the embodiment, the analysis result is obtained through the data cleaning module, the feature engineering module, the feature scaling module and the state prediction module in the target machine learning model, so that the quality of the 3D printing abnormity detection analysis result is effectively improved.
Referring to fig. 6, fig. 6 is a flowchart illustrating a 3D printing anomaly detection method according to a third embodiment of the present invention.
Based on the foregoing embodiments, in this embodiment, in order to enable the 3D printers of different models to achieve the anomaly detection and improve the comprehensiveness of the anomaly detection, before step S10, the method further includes:
step S01: and acquiring abnormal image data and standard image data of 3D printers of different models.
It should be noted that, the models printed by 3D printers of different models have differences due to different parameters of the 3D printers, and therefore, abnormal image data and standard image data of the 3D printers of different models during the 3D printing process need to be obtained.
It can be understood that the abnormal image data and the standard image data can be acquired by the 3D printer through the image sensor after the 3D printing is finished and stored in the detection device, and can be called at any time when the abnormal image data and the standard image data are required to be used.
Step S02: and constructing a model training sample according to the abnormal image data and the standard image data.
It should be noted that the model training sample can be constructed by a large amount of the above abnormal image data and standard image data, so as to improve the comprehensive type of the sample data.
It can be understood that the model training samples can be processed according to the methods corresponding to the data cleaning module, the feature engineering module, the feature scaling module and the state prediction module mentioned in the above embodiments, so as to improve the quality of the sample data.
Step S03: and performing iterative training on an initial machine learning model according to the model training sample to obtain a preset machine learning model, wherein the initial machine learning model comprises a neural network algorithm model, a guided aggregation algorithm model or a random forest algorithm model.
It should be noted that the initial machine learning model may be a model with learning algorithm, which may be used for abnormal states, such as neural network algorithm, guided aggregation algorithm, or random forest algorithm.
It can be understood that the preset machine learning model may be a model generated by processing data in a model training sample through the algorithm of the initial machine learning model.
In a specific implementation, the detection device may divide data in the model training sample into training set data, test set data, and validation set data according to a certain proportion (e.g. 2.
Step S04: and associating the preset machine learning model with the corresponding 3D printer model information.
It should be noted that the 3D printer model information may be information corresponding to a 3D printer model, and the 3D printing model information may include a 3D printing technology principle (such as FDM), a printing size, an image sensor position, a 3D printing direction, a 3D printing platform, and the like.
In specific implementation, machine information models under various 3D printer models can be obtained by correlating preset machine information models with corresponding 3D printer model information, and the machine information models are stored in detection equipment or a cloud server.
According to the embodiment, abnormal image data and standard image data of 3D printers of different models are obtained and used as model training samples, iterative training is carried out on an initial machine information model according to the model training samples to obtain a preset machine information model, and then correlation is carried out on the preset machine information model and the model information of the 3D printers, so that abnormal detection can be achieved for the 3D printers of different models, and the comprehensiveness of the abnormal detection is improved.
Furthermore, an embodiment of the present invention further provides a storage medium, where a 3D printing anomaly detection program is stored, and when being executed by a processor, the 3D printing anomaly detection program implements the steps of the 3D printing anomaly detection method as described above.
Referring to fig. 7, fig. 7 is a block diagram illustrating a structure of a 3D printing abnormality detection apparatus according to a first embodiment of the present invention.
As shown in fig. 7, the 3D printing abnormality detection apparatus according to the embodiment of the present invention includes:
an image acquisition unit 501 for acquiring 3D printing image data in a 3D printing process through an image sensor provided at a preset position of a 3D printer;
a machine learning model searching unit 502, configured to obtain model information of a 3D printer, and search a target machine learning model according to the model information of the 3D printer;
an analysis unit 503, configured to input the 3D printing image data to the target machine learning model for analysis, so as to obtain an analysis result;
an anomaly detection unit 504, configured to perform anomaly detection based on the analysis result, and obtain a detection result.
This embodiment acquires 3D through the image sensor who sets up in 3D printer preset position department and prints the 3D image data that prints in-process, then acquires 3D printer model information to look for target machine learning model according to 3D printer model information, print image data input to target machine learning model with 3D again and carry out the analysis, obtain the analysis result, carry out abnormal detection based on the analysis result at last, obtain the testing result. Because the 3D printing image data in the 3D printing process is subjected to abnormity detection through the target machine learning model corresponding to the model information of the 3D printer, compared with the mode that the abnormity in the 3D printing process is detected through manual observation in the prior art, the 3D printing abnormity detection device effectively reduces the labor cost.
Other embodiments or specific implementation manners of the 3D printing abnormality detection apparatus according to the present invention may refer to the above method embodiments, and are not described herein again.
It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or system. Without further limitation, an element defined by the phrases "comprising a," "8230," "8230," or "comprising" does not exclude the presence of other like elements in a process, method, article, or system comprising the element.
The above-mentioned serial numbers of the embodiments of the present invention are merely for description and do not represent the merits of the embodiments.
Through the above description of the embodiments, those skilled in the art will clearly understand that the method of the above embodiments can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware, but in many cases, the former is a better implementation manner. Based on such understanding, the technical solutions of the present invention or portions thereof contributing to the prior art may be embodied in the form of a software product, where the computer software product is stored in a storage medium (such as a rom/ram, a magnetic disk, and an optical disk), and includes several instructions for enabling a terminal device (such as a mobile phone, a computer, a server, an air conditioner, or a network device) to execute the methods according to the embodiments of the present invention.
The above description is only a preferred embodiment of the present invention, and not intended to limit the scope of the present invention, and all modifications of equivalent structures and equivalent processes, which are made by using the contents of the present specification and the accompanying drawings, or directly or indirectly applied to other related technical fields, are included in the scope of the present invention.

