CN116755648A - Printer abnormal state diagnosis method and system - Google Patents
Printer abnormal state diagnosis method and system Download PDFInfo
- Publication number
- CN116755648A CN116755648A CN202310812700.XA CN202310812700A CN116755648A CN 116755648 A CN116755648 A CN 116755648A CN 202310812700 A CN202310812700 A CN 202310812700A CN 116755648 A CN116755648 A CN 116755648A
- Authority
- CN
- China
- Prior art keywords
- signal
- printer
- humidity
- audio signal
- vibration
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000003745 diagnosis Methods 0.000 title claims abstract description 129
- 230000002159 abnormal effect Effects 0.000 title claims abstract description 88
- 238000000034 method Methods 0.000 title claims abstract description 51
- 230000005236 sound signal Effects 0.000 claims abstract description 143
- 230000004927 fusion Effects 0.000 claims abstract description 79
- 239000013598 vector Substances 0.000 claims abstract description 58
- 238000012706 support-vector machine Methods 0.000 claims abstract description 32
- 238000012549 training Methods 0.000 claims abstract description 16
- 230000009466 transformation Effects 0.000 claims abstract description 11
- 238000012544 monitoring process Methods 0.000 claims abstract description 4
- 238000010606 normalization Methods 0.000 claims description 18
- 238000012545 processing Methods 0.000 claims description 17
- 230000008569 process Effects 0.000 claims description 15
- 230000006870 function Effects 0.000 claims description 11
- 238000000354 decomposition reaction Methods 0.000 claims description 10
- 238000012546 transfer Methods 0.000 claims description 10
- 230000003044 adaptive effect Effects 0.000 claims description 9
- 238000004364 calculation method Methods 0.000 claims description 5
- 230000008602 contraction Effects 0.000 claims description 5
- 238000012937 correction Methods 0.000 claims description 4
- 230000002028 premature Effects 0.000 claims description 4
- 230000001174 ascending effect Effects 0.000 claims description 3
- 238000010276 construction Methods 0.000 claims description 3
- 238000011156 evaluation Methods 0.000 claims description 3
- 238000012795 verification Methods 0.000 claims description 3
- 238000005469 granulation Methods 0.000 claims description 2
- 230000003179 granulation Effects 0.000 claims description 2
- 238000005259 measurement Methods 0.000 claims description 2
- 238000012423 maintenance Methods 0.000 description 8
- 230000005856 abnormality Effects 0.000 description 7
- 238000004458 analytical method Methods 0.000 description 6
- 230000008859 change Effects 0.000 description 5
- 238000012790 confirmation Methods 0.000 description 5
- 230000000694 effects Effects 0.000 description 5
- 230000009286 beneficial effect Effects 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000004891 communication Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000007639 printing Methods 0.000 description 2
- 230000007547 defect Effects 0.000 description 1
- 239000006260 foam Substances 0.000 description 1
- 230000004807 localization Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000013519 translation Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/12—Digital output to print unit, e.g. line printer, chain printer
- G06F3/1201—Dedicated interfaces to print systems
- G06F3/1202—Dedicated interfaces to print systems specifically adapted to achieve a particular effect
- G06F3/121—Facilitating exception or error detection and recovery, e.g. fault, media or consumables depleted
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/2433—Single-class perspective, e.g. one-against-all classification; Novelty detection; Outlier detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
- G06F18/253—Fusion techniques of extracted features
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/12—Digital output to print unit, e.g. line printer, chain printer
- G06F3/1201—Dedicated interfaces to print systems
- G06F3/1223—Dedicated interfaces to print systems specifically adapted to use a particular technique
- G06F3/1229—Printer resources management or printer maintenance, e.g. device status, power levels
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/08—Feature extraction
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/12—Classification; Matching
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Human Computer Interaction (AREA)
- Control Or Security For Electrophotography (AREA)
Abstract
The invention discloses a printer abnormal state diagnosis method and system, belonging to the technical field of laser printers, wherein the method comprises the following steps: acquiring a paper image and identifying the image state of the paper image; performing preliminary diagnosis according to the image state of the paper output image; setting the weight of the signal during the subsequent feature fusion according to the initial diagnosis result; acquiring a status signal of a printer; normalizing the temperature signal, the humidity signal, the audio signal and the vibration signal; decomposing and denoising the signals by adopting wavelet packet transformation; summarizing the permutation entropy values under the optimal scale factors to obtain characteristic values, and constructing characteristic vectors; feature fusion is carried out according to the weight of each signal; inputting the feature fusion vector into a support vector machine for training; monitoring the abnormal state of the printer in real time through a trained support vector machine; and comparing the actual state with the diagnosis state of the printer, and correcting the weights of the temperature signal, the humidity signal, the audio signal and the vibration signal.
Description
Technical Field
The invention belongs to the technical field of laser printers, and particularly relates to a printer abnormal state diagnosis method and system.
Background
The printer is one of important office equipment indispensable in daily work and study, can print data on a terminal with a communication relationship on related media, such as paper, foam, text and the like, and provides great convenience for real life.
During actual use of the printer, various print failures are often unavoidable. If the faults cannot be removed in time, normal printing office work is affected, and inconvenience is brought to practical application.
When facing printer trouble at every turn, because the user is not professional printer maintainer, often adopts the manual disassembly's mode to examine, and the accuracy of finding out printer trouble is lower, even probably leads to the secondary damage of printer at violent in-process of dismantling. In real life, the operator needs to wait for the door-to-door confirmation of the maintenance personnel with abundant experience, a great deal of time is wasted in the waiting process, and even the maintenance personnel must check on site, so that the diagnosis efficiency is low.
Disclosure of Invention
The invention provides a printer abnormal state diagnosis method and system, which aims to solve the technical problems that in the prior art, when a printer is in fault, the accuracy is low when a user self-disassembles and checks the printer, even secondary damage of the printer is possibly caused in the violent disassembling process, and when waiting for the door-to-door confirmation of a maintenance person with abundant experience, a great deal of time is wasted.
First aspect
The invention provides a printer abnormal state diagnosis method, which comprises the following steps:
s101: acquiring a paper output condition of a printer, acquiring a paper output image under the condition that the printer can output paper, and identifying an image state of the paper image;
s102: performing preliminary diagnosis according to the image state of the paper output image;
s103: according to the preliminary diagnosis result, the weights of a temperature signal, a humidity signal, an audio signal and a vibration signal during the subsequent feature fusion are set;
s104: acquiring a status signal of the printer, wherein the status signal comprises a temperature signal, a humidity signal, an audio signal and a vibration signal;
s105: normalizing the temperature signal, the humidity signal, the audio signal and the vibration signal;
s106: decomposing and denoising temperature signals, humidity signals, audio signals and vibration signals by adopting wavelet packet transformation, and decomposing the temperature signals, the humidity signals, the audio signals and the vibration signals into a plurality of sub-band sequences;
s107: calculating arrangement entropy values of each frequency band under different scale factors, and respectively selecting optimal scale factors of a temperature signal, a humidity signal, an audio signal and a vibration signal;
s108: summarizing the arrangement entropy values of the temperature signal, the humidity signal, the audio signal and the vibration signal under the condition that the optimal scale factors are obtained respectively, obtaining the characteristic values of the temperature signal, the humidity signal, the audio signal and the vibration signal, and constructing a characteristic vector;
S109: according to the weights of the temperature signal, the humidity signal, the audio signal and the vibration signal, feature fusion is carried out on feature vectors of the temperature signal, the humidity signal, the audio signal and the vibration signal, and feature fusion vectors are obtained;
s110: adding a classification label for the feature fusion vector, inputting the feature fusion vector into a support vector machine, outputting a fault diagnosis result, and training the support vector machine by comparing whether the fault diagnosis result is consistent with the label;
s111: the abnormal state of the printer is monitored in real time through the trained support vector machine, an alarm is sent out when the printer is judged to be abnormal, and a specific fault type is output;
s112: and comparing the actual state with the diagnosis state of the printer, and correcting the weights of the temperature signal, the humidity signal, the audio signal and the vibration signal.
