CN109840900B - Fault online detection system and detection method applied to intelligent manufacturing workshop - Google Patents
Fault online detection system and detection method applied to intelligent manufacturing workshop Download PDFInfo
- Publication number
- CN109840900B CN109840900B CN201811651018.2A CN201811651018A CN109840900B CN 109840900 B CN109840900 B CN 109840900B CN 201811651018 A CN201811651018 A CN 201811651018A CN 109840900 B CN109840900 B CN 109840900B
- Authority
- CN
- China
- Prior art keywords
- image
- manufacturing
- area
- neural network
- workpiece
- 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.)
- Active
Links
- 238000004519 manufacturing process Methods 0.000 title claims abstract description 124
- 238000001514 detection method Methods 0.000 title claims abstract description 103
- 230000007547 defect Effects 0.000 claims abstract description 71
- 238000012545 processing Methods 0.000 claims abstract description 55
- 238000013528 artificial neural network Methods 0.000 claims abstract description 47
- 238000012549 training Methods 0.000 claims abstract description 19
- 238000012423 maintenance Methods 0.000 claims abstract description 17
- 238000005516 engineering process Methods 0.000 claims abstract description 9
- 239000013598 vector Substances 0.000 claims description 42
- 238000000605 extraction Methods 0.000 claims description 20
- 238000006243 chemical reaction Methods 0.000 claims description 19
- 239000011159 matrix material Substances 0.000 claims description 15
- 238000000034 method Methods 0.000 claims description 13
- 238000003708 edge detection Methods 0.000 claims description 11
- 238000004891 communication Methods 0.000 claims description 8
- 239000013307 optical fiber Substances 0.000 claims description 8
- 238000013135 deep learning Methods 0.000 claims description 5
- 238000005286 illumination Methods 0.000 claims description 5
- 238000004364 calculation method Methods 0.000 claims description 3
- 238000005192 partition Methods 0.000 claims description 3
- 238000013459 approach Methods 0.000 claims description 2
- 230000000007 visual effect Effects 0.000 claims description 2
- 238000004458 analytical method Methods 0.000 abstract description 4
- 238000003062 neural network model Methods 0.000 abstract description 2
- 238000003745 diagnosis Methods 0.000 description 3
- 238000003709 image segmentation Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 2
- 238000007689 inspection Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008033 biological extinction Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 230000002349 favourable effect Effects 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 238000009499 grossing Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000012797 qualification Methods 0.000 description 1
- 238000010223 real-time analysis Methods 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 230000001960 triggered effect Effects 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
Landscapes
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The invention relates to a fault online detection system and a fault online detection method applied to an intelligent manufacturing workshop, which construct a manufacturing defect prediction model based on a deep neural network, effectively utilize a large number of typical manufacturing defects obtained in the manufacturing process of the actual intelligent manufacturing workshop, and perform training and learning on a standard sample image library formed by typical manufacturing defect images by combining image acquisition and image processing technology, so that the deep neural network model can be used for identifying and classifying the manufacturing defects in real time, dynamically obtain manufacturing information of the product through real-time image acquisition, processing and analysis in the manufacturing process of the product, and provide and automatically execute maintenance strategies of the manufacturing defects by combining a PLC (programmable logic controller) and a historical maintenance database, and can provide more accurate reference information for product manufacturing precision and manufacturing information collection and analysis of the intelligent manufacturing workshop.
Description
Technical Field
The invention belongs to the technical field of intelligent manufacturing, and particularly relates to a fault online detection system and a fault online detection method applied to an intelligent manufacturing workshop.
Background
The existing intelligent manufacturing workshops have high-speed and high-precision numerical control machines, so that the allowance errors of manufactured products are small, and the intelligent manufacturing workshops have high requirements on comprehensive quality and environment of personnel, so that manufacturing defects in the manufacturing process of the manufacturing workshops are difficult to discover in time manually, and are generally discovered only through manual inspection of subsequent procedures, so that problems of increased waste products, reduced productivity and the like of intelligent manufacturing are caused, and even scrapping of subsequent manufacturing machines is caused in serious cases, and the problems occur in the intelligent manufacturing assembly interval, therefore, dynamic manufacturing defect inspection of the products in the manufacturing workshops is diagnosed, and the intelligent manufacturing workshops are one of the impeding factors influencing the further development of intelligent manufacturing, and are widely studied by industry students nowadays.
Disclosure of Invention
The invention aims to provide a fault online detection system and a fault online detection method applied to an intelligent manufacturing workshop.