Claims (10)

1. A3D printing anomaly detection method is characterized by comprising the following steps:
acquiring 3D printing image data in a 3D printing process through an image sensor arranged at a preset position of a 3D printer;
acquiring model information of a 3D printer, and searching a target machine learning model according to the model information of the 3D printer;
inputting the 3D printing image data into the target machine learning model for analysis to obtain an analysis result;
and carrying out anomaly detection based on the analysis result to obtain a detection result.
2. The 3D printing anomaly detection method according to claim 1, wherein the target machine learning model includes a data cleaning module, a feature engineering module, a feature scaling module and a state prediction module, and the step of inputting the 3D printing image data into the target machine learning model for analysis to obtain an analysis result includes:
inputting the 3D printing image data into the data cleaning module for data cleaning to obtain data to be predicted;
inputting the data to be predicted into the feature engineering module for vectorization to obtain a feature vector;
inputting the feature vector to the feature scaling module for preprocessing to obtain a target test vector;
and inputting the target test vector to the state prediction module for abnormal prediction, and taking a prediction result as an analysis result.
3. The 3D printing anomaly detection method according to claim 2, wherein the step of inputting the 3D printing image data to the data cleaning module for data cleaning to obtain data to be predicted comprises:
and inputting the 3D printing image data into the data cleaning module so that the data cleaning module deletes data which does not accord with the standard of the standard data in the 3D printing image data based on the standard of the standard data to obtain data to be predicted.
4. The 3D printing anomaly detection method according to claim 2, wherein the preprocessing includes normalization processing, and the step of inputting the feature vector to the feature scaling module for preprocessing to obtain a target test vector includes:
and when the number of the feature vectors is lower than a preset number threshold, inputting the feature vectors into the feature scaling module for normalization processing to obtain a target test vector under a preset dimension.
5. The 3D printing anomaly detection method according to claim 2, wherein the preprocessing includes a normalization processing, and the step of inputting the feature vector to the feature scaling module for preprocessing to obtain a target test vector includes:
and when the number of the feature vectors reaches a preset number threshold, inputting the feature vectors into the feature scaling module for standardization, and obtaining target test vectors in a preset limited interval.
6. The 3D printing abnormality detection method according to any one of claims 1 to 5, wherein the step of performing abnormality detection based on the analysis result to obtain a detection result includes:
reading abnormal data proportion in the analysis result;
and when the abnormal data ratio reaches a preset ratio, judging that abnormality exists in the 3D printing process, and taking the judgment result as a detection result.
7. The 3D printing abnormality detection method according to claim 1, characterized in that before the step of acquiring 3D printing image data during 3D printing by an image sensor provided at a preset position of a 3D printer, further comprising:
acquiring abnormal image data and standard image data of 3D printers of different models;
constructing a model training sample according to the abnormal image data and the standard image data;
performing iterative training on an initial machine learning model according to the model training sample to obtain a preset machine learning model, wherein the initial machine learning model comprises a neural network algorithm model, a guided aggregation algorithm model or a random forest algorithm model;
and associating the preset machine learning model with the corresponding 3D printer model information.
8. A 3D printing anomaly detection device, characterized in that said device comprises:
the image acquisition unit is used for acquiring 3D printing image data in a 3D printing process through an image sensor arranged at a preset position of the 3D printer;
the machine learning model searching unit is used for acquiring the model information of the 3D printer and searching a target machine learning model according to the model information of the 3D printer;
the analysis unit is used for inputting the 3D printing image data into the target machine learning model for analysis to obtain an analysis result;
and the abnormality detection unit is used for carrying out abnormality detection based on the analysis result to obtain a detection result.
9. A 3D printing abnormality detection apparatus, characterized in that the apparatus comprises: a memory, a processor and a 3D printing anomaly detection program stored on the memory and executable on the processor, the 3D printing anomaly detection program being configured to implement the steps of the 3D printing anomaly detection method according to any one of claims 1 to 7.
10. A storage medium having stored thereon a 3D printing anomaly detection program, the 3D printing anomaly detection program when executed by a processor implementing the steps of the 3D printing anomaly detection method according to any one of claims 1 to 7.
CN202210784999.8A 2022-07-05 2022-07-05 3D printing abnormity detection method, device, equipment and storage medium Pending CN115239645A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210784999.8A CN115239645A (en) 2022-07-05 2022-07-05 3D printing abnormity detection method, device, equipment and storage medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210784999.8A CN115239645A (en) 2022-07-05 2022-07-05 3D printing abnormity detection method, device, equipment and storage medium

