CN113071218B - Quality monitoring method for flexible sensor-oriented ink-jet printing process - Google Patents
Quality monitoring method for flexible sensor-oriented ink-jet printing process Download PDFInfo
- Publication number
- CN113071218B CN113071218B CN202110320664.6A CN202110320664A CN113071218B CN 113071218 B CN113071218 B CN 113071218B CN 202110320664 A CN202110320664 A CN 202110320664A CN 113071218 B CN113071218 B CN 113071218B
- Authority
- CN
- China
- Prior art keywords
- ink
- vibration signal
- fault
- signal
- sensor
- 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
- 238000000034 method Methods 0.000 title claims abstract description 103
- 238000007641 inkjet printing Methods 0.000 title claims abstract description 79
- 230000008569 process Effects 0.000 title claims abstract description 52
- 238000012544 monitoring process Methods 0.000 title claims abstract description 32
- 238000007639 printing Methods 0.000 claims abstract description 47
- 230000002159 abnormal effect Effects 0.000 claims abstract description 19
- 238000012800 visualization Methods 0.000 claims abstract description 13
- 238000006073 displacement reaction Methods 0.000 claims description 39
- 239000000758 substrate Substances 0.000 claims description 31
- 230000000007 visual effect Effects 0.000 claims description 30
- 238000012549 training Methods 0.000 claims description 25
- 239000000919 ceramic Substances 0.000 claims description 24
- 238000012545 processing Methods 0.000 claims description 22
- 238000010168 coupling process Methods 0.000 claims description 15
- 238000005859 coupling reaction Methods 0.000 claims description 15
- 230000008878 coupling Effects 0.000 claims description 14
- 230000004927 fusion Effects 0.000 claims description 12
- 230000001360 synchronised effect Effects 0.000 claims description 9
- 238000012935 Averaging Methods 0.000 claims description 6
- 238000005457 optimization Methods 0.000 claims description 5
- 239000007921 spray Substances 0.000 claims description 5
- 239000000463 material Substances 0.000 abstract description 10
- 239000002699 waste material Substances 0.000 abstract description 5
- 238000004458 analytical method Methods 0.000 description 23
- 239000013598 vector Substances 0.000 description 23
- 238000001228 spectrum Methods 0.000 description 12
- 238000004364 calculation method Methods 0.000 description 10
- 238000005507 spraying Methods 0.000 description 9
- 238000012360 testing method Methods 0.000 description 9
- 230000001105 regulatory effect Effects 0.000 description 7
- 238000003491 array Methods 0.000 description 4
- 238000002360 preparation method Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 238000004519 manufacturing process Methods 0.000 description 3
- 238000003860 storage Methods 0.000 description 3
- 230000004913 activation Effects 0.000 description 2
- 230000001276 controlling effect Effects 0.000 description 2
- 238000009826 distribution Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000007689 inspection Methods 0.000 description 2
- 210000002569 neuron Anatomy 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 230000003595 spectral effect Effects 0.000 description 2
- 230000002950 deficient Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000004146 energy storage Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000010191 image analysis Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 238000004886 process control Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 238000010926 purge Methods 0.000 description 1
- 230000008439 repair process Effects 0.000 description 1
- 238000012163 sequencing technique Methods 0.000 description 1
- 238000003892 spreading Methods 0.000 description 1
Images
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B41—PRINTING; LINING MACHINES; TYPEWRITERS; STAMPS
- B41J—TYPEWRITERS; SELECTIVE PRINTING MECHANISMS, i.e. MECHANISMS PRINTING OTHERWISE THAN FROM A FORME; CORRECTION OF TYPOGRAPHICAL ERRORS
- B41J2/00—Typewriters or selective printing mechanisms characterised by the printing or marking process for which they are designed
- B41J2/005—Typewriters or selective printing mechanisms characterised by the printing or marking process for which they are designed characterised by bringing liquid or particles selectively into contact with a printing material
- B41J2/01—Ink jet
- B41J2/015—Ink jet characterised by the jet generation process
- B41J2/04—Ink jet characterised by the jet generation process generating single droplets or particles on demand
- B41J2/045—Ink jet characterised by the jet generation process generating single droplets or particles on demand by pressure, e.g. electromechanical transducers
- B41J2/04501—Control methods or devices therefor, e.g. driver circuits, control circuits
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01D—MEASURING NOT SPECIALLY ADAPTED FOR A SPECIFIC VARIABLE; ARRANGEMENTS FOR MEASURING TWO OR MORE VARIABLES NOT COVERED IN A SINGLE OTHER SUBCLASS; TARIFF METERING APPARATUS; MEASURING OR TESTING NOT OTHERWISE PROVIDED FOR
- G01D21/00—Measuring or testing not otherwise provided for
- G01D21/02—Measuring two or more variables by means not covered by a single other subclass
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Ink Jet (AREA)
Abstract
本发明公开了一种面向柔性传感器喷墨打印过程的质量监控方法，采集柔性传感器喷墨打印过程中喷头的振动信号；采集柔性传感器喷墨打印过程中墨滴可视化信息；根据采集获得的所述振动信号和所述墨滴可视化信息获得打印质量及判断喷墨打印是否运行正常，即本申请通过打印过程中对振动信号和墨滴可视化信息进行监控，降低了不正常运行打印的可能性，提高了打印质量，同时减少了材料浪费。
The invention discloses a quality monitoring method oriented to the inkjet printing process of a flexible sensor. The vibration signal of a nozzle in the inkjet printing process of the flexible sensor is collected; the visualization information of ink droplets in the inkjet printing process of the flexible sensor is collected; The vibration signal and the ink drop visualization information are used to obtain the printing quality and determine whether the inkjet printing is running normally, that is, the present application monitors the vibration signal and the ink drop visualization information during the printing process, thereby reducing the possibility of abnormal operation of printing and improving the Improve print quality while reducing material waste.