According to the above object of the present invention, there is provided a fault on-line detection system comprising a workpiece detection platform, an image acquisition unit, an image processing unit, a feature vector extraction unit, a deep neural network unit, and a computer control unit, wherein,
the workpiece detection platform comprises a detection station, and an area array camera in the image acquisition unit acquires images of the workpiece detection platform on the detection station and sends the images to the image processing unit;
the image processing unit performs resolution scanning on the received image to obtain a sensitive area image of the current detection station, denoising the sensitive area image, and then sending the denoised sensitive area image to the feature vector extraction unit;
the feature vector extraction unit performs edge detection on the sensitive area image to form a target area, calculates and obtains the edge area, the edge shape factor and the average radius of the target area through formulas (1) to (3), and adds the Hu invariant moment of the previous 3 dimensions to form feature vectors of the sensitive area with four feature variables so as to reflect the workpiece quality information of the current workpiece detection platform, and the feature vectors are used as input layers to be sent to the deep neural network unit;
in the above formula, the parameters M and N are the number of edge points of the target area,wherein t (x, y) is the gray value of each edge point; the parameter L is the perimeter of the target area, which can be obtained by calculating by adopting a chain code method in an image processing technology, and the reference K is the number of edge points on the boundary of the target area, (x) k ,y k ) Representing pixel coordinates located on the boundary of the target area, is->The centroid coordinates representing the target region can be calculated by the following formula:
wherein the parameter a represents the area of the sensitive area and is adapted to obtain its size when the sensitive area is identified in the image processing;
the deep neural network unit builds a manufacturing defect prediction model based on a neural network algorithm, trains, learns and classifies the image feature vectors of the workpiece detection platform, identifies the type of the manufacturing defect of the workpiece to be detected on the current workpiece detection platform, and feeds the classification result back to the computer control unit.
In still another aspect, the present invention further provides a fault online detection method applied to an intelligent manufacturing shop, including:
step 1: constructing a manufacturing defect prediction model based on a deep neural network, and training and learning the deep neural network through sample images;
step 2: the workpiece detection platform drives a workpiece to be detected to move on a detection station under the control instruction of the computer control module to reach a preset detection position, triggers the proximity switch and sends a trigger signal to the computer control unit;
step 3: the computer control unit respectively sends instructions to the LED area light source matrix and the image acquisition unit, the LED area light source matrix turns on illumination, the image acquisition unit shoots a workpiece to be detected, and the generated image is sent to the image processing unit;
step 4: the image processing unit identifies and partitions a sensitive area of the current detection station, performs image denoising processing on the local area, and sends the sensitive area image to the feature vector extraction unit after denoising;
step 5: the feature vector extraction unit performs edge detection on the sensitive area to form a target area, calculates and obtains the edge area, the edge shape factor and the average radius of the target area, and combines the Hu invariant moment of the previous 3 dimensions to form the feature vector of the sensitive area with four feature variables;
step 6: and diagnosing the manufacturing information of the feature vector based on the trained deep neural network, predicting and classifying the manufacturing defects of the workpiece to be detected, and feeding back the classification result to the computer control unit.
The beneficial effects of the invention are as follows:
(1) The invention applies the image acquisition and image processing technology to the identification and diagnosis of the product manufacturing defect information of the intelligent manufacturing workshop for the first time, and carries out real-time image acquisition, processing and feature extraction on the product of the intelligent manufacturing workshop through the manufacturing defect prediction model constructed based on the deep neural network to obtain the feature vector representation of the product manufacturing information, thereby realizing the display of the product manufacturing information, the diagnosis of the defect and the automatic generation of the maintenance strategy in the intelligent manufacturing process, forming an organic integrated system for the real-time monitoring of the product manufacturing, the real-time analysis of the defect conclusion and the automatic implementation of the maintenance strategy, and effectively solving one of the factors restricting the wide application of the existing intelligent manufacturing technology;
(2) In order to obtain a precise image processing result, the invention effectively solves the problems of image acquisition precision and lens deflection of an area array camera caused by the possible micro vibration of a machine tool and a detection station in an intelligent manufacturing workshop by applying a plurality of optical fiber slip rings and applying a resolution scanning technology in the image processing process, and improves the precision of on-line fault defect diagnosis and identification of the product manufacturing of the invention;
(3) In order to reduce the time consumption of image processing, the method and the system for predicting the online faults of the product are capable of setting the marking point positions of the background targets in a targeted manner based on typical manufacturing defects possibly occurring on the current detection station of the workpiece to be detected, so that the sensitive area image of the current detection station can be directly positioned in the subsequent image processing, the image processing and the extraction and calculation of the characteristic variables are carried out on the local area, the image processing efficiency is greatly improved, and compared with the traditional method for processing the whole image and extracting the characteristic variables, the online fault predicting system and the online fault predicting method of the product are more real-time and efficient, and the application value of the product in an actual intelligent manufacturing workshop is greatly improved.