Publications (1)

Publication Number Publication Date
CN115239645A true CN115239645A (en) 2022-10-25

Family

ID=83671900

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210784999.8A Pending CN115239645A (en) 2022-07-05 2022-07-05 3D printing abnormity detection method, device, equipment and storage medium

Country Status (1)

Country Link
CN (1) CN115239645A (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115674688A (en) * 2022-10-27 2023-02-03 重庆电子工程职业学院 High-precision bionic bone 3D printing system and printing method thereof
CN117392471A (en) * 2023-12-12 2024-01-12 深圳市智能派科技有限公司 3D printing monitoring method and system based on multi-parameter cooperation

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115674688A (en) * 2022-10-27 2023-02-03 重庆电子工程职业学院 High-precision bionic bone 3D printing system and printing method thereof
CN117392471A (en) * 2023-12-12 2024-01-12 深圳市智能派科技有限公司 3D printing monitoring method and system based on multi-parameter cooperation
CN117392471B (en) * 2023-12-12 2024-03-26 深圳市智能派科技有限公司 3D printing monitoring method and system based on multi-parameter cooperation

Similar Documents

Publication Publication Date Title
CN115239645A (en) 3D printing abnormity detection method, device, equipment and storage medium
Schlegel et al. Towards a rigorous evaluation of XAI methods on time series
CN109886928B (en) Target cell marking method, device, storage medium and terminal equipment
US8249830B2 (en) Method and system for automatically diagnosing faults in rendering devices
KR102108956B1 (en) Apparatus for Performing Inspection of Machine Vision and Driving Method Thereof, and Computer Readable Recording Medium
JP7106391B2 (en) Image determination method, image determination device, and image determination program
US20150281470A1 (en) Method for automatically selecting test parameters of an image inspection system and image inspection system for implementing the method
CN112102263A (en) Defect detection model generation system, method and device and computer equipment
JP2019191117A (en) Image processing device, image processing method, and program
CN113220537A (en) Software monitoring method, device, equipment and readable storage medium
CN112712513A (en) Product defect detection method, device, equipment and computer readable storage medium
US11562479B2 (en) Inspection apparatus, inspection method, and non-volatile storage medium
CN112085078A (en) Image classification model generation system, method and device and computer equipment
EP3712846A1 (en) Recoater automated monitoring systems and methods for additive manufacturing machines
Sampedro et al. Fused deposition modeling 3d printing fault diagnosis using temporal convolutional network
CN113263728A (en) Alarm method and device of 3D printer, 3D printer and storage medium
US10657626B2 (en) Image processing apparatus, image processing system, and recording medium
US20230230337A1 (en) A Method for Testing an Embedded System of a Device, a Method for Identifying a State of the Device and a System for These Methods
CN111169164B (en) Apparatus and method for inspecting printing of articles
JP6977327B2 (en) Image formation system, management equipment and programs
CN112052150A (en) Page loading time detection method, equipment, storage medium and device
CN112633194A (en) Method and device for detecting fingerprint in screen
US11070716B2 (en) Image sensor system, image sensor, data generation method of image sensor in image sensor system, and non-transitory computer-readable recording medium
JP7024906B2 (en) Image formation system, management equipment and programs
CN115035101B (en) Defect verification method and detection method of semiconductor chip carrier plate and AOI equipment

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