CN115741218A

CN115741218A - Machine tool fault early warning method and system based on processing portrait analysis

Info

Publication number: CN115741218A
Application number: CN202310011878.4A
Authority: CN
Inventors: 王钢强
Original assignee: Shan County Asia Pacific Paper Products Co ltd
Current assignee: Shan County Asia Pacific Paper Products Co ltd
Priority date: 2023-01-05
Filing date: 2023-01-05
Publication date: 2023-03-07
Anticipated expiration: 2043-01-05
Also published as: CN115741218B

Abstract

The invention relates to the field of image processing, in particular to a machine tool fault early warning method and a machine tool fault early warning system based on processing image analysis.

Description

Machine tool fault early warning method and system based on processing portrait analysis

Technical Field

The invention relates to the field of image processing, in particular to a machine tool fault early warning method and system based on processing portrait analysis.

Background

In the field of modern machining, numerical control machine tool machining already occupies a dominating position, and particularly for machining parts with large batch and high-precision requirements, various numerical control machine tools play a role of no substitution, so that the working stability and reliability of the numerical control machine tool have very important significance for improving the production efficiency of enterprises.

Chinese patent publication No. CN112650146B discloses a fault diagnosis optimization method, system and device under multiple conditions of a numerically-controlled machine tool, which improves the effectiveness of data acquisition and utilization by optimizing the arrangement of sensors of the numerically-controlled machine tool, extracts characteristic information of the numerically-controlled machine tool representing different states at different time scales by using an improved multi-scale entropy algorithm, excavates deep-level characteristic information, and improves the differentiation of characteristics between different states; on the basis, the working condition of the numerical control machine tool is identified by using ITML-K mean value clustering so as to solve the problem that the identification effect of the traditional clustering method is poor under the condition of multi-working condition boundary fuzzy; and finally, solving the over-fitting problem of the data-driven numerical control machine tool fault diagnosis model during construction by utilizing the regularization function based on the entropy so as to improve the generalization and accuracy of the numerical control machine tool fault diagnosis model and realize the optimization of the numerical control machine tool fault diagnosis model, thereby having important help for improving the operation safety and reliability of the numerical control machine tool and improving the fault diagnosis rate of the numerical control machine tool.

Therefore, according to the technical scheme, the machine tool fault diagnosis model is established to identify the machine tool fault, and the type of the machine tool fault still cannot be accurately distinguished, so that the machine tool fault diagnosis is high in complexity and poor in accuracy.

Disclosure of Invention

Therefore, the invention provides a machine tool fault early warning method and system based on processing portrait analysis, which can solve the technical problems of high complexity and poor accuracy of machine tool fault diagnosis in the prior art.

In order to achieve the purpose, the invention provides a machine tool fault early warning method based on processing image analysis, which comprises the following steps:

acquiring processing temperature information and processing image information of a machine tool according to a standard transmission speed, determining the safety level of the machine tool according to the processing temperature information, and determining an alarm mode according to the safety level;

when the machine tool meets a preset safety level, acquiring the number of pixels of a processed image, determining the image level of the processed image according to the number of pixels, and processing the processed image according to the image level;

when the processed image meets the preset image grade, restoring the processed image to obtain a standard processed image;

carrying out time-frequency transformation on the standard processing image and adjusting in a frequency domain to obtain an adjusted processing image;

storing the gray value of the adjusted processing image in a target linked list, calculating the similarity of the target linked list and a preset standard linked list, judging the abnormal condition of the machine tool according to the similarity, and when the similarity meets the preset similarity condition, performing edge extraction on the adjusted processing image to obtain an effective processing image;

calculating the similarity between the effective processing image and a preset risk image to serve as rechecking similarity, and judging the abnormal condition of the machine tool according to the rechecking similarity;

and acquiring an effective processing image for judging that the machine tool has a fault, taking the effective processing image as a judging processing image, generating a three-dimensional track model according to the judging processing image and time, and judging the abnormal condition of the machine tool according to the three-dimensional track model.

Further, generating a machining thermal image according to the machining temperature information, acquiring a gray level image of the machining thermal image, generating a gray level frequency histogram according to the gray level image, dividing the frequency histogram into three parts along a direction parallel to a vertical coordinate at positions with abscissa 85 and 175, wherein the three parts comprise a first part, a second part and a third part, determining a safety level of the machine tool according to the frequency area of the three parts, and determining three types of alarm modes for the machine tool according to the safety level;

the first type of alarm mode is that when the frequency area of the first part meets a first area condition, the machine tool is judged to be at a first safety level, and no alarm is given;

the second type of alarm mode is that when the frequency area of the second part meets a second area condition, the machine tool is judged to be in a second safety level, and the first alarm mode is used for alarming;

the third type of alarm mode is that when the frequency area of the third part meets a third area condition, the machine tool is judged to be in a third safety level, the alarm is given by using the second alarm mode, and the machine tool is emergently braked;

the first area condition includes that the frequency area of the first portion is highest in the frequency area of the third portion, the second area condition includes that the frequency area of the second portion is highest in the frequency area of the third portion, the third area condition includes that the frequency area of the third portion is highest in the frequency area of the third portion, the safety of the first safety level is greater than the safety of the second safety level, the safety of the second safety level is greater than the safety of the third safety level, and the urgency degree of the first alarm mode is less than the urgency degree of the second alarm mode.

Further, when the machine tool meets a preset first safety level, acquiring the pixel number of the processed image according to the processed image information, determining the image level of the processed image according to the pixel number, and determining two types of processing modes according to the image level to obtain a standard processed image;

the first type of processing mode is that when the number of the pixels meets a first pixel condition, the processed image is judged to be at a first image grade, and the processed image is subjected to image restoration to obtain the standard processed image;

the second processing mode is that when the number of the pixels meets a second pixel condition, the processed image is judged to be a second image grade, the processed image is directly used as the standard processed image, and the standard transmission speed is corrected by using a speed correction coefficient;

the first pixel condition comprises that the number of the pixels is smaller than a preset number of pixels, the first pixel condition comprises that the number of the pixels is larger than or equal to the preset number of the pixels, and the definition of the first image level is smaller than that of the second image level.