Drawings
The invention will be further described with reference to the drawings and examples.
FIG. 1 is a schematic diagram of the fault on-line detection system of the present invention.
Detailed Description
The invention will now be described in further detail with reference to the accompanying drawings. The drawings are simplified schematic representations which merely illustrate the basic structure of the invention and therefore show only the structures which are relevant to the invention.
The invention provides a fault online detection system and a fault online detection method applied to an intelligent manufacturing workshop, which are used for constructing a manufacturing defect prediction model based on a deep neural network, effectively utilizing a large number of typical manufacturing defects obtained in the manufacturing process of the actual intelligent manufacturing workshop, combining image acquisition and image processing technology, training and learning a standard sample image library formed by typical manufacturing defect images, so that the deep neural network model can be used for identifying and classifying the manufacturing defects in real time, dynamically obtaining manufacturing information of the product through real-time image acquisition, processing and analysis in the manufacturing process of the product, and giving and automatically executing a maintenance strategy of the manufacturing defects by combining a PLC (programmable logic controller) and a historical maintenance database, and providing more accurate reference information for the product manufacturing precision and the manufacturing information collection and analysis of the intelligent manufacturing workshop.
The implementation of the present invention is discussed in detail by the following examples.
Referring to fig. 1, the invention provides a fault online detection system applied to an intelligent manufacturing workshop, which comprises a product detection platform 1, an image acquisition unit 2, an image processing unit 3, a feature vector extraction unit 4, a deep neural network unit 5 and a computer control unit 6.
In a preferred embodiment, the number of product detection platforms 1 may be several, and each product detection platform 1 corresponds to one image acquisition unit 2, and the image processing unit 3, the feature vector extraction unit 4, and the deep neural network unit 5 may be integrated as a software system in the computer control unit 6.
The product detection platform 1 includes a detection station 11, a proximity switch 12, and an intelligent positioning and releasing mechanism 13, in a preferred embodiment, the proximity switch 12, the intelligent positioning and releasing mechanism 13 are all in communication connection with the computer control unit 6, wherein the intelligent positioning and releasing mechanism 13 receives a control instruction of the computer control unit 6, drives a workpiece to be detected to move on the detection station 11 to reach a preset detection position, triggers the proximity switch 12 and sends a trigger signal to the computer control unit 6, and the computer control unit 6 starts the image acquisition unit 2 to acquire images.
On the other hand, a background target 14 is arranged on one side of the detection station 11 continuously along the flexible production line, a plurality of marking points corresponding to the current detection station 11 are arranged on the background target 14, and in a preferred embodiment, the positions of the marking points on the background target 14 are set based on typical manufacturing defect information possibly occurring on a workpiece to be detected of the current detection station so as to facilitate the identification and segmentation of a sensitive area by subsequent image processing. In addition, the opposite side of detection station 11 is provided with LED area light source matrix 15, this LED area light source matrix 15 and computer control unit 6 communication connection, computer control unit 6 can realize the illumination and the extinction of the LED area light source matrix 15 on the whole production line of intelligent control, when the work piece that awaits measuring moves to the preset position of current detection station 11, computer control unit 6 sends the instruction and gives the LED area light source matrix 15 that corresponds this detection station 11, make it light, when the work piece that awaits measuring is carried to next detection station, the LED area light source matrix 15 of current detection station 11 is extinguished in control, background target 14 and LED area light source matrix 15 constitute backlight environment jointly, be favorable to the work piece that awaits measuring on the image with the grey scale contrast of background target 14, edge detection time consuming time when reducing image processing.
The image acquisition unit 2 comprises an area-array camera 21, a first photoelectric conversion element 22, an optical fiber slip ring 23 and a second photoelectric conversion element 24 which are sequentially connected in a communication way, wherein when a workpiece to be detected moves to a preset detection position on the detection station 11 and approaches the center of a shooting visual field of the area-array camera 21, the proximity switch 12 is triggered and a trigger signal is sent to the computer control unit 6; the axis of the photosensitive lens of the area camera 21 is perpendicular to the circulation direction of the detection station, and considering that in an intelligent manufacturing workshop, the operation of equipment can be vibrated to a certain extent by the detection station, so that the shooting precision of the area camera 21 fixedly connected with the detection station is affected.