Second aspect
The invention provides a printer abnormal state diagnosis system, comprising:
the first acquisition module is used for acquiring the paper output condition of the printer, acquiring a paper image under the condition that the printer can output paper, and identifying the image state of the paper image;
the preliminary diagnosis module is used for performing preliminary diagnosis according to the image state of the paper output image;
The setting module is used for setting weights of a temperature signal, a humidity signal, an audio signal and a vibration signal during subsequent feature fusion according to the initial diagnosis result;
the second acquisition module is used for acquiring a state signal of the printer, wherein the state signal comprises a temperature signal, a humidity signal, an audio signal and a vibration signal;
the normalization module is used for performing normalization processing on the temperature signal, the humidity signal, the audio signal and the vibration signal;
the decomposition module is used for decomposing and denoising the temperature signal, the humidity signal, the audio signal and the vibration signal by adopting wavelet packet transformation, and decomposing the temperature signal, the humidity signal, the audio signal and the vibration signal into a plurality of sub-band sequences;
the calculation module is used for calculating the arrangement entropy values of each frequency band under different scale factors and respectively selecting the optimal scale factors of the temperature signal, the humidity signal, the audio signal and the vibration signal;
the construction module is used for summarizing the arrangement entropy values of the temperature signal, the humidity signal, the audio signal and the vibration signal under the condition that the optimal scale factors are obtained respectively, obtaining the characteristic values of the temperature signal, the humidity signal, the audio signal and the vibration signal, and constructing the characteristic vector;
the fusion module is used for carrying out feature fusion on the feature vectors of the temperature signal, the humidity signal, the audio signal and the vibration signal according to the weights of the temperature signal, the humidity signal, the audio signal and the vibration signal to obtain feature fusion vectors;
The training module is used for adding a classification label for the feature fusion vector, inputting the feature fusion vector into the support vector machine, outputting a fault diagnosis result, and training the support vector machine by comparing whether the fault diagnosis result is consistent with the label or not;
the diagnosis module is used for monitoring the abnormal state of the printer in real time through the trained support vector machine, sending out an alarm when judging that the printer is abnormal, and outputting a specific fault type;
and the correction module is used for comparing the actual state with the diagnosis state of the printer and correcting the weights of the temperature signal, the humidity signal, the audio signal and the vibration signal.
Compared with the prior art, the invention has at least the following beneficial technical effects:
(1) In the invention, the real-time state of the printer is automatically diagnosed through the support vector machine, an alarm is sent out when the printer is judged to be abnormal, a specific fault type is output, the user does not need to disassemble and check the printer by itself, a great deal of time is not wasted to wait for the entrance confirmation of maintenance personnel with abundant experience, the diagnosis accuracy is high, and the time and the labor are saved.
(2) When the printer is in fault, the conditions of temperature rise, abnormal humidity, abnormal sound emission, severe vibration and the like of the layout device are often accompanied, and in the invention, the state change possibly caused by the printer in fault is comprehensively considered, the temperature signal, the humidity signal, the audio signal and the vibration signal are subjected to multi-mode fusion, and the final diagnosis result is given after the temperature signal, the humidity signal, the audio signal and the vibration signal are comprehensively considered, so that the accuracy of the fault diagnosis result is further improved.
(3) In the invention, wavelet packets are adopted to decompose temperature signals, humidity signals, audio signals and vibration signals, noise is removed, detail characteristics of the temperature signals, the humidity signals, the audio signals and the vibration signals are deeply excavated, permutation entropy values under different scale factors are calculated, the permutation entropy values under the optimal scale factors are summarized to be used as characteristic values, the degree of distinction among different fault types is improved, and the accuracy of fault diagnosis results is further improved.
(4) For printer faults, observing the paper-out image can obtain very important conclusions, for example, when the paper-out image is completely black, the fault is often caused by hardware circuits, so that the fault range can be preliminarily determined by analyzing the paper-out image in advance. Aiming at the faults and problems of the types, the weight of a specific temperature signal, a specific humidity signal, a specific audio signal and a specific vibration signal is selected in the diagnosis process, so that the distinction degree among the fault types is further enlarged in the diagnosis process after the feature fusion, and the accuracy of the fault diagnosis result is improved.
(5) In the present invention, the weights of the temperature signal, the humidity signal, the audio signal, and the vibration signal are corrected by comparing the actual state and the diagnostic state of the printer to balance the loss caused by diagnosing the normal as abnormal with the loss caused by diagnosing the abnormal as normal.
Drawings
The above features, technical features, advantages and implementation of the present invention will be further described in the following description of preferred embodiments with reference to the accompanying drawings in a clear and easily understood manner.
FIG. 1 is a flow chart of a method for diagnosing abnormal printer conditions according to the present invention;
FIG. 2 is a schematic diagram of a laser printer according to the present invention;
fig. 3 is a schematic diagram of a printer abnormality diagnosis system according to the present invention.
Detailed Description
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following description will explain the specific embodiments of the present invention with reference to the accompanying drawings. It is evident that the drawings in the following description are only examples of the invention, from which other drawings and other embodiments can be obtained by a person skilled in the art without inventive effort.
For simplicity of the drawing, only the parts relevant to the invention are schematically shown in each drawing, and they do not represent the actual structure thereof as a product. Additionally, in order to simplify the drawing for ease of understanding, components having the same structure or function in some of the drawings are shown schematically with only one of them, or only one of them is labeled. Herein, "a" means not only "only this one" but also "more than one" case.
It should be further understood that the term "and/or" as used in the present specification and the appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes such combinations.
In this context, it should be noted that the terms "mounted," "connected," and "connected" are to be construed broadly, and may be, for example, fixedly connected, detachably connected, or integrally connected, unless explicitly stated or limited otherwise; can be mechanically or electrically connected; can be directly connected or indirectly connected through an intermediate medium, and can be communication between two elements. The specific meaning of the above terms in the present invention will be understood in specific cases by those of ordinary skill in the art.
In addition, in the description of the present invention, the terms "first," "second," and the like are used merely to distinguish between descriptions and are not to be construed as indicating or implying relative importance.
Example 1
In one embodiment, referring to fig. 1 of the specification, the present invention provides a flow chart of a method for diagnosing abnormal status of a printer.
Referring to fig. 2 of the specification, the invention provides a schematic structure of a laser printer.
A laser printer is a print output device that combines laser scanning technology and electrophotographic technology. The basic working principle is that binary data information transmitted by a computer is converted into video signals through a video controller, then the video signals are converted into laser driving signals through a video interface/control system, then a laser scanning system generates laser beams carrying character information, and finally an electronic photographing system images and transfers the laser beams onto paper.
As shown in fig. 2, the laser printer mainly includes components such as a fixing roller, a transfer roller, a charging roller, a feeding roller, a developing roller, a nip roller, a laser, a photosensitive drum, a reflecting mirror, and an ink cartridge, and it is understood that damage to any one of the critical components during daily use may cause abnormal conditions of the printer.
Further, the basic principle of the laser printer is already the prior art, and the invention is not repeated.
The invention provides a printer abnormal state diagnosis method, which comprises the following steps:
s101: the paper output condition of the printer is acquired, and when the printer is capable of outputting paper, a paper image is acquired, and the image state of the paper image is recognized.
The paper output condition refers to whether paper can be printed out smoothly. If the printer cannot print out paper, it may be due to paper jams or hardware damage, etc. If the printer can print out the paper smoothly, but the image on the paper is abnormal, some preliminary judgment can be obtained through the paper image.
S102: and performing preliminary diagnosis according to the image state of the paper output image.
For printer faults, observing the paper-out image can obtain very important conclusions, for example, when the paper-out image is completely black, the fault is often caused by hardware circuits, so that the fault range can be preliminarily determined by analyzing the paper-out image in advance. To facilitate subsequent further analysis.
In one possible implementation, S102 specifically includes substeps S1021 to S1024:
s1021: under the condition that the paper output image is completely black, the hardware circuit of the printer is primarily diagnosed to have faults.
S1022: in the case where the overall gradation of the paper-output image is low or the partial gradation is unbalanced, the failure of the ink cartridge, the reflecting mirror, or the photoconductor of the printer is diagnosed preliminarily.
S1023: in the case where there are regular spots in the paper-out image, the failure of the roll of the printer is preliminarily diagnosed.
S1024: under the condition that vertical white stripes exist in the paper-out image, the failure of a transfer electrode or a charging electrode of the printer is diagnosed preliminarily.
It should be noted that the problems in the paper output image are not only the four cases of the present invention, but the present invention is merely an example of the problems in the paper output image and the preliminary diagnosis result, and does not limit the protection scope of the present invention.
S103: and setting weights of the temperature signal, the humidity signal, the audio signal and the vibration signal during subsequent feature fusion according to the preliminary diagnosis result.
It should be noted that, the preliminary diagnosis result gives a general direction, and for a common fault corresponding to a specific situation, the method has a direct characteristic, and needs to be classified and diagnosed more accurately by a subsequent support vector machine, so that weights need to be given to a temperature signal, a humidity signal, an audio signal and a vibration signal.