Description
Technical Field
The invention relates to the technical field of ink-jet printing, in particular to a quality monitoring method for a flexible sensor ink-jet printing process.
Background
At present, quality problems occurring in the ink-jet printing preparation process in the market can be discovered only by the last quality inspection step, but are difficult to discover in the ink-jet printing preparation process, so that the subsequent steps are influenced, and a large amount of materials, equipment and energy are wasted.
A typical process for making a flexible sensor printer using inkjet printing technology is: sensor pattern/structure design, material preparation and pretreatment, ink-jet printing manufacturing, sensor prototype post-treatment, manufacturing quality inspection, and defective repair/qualified shipment. Most flexible sensor printer manufacturing adopts the method, but the method causes a great deal of waste of materials, equipment and energy.
Disclosure of Invention
The invention aims to provide a quality monitoring method for a flexible sensor ink-jet printing process, which improves the printing quality and reduces the material waste.
In order to achieve the purpose, the invention provides the following scheme:
a method of quality monitoring for a flexible sensor-oriented inkjet printing process, the method comprising:
acquiring a vibration signal of a nozzle in the ink-jet printing process of the flexible sensor;
collecting visual information of ink drops in the ink-jet printing process of the flexible sensor;
and obtaining printing quality according to the acquired vibration signal and the visual information of the ink drop and judging whether the ink-jet printing normally operates.
Optionally, the vibration signals include an ink bag pressure wave, an ink drop ejection wave, and an ink drop impact wave.
Optionally, the ink drop visualization information includes ink drop speed, ink drop column number and ejection base distance; the distance between the spraying bases is the distance between the spraying head and the base;
optionally, the acquiring, according to the vibration signal and the visual information of the ink droplet, print quality and determining whether inkjet printing is normal includes:
inputting the vibration signal into a first fault classifier, and judging whether a first fault occurs;
inputting the visual information of the ink drops into a second fault classifier, and judging whether a second fault occurs;
and if the first fault or the second fault occurs, judging that the ink jet printing is abnormal.
Optionally, the method further comprises:
collecting a vibration signal during normal printing and a vibration signal during fault printing;
and training a minimum distance classifier by taking the vibration signal during normal printing and the vibration signal during fault printing as first training samples to obtain a first fault classifier.
Optionally, the method further comprises:
collecting visual information of ink drops during normal printing and visual information of ink drops during fault printing;
and training a minimum distance classifier by taking the visual information of the ink drops during normal printing and the visual information of the ink drops during fault printing as a first training sample to obtain a second fault classifier.
Optionally, the method further comprises:
coupling the first fault signal and the second fault signal to obtain a characteristic curve of multivariate coupling; the first fault signal is a fault signal obtained when a first fault occurs, and the second fault signal is a fault signal obtained when a second fault occurs;
and comparing the characteristic curve of the multivariate coupling with a corresponding preset characteristic curve to obtain the printing quality.
Optionally, the acquiring a vibration signal of a nozzle in an inkjet printing process of the flexible sensor specifically includes:
obtaining the ink bag pressure wave by adsorbing a first displacement sensor on an ink bag of the flexible sensor;
acquiring the ink drop ejection wave by adsorbing a second displacement sensor on the piezoelectric ceramic of the flexible sensor;
and acquiring the ink drop impact wave by adsorbing a third displacement sensor on the flexible substrate of the flexible sensor.
Optionally, the obtaining the ink bag pressure wave by adsorbing a first displacement sensor on an ink bag of the flexible sensor; acquiring the ink drop ejection wave by adsorbing a second displacement sensor on the piezoelectric ceramic of the flexible sensor; the ink drop impact wave is obtained by adsorbing a third displacement sensor on a flexible substrate of the flexible sensor, and the method specifically comprises the following steps:
the vibration signals collected by the first displacement sensor, the second displacement sensor and the third displacement sensor are subjected to fusion collision processing, the vibration signals collected by the first displacement sensor are recorded as vibration signals of an ink sac, the vibration signals collected by the second displacement sensor are recorded as vibration signals of piezoelectric ceramics, the vibration signals collected by the third displacement sensor are recorded as vibration signals of a flexible substrate, and a signal optimization model for the fusion collision processing is expressed as follows:
f1＝f1-a1f2-b1f3-c1f0
f2＝f2-a2f1-b2f3-c2f0；
f3＝f3-a3f1-b3f2-c3f0
wherein f is1As a vibration signal of the ink bag, f2Is a vibration signal of the piezoelectric ceramic, f3Is a vibration signal of the flexible substrate, f0Is a vibration interference signal; a1, a2, a3, b1, b2, b3, c1, c2 and c3 are all proportional parameters; f1 is the optimized vibration signal of the ink bag, f2 is the optimized vibration signal of the piezoelectric ceramic, and f3 is the optimized vibration signal of the flexible substrate;
processing the optimized vibration signal of the piezoelectric ceramic by a time domain synchronous averaging method to obtain the ink sac pressure wave;
carrying out time domain synchronous averaging processing on the optimized vibration signal of the ceramic wafer to obtain the ink drop ejection wave;
and carrying out time domain synchronous average processing on the optimized vibration signal of the flexible substrate to obtain the ink drop impact wave.