Specifically, the first photoelectric conversion element 22 and the second photoelectric conversion element 24 are each provided with the same number of input terminals and a plurality of output terminals, the number of optical fiber slip rings corresponds to the same, each input terminal of the first photoelectric conversion element 22 is connected to the area camera, each output terminal is connected to the input terminal of the second photoelectric conversion element 24 through one optical fiber slip ring 23, and each output terminal of the second photoelectric conversion element 24 is connected to the image processing unit 3.
After receiving the image subjected to photoelectric conversion, the image processing unit 3 firstly obtains the sensitive area image of the current detection station through resolution scanning and performs image segmentation, then only denoising and edge detection are performed on the obtained sensitive area, and the traditional image processing is to perform denoising and then image segmentation processing on the whole image.
Specifically, the image processing unit 3 automatically locates the center positions of a plurality of marking points on the image, determines the included angle between the background target 14 and the horizontal direction, and calculates the deflection angle between the area array camera 21 and the background target 14, the image processing unit 3 controls the image to perform resolution scanning along the deflection angle, so as to complete the identification of the sensitive area of the current detection station, and after the sensitive area is identified, the image processing unit 3 performs image segmentation operation on the image, so as to obtain a new image to be processed.
When denoising the segmented sensitive area image, in order to enable the sensitive area image to have a more natural smoothing effect and enhance the processing effect of random noise on the sensitive area image, the invention adopts a Gaussian filtering method to denoise the sensitive area image; after the smooth sensitive area image is obtained, the feature vector extraction unit further processes the processed sensitive area image.
Specifically, the purpose of the feature vector extraction unit is to reduce the high-dimensional image information featuring the pixel set to the low-dimensional image information featuring the vector set, so as to facilitate the processing of a computer and ensure the accuracy of the classification of the deep neural network unit; in the invention, based on the flexible production characteristics of intelligent manufacturing, the intelligent manufacturing information of the workpiece to be detected is obtained by adopting the core image characteristic of the shape characteristic, so that the manufacturing information of the sensitive area image of the workpiece to be detected is comprehensively captured, the external edge information and the internal area information of the target area are simultaneously considered, the edge area of the target area, the edge shape factor, the average radius of the target area and the Hu invariant moment of the previous 3 dimension are taken as characteristic variables for representing the sensitive area image, and the characteristic variables are taken as characteristic vectors of the sensitive area image and are input into the depth neural network unit 5.
In a preferred embodiment, based on the edge attribute of the sensitive area image, the feature vector extraction unit firstly performs edge detection on the sensitive area image to obtain a target area, and calculates and obtains the edge area, the edge shape factor and the average radius of the target area through formulas (1) - (3), and adds the Hu invariant moment of the previous 3 dimensions to form the feature vector of the sensitive area with four feature variables so as to reflect manufacturing quality information such as processing or assembly of the current product detection platform, and the feature vector is used as an input layer to be sent to the deep neural network unit 5;
in the above formula, the parameters M and N are the number of edge points of the target area,wherein t (x, y) is the gray value of each edge point; the parameter L is the perimeter of the target area, which can be obtained by calculating by adopting a chain code method in an image processing technology, and the reference K is the number of edge points on the boundary of the target area, (x) k ,y k ) Representing pixel coordinates located on the boundary of the target area, is->The centroid coordinates representing the target region can be calculated by the following formula:
wherein the parameter a represents the area of the sensitive area, the size of which can be obtained when the sensitive area is identified in the image processing.
In addition, the Hu invariant moment is taken as an important global feature of the image, is not influenced by light and noise, has good geometry and does not deform, can effectively describe an object image with a relatively complex shape, takes the typical manufacturing defect property of intelligent manufacturing and the image processing efficiency into consideration, and the selection of the front 3-dimensional Hu invariant moment as one of feature variables of the image of the sensitive area of the workpiece to be detected is effective, and a specific calculation mode can refer to the conventional practice in the image processing technology and is not repeated herein.
In a preferred embodiment, the deep neural network unit 5 of the present invention is specifically a manufacturing defect prediction model based on a deep neural network, and includes three layers of neural networks, that is, an input layer, an hidden layer and an output layer, where the input layer and the output layer have the same scale, the input layer is used as an input interface for manufacturing the defect prediction model, receives feature vectors of an image of a workpiece to be measured, and through information encoding, reaches the hidden layer, and then is transformed to the output layer through information decoding; before the deep neural network unit 5 formally classifies manufacturing defects of the workpiece to be detected, learning and training are required, specifically, a standard sample image library is established for typical manufacturing defects possibly occurring in the workpiece to be detected at the current detection station, and in a preferred embodiment, the standard sample image library may include four sample library types including manufacturing qualification, manufacturing defect I, manufacturing defect II, manufacturing defect III, and the like, which are used as training sample libraries of the deep neural network; similar to the extraction of the feature vector of the workpiece to be detected, the invention carries out edge detection on the images in the standard sample image library, and sequentially extracts the feature variables such as the edge area, the edge standard deviation, the shape factor, the Hu invariant moment and the like of the images to form the feature vector of the training sample library; finally, the input layer of the deep neural network unit 5 reads the feature vectors in the training sample library, and based on the coding and decoding of the deep neural network unit 5, deep learning is carried out on the manufacturing information corresponding to the images in each standard sample image library, so that the manufacturing defect prediction model of the current detection station is obtained.