For example, in the case of a completely black paper-out image, the hardware circuit of the printer is primarily diagnosed as having a fault, and abnormal sound and vibration are critical features for the hardware circuit problem, and at this time, the weights of the audio signal and the vibration signal can be set higher than the weights of the temperature signal and the humidity signal.
For another example, when there is a vertical white stripe in the paper output image, the transfer electrode or the charging electrode of the printer is primarily diagnosed to have a fault, and when there is a fault on the transfer electrode or the charging electrode, the voltage or the circuit is often abnormal greatly, so that local temperature rise can be caused, and at this time, the temperature is a critical feature, and the weight of the temperature signal can be set to be higher than that of the audio signal, the vibration signal and the humidity signal.
The method is characterized in that the method is used for diagnosing faults and problems, and the method is used for selecting the weights of specific temperature signals, humidity signals, audio signals and vibration signals in the diagnosis process, so that the distinction degree among fault types can be further enlarged in the diagnosis process after feature fusion, and the accuracy of fault diagnosis results is improved.
Specifically, S103 includes:
s1031: under the condition that the hardware circuit of the printer is diagnosed to be faulty preliminarily, setting the weights of the temperature signal, the humidity signal, the audio signal and the vibration signal in the subsequent feature fusion as a first weight combination;
s1032: under the condition that the ink box of the preliminary diagnosis printer has faults, setting the weights of the temperature signal, the humidity signal, the audio signal and the vibration signal in the subsequent feature fusion as a second weight combination;
s1033: under the condition that the reflector of the preliminary diagnosis printer has faults, setting the weights of the temperature signal, the humidity signal, the audio signal and the vibration signal in the subsequent feature fusion as a third weight combination;
s1034: under the condition that the light guide body of the preliminary diagnosis printer has faults, setting the weights of the temperature signal, the humidity signal, the audio signal and the vibration signal in the subsequent feature fusion as a fourth weight combination;
S1035: under the condition that the fault exists in the roller of the preliminary diagnosis printer, setting the weights of the temperature signal, the humidity signal, the audio signal and the vibration signal in the subsequent feature fusion as a fifth weight combination;
s1036: in the case where the failure of the transfer electrode or the charging electrode of the printer is primarily diagnosed, the weights of the temperature signal, the humidity signal, the audio signal, and the vibration signal at the time of the subsequent feature fusion are set to be the sixth weight combination.
S104: status signals of the printer are acquired, wherein the status signals comprise a temperature signal, a humidity signal, an audio signal and a vibration signal.
The temperature signal can be collected by a temperature sensor, and the setting position of the temperature sensor can be a part of a device with higher damage rate. The humidity signal may be collected by a humidity sensor. The audio signal may be collected by an audio collection device such as a microphone. The vibration signal may be collected by a vibration sensor.
It should be noted that, the frequency consistency in acquiring the temperature signal, the humidity signal, the audio signal and the vibration signal should be maintained as much as possible to facilitate the subsequent analysis.
S105: and carrying out normalization processing on the temperature signal, the humidity signal, the audio signal and the vibration signal.
The temperature signal, the humidity signal, the audio signal and the vibration signal can be kept to be between 0 and 1 in amplitude by carrying out normalization processing on the temperature signal, the humidity signal, the audio signal and the vibration signal, so that subsequent feature fusion is facilitated.
Further, the normalization processing can improve the convergence speed of the whole algorithm and can also improve the accuracy of the algorithm.
In one possible implementation, S105 is specifically:
the temperature signal, the humidity signal, the audio signal and the vibration signal are normalized by adopting the following formula:
where f' (x) represents the signal after normalization, f (x) represents the signal before normalization, min [ f (x) ] represents the minimum value in f (x), and max [ f (x) ] represents the maximum value in f (x).
S106: the temperature signal, the humidity signal, the audio signal and the vibration signal are decomposed and denoised by wavelet packet transformation, and are decomposed into a plurality of sub-band sequences.
The wavelet packet transformation is a new transformation analysis method, inherits and develops the concept of short-time Fourier transformation localization, overcomes the defects that the window size does not change along with frequency and the like, can provide a time-frequency window which changes along with frequency, and is an ideal tool for carrying out signal time-frequency analysis and processing. The method is mainly characterized in that the characteristics of certain aspects of the problems can be fully highlighted through transformation, the local analysis of time (space) frequency can be realized, the multi-scale refinement of the signals (functions) is gradually carried out through telescopic translation operation, finally, the time subdivision at high frequency and the frequency subdivision at low frequency are finally achieved, the requirement of time-frequency signal analysis can be automatically met, and therefore, any details of the signals can be focused.
In the invention, wavelet packets are adopted to decompose temperature signals, humidity signals, audio signals and vibration signals, noise is removed, detail characteristics of the temperature signals, the humidity signals, the audio signals and the vibration signals are deeply excavated, permutation entropy values under different scale factors are calculated, the permutation entropy values under the optimal scale factors are summarized to be used as characteristic values, the degree of distinction among different fault types is improved, and the accuracy of fault diagnosis results is further improved.
In one possible embodiment, S106 specifically includes substeps S1061 to S1063:
s1061: for signals to be processedSelecting a group of orthogonal wavelet packet basis functions psi (t):
wherein h is k Representation ofG, g k Low-pass filter denoted psi (t), h k And g k Is a pair of conjugate orthogonal real coefficient filters, t represents a time variable, and k represents the number of wavelet packet decomposition layers.
Wherein the signal to be processedIs one of a temperature signal, a humidity signal, an audio signal, and a vibration signal.
S1062: recursively processing the formula and processing the signal to be processedDecomposition into quadrature wavelet packet signals { mu } n (t) } each orthogonal wavelet packet signal represents a frequency band:
wherein mu 2n (t) represents low-band coefficients, μ 2n+1 (t) represents a high-band coefficient.
S1063: the orthogonal wavelet packet signals with band coefficients exceeding the normal coefficient range are removed.
Specifically, the coefficient of the orthogonal wavelet packet signal whose band coefficient exceeds the normal coefficient range may be set to 0 to achieve the effect of removal.
The method comprises the steps of obtaining a normal coefficient range of a frequency band coefficient, wherein the frequency band coefficient is removed from orthogonal wavelet packet signals exceeding the normal coefficient range, so that interference signals can be eliminated, and the accuracy of a subsequent fault diagnosis result is improved.
S107: and calculating arrangement entropy values of each frequency band under different scale factors, and respectively selecting optimal scale factors of a temperature signal, a humidity signal, an audio signal and a vibration signal.
The permutation entropy is a measurement index for measuring the complexity of the time sequence, and compared with other traditional methods for processing nonlinear problems, the permutation entropy has strong stability, high calculation speed and strong anti-interference capability.
The permutation entropy is mainly related to three parameters of the original signal length N, the delay time tau and the embedding dimension m.
In the existing permutation entropy algorithm, only three parameters of an original signal length N, a delay time tau and an embedding dimension m are often considered, and entropy calculation is only carried out under a single scale, so that extracted features cannot fully reflect feature information. For each scale, the subsequence is decomposed into discrete symbol sequences, and the permutation entropy of each symbol sequence is then calculated. And then, selecting the optimal scale factors, and summarizing the arrangement entropy values under the optimal scale factors to be used as the characteristic values of the signals, so that the extracted characteristics can more accurately reflect the characteristic information.
In one possible embodiment, S107 specifically includes substeps S1071 to S1079:
s1071: the time series { x (1), x (2), …, x (N) } obtained by decomposing the wavelet packet is subjected to coarse graining treatment to obtain a time series y after coarse graining treatment (s) (j)：
Wherein s represents a scale factor,representation pair->Rounding, S represents the total length of the time series after the coarse-grain processing, and N represents the original signal length.
S1072: for y (s) (j) And (3) performing space reconstruction:
y (s) (i)＝{y (s) (i),y (s) (i+τ),…,y (s) (i+(m-1)τ)}
where τ represents the delay time and m represents the embedding dimension.
S1073: rearranging the ith individual reconstruction component Y (j) in ascending order:
y (s) [i+(j 1 -1)τ]≤y (s) [i+(j 2 -1)τ]≤…≤y (s) [i+(j m -1)τ]
wherein j is m Representing the position dimension.