Further, when the processed image meets a preset first image grade, calculating the image variance of the gray value of the processed image, dividing the processed image into at least two local images, calculating the local variance of the gray value of the local images, and determining three image restoration modes of the processed image according to the image variance and the local images;

the first type of image restoration method is to restore the processed image by using a first restoration method when the image variance and the local variance meet a first restoration condition;

the second type of image restoration method is that when the image variance and the local variance accord with a second restoration condition, a second restoration method is used for restoring the processed image;

a third type of image restoration method is to restore the processed image by using a third restoration method when the image variance and the local variance meet a third restoration condition;

the first recovery condition includes that the local variance is 0, the second recovery condition includes that the local variance is not 0 and the local variance is not equal to the image variance, the third recovery condition includes that the local variance is not 0 and the local variance is equal to the image variance, the first recovery mode includes not changing the gray value of the local image, the second recovery mode includes correcting the gray value of the local image by using a gray value correction coefficient, and the third recovery mode includes changing the gray value of the local image to the local variance.

Further, performing fourier transform on the standard processing image to obtain a processing spectrogram, determining three types of frequency adjustment modes for the processing spectrogram according to the processing spectrogram, adjusting frequencies in the processing spectrogram according to the frequency adjustment modes to obtain an adjusted spectrogram, and performing inverse fourier transform on the adjusted spectrogram to obtain an adjusted processing image;

the first type of frequency adjustment mode is that when the processing spectrogram meets a first frequency condition, the processing spectrogram is adjusted by using a first adjustment mode;

the second type of frequency adjustment mode is that when the processing spectrogram meets a second frequency condition, the processing spectrogram is adjusted by using a second adjustment mode;

a third frequency adjustment mode is that when the processing spectrogram meets a third frequency condition, the processing spectrogram is adjusted by using a third adjustment mode;

the first frequency condition includes that the frequency is smaller than a first preset frequency, the second frequency condition includes that the frequency is greater than or equal to the first preset frequency and smaller than a second preset frequency, the third frequency condition includes that the frequency is greater than or equal to the second preset frequency, the first preset frequency is smaller than the second preset frequency, the first adjustment mode includes that the frequency is denoised by using a denoising formula, and the denoising formula is that the frequency is denoised by using a denoising formula

X is an abscissa value of the processing spectrogram, y is an ordinate value of the processing spectrogram, D (x, y) is a distance from a point with the spectral coordinates (x, y) to the center of the spectrogram, and D1 is a cut-off frequency; the second adjustment mode comprises not adjusting the frequency, the third adjustment mode comprises enhancing the frequency by using an enhancement formula

。

Further, obtaining a gray value of each pixel point in the adjusted and processed image according to a preset time interval, taking the gray value as a target gray value, setting a storage bit corresponding to each pixel point in a preset linked list, storing the target gray value in the storage bit to obtain a target linked list, calculating the similarity between the target linked list and a preset standard linked list, and determining three types of execution modes according to the similarity;

the first type of execution mode is that when the similarity accords with a first similarity condition, the machine tool is judged to have a type of fault, and a first alarm mode is used for alarming;

the second type of execution mode is that when the similarity accords with a second similarity condition, a preset adjustment mode is used for adjusting the adjusted processing image so as to obtain an effective processing image;

the third type of execution mode is that when the similarity accords with a third similarity condition, the machine tool is judged to be normal, and a first correction coefficient is used for correcting the preset time interval;

wherein the first similarity condition includes that the similarity is less than a first preset similarity, the second similarity condition includes that the similarity is greater than or equal to a first preset similarity and less than a second preset similarity, the third similarity condition includes that the similarity is greater than or equal to the second preset similarity, the first preset similarity is less than the second preset similarity, the type of fault includes deformation of a machine tool element, the preset adjustment mode includes edge extraction of the adjusted and processed image by using an edge extraction formula, and the edge extraction formula is

Wherein m is the number of rows of the pixel points, n is the number of columns of the pixel points, G (m, n) is the gray value of the mth row and the nth column in the adjusted and processed image, and σ is the mean value of the pixel points.

Further, calculating the similarity between the effective processing image and a preset risk image based on a neural network algorithm to serve as rechecking similarity, and determining two types of control modes according to the rechecking similarity;

the first type of control mode is that when the rechecking similarity accords with a first rechecking condition, the machine tool is judged to have a type of fault, and a first alarm mode is used for alarming;

the second type of control mode is that when the rechecking similarity accords with a second rechecking condition, the machine tool is judged to be normal, and a second correction coefficient is used for correcting the preset time interval;

the first rechecking condition comprises that the rechecking similarity is smaller than a third preset similarity, and the second rechecking condition comprises that the rechecking similarity is larger than or equal to the third preset similarity.

Further, an effective processing image for judging that one type of fault exists in the machine tool is obtained and used as a judging processing image, the center coordinate of an abnormal area of the judging processing image is obtained, an abnormal element track is generated according to the motion condition of the center coordinate, a three-dimensional track model is generated according to the abnormal element track and time, track images corresponding to a preset number of time points in the three-dimensional track model are selected, the track images are compared with preset model images to obtain abnormal proportion, and three types of management modes are determined according to the abnormal proportion;

the first type of management mode is that when the abnormal proportion meets a first abnormal condition, the tool is judged to work normally;

the second type of management mode is that when the abnormal proportion accords with a second abnormal condition, the tool is judged to have abnormal risk in working, and a third correction coefficient is used for correcting the preset quantity;

the third type of management mode is that when the abnormal proportion accords with a third abnormal condition, the machine tool is judged to have second type of faults, and a fourth correction coefficient is used for correcting the preset quantity;

the preset number is greater than or equal to two, the first abnormal condition comprises that the abnormal proportion is equal to zero, the second abnormal condition comprises that the abnormal proportion is smaller than a first preset proportion, the third abnormal condition comprises that the abnormal proportion is greater than or equal to the first preset proportion, and the second type of fault comprises that a machine tool element is clamped.