Optionally, the acquiring visualized information of ink droplets in the inkjet printing process of the flexible sensor specifically includes:
acquiring image information of ink droplet running through a CCD image sensor;
and acquiring the ink drop speed, the ink drop column number and the ejection base distance according to the image information.
According to the specific embodiment provided by the invention, the invention discloses the following technical effects:
the invention discloses a quality monitoring method for a flexible sensor ink-jet printing process, which comprises the steps of collecting vibration signals of a nozzle in the flexible sensor ink-jet printing process; collecting visual information of ink drops in the ink-jet printing process of the flexible sensor; obtain according to the collection vibration signal with the visual information acquisition of ink droplet prints the quality and judges whether ink jet printing operates normally, and this application is monitored vibration signal and the visual information of ink droplet through printing the in-process promptly, has reduced the possibility that abnormal operation was printed, has improved the printing quality, has reduced the material simultaneously and has wasted.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings needed to be used in the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings without inventive exercise.
FIG. 1 is a schematic flow chart of a quality monitoring method for a flexible sensor-oriented inkjet printing process according to the present invention;
FIG. 2 is a schematic diagram of a system architecture for a quality monitoring method based on a flexible sensor-oriented inkjet printing process according to the present invention;
fig. 3 is a detailed flow chart of a quality monitoring method for a flexible sensor-oriented inkjet printing process according to the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The invention aims to provide a quality monitoring method for a flexible sensor ink-jet printing process, which improves the printing quality and reduces the material waste.
In order to make the aforementioned objects, features and advantages of the present invention comprehensible, embodiments accompanied with figures are described in further detail below.
Fig. 1 is a schematic flow chart of a quality monitoring method for a flexible sensor-oriented inkjet printing process according to the present invention, and as shown in fig. 1, the quality monitoring method for the flexible sensor-oriented inkjet printing process includes:
step 101: and acquiring a vibration signal of a nozzle in the ink-jet printing process of the flexible sensor.
Gather flexible sensor inkjet printing in-process printing apparatus's vibration waveform, specifically include:
and acquiring an ink bag pressure wave, an ink drop ejection wave and an ink drop impact wave in the ink jet printing process of the flexible sensor.
Gather flexible sensor inkjet printing in-process ink sac pressure wave, ink droplet ejection wave and ink droplet impact wave, specifically include:
obtaining the ink bag pressure wave by adsorbing a first displacement sensor on an ink bag of the flexible sensor;
acquiring the ink drop ejection wave by adsorbing a second displacement sensor on the piezoelectric ceramic of the flexible sensor;
and acquiring the ink drop impact wave by adsorbing a third displacement sensor on the flexible substrate of the flexible sensor.
The ink bag pressure wave is obtained by adsorbing a first displacement sensor on an ink bag of the flexible sensor; acquiring the ink drop ejection wave by adsorbing a second displacement sensor on the piezoelectric ceramic of the flexible sensor; the ink drop impact wave is obtained by adsorbing a third displacement sensor on a flexible substrate of the flexible sensor, and the method specifically comprises the following steps:
the vibration signals collected by the first displacement sensor, the second displacement sensor and the third displacement sensor are subjected to fusion collision processing, the vibration signals collected by the first displacement sensor are recorded as vibration signals of an ink sac, the vibration signals collected by the second displacement sensor are recorded as vibration signals of piezoelectric ceramics, the vibration signals collected by the third displacement sensor are recorded as vibration signals of a flexible substrate, and a signal optimization model for the fusion collision processing is expressed as follows:
f1＝f1-a1f2-b1f3-c1f0
f2＝f2-a2f1-b2f3-c2f0；
f3＝f3-a3f1-b3f2-c3f0
wherein f is1As a vibration signal of the ink bag, f2Is a vibration signal of the piezoelectric ceramic, f3Is a vibration signal of the flexible substrate, f0Is a vibration interference signal; a1, a2, a3, b1, b2, b3, c1, c2 and c3 are all proportional parameters; f1 is the optimized vibration signal of the ink bag, f2 is the optimized vibration signal of the piezoelectric ceramic, and f3 is the optimized vibration signal of the flexible substrate.
And carrying out time domain synchronous average processing on the optimized vibration signal of the piezoelectric ceramic to obtain the ink sac pressure wave.
And carrying out time domain synchronous averaging processing on the optimized vibration signal of the ceramic wafer to obtain the ink drop ejection wave.
And carrying out time domain synchronous average processing on the optimized vibration signal of the flexible substrate to obtain the ink drop impact wave.
The first displacement sensor, the second displacement sensor and the third displacement sensor are all laser displacement sensors.
Step 102: and collecting visual information of ink drops in the ink-jet printing process of the flexible sensor.
The ink drop visualization information comprises ink drop speed, ink drop column number and ejection base distance. And the spraying base interval is the interval between the spray head and the substrate.
Gather flexible sensor ink-jet printing in-process ink droplet speed, ink droplet column number and spout basic spacing, specifically include:
acquiring image information of ink droplet running through a CCD image sensor;
and acquiring the ink drop speed, the ink drop column number and the ejection base distance according to the image information.
The obtaining of the ink drop speed, the ink drop column number and the ejection base distance according to the image information specifically includes: and obtaining the ink drop speed, the ink drop column number and the ejection base spacing according to an image sorting calculation model, wherein the image sorting calculation model is expressed as:
Di＝Di+T
T＝t1+t2+t3
m＝kT
wherein T1 is the time corresponding to the ink droplet collecting speed, T2 is the time corresponding to the number of columns of the collected ink droplets, T3 is the time corresponding to the height of the collected ejection base, T is the collecting period, and m is the total collecting time; diThe ink drop distance corresponding to the time t1 is represented by i, k is an integer, V is the ink drop velocity number, and N is the ink drop column number; h is the jet base spacing.
The CCD image sensor is an MVC13000F-M00CCD image sensor.
Step 103: and obtaining printing quality according to the acquired vibration signal and the visual information of the ink drop and judging whether the ink-jet printing normally operates. The method specifically comprises the following steps: and obtaining the printing quality according to the vibration signal and the visual information of the ink drop obtained by collection, and judging whether the ink-jet printing normally operates according to the vibration signal and the visual information of the ink drop obtained by collection.
As a specific embodiment, judging whether inkjet printing normally operates according to the acquired vibration signal and the visualized information of the ink droplet specifically includes:
and inputting the vibration signal into a first fault classifier, and judging whether a first fault occurs.
And inputting the visual information of the ink drops into a second fault classifier, and judging whether a second fault occurs.
And if the first fault or the second fault occurs, judging that the ink jet printing is abnormal. Abnormal ink jet printing is the fault of ink jet printing.
The first fault classifier and the second fault classifier output whether a fault and a fault type.
The method further comprises the following steps:
collecting a vibration signal during normal printing and a vibration signal during fault printing;
and training a minimum distance classifier by taking the vibration signal during normal printing and the vibration signal during fault printing as first training samples to obtain a first fault classifier.
The fault type output by the first fault classifier comprises
The method further comprises the following steps:
collecting visual information of ink drops during normal printing and visual information of ink drops during fault printing;
and training a minimum distance classifier by taking the visual information of the ink drops during normal printing and the visual information of the ink drops during fault printing as a first training sample to obtain a second fault classifier.
The fault types output by the second fault classifier include a drop velocity fault, a drop column number fault, and a jet pitch fault.
The method further comprises the following steps: obtaining printing quality according to the acquired vibration signal and the visualized information of the ink drop, and specifically comprising the following steps:
coupling the first fault signal and the second fault signal to obtain a characteristic curve of multivariate coupling; the first fault signal is a fault signal obtained when a first fault occurs, and the second fault signal is a fault signal obtained when a second fault occurs;
and comparing the characteristic curve of the multivariate coupling with a corresponding preset characteristic curve to obtain the printing quality.
As a specific embodiment, the determining whether inkjet printing normally operates according to the acquired vibration signal and the visualized information of the ink droplet specifically includes:
obtaining a preset vibration waveform according to the model of the flexible sensor, wherein the preset vibration waveform is a vibration waveform when the flexible sensor normally operates;
judging whether the difference between a preset vibration waveform and an acquired vibration waveform is within a preset waveform range or not;
if not, the ink-jet printing operation is abnormal;
obtaining a preset ink drop speed, a preset ink drop longitudinal number and a preset spraying base interval according to the model of the flexible sensor;
judging whether the difference between the preset ink drop speed and the collected ink drop speed is within a preset speed range or not;
if not, the ink-jet printing operation is abnormal;
judging whether the difference between the preset ink drop column number and the acquired ink drop column number is within a preset number range or not;
if not, the ink-jet printing operation is abnormal;
judging whether the difference between the preset spraying base distance and the collected spraying base distance is within a preset distance range or not;
if not, the ink-jet printing operation is abnormal;
if the difference between the preset vibration waveform and the collected vibration waveform is within the preset waveform range, the difference between the preset ink drop speed and the collected ink drop speed is within the preset speed range, the difference between the preset ink drop longitudinal column number and the collected ink drop longitudinal column number is within the preset number range, and the difference between the preset jet base distance and the collected jet base distance is within the preset distance range, the ink-jet printing operation is normal.
The method further comprises the following steps:
and if the ink-jet printing is abnormal, alarming.
The invention can automatically and accurately monitor the ink-jet printing preparation process of the flexible sensor, can timely find abnormal operation of ink-jet printing, reduces material waste and simultaneously improves printing quality.
Fig. 2 is a schematic structural diagram of a system of the quality monitoring method based on the flexible sensor-oriented inkjet printing process, and as shown in fig. 2, the system of the quality monitoring method based on the flexible sensor-oriented inkjet printing process includes: the system comprises a knowledge base module, a multi-texture acoustics module, a visual monitoring module, a disordered reserve pool analysis module and a display alarm module.
The knowledge base module is used for storing the model of the flexible sensor and six corresponding parameters in the normal ink-jet printing process: ink bag pressure waves, ink drop ejection waves, ink drop substrate impact waves (drop impact waves), drop velocity, number of columns of ink drops, and ejection substrate spacing.
The multi-grain acoustic module is used for detecting the vibration of the key process points of ink-jet printing and collecting three vibration waveforms in the ink-jet printing process: the ink bag pressure wave, the ink drop ejection wave and the ink drop substrate impact wave form a characteristic vector through data processing, and whether a device fails or not is judged in advance through a failure pre-classification module (a first failure classifier).
The visual monitoring module is used for detecting the ink drop speed, the ink drop longitudinal column number and the spraying base distance in the ink-jet printing process, carrying out fusion and conflict processing through image processing, and judging whether a device fails through the fault pre-classification module (a second fault classifier).
The unordered reserve pool analysis module is used for analyzing data calculated by the multi-texture acoustics module and the visual monitoring module, and the printing threshold value is calculated through input at two ends (the multi-texture acoustics module and the visual monitoring module), a middle unordered calculation unit and multi-source coupling so as to evaluate the printing quality in the ink-jet printing process.
And the display alarm module is used for displaying the device states corresponding to the detected six signals and judging whether the corresponding fault device can be automatically regulated and controlled.
Fig. 3 is a detailed flow diagram of a quality monitoring method for a flexible sensor inkjet printing process according to the present invention, and as shown in fig. 3, the method includes the steps of flexible sensor model selection, six parameter settings, vibration signal acquisition by a displacement sensor, fusion collision processing, data processing, optical signal acquisition by a CCD image sensor, fusion collision processing, image processing, disorder coupling analysis, database comparison, process control, abnormal signal display and alarm, etc.
Firstly, selecting the model of the flexible sensor through a knowledge base module, and determining corresponding six parameters in the printing process as comparison objects, wherein the six parameters comprise an ink sac pressure wave, an ink droplet ejection wave, an ink droplet impact wave, an ink droplet speed, the number of vertical columns of ink droplets and an ejection base distance.
And detecting the vibration of the key process points of the ink-jet printing through the multi-texture acoustics module, and transmitting the acquired and processed characteristic signal alpha to the disordered storage pool analysis module.
And respectively adsorbing the laser displacement sensor on the ink bag, the piezoelectric ceramic and the flexible substrate of the flexible sensor to preliminarily collect vibration signals of the three positions.
Then, the data processor is used for carrying out fusion conflict processing on the vibration signals, and in order to prevent interference of non-collected signals (vibration in a non-critical process and vibration of environmental noise) and hybrid interference or conflict between the collected signals, a signal optimization model is specially provided:
f1＝f1-a1f2-b1f3-c1f0
f2＝f2-a2f1-b2f3-c2f0；
f3＝f3-a3f1-b3f2-c3f0
wherein f is1As a vibration signal from the ink bag, f2As a vibration signal from the piezoelectric ceramic, f3F0 is the vibration signal from the flexible substrate, and the collected interference vibration signal; a1, a2, a3, b1, b2, b3, c1, c2 and c3 are all proportion parameters of interference signalsCounting; f1 is the vibration signal of the optimized ink bag, f2 is the vibration signal of the optimized piezoelectric ceramic, and f3 is the vibration signal of the optimized flexible substrate.
Further, 16384 frequency components are calculated by FFT (fast fourier transform), and 1-30 common eigenvectors are calculated by SMOFS-30-multi frequency analysis processing 16384 frequency components. The steps of extracting feature vectors by SMOFS-30-multi xpanded are as follows:
1) the frequency spectrum of the vibro-acoustic signal is used to form a vector. The spectrum of a good ink sac is defined as the following vector a ═ a1, a 2. The frequency spectrum of the failed ink bag is defined as vector B ═ B1, B2. A good piezoelectric ceramic spectrum is defined as vector C ═ C1, C2. The frequency spectrum of the failed piezoceramic is defined as vector D ═ D1, D2. The spectrum of a good flexible substrate is defined as the vector E ═ E1, E2.., E16384; the flexible substrate spectrum of the fault is defined as vector F ═ F1, F2.., F16384;
2) calculate the absolute value of the difference between the previously formed vectors: the formula I is a-b, a-c, a-d, a-e, a-f, b-c, b-d, b-e, b-f, c-d, c-e, c-f, d-e, d-f, e-f. a, B, C, D, E, F, C, D, E, and F.
3) The frequency component is selected using equation (1). Selecting a value greater than threshold ThrSelxThe frequency component of (a).
||FSA|-|FSB||＞ThrSelx (1)
Wherein ThrSelxIs the threshold for frequency component selection for the x-th iteration; | FSA|-|FSBI is the difference between the acoustic signal spectra of states A and B, FSAThe spectrum representing state A (16384 frequency components), FSBRepresenting the spectrum of state B (16384 frequency components). x ranges from 1 to 16384.
4) Calculating a variable (threshold for frequency component selection) ThrSel for the x-th iterationx. Variable ThrSelxIs represented by equation (2):
NoFCx≤30 (3)
wherein the variable NoFCxIs the number of selected frequency components for the x-th iteration (if the window length is equal to 32768, the calculated spectrum consists of 16384 frequency components, for the first iteration, NoFC016384, then the next NoFC is iteratively reduced0). If variable NoFCxAbove 30, the SMOFS-30MULTIEXPANDED calculation is shown in equation (2).
If variable NoFCxIf not greater than 30, the calculation will be interrupted. SMOFS-30-multi xpanded selects 1-30 frequency components. Number of iterations x and variable NoFCxThe value of (d) depends on the acoustic signal. Example 1 was analyzed with three acoustic signals of states A, B and C. The frequency components of the acoustic signals for states A and B (| a-B |) are calculated from SMOFS-30-MULTIXPANDED, with | a-B | being 200Hz, 300Hz, 240Hz, 260Hz, 280Hz, 300Hz, 320Hz, and 340 Hz. The frequency components of the acoustic signals for states A and C are calculated from SMOFS-30-MULTIEXPANDED (| a-C |) |, 210Hz, 230Hz, 250Hz, 270Hz, 290Hz, 310Hz, 330Hz, and 350 Hz. The frequency components of the acoustic signals for states B and C are calculated from SMOFS-30-MULTIEXPANDED (| B-C |) |, 215Hz, 300Hz, 305Hz, 230Hz, 235Hz, 240Hz, 245Hz, and 250 Hz.