S1074: acquiring a position dimension sequence pi corresponding to Y (j) i ：
π i ＝(j 1 ,j 2 ,…,j m )。
S1075: defining each symbol sequence pi i The probability of occurrence is:
wherein s (pi) i ) Represents pi i Number of occurrences.
S1076: calculating permutation entropy H of time sequence in different position dimensions p (m)：
S1077: for permutation entropy H p (m) carrying out normalization treatment:
s1078: and selecting a plurality of arrangement entropy values of delay time tau, embedding dimension m and sample length N under different scale factors s by a control variable method, and constructing an arrangement entropy curve.
The control variable method refers to a method of researching a variable parameter by changing only one of a plurality of parameters and keeping the other parameters unchanged.
S1079: and calculating the average distance between the arrangement entropy curves under different scale factors s, and taking the scale factor corresponding to the maximum average distance as the optimal scale factor.
S108: summarizing the arrangement entropy values of the temperature signal, the humidity signal, the audio signal and the vibration signal under the condition that the optimal scale factors are obtained respectively, obtaining the characteristic values of the temperature signal, the humidity signal, the audio signal and the vibration signal, and constructing the characteristic vector.
The arrangement entropy value under the optimal scale factors is summarized to be used as the characteristic value of the signal, so that the extracted characteristic can more accurately reflect the characteristic information, and the accuracy of subsequent abnormality diagnosis is further improved.
S109: and carrying out feature fusion on the feature vectors of the temperature signal, the humidity signal, the audio signal and the vibration signal according to the weights of the temperature signal, the humidity signal, the audio signal and the vibration signal to obtain feature fusion vectors.
In the invention, the state change possibly caused by the printer fault is comprehensively considered, the temperature signal, the humidity signal, the audio signal and the vibration signal are subjected to multi-mode fusion, and the final diagnosis result is given after the temperature signal, the humidity signal, the audio signal and the vibration signal are comprehensively considered, so that the accuracy of the fault diagnosis result is further improved.
In one possible implementation, S109 is specifically:
in the presence of the preliminary diagnosis result, assigning weights to the temperature signal, the humidity signal, the audio signal and the vibration signal determined in S103; and under the condition that the preliminary diagnosis result does not exist, adopting the initial weights preset by the temperature signal, the humidity signal, the audio signal and the vibration signal to assign the weights.
Feature fusion is carried out on feature vectors of the temperature signal, the humidity signal, the audio signal and the vibration signal, and a feature fusion vector q is obtained:
q＝α 1 ·p 1 +α 2 ·p 2 +α 3 ·p 3 +α 4 ·p 4
wherein p is 1 Feature vector, alpha, representing temperature signal 1 Weight, p, of the temperature signal 2 Feature vector, alpha, representing humidity signal 2 Weight, p, of the humidity signal 3 Feature vector, alpha, representing audio signal 3 Representing weights of audio signals, p 4 Characteristic vector, alpha, representing vibration signal 4 Representing the weight of the vibration signal.
It should be noted that, the absence of the preliminary diagnosis result mainly means that in S101, the printer cannot print out the paper smoothly, and the subsequent preliminary diagnosis in S102 cannot be performed, and at this time, the preliminary diagnosis result will not be present.
S110: adding a classification label for the feature fusion vector, inputting the feature fusion vector into a support vector machine, outputting a fault diagnosis result, and training the support vector machine by comparing whether the fault diagnosis result is consistent with the label or not.
The support vector machine is a generalized linear classifier for binary classification of data according to a supervised learning mode, and the decision boundary is an optimal hyperplane for solving a learning sample.
The core parameters of the support vector machine are penalty factor C and core parameter g.
In one possible implementation, S110 specifically includes sub-steps S1101 to S1104:
s1101: and adding a classification label for the feature fusion vector, and dividing a training set and a verification set.
S1102: and inputting the feature fusion vector q in the training set into a support vector machine.
S1103: the support vector machine constructs an objective function, selects an optimal hyperplane so as to maximize sample classification intervals among various types of samples, and the various types of data samples can be partitioned:
wherein N represents the number of samples, q i Feature fusion vector representing input, u i Classification label representing input, C represents penalty factor, ζ i Represents a relaxation variable, b represents an offset, K (qi. Q j ) Representing a kernel function representing qi and q j Similarity between the two, g represents the kernel parameter.
S1104: determining the optimal hyperplane as:
in one possible implementation, the optimal penalty factor C and the kernel parameter g are obtained to determine the optimal hyperplane by:
S110A: continuously changing the population searching speed v i And searching for a locationTo search for each penalty factor C and each kernel parameter g:
wherein w represents an inertial weight factor, r is a random number between 0 and 1,x i representing the current search position, e 1 And e 2 Represent learning factors, P i Representing global optimum parameters, G i Representing individual optimal parameters.
S110B: during the search, the inertial weight factor is set to an adaptive inertial weight factor to avoid premature convergence of the algorithm:
wherein h represents the target parameter value sought by the current algorithm, h min Represents the minimum value of the target parameter value, h avg Mean value of target parameter value, w max Represents the maximum value, w, of the adaptive inertial weight factor min Representing the minimum value of the adaptive inertial weight factor.
Where too large a w can lead to premature convergence of the algorithm and too small a w can lead to the algorithm falling into the layout optimally without reaching the desired finding effect. In the invention, the inertia weight factor is set as the self-adaptive inertia weight factor, so that the algorithm selects larger w in the early stage of iteration, and uses smaller w to perform finer layout search in the later stage of iteration, thereby improving the stability of searching the optimal penalty factor C and the kernel parameter g.
S110C: in the searching process, a contraction factor omega is introduced, and the searching range is continuously narrowed towards the global optimal position:
wherein E represents a balance factor.
The population diversity of the algorithm is maintained by looking at the contraction factor omega, so that the search range is prevented from being far from the optimal position, the search range is continuously reduced to the global optimal position, and the search accuracy is improved.
S110D: searching the penalty factor C and the core parameter g according to the updated search formula until the global optimal parameter and the individual optimal parameter related to the penalty factor C, and the global optimal parameter and the individual optimal parameter related to the core parameter g are searched.
S110E: and taking the penalty factors C at the positions meeting the global optimal parameters and the individual optimal parameters as optimal penalty factors, and taking the core parameters g at the positions meeting the global optimal parameters and the individual optimal parameters as optimal core parameters.
In the invention, the search range is in a better area by introducing the self-adaptive inertia weight factor in the process of searching the optimal penalty factor C and the kernel parameter g. And introducing a contraction factor omega, and continuously reducing the search range to the global optimal position so as to accelerate the search speed of the optimal penalty factor C and the kernel parameter g.
S111: the abnormal state of the printer is monitored in real time through the trained support vector machine, an alarm is sent out when the printer is judged to be abnormal, and a specific fault type is output.
The warning can be performed in a mode of emitting warning sound or in a mode of displaying warning marks on a display screen, and the specific mode of the warning is not limited. The specific fault type can be displayed through the display screen so as to remind the user to carry out corresponding maintenance processing. For example, when a paper jam occurs, the user is reminded to take out the jammed paper, when the ink box is abnormal, the user is reminded to replace the ink box, and the like.
S112: and comparing the actual state with the diagnosis state of the printer, and correcting the weights of the temperature signal, the humidity signal, the audio signal and the vibration signal.
The weight is corrected by comparing the actual state and the diagnosis state of the printer, so that the accuracy and the stability of the subsequent abnormal diagnosis can be ensured.
In one possible embodiment, S112 specifically includes substeps S1121 to S1122:
s1121: comparing the diagnosis state with the actual state, and evaluating the diagnosis result, wherein the evaluation result comprises: detecting the normal as an abnormal class result, detecting the abnormal as a normal class result, detecting the normal as a normal class result, and detecting the abnormal as an abnormal class result.
S1122: the ratio FY of the predicted errors of the normal sample is calculated by setting the number of times the normal class result is predicted as TX, the number of times the normal class result is predicted as FY, the number of times the abnormality is predicted as TY, and the number of times the abnormality is predicted as the normal class result as FX rate And the correct ratio FX of the abnormal sample is predicted rate The method comprises the following steps:
/>
let p be the cost of predicting normal as an abnormal class result, q be the cost of predicting abnormal as a normal class result, and correct the weights of the temperature signal, the humidity signal, the audio signal and the vibration signal so that:
the cost p for predicting the normal as the abnormal result can be estimated by economic loss caused by the fact that the file cannot be printed, and the cost q for predicting the abnormal as the normal result can be estimated by maintenance cost for causing the damage of the printer.