In a second aspect, the present invention provides a machine tool fault early warning system based on processing figure analysis, including:

the information acquisition module is used for acquiring processing temperature information and processing image information of a machine tool according to a standard transmission speed, determining the safety level of the machine tool according to the processing temperature information, and determining an alarm mode according to the safety level;

the safety confirmation module is used for acquiring the number of pixels of a processed image when the machine tool meets a preset safety level, determining the image level of the processed image according to the number of pixels and processing the processed image according to the image level;

the image processing module is used for restoring the processed image to obtain a standard processed image when the processed image meets the preset image grade;

the image preprocessing module is used for carrying out time-frequency transformation on the standard processing image and adjusting the standard processing image in a frequency domain to obtain an adjusted processing image;

the similarity judging module is used for storing the gray value of the adjusted processing image in a target linked list, calculating the similarity of the target linked list and a preset standard linked list, judging the abnormal condition of the machine tool according to the similarity, and when the similarity meets the preset similarity condition, performing edge extraction on the adjusted processing image to obtain an effective processing image;

the rechecking module is used for calculating the similarity between the effective processing image and a preset risk image to serve as the rechecking similarity, and judging the abnormal condition of the machine tool according to the rechecking similarity;

and the track judging module is used for acquiring an effective processing image for judging that the machine tool has faults, generating a three-dimensional track model according to the judging processing image and time and judging the abnormal condition of the machine tool according to the three-dimensional track model.

Further, the safety confirmation module is configured to generate a machining thermal image according to the machining temperature information, acquire a grayscale image of the machining thermal image, generate a grayscale frequency histogram according to the grayscale image, divide the frequency histogram into three parts in a direction parallel to an ordinate at positions where abscissa is 85 and 175, where the three parts include a first part, a second part, and a third part, determine a safety level of the machine tool according to a frequency area of the three parts, and determine three types of alarm modes for the machine tool according to the safety level;

the second type of alarm mode is that when the frequency area of the second part meets a second area condition, the machine tool is judged to be at a second safety level, and the first alarm mode is used for alarming;

the first area condition includes that the frequency area of the first portion is the highest among the frequency areas of the three portions, the second area condition includes that the frequency area of the second portion is the highest among the frequency areas of the three portions, the third area condition includes that the frequency area of the third portion is the highest among the frequency areas of the three portions, the safety of the first safety level is higher than the safety of the second safety level, the safety of the second safety level is higher than the safety of the third safety level, and the urgency degree of the first alarm mode is lower than the urgency degree of the second alarm mode.

Compared with the prior art, the method has the advantages that the processing temperature information and the processing image information of the machine tool are obtained according to the standard transmission speed; determining the safety level of the machine tool according to the machining temperature information, determining an alarm mode for the machine tool according to the safety level, and judging the most urgent temperature fault firstly, so that the safety of the machine tool is ensured; when the machine tool meets a preset safety level, determining the image level of the processed image according to the processed image information, analyzing and processing according to the image level, obtaining standard image information, and carrying out enhanced denoising on an unclear image, so that the identification accuracy is improved; performing image preprocessing on the standard image information to obtain an adjusted and processed image, and improving the quality of all images; storing the adjusted and processed image in a target linked list, calculating the similarity of the target linked list and a preset standard linked list, and analyzing the abnormal condition of the element according to the similarity, thereby realizing the rapid judgment of the abnormal condition; adjusting the adjusted processing image which meets the preset similarity condition to obtain an effective processing image, rechecking the abnormal condition of the element, and improving the accuracy of fault identification through secondary judgment; and determining the abnormal condition of the element track of the machine tool according to the effective processing image, and identifying various abnormal scenes.

Further, generating a machining thermal image according to the machining temperature information, acquiring a grayscale image of the machining thermal image, generating a grayscale frequency histogram according to the grayscale image, dividing the frequency histogram into three parts along a direction parallel to an ordinate at positions with abscissa 85 and abscissa 175, wherein the three parts comprise a first part, a second part and a third part, determining the safety level of the machine tool according to the frequency area of the three parts, and determining three types of alarm modes for the machine tool according to the safety level; the first type of alarm mode is that when the frequency area of the first part meets a first area condition, the machine tool is judged to be at a first safety level, and no alarm is given; the second type of alarm mode is that when the frequency area of the second part meets a second area condition, the machine tool is judged to be in a second safety level, and the first alarm mode is used for alarming; the third type of alarm mode is that when the frequency area of the third part meets a third area condition, the machine tool is judged to be in a third safety level, the alarm is given by using the second alarm mode, and the machine tool is emergently braked; the first area condition comprises that the frequency area of the first part is highest in the frequency area of the third part, the second area condition comprises that the frequency area of the second part is highest in the frequency area of the third part, the third area condition comprises that the frequency area of the third part is highest in the frequency area of the third part, the safety of the first safety level is higher than the second safety level, the safety of the second safety level is higher than the third safety level, the emergency degree of the first alarm mode is lower than the emergency degree of the second alarm mode, safety accidents of the machine tool are effectively avoided, the most emergency temperature fault is judged firstly, the safety of the machine tool is guaranteed, and the technical effect of accurately and timely detecting the faults of the machine tool is achieved.

Further, when the machine tool meets a preset first safety level, acquiring the pixel number of the processed image according to the processed image information, determining the image level of the processed image according to the pixel number, and determining two types of processing modes according to the image level to obtain a standard processed image; the first type of processing mode is that when the number of the pixels meets a first pixel condition, the processed image is judged to be in a first image grade, and the processed image is subjected to image restoration to obtain the standard processed image; the second type of processing mode is that when the number of pixels meets a second pixel condition, the processed image is judged to be a second image grade, the processed image is directly used as the standard processed image, and a speed correction coefficient is used for correcting the standard transmission speed; the first pixel condition comprises that the number of the pixels is smaller than the preset number of the pixels, the first pixel condition comprises that the number of the pixels is larger than or equal to the preset number of the pixels, the definition of the first image grade is smaller than that of the second image grade, and the unclear image is subjected to enhanced denoising, so that the identification accuracy is improved, the safety of a machine tool is ensured, and the technical effect of accurately and timely detecting the fault of the machine tool is realized.