The acoustic signals of states A, B and C have no common frequency components. The frequency components 300Hz, 230Hz, 240Hz and 250Hz are found 2 times. Using the found frequency components is a good way, but in order to solve the problem of more acoustic signals to be considered, a parameter called TCFC-multiple (threshold value multiple x d extension of common frequency components) was introduced to analyze a larger number of acoustic signals to be considered.
5) The parameter TCFC-MULTII is set. It is expressed as: TCFC-MULTI ═ (number of common frequency components needed for the training set under consideration)/(number of calculated differences). The parameter TCFC-MULTI is crucial for the selection of the final common frequency component.
Considering example 2, there are 4 training sets each having 6 acoustic signals, and the acoustic signals of the 4 training sets are (FA1, FB1, FC1, FD1, FE1, FF1), (FA2, FB2, FC2, FD2, FE2, FF2), (FA3, FB3, FC3, FD3, FE3, FF3), and (FA4, FB 63 4, FC4, FD4, FE4, FF 4). SMOFS-30-MULTIEXPANDED selects a frequency component for each difference in a training set: (| FA1-FB1|), (| FA1-FC1|), (| FA1-FD1|), (| FA1-FE1|), (| FA1-FF1|), (| FB1-FC1|), (| FB1-FD1|), (| FB1-FE1|), (| FB 1-1 |), (| FC1-FD1|), (|) FC1-FE1|), (|) 1-FF1|), (| FC1-FE1|), (| FA1-FF1|), | FE 1-FE1|), (| FA1-FF1|), (|), (| FA 1-1 |), (| FB 1-1 |), (| 1-1 |), (| FA 1). Differences between the frequency spectra of acoustic signals (vibration signals) FA1, FB1, FC1, FD1, fe1.. FA4, FB4, FC4, and FD4, FE 4-6 sound wave states (a, B, C, D, E). If the parameter TCFC-multiple 10/50 is 0.20, SMOFS-30-multiple xpanded selects the frequency component that is found in 10 differences (maximum number of 50). For example, frequency component 100Hz is found 10 times, frequency component 150Hz is found 20 times, frequency component 200Hz is found 25 times, SMOFS-30-multi xpanded selects 100Hz, 150Hz, and 200Hz (TCFCMULTI-10/40-0.25). If the parameter TCFC-multiple 18/50 is 0.36, SMOFS-30-multiple xpanded selects the frequency components 150Hz and 200 Hz. If the parameter TCFC-multiple 30/50 is 0.60, SMOFS-30-multiple xpanded selects 0 (TCFC-multiple should be set again).
6) 1-30 common frequency components are found.
7) The final feature vector (1-30 frequency components found) is formed.
And further, judging whether the fault occurs and the fault type through a minimum distance classifier. The minimum distance classifier trains the feature vectors, which are then used in the prediction phase. The minimum distance classifier searches for the most similar training and test vectors. Next, it assigns the test vector to the nearest class. The classifier traverses the entire test set and calculates d (similarity distance) between the test feature vector and each training feature vector. Finally, the test vector is assigned to the class with the closest training vector.
Training the characteristic vectors formed by the six sound wave signals by using a minimum distance classifier, creating a model, applying the model to the three detected sound wave signals for identification, judging whether a device fails in advance, if so, forming a characteristic signal corresponding to the failed signal, and transmitting the characteristic signal to a disordered storage pool analysis module for further analysis.
Meanwhile, the visual monitoring module comprises fusion conflict processing, image analysis and fault pre-classification of image signals and is used for judging whether the ink droplet forms are uniform and consistent or not and whether satellite droplets are formed or not, transmitting data to the fault pre-classification and judging whether a device fails or not in advance.
Specifically, a CCD image sensor is arranged at a corresponding position, the CCD image sensor is an MVC13000F-M00CCD image sensor, a fixed ruler is arranged on a workbench at an initial position to serve as a standard component, the size of the CCD is calibrated, and the camera pixel equivalent d is calibrated. Firstly, converting optical signals of the distance between an ink droplet on a workbench, a spray head and a substrate into analog current signals by a CCD camera; then the analog current signal is converted into a digital image signal by an image acquisition card and transmitted to an image processor.
Furthermore, in order to prevent the fusion conflict of the acquired image signals and enhance the corresponding relation, an image sequencing calculation model is particularly provided:
Di＝Di+T
T＝t1+t2+t3
m＝kT
wherein t1, t2 and t3 are the corresponding positions of the speed of collecting ink drop, the number of columns of ink drop and the height of spraying baseAt the corresponding moment, the three moments are taken as a period T to circularly acquire images, so that the orderliness is provided for subsequent calculation; m is the total time collected; diThe ink drop distance corresponding to the t1 moment, V is the average ink drop speed of 1/3m collecting time period; k is an integer; n is the average ink drop column number of the acquisition time period of 1/3 m; h is the average spray base height for the time period collected at 1/3 m.
Further, the period T is divided into 100 segments, and six corresponding characteristic arrays are formed, that is, the ink drop velocity array at the normal voltage is defined as O ═ O1, O2.., O100; the ink drop velocity array at abnormal voltage is defined as P ═ P1, P2.., P100; the number array of columns of ink droplets with non-blocked orifices is defined as Q ═ Q1, Q2,. multidot.q 100; the number array of columns of ink droplets with blocked orifices is defined as R ═ R1, R2.., R100; the normal-height nozzle-substrate spacing array is defined as S ═ S1, S2,., S100; the nozzle-substrate spacing array of abnormal height is defined as Z ═ Z1, Z2,.., Z100; 4 training sets were created, each with six arrays (O1, P1, Q1, R1, S1, Z1), (O2, P2, Q2, R2, S2, Z2), (O3, P3, Q3, R3, S3, Z3), (O4, P4, Q4, R4, S4, Z4).
The training set is transmitted to a minimum distance classifier for training to create a model, which is then used in the prediction phase. Next, it assigns the test array to the nearest class. The classifier traverses the entire test set and calculates d (similarity distance) between the test feature array and each training feature array. Finally, the test array is assigned to the class with the closest training array.
And training the feature arrays formed by the six states by using a minimum distance classifier, creating a model, applying the model to the three detected arrays for identification, judging whether a device fails in advance, if so, forming a feature signal by the array corresponding to the failure, and transmitting the feature signal to the unordered reserve pool analysis module for further analysis.
Furthermore, through the unordered reserve pool analysis module, firstly, coupling analysis is carried out on the alpha of the characteristic signal of the multi-texture acoustics module and the beta of the characteristic signal of the visual monitoring module to form characteristic curve data of multi-coupling, then, comparison analysis is carried out on the characteristic curve data and the characteristic curve in the database, when the average difference value of the characteristic curve data and the characteristic curve data or the difference value in a unit period exceeds a set threshold value, the display alarm module is activated, the workload of comparison analysis calculation is reduced, the printing quality judgment efficiency is improved, and the power consumption is reduced.
Specifically, the unordered reservoir analysis module is composed of three different groups of neurons, namely an input layer, a reservoir layer and an output layer. The reserve layer may perform a non-linear time-spreading mapping on the input signal. The input signal represents g ═ { α, β }, and the state of the energy storage in the reserve layer is represented as r ═ { r ═ r }1，r2,,...,rHWherein r in r1、r2Equal represents the state parameter and the output is denoted v ═ { v1, v 2.., vo }, respectively, where v1, v2 represent the component data in the characteristic curve v. The following formula is used to control the status update of the network:
r＝n(W1g+W2r),
v＝W3[r；g],
wherein W1∈RH×S,W2∈RH×HAnd W3∈RO×(H+S)Represents the input layer weight, reserve layer weight and output layer weight, respectively, p represents the input data set of the activation function, vector [ r; g]Concatenating the reserve layer and the input state always refers to the extended state vector. A sigmoid function is selected in the equation as the activation function n ().
Separately determining and optimizing W in an unordered pool analysis module1，W2And W3. To ensure the validity of the principle, the reservoir weight W must be determined2It has two properties: echo state and separability. The echo state attribute means that the reservoir will asymptotically purge all information from the initial conditions. W2The spectral radius of (a), i.e. the absolute value of its maximum eigenvalue, is typically used to ensure this property. According to experience, ifA spectral radius less than 1 may confer an echo state attribute. Separability attributes require different input states to be generated from different inputs. Generally, to ensure this property, we should take W from the standard normal distribution sample2A sparse matrix is initialized and the container needs to contain a sufficient number of neurons. For an input weight vector W1Also using W2The same distribution of which initializes it. The difference between them is that it needs to be in W1In the dense connection, and in W2Is not required. By determining W1And W2After obtaining a stable reservoir, the weight vector W can be optimized by solving the following optimization problem3：
W3＝argminMSE(vexpect,v),
Wherein v isexpectRepresenting the desired fault signature sequence, v representing a predicted training data set, N representing the total number of samples, and MSE () representing the mean square error calculation.
Furthermore, the results of the fault pre-classification and the results of the unordered reserve pool analysis module are displayed through the display alarm module, whether the corresponding fault device can be automatically regulated and controlled is judged according to the abnormal signals, and if the corresponding fault device can be automatically regulated and controlled, the process regulation and control module is used for regulating and controlling; if not, a warning is sent out and the abnormal signal is displayed in red, so that the staff can adjust the corresponding device in time, and the loss of materials and time is reduced.
In summary, the invention provides a process quality monitoring method for flexible sensor inkjet printing, which selects the model of a sensor and determines parameters through a knowledge base module; through many ripples acoustic science module for detect the vibrations of inkjet printing key technology point, will gather three kinds of vibration waveforms in the inkjet printing process: the ink sac pressure wave, the ink droplet ejection wave and the ink droplet substrate impact wave are transmitted and processed to the disordered storage pool analysis module; the visual monitoring module is used for detecting the speed, the longitudinal number and the spraying base spacing of ink drops in the ink-jet printing process, and transmitting the ink drops to the unordered reserve pool analysis module after the ink drops are processed by collecting images; coupling analysis is carried out on alpha of the characteristic signal of the multi-texture acoustics module and beta of the characteristic signal of the visual monitoring module through a disorder reserve pool analysis module to form characteristic curve data of multivariate coupling, then comparison analysis is carried out on the characteristic curve data and the characteristic curve data in a database, and an analysis result is transmitted to a display alarm module, so that the workload of comparison analysis calculation is reduced, the printing quality judgment efficiency is improved, and the power consumption is reduced; the six detection result signals are displayed on the display alarm module, whether the corresponding fault device can be automatically regulated and controlled is judged according to the abnormal signals, and if the corresponding fault device can be automatically regulated and controlled, the process regulation and control module is used for regulating and controlling; if not, a warning is sent out and the abnormal signal is displayed in red, so that the staff can adjust the corresponding device in time, and the loss of materials and time is reduced.
The embodiments in the present description are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments are referred to each other.
The principles and embodiments of the present invention have been described herein using specific examples, which are provided only to help understand the method and the core concept of the present invention; meanwhile, for a person skilled in the art, according to the idea of the present invention, the specific embodiments and the application range may be changed. In view of the above, the present disclosure should not be construed as limiting the invention.
Claims (7)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110320664.6A CN113071218B (en) | 2021-03-25 | 2021-03-25 | Quality monitoring method for flexible sensor-oriented ink-jet printing process |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110320664.6A CN113071218B (en) | 2021-03-25 | 2021-03-25 | Quality monitoring method for flexible sensor-oriented ink-jet printing process |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113071218A CN113071218A (en) | 2021-07-06 |
| CN113071218B true CN113071218B (en) | 2022-03-18 |
Family
ID=76611610
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110320664.6A Active CN113071218B (en) | 2021-03-25 | 2021-03-25 | Quality monitoring method for flexible sensor-oriented ink-jet printing process |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113071218B (en) |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP5560253B2 (en) * | 2011-09-30 | 2014-07-23 | 富士フイルム株式会社 | Inkjet recording apparatus and method, and abnormal nozzle detection method |
| WO2013173495A1 (en) * | 2012-05-15 | 2013-11-21 | Corinthian Ophthalmic, Inc. | Ejector devices, methods, drivers, and circuits therefor |
| KR20140075118A (en) * | 2012-12-10 | 2014-06-19 | 삼성전기주식회사 | Ink circulation apparatus, inkjet printer having the same |
| JP7344704B2 (en) * | 2018-08-07 | 2023-09-14 | キヤノン株式会社 | Recording device and method for determining its nozzle discharge state |
| DE102019207279A1 (en) * | 2019-05-18 | 2020-11-19 | Robert Bosch Gmbh | Data fused sensor system |
| CN110293762B (en) * | 2019-06-20 | 2021-02-26 | 森大（深圳）技术有限公司 | Method and device for dynamically detecting abnormal nozzle in printing process and storage medium |
-
2021
- 2021-03-25 CN CN202110320664.6A patent/CN113071218B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN113071218A (en) | 2021-07-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111942022B (en) | Information processing apparatus, printing apparatus, learning apparatus, and information processing method | |
| CN101670705B (en) | Liquid ejecting apparatus and ejection inspecting method | |
| CN111942023B (en) | Information processing apparatus, printing apparatus, learning apparatus, and information processing method | |
| EP2842752B1 (en) | Ink jet print head health detection | |
| JP2012018066A (en) | Device for inspecting abnormality | |
| Hickman et al. | Application of smart structures to aircraft health monitoring | |
| EP3631592A1 (en) | Method and system for determining a remaining service life of a process device through which fluid flows | |
| CN113071218B (en) | Quality monitoring method for flexible sensor-oriented ink-jet printing process | |
| EP2328757B1 (en) | Jet performance | |
| CN110370648A (en) | The identification of multichannel piezoelectric type 3D printing faulty nozzle and condition monitoring system and method | |
| CN117400633A (en) | Spray nozzle state evaluation method and device based on spray hole sampling detection | |
| CN109087882A (en) | A preparation method of a flexible electronic device and its product and system | |
| Topac et al. | Hybrid models for situational awareness of an aerial vehicle from multimodal sensing | |
| US11679589B2 (en) | Information processing system, learning device, and information processing method | |
| KR20240071945A (en) | Deep learning based inkjet droplet object detection and jetting characterization apparatus and the method thereof | |
| JP2010036388A (en) | Droplet amount measuring method and droplet discharging system employing thereof | |
| Chen et al. | Application of wavelet analysis and decision tree in UTDR data for diagnosis of membrane filtration | |
| Topac et al. | Flight State Monitoring of a Morphing Wing using Multimodal Distributed Sensors-informed Prediction Models | |
| JP2001209628A (en) | Pattern matching method | |
| Chivate et al. | Investigating Jet Stability in Inkjet Printing Through a Novel Sensing Modality | |
| CN210113090U (en) | Parallel arm three-dimensional printer based on fault diagnosis system | |
| van Peijpe | Fault Detection and Isolation for High-End Industrial Printers | |
| Ye | Identification and feedforward control of a drop-on-demand inkjet printhead | |
| JP6983174B2 (en) | How to cancel electric crosstalk on the printhead | |
| Hacking | Classification of jetting behavior based on self-sensing piezo actuators |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20240626 Address after: 230000 Anhui Hefei high tech Zone Innovation Industrial Park two phase J2 District C block 18 floor. Patentee after: HEFEI LONGTUTEM INFORMATION TECHNOLOGY Co.,Ltd. Country or region after: China Address before: 100193 No. 2 Old Summer Palace West Road, Beijing, Haidian District Patentee before: CHINA AGRICULTURAL University Country or region before: China |
|
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20240823 Address after: 528403 Zone A, 1st Floor, No. 130 Qiwan Road, East District, Zhongshan City, Guangdong Province, China Patentee after: Zhongshan Yinyi Express Digital Graphic and Text Co.,Ltd. Country or region after: China Address before: 230000 Anhui Hefei high tech Zone Innovation Industrial Park two phase J2 District C block 18 floor. Patentee before: HEFEI LONGTUTEM INFORMATION TECHNOLOGY Co.,Ltd. Country or region before: China |