The balance is p, the cost of the normal prediction is an abnormal result, and q, the cost of the abnormal prediction is an abnormal result, so that the weights of the temperature signal, the humidity signal, the audio signal and the vibration signal are in reasonable values, and not only can excessive early warning be avoided, but also false early warning can be avoided.
In the present invention, the weight is adjusted twice. The primary adjustment of the weight of each signal according to the result of the preliminary diagnosis occurs when the preliminary diagnosis is performed on the paper-out image, and the weight is adjusted to adapt to the result of the preliminary diagnosis because the preliminary diagnosis result gives a rough direction, so that the accuracy of the subsequent abnormal diagnosis is improved. The other is that the cost of the normal prediction as the abnormal result is p and the cost of the abnormal prediction as the normal result is q, so as to avoid the abnormal printing caused by excessive early warning, and avoid the printer damage caused by error early warning, and the balance is found between the excessive early warning and the error early warning.
Compared with the prior art, the invention has at least the following beneficial technical effects:
(1) In the invention, the real-time state of the printer is automatically diagnosed through the support vector machine, an alarm is sent out when the printer is judged to be abnormal, a specific fault type is output, the user does not need to disassemble and check the printer by itself, a great deal of time is not wasted to wait for the entrance confirmation of maintenance personnel with abundant experience, the diagnosis accuracy is high, and the time and the labor are saved.
(2) When the printer is in fault, the conditions of temperature rise, abnormal humidity, abnormal sound emission, severe vibration and the like of the layout device are often accompanied, and in the invention, the state change possibly caused by the printer in fault is comprehensively considered, the temperature signal, the humidity signal, the audio signal and the vibration signal are subjected to multi-mode fusion, and the final diagnosis result is given after the temperature signal, the humidity signal, the audio signal and the vibration signal are comprehensively considered, so that the accuracy of the fault diagnosis result is further improved.
(3) In the invention, wavelet packets are adopted to decompose temperature signals, humidity signals, audio signals and vibration signals, noise is removed, detail characteristics of the temperature signals, the humidity signals, the audio signals and the vibration signals are deeply excavated, permutation entropy values under different scale factors are calculated, the permutation entropy values under the optimal scale factors are summarized to be used as characteristic values, the degree of distinction among different fault types is improved, and the accuracy of fault diagnosis results is further improved.
(4) For printer faults, observing the paper-out image can obtain very important conclusions, for example, when the paper-out image is completely black, the fault is often caused by hardware circuits, so that the fault range can be preliminarily determined by analyzing the paper-out image in advance. Aiming at the faults and problems of the types, the weight of a specific temperature signal, a specific humidity signal, a specific audio signal and a specific vibration signal is selected in the diagnosis process, so that the distinction degree among the fault types is further enlarged in the diagnosis process after the feature fusion, and the accuracy of the fault diagnosis result is improved.
(5) In the present invention, the weights of the temperature signal, the humidity signal, the audio signal, and the vibration signal are corrected by comparing the actual state and the diagnostic state of the printer to balance the loss caused by diagnosing the normal as abnormal with the loss caused by diagnosing the abnormal as normal.
Example two
In one embodiment, referring to fig. 3 of the specification, the present invention provides a structural schematic diagram of a printer abnormality diagnosis system.
The present invention provides a printer abnormal state diagnosis system 30, comprising:
a first obtaining module 301, configured to obtain a paper output situation of the printer, obtain a paper image when the printer is capable of outputting paper, and identify an image state of the paper image;
the preliminary diagnosis module 302 is configured to perform preliminary diagnosis according to an image state of the paper output image;
the setting module 303 is configured to set weights of a temperature signal, a humidity signal, an audio signal and a vibration signal during subsequent feature fusion according to the preliminary diagnosis result;
a second obtaining module 304, configured to obtain a status signal of the printer, where the status signal includes a temperature signal, a humidity signal, an audio signal, and a vibration signal;
a normalization module 305, configured to normalize the temperature signal, the humidity signal, the audio signal, and the vibration signal;
the decomposition module 306 is configured to decompose and denoise the temperature signal, the humidity signal, the audio signal, and the vibration signal by using wavelet packet transformation, and decompose the temperature signal, the humidity signal, the audio signal, and the vibration signal into a plurality of subband sequences;
A calculation module 307, configured to calculate permutation entropy values of each frequency band under different scale factors, and select optimal scale factors of the temperature signal, the humidity signal, the audio signal, and the vibration signal respectively;
the construction module 308 is configured to aggregate the arrangement entropy values of the temperature signal, the humidity signal, the audio signal, and the vibration signal under the optimal scale factors, obtain the feature values of the temperature signal, the humidity signal, the audio signal, and the vibration signal, and construct feature vectors;
the fusion module 309 is configured to perform feature fusion on feature vectors of the temperature signal, the humidity signal, the audio signal, and the vibration signal according to weights of the temperature signal, the humidity signal, the audio signal, and the vibration signal, to obtain feature fusion vectors;
the training module 310 is configured to add a classification label to the feature fusion vector, input the feature fusion vector to the support vector machine, output a fault diagnosis result, and train the support vector machine by comparing whether the fault diagnosis result is consistent with the label;
the diagnosis module 311 is used for monitoring the abnormal state of the printer in real time through the trained support vector machine, sending out an alarm when judging that the printer is abnormal, and outputting a specific fault type;
The correction module 312 is configured to compare the actual status of the printer with the diagnostic status, and correct the weights of the temperature signal, the humidity signal, the audio signal, and the vibration signal.
In one possible implementation, the preliminary diagnostic module 302 is specifically configured to:
under the condition that the paper output image is completely black, primarily diagnosing that a hardware circuit of the printer has faults;
under the condition that the overall gray level of the paper output image is low or the local gray level is unbalanced, the failure of the ink box, the reflecting mirror or the photoconductor of the printer is diagnosed preliminarily;
under the condition that regular spots exist in the paper-out image, primarily diagnosing that faults exist in rollers of the printer;
under the condition that vertical white stripes exist in the paper-out image, primarily diagnosing faults of a transfer electrode or a charging electrode of the printer;
the setting module 303 is specifically configured to:
under the condition that the hardware circuit of the printer is initially diagnosed to have faults, setting the weights of a temperature signal, a humidity signal, an audio signal and a vibration signal in the subsequent feature fusion as a first weight combination;
under the condition of preliminary diagnosis of the failure of the ink box of the printer, setting the weights of the temperature signal, the humidity signal, the audio signal and the vibration signal in the subsequent feature fusion as a second weight combination;
Under the condition that the fault exists in the reflecting mirror of the printer in the preliminary diagnosis, setting the weights of the temperature signal, the humidity signal, the audio signal and the vibration signal in the subsequent feature fusion as a third weight combination;
under the condition of preliminary diagnosis of the fault of the photoconductor of the printer, setting the weights of the temperature signal, the humidity signal, the audio signal and the vibration signal in the subsequent feature fusion as a fourth weight combination;
under the condition of preliminary diagnosis of the fault of a roller of the printer, setting weights of a temperature signal, a humidity signal, an audio signal and a vibration signal in subsequent feature fusion as a fifth weight combination;
and under the condition of preliminary diagnosis of faults of a transfer electrode or a charging electrode of the printer, setting weights of a temperature signal, a humidity signal, an audio signal and a vibration signal in subsequent feature fusion as a sixth weight combination.
In one possible implementation, the normalization module 305 is specifically configured to:
the temperature signal, the humidity signal, the audio signal and the vibration signal are normalized by adopting the following formula:
where f' (x) represents the signal after normalization, f (x) represents the signal before normalization, min [ f (x) ] represents the minimum value in f (x), and max [ f (x) ] represents the maximum value in f (x).
In one possible implementation, the decomposition module 306 is specifically configured to:
for signals to be processedSelecting a group of orthogonal wavelet packet basis functions psi (t):
wherein h is k Representation ofG, g k Low-pass filter denoted psi (t), h k And g k Is a pair of conjugate orthogonal real coefficient filters, t represents a time variable, and k represents the number of wavelet packet decomposition layers;
wherein the signal to be processedOne of a temperature signal, a humidity signal, an audio signal, and a vibration signal;
recursively processing the formula and processing the signal to be processedDecomposition into quadrature wavelet packet signals { mu } n (t) } each orthogonal wavelet packet signal represents a frequency band:
wherein mu 2n (t) represents low-band coefficients, μ 2n+1 (t) represents a high-band coefficient.
The orthogonal wavelet packet signals with band coefficients exceeding the normal coefficient range are removed.