After the deep neural network unit 5 performs learning training on the training sample library, the training sample library can be used for classifying and predicting manufacturing defect information of the workpiece to be detected in the current detection station 11, so that the type of the manufacturing defect of the workpiece to be detected on the current product detection platform 1 is identified, the manufacturing defect information of the workpiece to be detected is further sent to the computer control unit 6, and the computer control unit 6 is used for maintaining and processing the manufacturing defect.
In a preferred embodiment, the computer control unit 6 includes a PLC controller 61 and is respectively in communication connection with the product detection platform 1, the image acquisition unit 2 and the deep neural network unit 5, and the computer control unit 6 determines resources required for repairing the manufacturing defect according to manufacturing defect information indicated by the classification result of the deep neural network unit 5 in combination with a historical repair database, automatically generates a repair policy for repairing the defect in terms of the defect type, the defect position, the defect degree, the maintainer selection, and the like, forms a corresponding working instruction, sends the working instruction to the PLC controller 61, performs a specific repair operation by the PLC controller 61, and updates and adds a repair record for the execution condition in the historical repair database.
Second embodiment
The invention further provides a fault online detection method applied to the intelligent manufacturing workshop, which adopts the intelligent manufacturing workshop fault online detection system and comprises the following steps:
step 1: constructing a manufacturing defect prediction model based on a deep neural network, and training and learning the deep neural network through sample images;
step 2: the intelligent positioning and releasing mechanism drives the workpiece to be tested to move on the detection station under the control instruction of the computer control module to enable the workpiece to be tested to reach a preset detection position, triggers the proximity switch and sends a trigger signal to the computer control unit;
step 3: the computer control unit respectively sends instructions to the LED area light source matrix and the image acquisition unit, the LED area light source matrix turns on illumination, the image acquisition unit shoots a workpiece to be detected according to a preset program and parameters, and after photoelectric conversion, the generated image is sent to the image processing unit;
step 4: the image processing unit identifies and partitions a sensitive area of the current detection station, performs image denoising processing on the local area, and sends the sensitive area image to the feature vector extraction unit after denoising;
step 5: the feature vector extraction unit performs edge detection on the sensitive area to form a target area, calculates the edge area, the edge shape factor and the average radius of the target area through formulas (1) - (3), and combines the Hu invariant moment of the previous 3 dimensions to form the feature vector of the sensitive area with four feature variables;
step 6: diagnosing the manufacturing information of the feature vector based on the trained deep neural network, predicting and classifying manufacturing defects of the workpiece to be detected, and feeding back the classification result to the computer control unit;
step 7: the computer control unit determines a maintenance strategy for maintaining the manufacturing defect based on the classification result and the historical maintenance database information, and sends a working instruction to the PLC controller, and the PLC controller executes specific maintenance operation.
In a preferred embodiment, the step 1 specifically includes:
step 1.1: constructing a manufacturing defect prediction model based on a deep neural network, wherein the deep neural network is a 3-layer neural network and comprises an input layer, an output layer and an hidden layer, and the input layer and the output layer have the same scale;
step 1.2: aiming at typical manufacturing defects possibly occurring in the current detection station, a standard sample image library is established, wherein the standard sample image library comprises four sample library types including qualified manufacturing, manufacturing defect I, manufacturing defect II and manufacturing defect III, and the standard sample image library is used as a training sample library of a deep neural network;
step 1.3: carrying out edge detection on images in a standard sample image library, sequentially extracting characteristic variables such as edge area, edge standard deviation, shape factor, hu invariant moment and the like of the images, and forming characteristic vectors of a training sample library;
step 1.4: and the input layer of the deep learning network unit reads the feature vectors in the training sample library, and deep learning is carried out on manufacturing information corresponding to the images in each standard sample image library, so that a manufacturing defect prediction model of the current detection station is obtained.
With the above-described preferred embodiments according to the present invention as an illustration, the above-described descriptions can be used by persons skilled in the relevant art to make various changes and modifications without departing from the scope of the technical idea of the present invention. The technical scope of the present invention is not limited to the description, but must be determined according to the scope of claims.