Further, when the processed image meets a preset first image grade, calculating an image variance of a gray value of the processed image, dividing the processed image into at least two local images, calculating a local variance of the gray value of the local images, and determining three types of image restoration modes of the processed image according to the image variance and the local images; the first type of image restoration method is to restore the processed image by using a first restoration method when the image variance and the local variance meet a first restoration condition; a second type of image restoration method is to restore the processed image by using a second restoration method when the image variance and the local variance meet a second restoration condition; a third type of image restoration method is to restore the processed image by using a third restoration method when the image variance and the local variance meet a third restoration condition; the first recovery condition comprises that the local variance is 0, the second recovery condition comprises that the local variance is not 0 and the local variance is not equal to the image variance, the third recovery condition comprises that the local variance is not 0 and the local variance is equal to the image variance, the first recovery mode comprises not changing the gray value of the local image, the second recovery mode comprises correcting the gray value of the local image by using a gray value correction coefficient, and the third recovery mode comprises changing the gray value of the local image to the local variance, so that the quality of all images is improved, the safety of a machine tool is guaranteed, and the technical effect of accurately and timely detecting the fault of the machine tool is realized.

Further, performing fourier transform on the standard processing image to obtain a processing spectrogram, determining three types of frequency adjustment modes for the processing spectrogram according to the processing spectrogram, adjusting frequencies in the processing spectrogram according to the frequency adjustment modes to obtain an adjusted spectrogram, and performing inverse fourier transform on the adjusted spectrogram to obtain an adjusted processing image; the first type of frequency adjustment mode is that when the processing spectrogram meets a first frequency condition, the processing spectrogram is adjusted by using a first adjustment mode; the second type of frequency adjustment mode is that when the processing spectrogram meets a second frequency condition, the processing spectrogram is adjusted by using a second adjustment mode; a third frequency adjustment mode is that when the processing spectrogram meets a third frequency condition, the processing spectrogram is adjusted by using a third adjustment mode; the first frequency condition includes that the frequency is less than a first preset frequency, the second frequency condition includes that the frequency is greater than or equal to the first preset frequency and less than a second preset frequency, the third frequency condition includes that the frequency is greater than or equal to the second preset frequency, the first preset frequency is less than the second preset frequency, the first adjustment mode includes using a denoising formula to denoise the frequency, and the denoising formula is that the frequency is denoised

X is an abscissa value of the processing spectrogram, y is an ordinate value of the processing spectrogram, D (x, y) is a distance from a point with the frequency spectrum coordinate (x, y) to the center of the spectrogram, and D1 is a cut-off frequency; the second adjustment mode comprises not adjusting the frequency, and the third adjustment mode comprisesIncludes performing enhancement processing on the frequency by using an enhancement formula

The method and the device have the advantages that the abnormal conditions can be judged quickly, the safety of the machine tool is guaranteed, and the technical effect of accurately and timely detecting the faults of the machine tool is achieved.

Further, obtaining a gray value of each pixel point in the adjusted and processed image according to a preset time interval, taking the gray value as a target gray value, setting a storage bit corresponding to each pixel point in a preset linked list, storing the target gray value in the storage bit to obtain a target linked list, calculating the similarity between the target linked list and a preset standard linked list, and determining three types of execution modes according to the similarity; the first type of execution mode is that when the similarity accords with a first similarity condition, the machine tool is judged to have a type of fault, and a first alarm mode is used for alarming; the second type of execution mode is that when the similarity accords with a second similarity condition, a preset adjustment mode is used for adjusting the adjusted processing image so as to obtain an effective processing image; the third type of execution mode is that when the similarity accords with a third similarity condition, the machine tool is judged to be normal, and a first correction coefficient is used for correcting the preset time interval; wherein the first similarity condition includes that the similarity is smaller than a first preset similarity, the second similarity condition includes that the similarity is greater than or equal to a first preset similarity and smaller than a second preset similarity, the third similarity condition includes that the similarity is greater than or equal to a second preset similarity, the first preset similarity is smaller than the second preset similarity, the one-class fault includes deformation of a machine tool element, the preset adjusting mode includes edge extraction of the adjusted and processed image by using an edge extraction formula, and the edge extraction formula is that the edge extraction formula is

Wherein m is the number of rows of the pixel points, n is the number of columns of the pixel points, and G (m, n) is the gray value of the mth row and the nth column in the adjustment processing imageAnd sigma is the average value of the pixel points, the secondary judgment improves the accuracy of fault identification, the safety of the machine tool is ensured, and the technical effect of accurately and timely detecting the fault of the machine tool is realized.

Further, calculating the similarity between the effective processing image and a preset risk image based on a neural network algorithm to serve as rechecking similarity, and determining two types of control modes according to the rechecking similarity; the first type of control mode is that when the rechecking similarity accords with a first rechecking condition, the machine tool is judged to have a type of fault, and a first alarm mode is used for alarming; the second type of control mode is that when the rechecking similarity accords with a second rechecking condition, the machine tool is judged to be normal, and a second correction coefficient is used for correcting the preset time interval; the first rechecking condition comprises that the rechecking similarity is smaller than a third preset similarity, and the second rechecking condition comprises that the rechecking similarity is larger than or equal to the third preset similarity, so that the safety of the machine tool is ensured, and the technical effect of accurately and timely detecting the fault of the machine tool is realized.