In one possible implementation, the computing module 307 is specifically configured to:
the time series { x (1), x (2), …, x (N) } obtained by decomposing the wavelet packet is subjected to coarse graining treatment to obtain a time series y after coarse graining treatment (s) (j)：
/>
Wherein s represents a scale factor,representation pair->Rounding, S represents the total length of the time sequence after coarse granulation processing, and N represents the length of the original signal;
for y (s) (j) And (3) performing space reconstruction:
y (s) (i)＝{y (s) (i),y (s) (i+τ),…,y (s) (i+(m-1)τ)}
Where τ represents the delay time and m represents the embedding dimension;
rearranging the ith individual reconstruction component Y (j) in ascending order:
y (s) [i+(j 1 -1)τ]≤y (s) [i+(j 2 -1)τ]≤…≤y (s) [i+(j m -1)τ]
wherein j is m Representing a location dimension;
acquiring a position dimension sequence pi corresponding to Y (j) i ：
π i ＝(j 1 ,j 2 ,…,j m )；
Defining each symbol sequence pi i The probability of occurrence is:
wherein s (pi) i ) Represents pi i The number of occurrences;
calculating the time sequence inPermutation entropy H of different position dimensions p (m)：
For permutation entropy H p (m) carrying out normalization treatment:
selecting a plurality of arrangement entropy values of delay time tau, embedding dimension m and sample length N under different scale factors s by a control variable method, and constructing an arrangement entropy curve;
and calculating the average distance between the arrangement entropy curves under different scale factors s, and taking the scale factor corresponding to the maximum average distance as the optimal scale factor.
In one possible implementation, the fusion module 309 is specifically configured to:
in the case that the preliminary diagnosis result exists, the weights of the temperature signal, the humidity signal, the audio signal and the vibration signal determined by the setting module 303 are adopted to assign weights; under the condition that the preliminary diagnosis result does not exist, adopting the initial weights preset by the temperature signal, the humidity signal, the audio signal and the vibration signal to assign weights;
Feature fusion is carried out on feature vectors of the temperature signal, the humidity signal, the audio signal and the vibration signal, and a feature fusion vector q is obtained:
q＝α 1 ·p 1 +α 2 ·p 2 +α 3 ·p 3 +α 4 ·p 4
wherein p is 1 Feature vector, alpha, representing temperature signal 1 Weight, p, of the temperature signal 2 Feature vector, alpha, representing humidity signal 2 Weight, p, of the humidity signal 3 Feature vector, alpha, representing audio signal 3 Representing weights of audio signals，p 4 Characteristic vector, alpha, representing vibration signal 4 A weight representing the vibration signal;
wherein, in the case that the preliminary diagnosis result exists, the weights of the temperature signal, the humidity signal, the audio signal and the vibration signal determined in S103 are adopted; in the absence of the preliminary diagnostic result, the initial weights preset for the temperature signal, the humidity signal, the audio signal, and the vibration signal are adopted.
In one possible implementation, the training module 310 is specifically configured to:
adding a classification label for the feature fusion vector, and dividing a training set and a verification set;
inputting the feature fusion vector q in the training set to a support vector machine;
the support vector machine constructs an objective function, selects an optimal hyperplane so as to maximize sample classification intervals among various types of samples, and the various types of data samples can be partitioned:
Wherein N represents the number of samples, q i Feature fusion vector representing input, u i Classification label representing input, C represents penalty factor, ζ i Represents a relaxation variable, b represents an offset, K (qi. Q j ) Represents a kernel function, represents q i And q j Similarity between the two, g represents a kernel parameter;
determining the optimal hyperplane as:
in one possible implementation, the optimal penalty factor C and the kernel parameter g are obtained to determine the optimal hyperplane by:
continuously changing the population searching speed v i And searching for a locationTo search for each penalty factor C and each kernel parameter g:
wherein w represents an inertial weight factor, r is a random number between 0 and 1, x i Representing the current search position, e 1 And e 2 Represent learning factors, P i Representing global optimum parameters, G i Representing individual optimal parameters;
during the search, the inertial weight factor is set to an adaptive inertial weight factor to avoid premature convergence of the algorithm:
wherein h represents the target parameter value sought by the current algorithm, h min Represents the minimum value of the target parameter value, h avg Mean value of target parameter value, w max Represents the maximum value, w, of the adaptive inertial weight factor min Representing the minimum value of the adaptive inertial weight factor;
In the searching process, a contraction factor omega is introduced, and the searching range is continuously narrowed towards the global optimal position:
wherein E represents a balance factor;
searching the penalty factor C and the kernel parameter g according to the updated search formula until the global optimal parameter and the individual optimal parameter related to the penalty factor C, and the global optimal parameter and the individual optimal parameter related to the kernel parameter g are searched;
and taking the penalty factors C at the positions meeting the global optimal parameters and the individual optimal parameters as optimal penalty factors, and taking the core parameters g at the positions meeting the global optimal parameters and the individual optimal parameters as optimal core parameters.
In one possible implementation, the correction module 312 is specifically configured to:
s1121: comparing the diagnosis state with the actual state, and evaluating the diagnosis result, wherein the evaluation result comprises: detecting the normal as an abnormal class result, detecting the abnormal as a normal class result, detecting the normal as a normal class result, and detecting the abnormal as an abnormal class result;
s1122: the ratio FY of the predicted errors of the normal sample is calculated by setting the number of times the normal class result is predicted as TX, the number of times the normal class result is predicted as FY, the number of times the abnormality is predicted as TY, and the number of times the abnormality is predicted as the normal class result as FX rate And the correct ratio FX of the abnormal sample is predicted rate The method comprises the following steps:
let p be the cost of predicting normal as an abnormal class result, q be the cost of predicting abnormal as a normal class result, and correct the weights of the temperature signal, the humidity signal, the audio signal and the vibration signal so that:
compared with the prior art, the invention has at least the following beneficial technical effects:
(1) In the invention, the real-time state of the printer is automatically diagnosed through the support vector machine, an alarm is sent out when the printer is judged to be abnormal, a specific fault type is output, the user does not need to disassemble and check the printer by itself, a great deal of time is not wasted to wait for the entrance confirmation of maintenance personnel with abundant experience, the diagnosis accuracy is high, and the time and the labor are saved.
(2) When the printer is in fault, the conditions of temperature rise, abnormal humidity, abnormal sound emission, severe vibration and the like of the layout device are often accompanied, and in the invention, the state change possibly caused by the printer in fault is comprehensively considered, the temperature signal, the humidity signal, the audio signal and the vibration signal are subjected to multi-mode fusion, and the final diagnosis result is given after the temperature signal, the humidity signal, the audio signal and the vibration signal are comprehensively considered, so that the accuracy of the fault diagnosis result is further improved.
(3) In the invention, wavelet packets are adopted to decompose temperature signals, humidity signals, audio signals and vibration signals, noise is removed, detail characteristics of the temperature signals, the humidity signals, the audio signals and the vibration signals are deeply excavated, permutation entropy values under different scale factors are calculated, the permutation entropy values under the optimal scale factors are summarized to be used as characteristic values, the degree of distinction among different fault types is improved, and the accuracy of fault diagnosis results is further improved.
(4) For printer faults, observing the paper-out image can obtain very important conclusions, for example, when the paper-out image is completely black, the fault is often caused by hardware circuits, so that the fault range can be preliminarily determined by analyzing the paper-out image in advance. Aiming at the faults and problems of the types, the weight of a specific temperature signal, a specific humidity signal, a specific audio signal and a specific vibration signal is selected in the diagnosis process, so that the distinction degree among the fault types is further enlarged in the diagnosis process after the feature fusion, and the accuracy of the fault diagnosis result is improved.
(5) In the present invention, the weights of the temperature signal, the humidity signal, the audio signal, and the vibration signal are corrected by comparing the actual state and the diagnostic state of the printer to balance the loss caused by diagnosing the normal as abnormal with the loss caused by diagnosing the abnormal as normal.
The printer abnormal state diagnosis system 30 provided by the present invention can implement each process implemented in the above method embodiment, and in order to avoid repetition, a detailed description is omitted here.
The virtual system provided by the invention can be a system, and can also be a component, an integrated circuit or a chip in a terminal.
The technical features of the above embodiments may be arbitrarily combined, and all possible combinations of the technical features in the above embodiments are not described for brevity of description, however, as long as there is no contradiction between the combinations of the technical features, they should be considered as the scope of the description.
The foregoing examples illustrate only a few embodiments of the invention, which are described in detail and are not to be construed as limiting the scope of the invention. It should be noted that it will be apparent to those skilled in the art that several variations and modifications can be made without departing from the spirit of the invention, which are all within the scope of the invention. Accordingly, the scope of protection of the present invention is to be determined by the appended claims.
Claims (10)
1. A printer abnormal state diagnosis method, characterized by comprising:
S101: acquiring a paper output condition of a printer, acquiring a paper output image under the condition that the printer can output paper, and identifying an image state of the paper output image;
s102: performing preliminary diagnosis according to the image state of the paper output image;
s103: according to the preliminary diagnosis result, the weights of a temperature signal, a humidity signal, an audio signal and a vibration signal during the subsequent feature fusion are set;
s104: acquiring a status signal of the printer, wherein the status signal comprises the temperature signal, the humidity signal, the audio signal and the vibration signal;
s105: normalizing the temperature signal, the humidity signal, the audio signal and the vibration signal;
s106: decomposing and denoising the temperature signal, the humidity signal, the audio signal and the vibration signal by adopting wavelet packet transformation, and decomposing the temperature signal, the humidity signal, the audio signal and the vibration signal into a plurality of sub-band sequences;
s107: calculating arrangement entropy values of each frequency band under different scale factors, and respectively selecting the optimal scale factors of the temperature signal, the humidity signal, the audio signal and the vibration signal;
s108: summarizing the arrangement entropy values of the temperature signal, the humidity signal, the audio signal and the vibration signal under the condition that the optimal scale factors are obtained respectively, obtaining the characteristic values of the temperature signal, the humidity signal, the audio signal and the vibration signal, and constructing a characteristic vector;
S109: according to the weights of the temperature signal, the humidity signal, the audio signal and the vibration signal, feature fusion is carried out on feature vectors of the temperature signal, the humidity signal, the audio signal and the vibration signal, so that feature fusion vectors are obtained;
s110: adding a classification label for the feature fusion vector, inputting the feature fusion vector into a support vector machine, outputting a fault diagnosis result, and training the support vector machine by comparing whether the fault diagnosis result is consistent with the label;
s111: the trained support vector machine monitors the abnormal state of the printer in real time, gives an alarm when judging that the printer is abnormal, and outputs a specific fault type;
s112: and comparing the actual state with the diagnosis state of the printer, and correcting the weights of the temperature signal, the humidity signal, the audio signal and the vibration signal.
2. The printer abnormal state diagnosis method according to claim 1, wherein S102 specifically comprises:
s1021: under the condition that the paper output image is completely black, primarily diagnosing that a hardware circuit of the printer has faults;
S1022: under the condition that the overall gray level of the paper output image is low or the local gray level is unbalanced, primarily diagnosing that the ink box, the reflecting mirror or the photoconductor of the printer has faults;
s1023: under the condition that regular spots exist in the paper-out image, primarily diagnosing that faults exist in rollers of the printer;
s1024: under the condition that vertical white stripes exist in the paper-out image, primarily diagnosing that faults exist in a transfer electrode or a charging electrode of the printer;
the step S103 specifically includes:
s1031: under the condition that the hardware circuit of the printer is initially diagnosed to have faults, setting the weights of a temperature signal, a humidity signal, an audio signal and a vibration signal in the subsequent feature fusion as a first weight combination;
s1032: under the condition of preliminary diagnosis of the failure of the ink box of the printer, setting the weights of the temperature signal, the humidity signal, the audio signal and the vibration signal in the subsequent feature fusion as a second weight combination;
s1033: under the condition that the fault exists in the reflecting mirror of the printer in the preliminary diagnosis, setting the weights of the temperature signal, the humidity signal, the audio signal and the vibration signal in the subsequent feature fusion as a third weight combination;
S1034: under the condition of preliminary diagnosis of the fault of the photoconductor of the printer, setting the weights of the temperature signal, the humidity signal, the audio signal and the vibration signal in the subsequent feature fusion as a fourth weight combination;
s1035: under the condition of preliminary diagnosis of the fault of a roller of the printer, setting weights of a temperature signal, a humidity signal, an audio signal and a vibration signal in subsequent feature fusion as a fifth weight combination;
s1036: and under the condition of preliminary diagnosis of faults of a transfer electrode or a charging electrode of the printer, setting weights of a temperature signal, a humidity signal, an audio signal and a vibration signal in subsequent feature fusion as a sixth weight combination.
3. The printer abnormal state diagnosis method according to claim 1, wherein S105 is specifically:
normalizing the temperature signal, the humidity signal, the audio signal and the vibration signal by adopting the following formula:
where f' (x) represents the signal after normalization, f (x) represents the signal before normalization, min [ f (x) ] represents the minimum value in f (x), and max [ f (x) ] represents the maximum value in f (x).
4. The printer abnormal state diagnosis method according to claim 1, wherein S106 specifically comprises:
S1061: for signals to be processedSelecting a group of orthogonal wavelet packet basis functions psi (t):
wherein h is k Representation ofG, g k Low-pass filter denoted psi (t), h k And g k Is a pair of conjugate orthogonal real coefficient filters, t represents a time variable, and k represents the number of wavelet packet decomposition layers;
wherein the signal to be processedIs one of the temperature signal, the humidity signal, the audio signal, and the vibration signal;
s1062: recursively processing the formula and processing the signal to be processedDecomposition into quadrature wavelet packet signals { mu } n (t) } each of the orthogonal wavelet packet signals represents a frequency band:
wherein mu 2n (t) represents low-band coefficients, μ 2n+1 (t) represents a high-band coefficient.
S1063: the orthogonal wavelet packet signals with band coefficients exceeding the normal coefficient range are removed.
5. The printer abnormal state diagnosis method according to claim 1, wherein S107 specifically comprises:
s1071: the time series { x (1), x (2), …, x (N) } obtained by decomposing the wavelet packet is subjected to coarse graining treatment to obtain a time series y after coarse graining treatment (s) (j)：
Wherein s represents a scale factor,representation pair->Rounding, S represents the total length of the time sequence after coarse granulation processing, and N represents the length of the original signal;
S1072: for y (s) (j) And (3) performing space reconstruction:
y (s) (i)＝{y (s) (i),y (s) (i+τ),…,y (s) (i+(m-1)τ)}
where τ represents the delay time and m represents the embedding dimension;
s1073: rearranging the ith individual reconstruction component Y (j) in ascending order:
y (s) [i+(j 1 -1)τ]≤y (s) [i+(j 2 -1)τ]≤…≤y (s) [i+(j m -1)τ]
wherein j is m Representing a location dimension;
s1074: acquiring a position dimension sequence pi corresponding to Y (j) i ：
π i ＝(j 1 ,j 2 ,…,j m )；
S1075: defining each symbol sequence pi i The probability of occurrence is:
wherein s (pi) i ) Represents pi i The number of occurrences;
s1076: calculating permutation entropy H of time sequence in different position dimensions p (m)：
S1077: for permutation entropy H p (m) carrying out normalization treatment:
s1078: selecting a plurality of arrangement entropy values of delay time tau, embedding dimension m and sample length N under different scale factors s by a control variable method, and constructing an arrangement entropy curve;
s1079: calculating the average distance between the arrangement entropy curves under different scale factors s, and taking the scale factor corresponding to the maximum average distance as the optimal scale factor.
6. The printer abnormal state diagnosis method according to claim 1, wherein S109 is specifically:
in the presence of the preliminary diagnosis result, assigning weights to the temperature signal, the humidity signal, the audio signal and the vibration signal determined in S103; under the condition that the preliminary diagnosis result does not exist, adopting the initial weights preset by the temperature signal, the humidity signal, the audio signal and the vibration signal to assign weights;
Feature fusion is carried out on the feature vectors of the temperature signal, the humidity signal, the audio signal and the vibration signal to obtain a feature fusion vector q:
q＝α 1 ·p 1 +α 2 ·p 2 +α 3 ·p 3 +α 4 ·p 4
wherein p is 1 A eigenvector, alpha, representing the temperature signal 1 Weight, p, representing the temperature signal 2 A feature vector, alpha, representing the humidity signal 2 Weights representing the humidity signal, p 3 Feature vector, alpha, representing the audio signal 3 Representing the weight, p, of the audio signal 4 Characteristic vector, alpha, representing the vibration signal 4 Representing the weight of the vibration signal.