Claims (7)
1. The fault on-line detection system is characterized by comprising a workpiece detection platform, an image acquisition unit, an image processing unit, a feature vector extraction unit, a deep neural network unit and a computer control unit, wherein,
the workpiece detection platform comprises a detection station, and an area array camera in the image acquisition unit acquires images of the workpiece detection platform on the detection station and sends the images to the image processing unit;
the image processing unit performs resolution scanning on the received image to obtain a sensitive area image of the current detection station, denoising the sensitive area image, and then sending the denoised sensitive area image to the feature vector extraction unit;
the feature vector extraction unit performs edge detection on the sensitive area image to form a target area, calculates and obtains the edge area, the edge shape factor and the average radius of the target area through formulas (1) to (3), and adds the Hu invariant moment of the previous 3 dimensions to form feature vectors of the sensitive area with four feature variables so as to reflect the workpiece quality information of the current workpiece detection platform, and the feature vectors are used as input layers to be sent to the deep neural network unit;
in the above formula, the parameters M and N are the number of edge points of the target area,wherein t (x, y) is the gray value of each edge point; the parameter L is the perimeter of the target area, the parameter L is obtained by calculation by adopting a chain code method in an image processing technology, the reference K is the number of edge points on the boundary of the target area, (x) k ,y k ) Representing pixel coordinates located on the boundary of the target area,representing the centroid coordinates of the target region, calculated by the following formula:
wherein the parameter a represents the area of the sensitive area and is adapted to obtain its size when the sensitive area is identified in the image processing;
the deep neural network unit builds a manufacturing defect prediction model based on a neural network algorithm, trains, learns and classifies the image feature vectors of the workpiece detection platform, identifies the type of the manufacturing defect of the workpiece to be detected on the current workpiece detection platform, and feeds the classification result back to the computer control unit;
the workpiece detection platform further comprises a proximity switch and an intelligent positioning and releasing mechanism, a background target plate which is continuously arranged along the flexible production line is arranged on one side of the detection station, a plurality of marking points corresponding to the current detection station are arranged on the background target plate, an LED area light source matrix is arranged on the other side of the detection station, the LED area light source matrix is in communication connection with the computer control unit, and the background target plate and the LED area light source matrix jointly form a backlight illumination environment; and
the proximity switch, the intelligent positioning and releasing mechanism of the workpiece detection platform are all in communication connection with the computer control unit, wherein the intelligent positioning and releasing mechanism receives a control instruction of the computer control unit, drives a workpiece to be detected to move on a detection station to reach a preset detection position and approach the shooting visual field center of the area array camera, triggers the proximity switch and sends a trigger signal to the computer control unit, and the computer control unit starts the image acquisition unit to acquire images;
the image processing unit automatically locates the central positions of a plurality of mark points on the image, and determines the included angle between the background target and the horizontal direction, so as to calculate the deflection angle between the area array camera and the background target, and the image processing unit controls the image to carry out resolution scanning along the deflection angle, so that the identification of the sensitive area of the current detection station is completed.
2. The fault on-line detection system of claim 1, wherein,
the image acquisition unit comprises a first photoelectric conversion element, an optical fiber slip ring and a second photoelectric conversion element which are sequentially in communication connection with the area array camera, wherein the axis of a photosensitive lens of the area array camera is perpendicular to the circulation direction of the detection station, and after the image acquired by the area array camera is subjected to photoelectric conversion, the second photoelectric conversion element sends an electric signal of the image to the image processing unit;
the first photoelectric conversion elements and the second photoelectric conversion elements of the image acquisition unit are respectively provided with the same number of input ends and a plurality of output ends, the number of the optical fiber slip rings corresponds to the same number of the optical fiber slip rings, each input end of the first photoelectric conversion elements is respectively connected with the area array camera, each output end of the first photoelectric conversion elements is respectively connected with the input end of the second photoelectric conversion elements through one optical fiber slip ring, and each output end of the second photoelectric conversion elements is connected with the image processing unit.
3. The fault on-line detection system of claim 1, wherein,
the computer control unit comprises a PLC controller and is respectively in communication connection with the workpiece detection platform, the image acquisition unit and the deep neural network unit, and after receiving the manufacturing defect information of the workpiece detection platform, the computer control unit automatically generates a maintenance strategy and sends a working instruction to the PLC controller, and the PLC controller executes specific maintenance operation.