Further, an effective processing image for judging that one type of fault exists in the machine tool is obtained and used as a judging processing image, the center coordinate of an abnormal area of the judging processing image is obtained, an abnormal element track is generated according to the motion condition of the center coordinate, a three-dimensional track model is generated according to the abnormal element track and time, track images corresponding to a preset number of time points in the three-dimensional track model are selected, the track images are compared with preset model images to obtain abnormal proportion, and three types of management modes are determined according to the abnormal proportion; the first type of management mode is that when the abnormal proportion meets a first abnormal condition, the tool is judged to work normally; the second type of management mode is that when the abnormal proportion accords with a second abnormal condition, the tool is judged to have abnormal risk in working, and a third correction coefficient is used for correcting the preset quantity; the third type of management mode is that when the abnormal proportion accords with a third abnormal condition, the machine tool is judged to have two types of faults, and a fourth correction coefficient is used for correcting the preset quantity; the preset number is greater than or equal to two, the first abnormal condition comprises that the abnormal proportion is equal to zero, the second abnormal condition comprises that the abnormal proportion is smaller than a first preset proportion, the third abnormal condition comprises that the abnormal proportion is greater than or equal to the first preset proportion, the second fault comprises that machine tool elements are clamped, various abnormal scenes can be identified, the safety of the machine tool is guaranteed, and the technical effect of accurately and timely detecting the machine tool fault is realized.

Drawings

FIG. 1 is a flowchart of a machine tool fault warning method based on processing image analysis according to an embodiment of the present invention;

fig. 2 is a schematic structural diagram of a machine tool fault early warning system based on processing image analysis according to an embodiment of the present invention.

Detailed Description

In order that the objects and advantages of the invention will be more clearly understood, the invention is further described below with reference to examples; it should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

Preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only for explaining the technical principles of the present invention, and do not limit the scope of the present invention.

It should be noted that the terms "comprises" and "comprising," and any variations thereof, in the description and claims of the present invention and the above-described drawings, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.

Referring to fig. 1, which is a flowchart of a machine tool fault warning method based on processing image analysis according to an embodiment of the present invention, the machine tool fault warning method based on processing image analysis includes:

s101, acquiring processing temperature information and processing image information of a machine tool according to a standard transmission speed, determining the safety level of the machine tool according to the processing temperature information, and determining an alarm mode according to the safety level;

s102, when the machine tool accords with a preset safety level, acquiring the pixel number of a processed image, determining the image level of the processed image according to the pixel number, and then processing the processed image according to the image level;

s103, when the processed image meets the preset image grade, restoring the processed image to obtain a standard processed image;

s104, performing time-frequency transformation on the standard processing image and adjusting in a frequency domain to obtain an adjusted processing image;

s105, storing the gray value of the adjusted processing image in a target linked list, calculating the similarity of the target linked list and a preset standard linked list, judging the abnormal condition of the machine tool according to the similarity, and when the similarity meets the preset similarity condition, performing edge extraction on the adjusted processing image to obtain an effective processing image;

s106, calculating the similarity of the effective processing image and a preset risk image to serve as rechecking similarity, and judging the abnormal condition of the machine tool according to the rechecking similarity;

s107, obtaining an effective processing image for judging the fault of the machine tool as a judging processing image, generating a three-dimensional track model according to the judging processing image and time, and judging the abnormal condition of the machine tool according to the three-dimensional track model.

The invention has the advantages that the processing temperature information and the processing image information of the machine tool are obtained according to the standard transmission speed; determining the safety level of the machine tool according to the machining temperature information, determining an alarm mode for the machine tool according to the safety level, and judging the most urgent temperature fault firstly, so that the safety of the machine tool is ensured; when the machine tool meets a preset safety level, determining the image level of the processed image according to the processed image information, analyzing and processing according to the image level, obtaining standard image information, and performing enhanced denoising on an unclear image, so that the recognition accuracy is improved; performing image preprocessing on the standard image information to obtain an adjusted and processed image, and improving the quality of all images; storing the adjusted and processed image in a target linked list, calculating the similarity of the target linked list and a preset standard linked list, and analyzing the abnormal condition of the element according to the similarity, thereby realizing the rapid judgment of the abnormal condition; adjusting the adjusted processing image which meets the preset similarity condition to obtain an effective processing image, rechecking the abnormal condition of the element, and improving the accuracy of fault identification through secondary judgment; and determining the abnormal condition of the element track of the machine tool according to the effective processing image, and identifying various abnormal scenes.

Specifically, a machining thermal image is generated according to the machining temperature information, a gray level image of the machining thermal image is acquired, a gray level frequency histogram is generated according to the gray level image, the frequency histogram is divided into three parts along the direction parallel to the ordinate at positions with the abscissa of 85 and the abscissa of 175, the three parts comprise a first part, a second part and a third part, the safety level of the machine tool is determined according to the frequency area of the three parts, and three types of alarm modes for the machine tool are determined according to the safety level;

optionally, the ordinate of the grayscale frequency histogram is frequency, the abscissa of the grayscale frequency histogram is grayscale, the grayscale of 0 to 255 is divided into three equal parts, and the larger the frequency area of the part with the higher grayscale is, the larger the temperature range is, the larger the probability of the fault caused by abnormal operation of the machine tool is.

the third type of alarm mode is that when the frequency area of the third part meets a third area condition, the machine tool is judged to be at a third safety level, the alarm is carried out by using the second alarm mode, and the emergency brake is carried out on the machine tool;

Specifically, when the machine tool meets a preset first safety level, the number of pixels of the processed image is obtained according to the processed image information, the image level of the processed image is determined according to the number of pixels, and two processing modes are determined according to the image level to obtain a standard processed image;

the first type of processing mode is that when the number of the pixels meets a first pixel condition, the processed image is judged to be in a first image grade, and the processed image is subjected to image restoration to obtain the standard processed image;

the second type of processing mode is that when the number of pixels meets a second pixel condition, the processed image is judged to be a second image grade, the processed image is directly used as the standard processed image, and a speed correction coefficient is used for correcting the standard transmission speed;

Specifically, when the processed image meets a preset first image level, calculating the image variance of the gray value of the processed image, dividing the processed image into at least two local images, calculating the local variance of the gray value of the local images, and determining three image restoration modes for the processed image according to the image variance and the local images;

Specifically, performing fourier transform on the standard processing image to obtain a processing spectrogram, determining three types of frequency adjustment modes for the processing spectrogram according to the processing spectrogram, adjusting the frequency in the processing spectrogram according to the frequency adjustment modes to obtain an adjusted spectrogram, and performing inverse fourier transform on the adjusted spectrogram to obtain an adjusted processing image;

the third frequency adjustment mode is that when the processing spectrogram meets a third frequency condition, the processing spectrogram is adjusted by using a third adjustment mode;

X is an abscissa value of the processing spectrogram, y is an ordinate value of the processing spectrogram, D (x, y) is a distance from a point with the frequency spectrum coordinate (x, y) to the center of the spectrogram, and D1 is a cut-off frequency; the second adjustment mode comprises not adjusting the frequency, the third adjustment mode comprises enhancing the frequency by using an enhancement formula

。

Specifically, the gray value of each pixel point in the adjusted and processed image is obtained according to a preset time interval and is used as a target gray value, a storage bit corresponding to each pixel point is set in a preset linked list, the target gray value is stored in the storage bit to obtain a target linked list, the similarity of the target linked list and a preset standard linked list is calculated, and three types of execution modes are determined according to the similarity;

wherein the first similarity condition includes that the similarity is smaller than a first preset similarity, the second similarity condition includes that the similarity is greater than or equal to a first preset similarity and smaller than a second preset similarity, the third similarity condition includes that the similarity is greater than or equal to a second preset similarity, the first preset similarity is smaller than the second preset similarity, the one-class fault includes deformation of a machine tool element, the preset adjusting mode includes edge extraction of the adjusted and processed image by using an edge extraction formula, and the edge extraction formula is that the edge extraction formula is

Specifically, the similarity between the effective processing image and a preset risk image is calculated based on a neural network algorithm to serve as rechecking similarity, and two control modes are determined according to the rechecking similarity;

Specifically, an effective processing image for judging that one type of fault exists in the machine tool is obtained and used as a judging processing image, the center coordinate of an abnormal area of the judging processing image is obtained, an abnormal element track is generated according to the motion condition of the center coordinate, a three-dimensional track model is generated according to the abnormal element track and time, track images corresponding to a preset number of time points in the three-dimensional track model are selected, the track images are compared with preset model images to obtain abnormal proportion, and three types of management modes are determined according to the abnormal proportion;

the first type of management mode is that when the abnormal proportion accords with a first abnormal condition, the machine tool element is judged to work normally;

the second type of management mode is that when the abnormal proportion accords with a second abnormal condition, the abnormal risk of the machine tool element during working is judged, and a third correction coefficient is used for correcting the preset quantity;

the preset number is greater than or equal to two, the first abnormal condition comprises that the abnormal proportion is equal to zero, the second abnormal condition comprises that the abnormal proportion is smaller than a first preset proportion, the third abnormal condition comprises that the abnormal proportion is greater than or equal to the first preset proportion, and the second fault comprises that a machine tool element is clamped.

Referring to fig. 2, a schematic structural diagram of a machine tool fault early warning system based on processing image analysis according to an embodiment of the present invention is shown, where the machine tool fault early warning system based on processing image analysis includes:

the information acquisition module 201 is used for acquiring processing temperature information and processing image information of a machine tool according to a standard transmission speed, determining the safety level of the machine tool according to the processing temperature information, and determining an alarm mode according to the safety level;

the safety confirmation module 202 is configured to, when the machine tool meets a preset safety level, acquire the number of pixels of a processed image, determine an image level of the processed image according to the number of pixels, and process the processed image according to the image level;

the image processing module 203 is configured to restore the processed image to obtain a standard processed image when the processed image meets a preset image grade;

the image preprocessing module 204 is configured to perform time-frequency transformation on the standard processing image and perform adjustment in a frequency domain to obtain an adjusted processing image;

a similarity judging module 205, configured to store the gray value of the adjusted processed image in a target linked list, calculate a similarity between the target linked list and a preset standard linked list, judge an abnormal condition of the machine tool according to the similarity, and perform edge extraction on the adjusted processed image when the similarity meets a preset similarity condition, so as to obtain an effective processed image;

the rechecking module 206 is configured to calculate a similarity between the effective processing image and a preset risk image, use the similarity as a rechecking similarity, and judge an abnormal condition of the machine tool according to the rechecking similarity;

and the track judging module 207 is used for acquiring an effective processing image for judging that the machine tool has a fault, generating a three-dimensional track model according to the judging processing image and time, and judging the abnormal condition of the machine tool according to the three-dimensional track model.

Further, the safety confirmation module 202 is configured to generate a machining thermal image according to the machining temperature information, acquire a grayscale image of the machining thermal image, generate a grayscale frequency histogram according to the grayscale image, divide the frequency histogram into three parts along a direction parallel to an ordinate at positions where abscissa is 85 and 175, where the three parts include a first part, a second part, and a third part, determine a safety level of the machine tool according to a frequency area of the three parts, and determine three types of alarm modes for the machine tool according to the safety level;

It should be noted that, in the embodiment of the machine tool fault early warning system based on processing image analysis, the included units and modules are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be realized; in addition, the specific names of the functional units are only for the convenience of distinguishing from each other, and are not used for limiting the protection scope of the present invention.

So far, the technical solutions of the present invention have been described in connection with the preferred embodiments shown in the drawings, but it is apparent to those skilled in the art that the scope of the present invention is not limited to these specific embodiments. Equivalent changes or substitutions of related technical features can be made by those skilled in the art without departing from the principle of the invention, and the technical scheme after the changes or substitutions can be within the protection scope of the invention.

The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention; various modifications and alterations to this invention will become apparent to those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A machine tool fault early warning method based on processing portrait analysis is characterized by comprising the following steps: acquiring processing temperature information and processing image information of a machine tool according to a standard transmission speed, determining the safety level of the machine tool according to the processing temperature information, and determining an alarm mode according to the safety level; when the machine tool meets a preset safety level, acquiring the number of pixels of a processed image, determining the image level of the processed image according to the number of pixels, and processing the processed image according to the image level; when the processed image meets the preset image grade, restoring the processed image to obtain a standard processed image; carrying out time-frequency transformation on the standard processing image and adjusting in a frequency domain to obtain an adjusted processing image; storing the gray value of the adjusted processing image in a target linked list, calculating the similarity of the target linked list and a preset standard linked list, judging the abnormal condition of the machine tool according to the similarity, and when the similarity meets the preset similarity condition, performing edge extraction on the adjusted processing image to obtain an effective processing image; calculating the similarity between the effective processing image and a preset risk image to serve as rechecking similarity, and judging the abnormal condition of the machine tool according to the rechecking similarity; and acquiring an effective processing image for judging that the machine tool has a fault, taking the effective processing image as a judging processing image, generating a three-dimensional track model according to the judging processing image and time, and judging the abnormal condition of the machine tool according to the three-dimensional track model.

2. The machine tool fault early warning method based on processing portrait analysis of claim 1, characterized in that, generating a processing thermal image according to the processing temperature information, acquiring a gray scale image of the processing thermal image, generating a gray scale frequency histogram according to the gray scale image, dividing the frequency histogram into three parts along a direction parallel to an ordinate at abscissa 85 and 175, wherein the three parts comprise a first part, a second part and a third part, determining the safety level of the machine tool according to the frequency area of the three parts, and determining three types of alarm modes for the machine tool according to the safety level; the first type of alarm mode is that when the frequency area of the first part meets a first area condition, the machine tool is judged to be at a first safety level, and no alarm is given; the second type of alarm mode is that when the frequency area of the second part meets a second area condition, the machine tool is judged to be in a second safety level, and the first alarm mode is used for alarming; the third type of alarm mode is that when the frequency area of the third part meets a third area condition, the machine tool is judged to be at a third safety level, the alarm is carried out by using the second alarm mode, and the emergency brake is carried out on the machine tool; the first area condition includes that the frequency area of the first portion is highest in the frequency area of the third portion, the second area condition includes that the frequency area of the second portion is highest in the frequency area of the third portion, the third area condition includes that the frequency area of the third portion is highest in the frequency area of the third portion, the safety of the first safety level is greater than the safety of the second safety level, the safety of the second safety level is greater than the safety of the third safety level, and the urgency degree of the first alarm mode is less than the urgency degree of the second alarm mode.

3. The machine tool fault early warning method based on processing portrait analysis of claim 2, characterized in that when the machine tool meets a preset first safety level, the number of pixels of the processing image is obtained according to the processing image information, the image level of the processing image is determined according to the number of pixels, and two types of processing modes are determined according to the image level to obtain a standard processing image; the first type of processing mode is that when the number of the pixels meets a first pixel condition, the processed image is judged to be in a first image grade, and the processed image is subjected to image restoration to obtain the standard processed image; the second type of processing mode is that when the number of pixels meets a second pixel condition, the processed image is judged to be a second image grade, the processed image is directly used as the standard processed image, and a speed correction coefficient is used for correcting the standard transmission speed; the first pixel condition comprises that the number of pixels is smaller than a preset number of pixels, the first pixel condition comprises that the number of pixels is larger than or equal to the preset number of pixels, and the definition of the first image level is smaller than that of the second image level.

4. The machine tool fault early warning method based on processing portrait analysis of claim 3, characterized in that when the processed image meets a preset first image level, the image variance of the gray value of the processed image is calculated, the processed image is divided into at least two partial images, the local variance of the gray value of the partial images is calculated, and three types of image restoration modes for the processed image are determined according to the image variance and the partial images; the first type of image restoration method is that when the image variance and the local variance accord with a first restoration condition, a first restoration method is used for restoring the processed image; a second type of image restoration method is to restore the processed image by using a second restoration method when the image variance and the local variance meet a second restoration condition; a third type of image restoration method is to restore the processed image by using a third restoration method when the image variance and the local variance meet a third restoration condition; the first recovery condition includes that the local variance is 0, the second recovery condition includes that the local variance is not 0 and the local variance is not equal to the image variance, the third recovery condition includes that the local variance is not 0 and the local variance is equal to the image variance, the first recovery mode includes not changing the gray value of the local image, the second recovery mode includes correcting the gray value of the local image by using a gray value correction coefficient, and the third recovery mode includes changing the gray value of the local image to the local variance.

5. According to the rightThe machine tool fault early warning method based on processing portrait analysis of claim 4, characterized in that, the standard processing image is Fourier-transformed to obtain a processing spectrogram, three types of frequency adjustment modes for the processing spectrogram are determined according to the processing spectrogram, the frequency in the processing spectrogram is adjusted according to the frequency adjustment modes to obtain an adjusted spectrogram, and the adjusted spectrogram is Fourier-inverted to obtain an adjusted processing image; the first type of frequency adjustment mode is that when the processing spectrogram meets a first frequency condition, the processing spectrogram is adjusted by using a first adjustment mode; the second type of frequency adjustment mode is that when the processing spectrogram meets a second frequency condition, the processing spectrogram is adjusted by using a second adjustment mode; the third frequency adjustment mode is that when the processing spectrogram meets a third frequency condition, the processing spectrogram is adjusted by using a third adjustment mode; the first frequency condition includes that the frequency is smaller than a first preset frequency, the second frequency condition includes that the frequency is greater than or equal to the first preset frequency and smaller than a second preset frequency, the third frequency condition includes that the frequency is greater than or equal to the second preset frequency, the first preset frequency is smaller than the second preset frequency, the first adjustment mode includes that the frequency is denoised by using a denoising formula, and the denoising formula is that the frequency is denoised by using a denoising formula

。

6. According toThe machine tool failure early warning method based on processing image analysis as claimed in claim 5, wherein the gray value of each pixel point in the adjusted processing image is obtained according to a preset time interval and is used as a target gray value, a storage bit corresponding to each pixel point is set in a preset linked list, the target gray value is stored in the storage bit to obtain a target linked list, the similarity of the target linked list and a preset standard linked list is calculated, and three types of execution modes are determined according to the similarity; the first type of execution mode is that when the similarity accords with a first similarity condition, the machine tool is judged to have a type of fault, and a first alarm mode is used for alarming; the second type of execution mode is that when the similarity accords with a second similarity condition, a preset adjustment mode is used for adjusting the adjusted processing image so as to obtain an effective processing image; the third type of execution mode is that when the similarity accords with a third similarity condition, the machine tool is judged to be normal, and a first correction coefficient is used for correcting the preset time interval; wherein the first similarity condition includes that the similarity is smaller than a first preset similarity, the second similarity condition includes that the similarity is greater than or equal to the first preset similarity and smaller than a second preset similarity, the third similarity condition includes that the similarity is greater than or equal to the second preset similarity, the first preset similarity is smaller than the second preset similarity, the first type of fault includes deformation of a machine tool element, the preset adjusting mode includes edge extraction of the adjusted and processed image by using an edge extraction formula, and the edge extraction formula is that

7. The machine tool fault early warning method based on processing portrait analysis according to claim 6, characterized in that, similarity of the effective processing image and the preset risk image is calculated based on neural network algorithm as recheck similarity, and two types of control modes are determined according to the recheck similarity; the first type of control mode is that when the rechecking similarity accords with a first rechecking condition, the machine tool is judged to have a type of fault, and a first alarm mode is used for alarming; the second type of control mode is that when the rechecking similarity accords with a second rechecking condition, the machine tool is judged to be normal, and a second correction coefficient is used for correcting the preset time interval; the first rechecking condition comprises that the rechecking similarity is smaller than a third preset similarity, and the second rechecking condition comprises that the rechecking similarity is larger than or equal to the third preset similarity.

8. The machine tool fault early warning method based on processing portrait analysis according to claim 7, wherein effective processing images for judging that one type of fault exists in the machine tool are obtained as judgment processing images, center coordinates of abnormal areas of the judgment processing images are obtained, abnormal element tracks are generated according to motion conditions of the center coordinates, a three-dimensional track model is generated according to the abnormal element tracks and time, track images corresponding to a preset number of time points in the three-dimensional track model are selected, the track images are compared with preset model images to obtain abnormal proportions, and three types of management modes are determined according to the abnormal proportions; the first type of management mode is that when the abnormal proportion accords with a first abnormal condition, the machine tool element is judged to work normally; the second type of management mode is that when the abnormal proportion accords with a second abnormal condition, the abnormal risk of the machine tool element is judged to exist, and a third correction coefficient is used for correcting the preset quantity; the third type of management mode is that when the abnormal proportion accords with a third abnormal condition, the machine tool is judged to have second type of faults, and a fourth correction coefficient is used for correcting the preset quantity; the preset number is an integer greater than or equal to 2, the first abnormal condition comprises that the abnormal proportion is equal to zero, the second abnormal condition comprises that the abnormal proportion is smaller than a first preset proportion, the third abnormal condition comprises that the abnormal proportion is greater than or equal to the first preset proportion, and the second type of fault comprises that a machine tool element is blocked.

9. The utility model provides a lathe trouble early warning system based on processing portrait analysis which characterized in that includes: the information acquisition module is used for acquiring processing temperature information and processing image information of a machine tool according to a standard transmission speed, determining the safety level of the machine tool according to the processing temperature information, and determining an alarm mode according to the safety level; the safety confirmation module is used for acquiring the number of pixels of a processed image when the machine tool meets a preset safety level, determining the image level of the processed image according to the number of pixels and processing the processed image according to the image level; the image processing module is used for restoring the processed image to obtain a standard processed image when the processed image meets the preset image grade; the image preprocessing module is used for carrying out time-frequency transformation on the standard processing image and adjusting the standard processing image in a frequency domain to obtain an adjusted processing image; the similarity judging module is used for storing the gray value of the adjusted processing image in a target linked list, calculating the similarity of the target linked list and a preset standard linked list, judging the abnormal condition of the machine tool according to the similarity, and when the similarity meets a preset similarity condition, performing edge extraction on the adjusted processing image to obtain an effective processing image; the rechecking module is used for calculating the similarity between the effective processing image and a preset risk image to serve as the rechecking similarity, and judging the abnormal condition of the machine tool according to the rechecking similarity; and the track judging module is used for acquiring an effective processing image for judging that the machine tool has faults, generating a three-dimensional track model according to the judging processing image and time and judging the abnormal condition of the machine tool according to the three-dimensional track model.

10. The machine tool failure early warning system based on processing image analysis according to claim 9, comprising: the safety confirmation module is used for generating a machining thermal image according to the machining temperature information, acquiring a gray level image of the machining thermal image, generating a gray level frequency histogram according to the gray level image, dividing the frequency histogram into three parts along a direction parallel to a vertical coordinate at positions with abscissa 85 and 175, wherein the three parts comprise a first part, a second part and a third part, determining the safety level of the machine tool according to the frequency area of the three parts, and determining three types of alarm modes for the machine tool according to the safety level; the first type of alarm mode is that when the frequency area of the first part meets a first area condition, the machine tool is judged to be at a first safety level, and no alarm is given; the second type of alarm mode is that when the frequency area of the second part meets a second area condition, the machine tool is judged to be in a second safety level, and the first alarm mode is used for alarming; the third type of alarm mode is that when the frequency area of the third part meets a third area condition, the machine tool is judged to be in a third safety level, the alarm is given by using the second alarm mode, and the machine tool is emergently braked; the first area condition includes that the frequency area of the first portion is highest in the frequency area of the third portion, the second area condition includes that the frequency area of the second portion is highest in the frequency area of the third portion, the third area condition includes that the frequency area of the third portion is highest in the frequency area of the third portion, the safety of the first safety level is greater than the safety of the second safety level, the safety of the second safety level is greater than the safety of the third safety level, and the urgency degree of the first alarm mode is less than the urgency degree of the second alarm mode.