7. The printer abnormal state diagnosis method according to claim 1, wherein S110 specifically comprises:
s1101: adding a classification label for the feature fusion vector, and dividing a training set and a verification set;
s1102: inputting the feature fusion vector q in the training set to a support vector machine;
s1103: the support vector machine constructs an objective function, selects an optimal hyperplane so as to maximize sample classification intervals among various samples, and the various samples of the data can be divided into:
wherein N represents the number of samples, q i Feature fusion vector representing input, u i Classification label representing input, C represents penalty factor, ζ i Represents a relaxation variable, b represents an offset, K (q i .q j ) Represents a kernel function, represents q i And q j Similarity between the two, g represents a kernel parameter;
s1104: determining the optimal hyperplane as:
8. the printer abnormal state diagnosis method according to claim 7, wherein the optimal hyperplane is determined by obtaining an optimal penalty factor C and a kernel parameter g by:
S110A: continuously changing the population searching speed v i And searching for a locationTo search for each penalty factor C and each kernel parameter g:
wherein w represents an inertial weight factor, r is a random number between 0 and 1, x i Representing the current search position, e 1 And e 2 Represent learning factors, P i Representing global optimum parameters, G i Representing individual optimal parameters;
S110B: during the search, the inertial weight factor is set to an adaptive inertial weight factor to avoid premature convergence of the algorithm:
wherein h represents the target parameter value sought by the current algorithm, h min Represents the minimum value of the target parameter value, h avg Mean value of target parameter value, w max Represents the maximum value, w, of the adaptive inertial weight factor min Representing the minimum value of the adaptive inertial weight factor;
S110C: in the searching process, a contraction factor omega is introduced, and the searching range is continuously narrowed towards the global optimal position:
Wherein E represents a balance factor;
S110D: searching the penalty factor C and the kernel parameter g according to the updated search formula until the global optimal parameter and the individual optimal parameter related to the penalty factor C, and the global optimal parameter and the individual optimal parameter related to the kernel parameter g are searched;
S110E: and taking the penalty factors C at the positions meeting the global optimal parameters and the individual optimal parameters as optimal penalty factors, and taking the core parameters g at the positions meeting the global optimal parameters and the individual optimal parameters as optimal core parameters.
9. The printer abnormal state diagnosis method according to claim 1, wherein S112 specifically comprises:
s1121: comparing the diagnosis state with the actual state, and evaluating the diagnosis result, wherein the evaluation result comprises: detecting the normal as an abnormal class result, detecting the abnormal as a normal class result, detecting the normal as a normal class result, and detecting the abnormal as an abnormal class result;
s1122: the number of times of predicting the normal as the normal class result is TX, the number of times of predicting the normal as the abnormal class result is FY, the number of times of predicting the abnormal as the abnormal class result is TY, the number of times of predicting the abnormal as the normal class result is FX, and a normal sample is pre-prepared Measurement of error ratio FY rate And the correct ratio FX of the abnormal sample is predicted rate The method comprises the following steps:
and correcting weights of the temperature signal, the humidity signal, the audio signal and the vibration signal so that the weight of the temperature signal, the humidity signal, the audio signal and the vibration signal is as follows:
10. a printer abnormal state diagnosis system, comprising:
the first acquisition module is used for acquiring the paper output condition of the printer, acquiring a paper output image under the condition that the printer can output paper, and identifying the image state of the paper output image;
the preliminary diagnosis module is used for performing preliminary diagnosis according to the image state of the paper output image;
the setting module is used for setting weights of a temperature signal, a humidity signal, an audio signal and a vibration signal during subsequent feature fusion according to the initial diagnosis result;
a second acquisition module configured to acquire a status signal of the printer, the status signal including the temperature signal, the humidity signal, the audio signal, and the vibration signal;
the normalization module is used for performing normalization processing on the temperature signal, the humidity signal, the audio signal and the vibration signal;
The decomposition module is used for decomposing and denoising the temperature signal, the humidity signal, the audio signal and the vibration signal by adopting wavelet packet transformation, and decomposing the temperature signal, the humidity signal, the audio signal and the vibration signal into a plurality of sub-band sequences;
the calculation module is used for calculating the arrangement entropy values of each frequency band under different scale factors and respectively selecting the optimal scale factors of the temperature signal, the humidity signal, the audio signal and the vibration signal;
the construction module is used for summarizing the arrangement entropy values of the temperature signal, the humidity signal, the audio signal and the vibration signal under the condition that the optimal scale factors are obtained respectively, obtaining the characteristic values of the temperature signal, the humidity signal, the audio signal and the vibration signal, and constructing a characteristic vector;
the fusion module is used for carrying out feature fusion on the feature vectors of the temperature signal, the humidity signal, the audio signal and the vibration signal according to the weights of the temperature signal, the humidity signal, the audio signal and the vibration signal to obtain feature fusion vectors;
the training module is used for adding a classification label to the feature fusion vector, inputting the feature fusion vector into a support vector machine, outputting a fault diagnosis result, and training the support vector machine by comparing whether the fault diagnosis result is consistent with the label;
The diagnosis module is used for monitoring the abnormal state of the printer in real time through the trained support vector machine, sending out an alarm when judging that the printer is abnormal, and outputting a specific fault type;
and the correction module is used for comparing the actual state with the diagnosis state of the printer and correcting the weights of the temperature signal, the humidity signal, the audio signal and the vibration signal.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310812700.XA CN116755648A (en) | 2023-07-04 | 2023-07-04 | Printer abnormal state diagnosis method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310812700.XA CN116755648A (en) | 2023-07-04 | 2023-07-04 | Printer abnormal state diagnosis method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116755648A true CN116755648A (en) | 2023-09-15 |
Family
ID=87949580
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310812700.XA Pending CN116755648A (en) | 2023-07-04 | 2023-07-04 | Printer abnormal state diagnosis method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116755648A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117289745A (en) * | 2023-11-27 | 2023-12-26 | 湖北华中电力科技开发有限责任公司 | Operation monitoring method for digital power distribution room |
-
2023
- 2023-07-04 CN CN202310812700.XA patent/CN116755648A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117289745A (en) * | 2023-11-27 | 2023-12-26 | 湖北华中电力科技开发有限责任公司 | Operation monitoring method for digital power distribution room |
| CN117289745B (en) * | 2023-11-27 | 2024-02-13 | 湖北华中电力科技开发有限责任公司 | Operation monitoring method for digital power distribution room |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US6862414B2 (en) | Automated banding defect analysis and repair for document processing systems | |
| Kwan et al. | A novel approach to fault diagnostics and prognostics | |
| US7773774B2 (en) | Dynamic test pattern composition for image-analysis based automatic machine diagnostics | |
| WO2018131219A1 (en) | Anomaly sensing device, anomaly sensing method, and storage medium | |
| JP2016085704A (en) | Information processing system, information processing device, information processing method, and program | |
| CN100474878C (en) | Image quality prediction method and apparatus and fault diagnosis system | |
| US20050262394A1 (en) | Failure diagnosis method, failure diagnosis apparatus, conveyance device, image forming apparatus, program, and storage medium | |
| EP1109395A2 (en) | Systems and methods for automated image quality based diagnostics and remediation of document processing systems | |
| CN116755648A (en) | Printer abnormal state diagnosis method and system | |
| JP2007323455A (en) | System and method for supporting failure preventive diagnosis | |
| JP7091174B2 (en) | System, system control method and program | |
| JP2007249975A (en) | Method and system for assisting user to correct device problem, and method for preparing system | |
| US9696947B1 (en) | Fault identification for a printing system | |
| US20100325487A1 (en) | Method and system for automatically diagnosing faults in rendering devices | |
| EP2410312A1 (en) | A method for computer-assisted analyzing of a technical system | |
| JP2015162140A (en) | Diagnostic system of plant and control device | |
| EP3761133A1 (en) | Diagnosis device and diagnosis method | |
| CN111271183B (en) | Method and system for self-adaptive online prediction of state of diesel engine | |
| Ghoshal et al. | On cost-sensitive calibrated uncertainty in deep learning: An application on covid-19 detection | |
| US8724168B2 (en) | Updating a smoothness constrained cluster model for color control in a color management system | |
| JP4631809B2 (en) | Defect classification system, image forming apparatus, and defect classification program | |
| CN112202630A (en) | Network quality abnormity detection method and device based on unsupervised model | |
| CN111814557A (en) | Action flow detection method, device, equipment and storage medium | |
| US11301181B2 (en) | Diagnosing an image forming apparatus based on sound signal analysis | |
| JP2022086643A (en) | Printer abnormality sign detection system, printer abnormality sign detection method and program |
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 |