4. A fault on-line detection system according to any one of claims 1 to 3, wherein,
the computer control unit determines resources required for maintaining the manufacturing defects according to the manufacturing defect information indicated by the classification result of the deep neural network unit and combines a historical maintenance database, and selects a maintenance strategy for automatically generating the defects for maintaining the defects according to the defect type, the defect position, the defect degree and maintenance personnel, forms a corresponding working instruction and sends the working instruction to the PLC, the PLC executes specific maintenance operation, and updates and adds a maintenance record for the execution condition in the historical maintenance database.
5. A fault on-line detection method for intelligent manufacturing plants using a fault on-line detection system according to claim 1, comprising:
step 1: constructing a manufacturing defect prediction model based on a deep neural network, and training and learning the deep neural network through sample images;
step 2: the workpiece detection platform drives a workpiece to be detected to move on a detection station under the control instruction of the computer control module to reach a preset detection position, triggers the proximity switch and sends a trigger signal to the computer control unit;
step 3: the computer control unit respectively sends instructions to the LED area light source matrix and the image acquisition unit, the LED area light source matrix turns on illumination, the image acquisition unit shoots a workpiece to be detected, and the generated image is sent to the image processing unit;
step 4: the image processing unit identifies and partitions a sensitive area of the current detection station, performs image denoising processing on the sensitive area, and sends the sensitive area image to the feature vector extraction unit after denoising;
step 5: the feature vector extraction unit performs edge detection on the sensitive area to form a target area, calculates and obtains the edge area, the edge shape factor and the average radius of the target area, and combines the Hu invariant moment of the previous 3 dimensions to form the feature vector of the sensitive area with four feature variables;
step 6: and diagnosing the manufacturing information of the feature vector based on the trained deep neural network, predicting and classifying the manufacturing defects of the workpiece to be detected, and feeding back the classification result to the computer control unit.
6. The method for online detection of faults as claimed in claim 5, wherein,
the fault on-line detection method is adapted to detect manufacturing defects of a workpiece using the fault on-line detection system of claim 1.
7. The method for on-line fault detection as claimed in claim 6, wherein,
the step 1: the method for constructing the manufacturing defect prediction model based on the deep neural network and training and learning the deep neural network through the sample image comprises the following steps:
step 1.1: constructing a manufacturing defect prediction model based on a deep neural network, wherein the deep neural network is a 3-layer neural network and comprises an input layer, an output layer and an hidden layer, and the input layer and the output layer have the same scale;
step 1.2: aiming at typical manufacturing defects possibly occurring in the current detection station, a standard sample image library is established, wherein the standard sample image library comprises four sample library types including qualified manufacturing, manufacturing defect I, manufacturing defect II and manufacturing defect III, and the standard sample image library is used as a training sample library of a deep neural network;
step 1.3: carrying out edge detection on images in a standard sample image library, and sequentially extracting edge area, edge standard deviation, shape factor and Hu invariant moment characteristic variables of the images to form characteristic vectors of a training sample library;
step 1.4: and the input layer of the deep learning network unit reads the feature vectors in the training sample library, and deep learning is carried out on manufacturing information corresponding to the images in each standard sample image library, so that a manufacturing defect prediction model of the current detection station is obtained.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811651018.2A CN109840900B (en) | 2018-12-31 | 2018-12-31 | Fault online detection system and detection method applied to intelligent manufacturing workshop |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811651018.2A CN109840900B (en) | 2018-12-31 | 2018-12-31 | Fault online detection system and detection method applied to intelligent manufacturing workshop |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109840900A CN109840900A (en) | 2019-06-04 |
| CN109840900B true CN109840900B (en) | 2023-12-19 |
Family
ID=66883679
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811651018.2A Active CN109840900B (en) | 2018-12-31 | 2018-12-31 | Fault online detection system and detection method applied to intelligent manufacturing workshop |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109840900B (en) |
Families Citing this family (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110989507A (en) * | 2019-11-02 | 2020-04-10 | 温州大学 | Automatic production digital twin workshop generation device that detects of circuit breaker |
| CN113217374B (en) * | 2020-02-05 | 2023-05-30 | 中国石油天然气股份有限公司 | Operation maintenance method and system for vertical well screw pump |
| CN111272775A (en) * | 2020-02-24 | 2020-06-12 | 上海感图网络科技有限公司 | Device and method for detecting defects of heat exchanger by using artificial intelligence |
| CN111583190B (en) * | 2020-04-16 | 2022-07-22 | 浙江浙能技术研究院有限公司 | Automatic identification method for hidden crack defect of internal cascade structure component |
| CN111721728B (en) * | 2020-07-16 | 2023-02-21 | 陈皓 | Fruit online detection device and use method thereof |
| CN112270284B (en) * | 2020-11-06 | 2021-12-03 | 奥斯福集团有限公司 | Lighting facility monitoring method and system and electronic equipment |
| CN113172422B (en) * | 2021-05-18 | 2022-04-12 | 中品智能机械有限公司 | Woodworking mechanical equipment assembly production line and assembly process thereof |
| CN113781448B (en) * | 2021-09-14 | 2024-01-23 | 国电四子王旗光伏发电有限公司 | Intelligent defect identification method for photovoltaic power station assembly based on infrared image analysis |
| CN113886627A (en) * | 2021-10-09 | 2022-01-04 | 陕西通信规划设计研究院有限公司 | Mobile communication system based on information synchronization |
| CN116805204B (en) * | 2023-08-24 | 2023-12-01 | 超网实业(成都)股份有限公司 | Intelligent plant monitoring method and system |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2016201947A1 (en) * | 2015-06-16 | 2016-12-22 | 华南理工大学 | Method for automated detection of defects in cast wheel products |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20180211373A1 (en) * | 2017-01-20 | 2018-07-26 | Aquifi, Inc. | Systems and methods for defect detection |
-
2018
- 2018-12-31 CN CN201811651018.2A patent/CN109840900B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2016201947A1 (en) * | 2015-06-16 | 2016-12-22 | 华南理工大学 | Method for automated detection of defects in cast wheel products |
Non-Patent Citations (3)
| Title |
|---|
| "基于机器视觉的彩钢板缺陷检测和智能分类研究";孙创开;《中国优秀硕士学位论文全文数据库 信息科技辑》;20170715(第07期);第二章、第三章 * |
| "触点零件形貌在线自学习视觉检测系统研究";戴舒文;《中国优秀硕士学位论文全文数据库 信息科技辑》;20090915(第09期);第36-46页 * |
| 基于BP神经网络的GIS缺陷图像识别系统的研究;万书亭等;《电力科学与工程》;20171128(第11期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109840900A (en) | 2019-06-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109840900B (en) | Fault online detection system and detection method applied to intelligent manufacturing workshop | |
| CN111272763B (en) | System and method for workpiece inspection | |
| EP3776462B1 (en) | System and method for image-based target object inspection | |
| CN111815601B (en) | Texture image surface defect detection method based on depth convolution self-encoder | |
| CN111929309B (en) | Cast part appearance defect detection method and system based on machine vision | |
| CN111507976B (en) | Defect detection method and system based on multi-angle imaging | |
| CN109693140B (en) | Intelligent flexible production line and working method thereof | |
| CN109978835B (en) | Online assembly defect identification system and method thereof | |
| CN109461149A (en) | The intelligent checking system and method for lacquered surface defect | |
| CN104992449A (en) | Information identification and surface defect on-line detection method based on machine visual sense | |
| CN102529019B (en) | Method for mould detection and protection as well as part detection and picking | |
| JP2021515885A (en) | Methods, devices, systems and programs for setting lighting conditions and storage media | |
| CN110186375A (en) | Intelligent high-speed rail white body assemble welding feature detection device and detection method | |
| CN114280075A (en) | Online visual inspection system and method for surface defects of pipe parts | |
| CN115330734A (en) | Automatic robot repair welding system based on three-dimensional target detection and point cloud defect completion | |
| US20230110396A1 (en) | Layup table for composite fabrication with ai quality control modules | |
| CN113592813B (en) | New energy battery welding defect detection method based on deep learning semantic segmentation | |
| CN116843615B (en) | Lead frame intelligent total inspection method based on flexible light path | |
| Brambilla et al. | Automated Vision Inspection of Critical Steel Components based on Signal Analysis Extracted form Images | |
| CN116840240A (en) | Visual detection system of power supply distributor | |
| Akbar et al. | The design and development of automated visual inspection system for press part sorting | |
| Lin et al. | Enhancing the quality inspection process in the food manufacturing industry through automation | |
| CN118070983B (en) | Industrial machinery production optimization method and system based on deep learning | |
| CN114943684B (en) | Curved surface anomaly detection method by using confrontation to generate self-coding neural network | |
| CN117314829A (en) | Industrial part quality inspection method and system based on computer vision |
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 | ||
| CB02 | Change of applicant information |
Address after: 213100 No.28, Mingxin Middle Road, Wujin District, Changzhou City, Jiangsu Province Applicant after: Changzhou Polytechnic Address before: 213100 No.28, Mingxin Middle Road, Wujin District, Changzhou City, Jiangsu Province Applicant before: Changzhou Institute of Industry Technology |
|
| CB02 | Change of applicant information | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |