CN117723739A - Quality analysis method and system for low-carbon lubricating oil - Google Patents
Quality analysis method and system for low-carbon lubricating oil Download PDFInfo
- Publication number
- CN117723739A CN117723739A CN202311715868.5A CN202311715868A CN117723739A CN 117723739 A CN117723739 A CN 117723739A CN 202311715868 A CN202311715868 A CN 202311715868A CN 117723739 A CN117723739 A CN 117723739A
- Authority
- CN
- China
- Prior art keywords
- image
- lubricating oil
- abnormal
- quality
- analysis
- 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
- 239000010687 lubricating oil Substances 0.000 title claims abstract description 180
- 238000004458 analytical method Methods 0.000 title claims abstract description 88
- 229910052799 carbon Inorganic materials 0.000 title claims abstract description 26
- 230000002159 abnormal effect Effects 0.000 claims abstract description 64
- 238000012545 processing Methods 0.000 claims abstract description 50
- 239000000126 substance Substances 0.000 claims abstract description 31
- 238000001514 detection method Methods 0.000 claims abstract description 26
- 238000004519 manufacturing process Methods 0.000 claims abstract description 25
- 238000005516 engineering process Methods 0.000 claims abstract description 17
- 238000012544 monitoring process Methods 0.000 claims abstract description 14
- 238000012795 verification Methods 0.000 claims abstract description 6
- 238000000034 method Methods 0.000 claims description 45
- 239000002245 particle Substances 0.000 claims description 37
- 238000009826 distribution Methods 0.000 claims description 24
- 230000005856 abnormality Effects 0.000 claims description 12
- 239000011159 matrix material Substances 0.000 claims description 10
- 238000012549 training Methods 0.000 claims description 10
- 238000004364 calculation method Methods 0.000 claims description 7
- 238000000605 extraction Methods 0.000 claims description 7
- 238000007781 pre-processing Methods 0.000 claims description 7
- 238000003708 edge detection Methods 0.000 claims description 6
- 230000000877 morphologic effect Effects 0.000 claims description 6
- 238000006243 chemical reaction Methods 0.000 claims description 5
- 238000012937 correction Methods 0.000 claims description 5
- 238000013135 deep learning Methods 0.000 claims description 5
- 239000003921 oil Substances 0.000 claims description 5
- 238000003672 processing method Methods 0.000 claims description 5
- 238000012935 Averaging Methods 0.000 claims description 4
- 239000002253 acid Substances 0.000 claims description 4
- 230000002708 enhancing effect Effects 0.000 claims description 4
- 238000003062 neural network model Methods 0.000 claims description 4
- 238000000513 principal component analysis Methods 0.000 claims description 4
- 238000005070 sampling Methods 0.000 claims description 4
- 230000006870 function Effects 0.000 claims description 3
- 238000002372 labelling Methods 0.000 claims description 3
- 239000000314 lubricant Substances 0.000 claims description 3
- 230000000875 corresponding effect Effects 0.000 description 6
- 238000004422 calculation algorithm Methods 0.000 description 3
- 230000008859 change Effects 0.000 description 3
- 238000002329 infrared spectrum Methods 0.000 description 3
- OKTJSMMVPCPJKN-UHFFFAOYSA-N Carbon Chemical compound [C] OKTJSMMVPCPJKN-UHFFFAOYSA-N 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 238000001228 spectrum Methods 0.000 description 2
- 241001584785 Anavitrinella pampinaria Species 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 230000000295 complement effect Effects 0.000 description 1
- 238000012790 confirmation Methods 0.000 description 1
- 230000007797 corrosion Effects 0.000 description 1
- 238000005260 corrosion Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000001066 destructive effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005265 energy consumption Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000010191 image analysis Methods 0.000 description 1
- 238000003709 image segmentation Methods 0.000 description 1
- 238000009776 industrial production Methods 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 230000004807 localization Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Landscapes
- Image Analysis (AREA)
Abstract
The invention discloses a quality analysis method and a system of low-carbon lubricating oil, wherein high-definition camera equipment is adopted to collect real-time images of the lubricating oil on a production line; processing the acquired image and extracting characteristic parameters of the lubricating oil image; establishing a reference model, and taking the image characteristics of normal lubricating oil as a standard; performing anomaly detection and classification on the lubricating oil image by using a computer vision technology, and finding out image features which are inconsistent with the reference model; combining the image processing analysis result with the chemical analysis data, further confirming the abnormal image, and verifying whether the lubricating oil quality problem exists; when the verification finds that the quality of the lubricating oil is abnormal, an alarm is sent to an operator or a monitoring system through an alarm device; the quality analysis method of the low-carbon lubricating oil can be used for monitoring in real time, extracting multidimensional characteristic parameters for analysis, combining image processing analysis with chemical analysis data, improving the judgment accuracy and reducing the false alarm rate.
Description
Technical Field
The invention relates to a quality analysis method and a quality analysis system for low-carbon lubricating oil.
Background
The low-carbon lubricating oil is environment-friendly and is mainly characterized by being capable of reducing negative influence on environment and reducing carbon emission in the use process. The low-carbon lubricating oil is widely applied in the fields of industrial production, transportation and the like, and is an important component for sustainable development and environmental protection. It can help reduce carbon emissions, reduce energy consumption, and have less impact on the environment.
The existing quality detection method of the lubricating oil is shown as a quick detection method of the quality of new lubricating oil disclosed in application number 201010136894.9, and the method disclosed by the invention comprises the following steps: (1) Collecting a representative lubricating oil sample as a training set; (2) determining the infrared spectrum of the training set lubricating oil sample; and performing corresponding pretreatment, wherein the pretreated spectrum data is used as a variable; (3) Selecting a proper multielement correction method, and establishing a relation model between a new oil quality index of the lubricating oil and a spectrum; (4) For quality detection of unknown lubricating oil samples, firstly, the infrared spectrum is measured, the same pretreatment is carried out, then the lubricating oil quality index is measured by utilizing a lubricating oil quality analysis model, the sample is required to be collected and then tested according to the scheme, real-time monitoring cannot be carried out, data analysis is carried out mainly by relying on the infrared spectrum, the characteristic latitude is relatively low, and the accuracy is insufficient.
Disclosure of Invention
Aiming at the defects in the prior art, the invention aims to provide the quality analysis method and the system for the low-carbon lubricating oil, which can be used for carrying out real-time monitoring, extracting multidimensional characteristic parameters for analysis, combining image processing analysis with chemical analysis data, improving the judgment accuracy and reducing the false alarm rate.
The technical scheme adopted for solving the technical problems is as follows:
a quality analysis method of low-carbon lubricating oil comprises the following steps:
adopting high-definition camera equipment to collect real-time images of lubricating oil on a production line;
processing the acquired image, including denoising and enhancing contrast, and extracting characteristic parameters of the lubricating oil image, including color distribution and particle shape;
establishing a reference model, and taking the image characteristics of normal lubricating oil as a standard;
using computer vision technology to detect and classify the lubricating oil image, finding out the image characteristics which are not in accordance with the reference model, including abnormal color and uneven particles;
combining the image processing analysis result with the chemical analysis data, further confirming the abnormal image, and verifying whether the lubricating oil quality problem exists;
when the verification finds that the quality of the lubricating oil is abnormal, an alarm is sent out to an operator or a monitoring system through an alarm device, and an abnormal image and an analysis result are fed back to production management staff.
Preferably, the method for extracting the characteristic parameters of the lubricating oil image comprises the following steps:
converting the image into a color space by using a color space conversion technology, and then calculating the pixel duty ratio of each color in each region by dividing or selecting the image, thereby obtaining the color distribution characteristics of the lubricating oil image;
processing the image by using a morphological processing method to highlight particles in the lubricating oil, and then extracting shape features of the lubricating oil particles by a contour analysis method;
the uniformity of the lubricating oil image is evaluated by calculating the gradient using the gray information of the image.
Preferably, the method for evaluating the uniformity of the lubricating oil image by calculating the gradient is as follows:
converting the lubricating oil image into a gray scale image;
calculating gradients in the horizontal direction and the vertical direction of the gray image by using a Sobel operator;
averaging or summing the gradient images in the horizontal direction and the vertical direction to obtain a comprehensive gradient image;
calculating the mean or variance of the comprehensive gradient image;
and judging the uniformity of the lubricating oil image according to the mean value or the variance.
Preferably, the method for calculating the gradient of the gray image in the horizontal direction and the vertical direction by using the Sobel operator is as follows:
defining two convolution kernels of the Sobel operator, wherein the two convolution kernels respectively represent gradient calculation in the horizontal direction and gradient calculation in the vertical direction, and the horizontal direction convolution kernel (Gx) and the vertical direction convolution kernel (Gy) are respectively:
Gx＝[[-1,0,1],
[-2,0,2],
[-1,0,1]]
Gy＝[[-1,-2,-1],
[0,0,0],
[1,2,1]]
performing two-dimensional convolution operation on the gray level image by using Gx and Gy respectively to obtain gradient approximate values in the horizontal direction and the vertical direction;
calculating the gradient amplitude in the horizontal and vertical directions for each pixel point, i.e
gradient_magnitude＝sqrt(Gx^2+Gy^2)
And gradient direction, i.e.
gradient_direction＝arctan2(Gy,Gx)
Wherein arctan2 is an arctangent function, the correct gradient direction can be obtained from the symbols of Gy and Gx.
Preferably, the method for establishing the reference model comprises the following steps:
collecting a plurality of representative lubricating oil image data, including a lubricating oil image in a normal state and an image under abnormal color and uneven particles;
preprocessing the acquired lubricating oil image data, including denoising, color correction and size standardization operation, so as to ensure the quality and consistency of the data;
extracting characteristic parameters including color distribution characteristics, particle shape characteristics and oil uniformity characteristics from the preprocessed image data by utilizing image processing and computer vision technologies;
labeling the extracted characteristic parameters, marking the lubricating oil image in a normal state as normal, and marking the image in an abnormal state as a corresponding abnormal type;
training by using the marked characteristic parameter data and using a deep learning neural network model, and establishing a reference model of the lubricating oil quality.
Preferably, the method for extracting the characteristic parameters from the preprocessed image data comprises the following steps:
extracting color information in the image using the color histogram;
analyzing texture information of the image by using the gray level co-occurrence matrix;
describing shape features of objects in the image using edge detection;
selecting and dimension-reducing the extracted features by using principal component analysis;
the extracted feature parameters are represented as a computer-processable feature matrix.
Preferably, the method for abnormality detection and classification of the lubricating oil image is as follows:
the extracted characteristic parameters are input into a reference model, the model is utilized to carry out anomaly detection, whether the anomaly exists or not is judged by comparing the difference degree between the characteristic parameters of the image to be detected and normal characteristic distribution, and the image is divided into a color anomaly image and a particle non-uniformity image according to the anomaly type.
Preferably, the method of combining the results of the image processing analysis with the chemical analysis data is:
sampling the lubricating oil with the abnormal image, and analyzing chemical components of the sampled lubricating oil to obtain detection data of viscosity, acid value and base number indexes of the lubricating oil;
comparing the detection data of the lubricating oil with the standard data, judging whether the detection data is in the standard data interval, and if not, determining that the index of the lubricating oil is abnormal.
Another technical problem to be solved by the present invention is to provide a quality analysis system for low-carbon lubricating oil, comprising:
the image acquisition module is used for acquiring the real-time image of the lubricating oil on the production line and comprises high-definition camera equipment;
the image processing module is used for preprocessing the acquired image and comprises a denoising and contrast enhancement module;
the feature extraction module is used for extracting feature parameters such as color distribution, particle shape and the like of the lubricating oil image;
the reference model building module is used for taking the image characteristics of the normal lubricating oil as a standard to build a reference model;
computer vision module for detecting and classifying the abnormality of lubricating oil image by computer vision technique to find out the image characteristics different from reference model including abnormal color and uneven particles
The image processing analysis module is used for combining the result of the image processing analysis with the chemical analysis data to further confirm the abnormal image;
and the feedback module is used for feeding back the abnormal image and the analysis result to production management personnel.
Preferably, an alarm device is also included to alert an operator or a monitoring system when the abnormal quality of the lubricating oil is verified.
The beneficial effects of the invention are as follows:
the high-definition camera equipment is utilized to collect real-time images of the lubricating oil on the production line, so that the state of the lubricating oil can be monitored at any time; the lubricating oil images are analyzed by image processing and computer vision techniques without the need for destructive sampling and chemical analysis. Thus, the interference and the cost to the production process can be reduced; the color distribution, particle shape and other multidimensional characteristic parameters of the lubricating oil image can be extracted, and the anomaly detection and classification can be carried out by assisting with the computer vision technology; and combining the result of the image processing analysis with the chemical analysis data to further confirm and verify the abnormal image. Thus, the accuracy of judgment can be improved, and the false alarm rate is reduced; when the quality of the lubricating oil is abnormal, an alarm can be sent out to an operator or a monitoring system in time through an alarm device, and an abnormal image and an analysis result are fed back to production management staff. This helps take measures in time, avoiding further quality problems.
Drawings
FIG. 1 is a schematic flow chart of a quality analysis method of low-carbon lubricating oil of the invention;
FIG. 2 is a schematic block diagram of a method for mass analysis of low-carbon lubricating oil according to the present invention.
Detailed Description
The principles and features of the present invention are described below with reference to the drawings, the examples are illustrated for the purpose of illustrating the invention and are not to be construed as limiting the scope of the invention. The invention is more particularly described by way of example in the following paragraphs with reference to the drawings. Advantages and features of the invention will become more apparent from the following description and from the claims. It should be noted that the drawings are in a very simplified form and are all to a non-precise scale, merely for convenience and clarity in aiding in the description of embodiments of the invention.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and/or" as used herein includes any and all combinations of one or more of the associated listed items.
Examples
Referring to fig. 1, a quality analysis method of low-carbon lubricating oil comprises the following steps:
adopting high-definition camera equipment to collect real-time images of lubricating oil on a production line;
processing the acquired image, including denoising and enhancing contrast, and extracting characteristic parameters of the lubricating oil image, including color distribution and particle shape;
establishing a reference model, and taking the image characteristics of normal lubricating oil as a standard;
using computer vision technology to detect and classify the lubricating oil image, finding out the image characteristics which are not in accordance with the reference model, including abnormal color and uneven particles;
combining the image processing analysis result with the chemical analysis data, further confirming the abnormal image, and verifying whether the lubricating oil quality problem exists;
when the verification finds that the quality of the lubricating oil is abnormal, an alarm is sent out to an operator or a monitoring system through an alarm device, and an abnormal image and an analysis result are fed back to production management staff.
And installing high-definition camera equipment on a production line, and collecting the real-time image of the lubricating oil. Ensuring that the position and angle of the image pickup apparatus can cover the entire area of the lubricating oil; processing the acquired images, including removing noise, enhancing contrast, etc., which can improve the accuracy of subsequent feature extraction and anomaly detection;
based on the processed image, characteristic parameters of the lubricating oil image, including color distribution, particle shape, and the like, are extracted. Features can be extracted by computer vision techniques such as image segmentation, edge detection, etc.; and taking the image characteristics of the normal lubricating oil as a reference, and establishing a reference model. The features may be trained and modeled using machine learning or deep learning methods for subsequent anomaly detection and classification;
and (3) performing anomaly detection and classification on the lubricating oil image by using a reference model and a computer vision technology. By comparing the difference between the collected image characteristics and the reference model, whether the lubricating oil has quality abnormality, such as abnormal color or uneven particles, can be judged; and fusing the image processing analysis result with the chemical analysis data. Chemical component analysis can be performed on lubricating oil by using a professional instrument, and chemical analysis data and an image analysis result are combined to improve the accuracy and reliability of judgment;
further confirmation and verification are performed on the lubricating oil image determined to be abnormal. If the verification finds that the lubricating oil has quality problems, an alarm device sends an alarm signal to an operator or a monitoring system so as to take measures in time to treat abnormal conditions; and feeding back the abnormal image and the analysis result to production management personnel. The manager can process and adjust in time according to the information, so as to ensure that the quality of the lubricating oil meets the requirements.
The method for extracting the characteristic parameters of the lubricating oil image comprises the following steps:
converting the image into a color space by using a color space conversion technology, and then calculating the pixel duty ratio of each color in each region by dividing or selecting the image, thereby obtaining the color distribution characteristics of the lubricating oil image;
processing the image by using a morphological processing method to highlight particles in the lubricating oil, and then extracting shape features of the lubricating oil particles by a contour analysis method;
the uniformity of the lubricating oil image is evaluated by calculating the gradient using the gray information of the image.
The quality condition of the lubricating oil image can be comprehensively reflected through the extraction of characteristic parameters in various aspects such as color distribution characteristics, particle shape characteristics, image uniformity and the like, and the quality of the lubricating oil can be comprehensively evaluated; the color space conversion technology and the morphological processing method are adopted, so that the method can adapt to different types of lubricating oil images, and can be adjusted and optimized according to different lubricating oil quality requirements; by utilizing the image processing and analyzing technology, the characteristic parameters can be accurately extracted and calculated, and the accurate evaluation of the image quality of the lubricating oil is improved.
The lube image is converted to a specific color space such as RGB, CMYK, or HSV. Then dividing or selecting the image, and calculating the pixel duty ratio of each color in each area, thereby obtaining the color distribution characteristics of the lubricating oil image.
The image is processed using morphological processing methods to highlight particles in the lubricating oil. Morphological processing operations such as corrosion, expansion, open operation or close operation can be adopted, so that particles are more outstanding and clear. And then extracting the shape characteristics of the lubricating oil particles by a contour analysis method.
The uniformity of the lubricating oil image is evaluated by calculating the gradient using the gray information of the image. A gradient calculation algorithm may be employed to analyze the gray scale variation of the image to evaluate the uniformity of the lubricating oil image.
The method for evaluating the uniformity of the lubricating oil image by calculating the gradient comprises the following steps:
converting the lubricating oil image into a gray scale image;
calculating gradients in the horizontal direction and the vertical direction of the gray image by using a Sobel operator;
averaging or summing the gradient images in the horizontal direction and the vertical direction to obtain a comprehensive gradient image;
calculating the mean or variance of the comprehensive gradient image;
and judging the uniformity of the lubricating oil image according to the mean value or the variance.
Preferably, the method for calculating the gradient of the gray image in the horizontal direction and the vertical direction by using the Sobel operator is as follows:
defining two convolution kernels of the Sobel operator, wherein the two convolution kernels respectively represent gradient calculation in the horizontal direction and gradient calculation in the vertical direction, and the horizontal direction convolution kernel (Gx) and the vertical direction convolution kernel (Gy) are respectively:
Gx＝[[-1,0,1],
[-2,0,2],
[-1,0,1]]
Gy＝[[-1,-2,-1],
[0,0,0],
[1,2,1]]
performing two-dimensional convolution operation on the gray level image by using Gx and Gy respectively to obtain gradient approximate values in the horizontal direction and the vertical direction;
calculating the gradient amplitude in the horizontal and vertical directions for each pixel point, i.e
gradient_magnitude＝sqrt(Gx^2+Gy^2)
And gradient direction, i.e.
gradient_direction＝arctan2(Gy,Gx)
Wherein arctan2 is an arctangent function, the correct gradient direction can be obtained from the symbols of Gy and Gx.
After the image is converted into a gray image, the gradient in the horizontal direction and the gradient in the vertical direction are calculated by using a Sobel operator, so that the amplitude of the color change in the image can be intuitively reflected, and the uniformity of the lubricating oil image is evaluated; the gradient amplitude and the gradient direction can be efficiently calculated by utilizing convolution operation and mathematical operation, and complex image processing algorithm and a large amount of calculation resources are not needed; by calculating the mean value or variance of the gradient amplitude, a quantitative index can be obtained to evaluate the uniformity of the lubricating oil image, and the comparison and judgment are facilitated.
The lubricating oil image is converted into a gray image, and a common gray conversion algorithm can be adopted, such as averaging three channel values of the RGB image.
And carrying out convolution operation on the gray level image by using a horizontal convolution kernel (Gx) and a vertical convolution kernel (Gy) of the Sobel operator to obtain gradient approximate values in the horizontal and vertical directions.
And counting the mean value or variance of the gradient amplitude values of all pixel points in the image, and taking the mean value or variance as an index for evaluating the uniformity of the lubricating oil image. The smaller the mean or the smaller the variance, the more uniform the color change of the image, and conversely, the non-uniform the color distribution.
The method for establishing the reference model comprises the following steps:
collecting a plurality of representative lubricating oil image data, including a lubricating oil image in a normal state and an image under abnormal color and uneven particles;
preprocessing the acquired lubricating oil image data, including denoising, color correction and size standardization operation, so as to ensure the quality and consistency of the data;
extracting characteristic parameters including color distribution characteristics, particle shape characteristics and oil uniformity characteristics from the preprocessed image data by utilizing image processing and computer vision technologies;
labeling the extracted characteristic parameters, marking the lubricating oil image in a normal state as normal, and marking the image in an abnormal state as a corresponding abnormal type;
training by using the marked characteristic parameter data and using a deep learning neural network model, and establishing a reference model of the lubricating oil quality.
By collecting representative lubricating oil image data, including images under normal and abnormal conditions, the actual service condition of lubricating oil can be better covered, and the generalization capability of the model is improved; the collected image data is subjected to preprocessing operations such as denoising, color correction, size standardization and the like, so that noise and unnecessary changes are eliminated, the quality and consistency of the data are ensured, and the model training effect is improved.
Color distribution characteristics, particle shape characteristics and oil uniformity characteristics are extracted by utilizing image processing and computer vision technology, and the characteristics can reflect the quality condition of lubricating oil more comprehensively, so that the accuracy of a model is improved; marking the extracted characteristic parameters, and marking normal and abnormal lubricating oil images, so that the model is facilitated to learn the characteristics under normal and abnormal conditions, and the classification and recognition capability of the model are improved; by utilizing the marked characteristic parameter data and training by using a deep learning neural network model, potential rules and characteristics in the data can be better mined, and a reference model of lubricating oil quality can be established.
The method for extracting the characteristic parameters from the preprocessed image data comprises the following steps:
extracting color information in the image using the color histogram;
analyzing texture information of the image by using the gray level co-occurrence matrix;
describing shape features of objects in the image using edge detection;
selecting and dimension-reducing the extracted features by using principal component analysis;
the extracted feature parameters are represented as a computer-processable feature matrix.
The color histogram, gray level co-occurrence matrix, edge detection and other methods are used, so that the color, texture and shape characteristics of the image can be extracted from different angles, and the characteristic parameters are more comprehensive and diversified; the color histogram may reflect the distribution of various colors in the image, the gray level co-occurrence matrix may describe texture information of the image, and the edge detection may capture shape features of objects in the image. Representing these feature parameters as a computer-processable feature matrix can provide rich information for subsequent model training and analysis.
The extracted features are selected and reduced in dimension by using methods such as principal component analysis and the like, so that redundancy and noise of the features can be reduced, and the effectiveness of the features and the training efficiency of the model are improved; because the extracted characteristic parameters are expressed as a characteristic matrix which can be processed by a computer, the parallel computing capacity of the computer can be fully utilized, and the efficiency of characteristic extraction and processing is improved; the feature extraction method can be suitable for different types of images, including color change, texture difference, shape features and the like in the lubricating oil image, and has strong adaptability and generalization capability.
The method for detecting and classifying the lubricating oil image comprises the following steps:
the extracted characteristic parameters are input into a reference model, the model is utilized to carry out anomaly detection, whether the anomaly exists or not is judged by comparing the difference degree between the characteristic parameters of the image to be detected and normal characteristic distribution, and the image is divided into a color anomaly image and a particle non-uniformity image according to the anomaly type.
By inputting the extracted characteristic parameters into the reference model, abnormality detection can be performed quickly and accurately. The model can learn the characteristic distribution of the normal lubricating oil image, and abnormal conditions can be effectively detected by comparing the difference between the characteristic of the image to be detected and the normal characteristic; based on the result of the comparison, the abnormal images can be classified into different categories such as color abnormality and particle unevenness. Such classification helps describe and identify anomalies in the lubricant images in more detail, providing more specific problem localization and resolution.
The method can flexibly classify the abnormality as required, and can define and distinguish different abnormality types according to actual conditions. Meanwhile, model training and optimization can be adjusted and expanded according to actual application scenes, so that the adaptability and generalization capability of the model are improved; by using the reference model to detect and classify the abnormality, the automatic processing and intelligent analysis of the lubricating oil image can be realized. Thus, the working efficiency can be greatly improved, the human error can be reduced, and the method is particularly important for processing large-scale data sets; the abnormal detection result based on the model can be visually displayed and interpreted, so that operators can be helped to more intuitively understand and analyze the abnormal condition in the lubricating oil image. This helps to provide basis for decision support and problem investigation.
The method for combining the results of the image processing analysis with the chemical analysis data comprises the following steps:
sampling the lubricating oil with the abnormal image, and analyzing chemical components of the sampled lubricating oil to obtain detection data of viscosity, acid value and base number indexes of the lubricating oil;
comparing the detection data of the lubricating oil with the standard data, judging whether the detection data is in the standard data interval, and if not, determining that the index of the lubricating oil is abnormal.
Index data such as viscosity, acid value, base number and the like of the lubricating oil can be obtained through chemical component analysis, and the indexes are very important for judging whether the lubricating oil works normally. The chemical analysis data and the image processing analysis result are combined, so that the state of the lubricating oil can be more comprehensively known, and the lubricating oil can be evaluated from different angles; comparing the image processing analysis result with the chemical analysis data can more accurately judge whether the lubricating oil is abnormal or not. Image processing can capture subtle changes in color and particle distribution, while chemical analysis can provide more specific numerical indicators. By comparison, the abnormal image and the specific chemical index can be related, and the accuracy of abnormal detection is improved.
By combining the image processing analysis results with the chemical analysis data, the state of the lubricating oil can be monitored in real time and potential problems can be found in time. If a certain chemical index exceeds a standard data interval, judging that the lubricating oil is abnormal in the aspect, and taking corresponding measures to avoid equipment faults or accidents; image processing analysis may provide visual information, while chemical composition analysis may provide specific numerical indicators. The two are combined for use, so that quantitative index information which cannot be obtained by image processing can be supplemented, and the evaluation is more accurate and comprehensive; image processing and chemical analysis are two different methods, each with its advantages and limitations. The combination of the two components can complement each other, thereby improving the reliability of the result. If the conclusion drawn by the image processing and the chemical analysis is consistent, the judgment of the abnormality of the lubricating oil is more reliable.
Referring to fig. 2, a mass analysis system for low-carbon lubricating oil includes:
the image acquisition module is used for acquiring the real-time image of the lubricating oil on the production line and comprises high-definition camera equipment;
the image processing module is used for preprocessing the acquired image and comprises a denoising and contrast enhancement module;
the feature extraction module is used for extracting feature parameters such as color distribution, particle shape and the like of the lubricating oil image;
the reference model building module is used for taking the image characteristics of the normal lubricating oil as a standard to build a reference model;
computer vision module for detecting and classifying the abnormality of lubricating oil image by computer vision technique to find out the image characteristics different from reference model including abnormal color and uneven particles
The image processing analysis module is used for combining the result of the image processing analysis with the chemical analysis data to further confirm the abnormal image;
and the feedback module is used for feeding back the abnormal image and the analysis result to production management personnel.
The system can acquire images of the lubricating oil on the production line in real time, and perform real-time anomaly detection and classification through a computer vision technology, so that the quality problem of the lubricating oil can be found in time; the system can automatically extract characteristic parameters of the lubricating oil image by utilizing image processing and computer vision technology, and perform anomaly detection and classification, so that the burden of manual analysis is reduced, and the analysis efficiency is improved.
The system can take the image characteristics of normal lubricating oil as a standard through the reference model building module to build a reference model, so that abnormal conditions can be identified more accurately; the image processing analysis module of the system combines the result of image processing analysis with chemical analysis data, so that abnormal images can be further confirmed, and the comprehensiveness and accuracy of analysis are improved; the system is provided with a feedback module, and can timely feed back the abnormal image and the analysis result to production management personnel, so that the production management personnel can quickly take corresponding measures, and the influence of the quality problem of the lubricating oil on production is prevented.
And the system also comprises an alarm device which gives an alarm to an operator or a monitoring system when the abnormal quality of the lubricating oil is verified.
When the system verifies that the quality of the lubricating oil is abnormal, the alarm device can immediately give an alarm to an operator or a monitoring system, so that the operator or the monitoring system can take corresponding actions in time; the abnormal quality of the lubricating oil is timely found through the alarm device, so that the failure or damage of machine equipment caused by using low-quality lubricating oil can be avoided, and the loss caused by production line shutdown and maintenance is reduced; the timely response of the alarm device can help operators to quickly find problems, reduce production process interruption caused by quality problems, and improve production efficiency.
The above-mentioned embodiments of the present invention are not intended to limit the scope of the present invention, and the embodiments of the present invention are not limited thereto, and all kinds of modifications, substitutions or alterations made to the above-mentioned structures of the present invention according to the above-mentioned general knowledge and conventional means of the art without departing from the basic technical ideas of the present invention shall fall within the scope of the present invention.
Claims (10)
1. The quality analysis method of the low-carbon lubricating oil is characterized by comprising the following steps of:
adopting high-definition camera equipment to collect real-time images of lubricating oil on a production line;
processing the acquired image, including denoising and enhancing contrast, and extracting characteristic parameters of the lubricating oil image, including color distribution and particle shape;
establishing a reference model, and taking the image characteristics of normal lubricating oil as a standard;
using computer vision technology to detect and classify the lubricating oil image, finding out the image characteristics which are not in accordance with the reference model, including abnormal color and uneven particles;
combining the image processing analysis result with the chemical analysis data, further confirming the abnormal image, and verifying whether the lubricating oil quality problem exists;
when the verification finds that the quality of the lubricating oil is abnormal, an alarm is sent out to an operator or a monitoring system through an alarm device, and an abnormal image and an analysis result are fed back to production management staff.
2. The method for analyzing the quality of low-carbon lubricating oil according to claim 1, wherein the method for extracting the characteristic parameters of the lubricating oil image is as follows:
converting the image into a color space by using a color space conversion technology, and then calculating the pixel duty ratio of each color in each region by dividing or selecting the image, thereby obtaining the color distribution characteristics of the lubricating oil image;
processing the image by using a morphological processing method to highlight particles in the lubricating oil, and then extracting shape features of the lubricating oil particles by a contour analysis method;
the uniformity of the lubricating oil image is evaluated by calculating the gradient using the gray information of the image.
3. The method for mass analysis of low-carbon lubricating oil according to claim 2, wherein the method for evaluating uniformity of a lubricating oil image by calculating gradient is:
converting the lubricating oil image into a gray scale image;
calculating gradients in the horizontal direction and the vertical direction of the gray image by using a Sobel operator;
averaging or summing the gradient images in the horizontal direction and the vertical direction to obtain a comprehensive gradient image;
calculating the mean or variance of the comprehensive gradient image;
and judging the uniformity of the lubricating oil image according to the mean value or the variance.
4. The method for analyzing the quality of low-carbon lubricant according to claim 3, wherein the method for calculating the gradient of the gray image in the horizontal direction and the vertical direction using the Sobel operator is as follows:
defining two convolution kernels of the Sobel operator, wherein the two convolution kernels respectively represent gradient calculation in the horizontal direction and gradient calculation in the vertical direction, and the horizontal direction convolution kernel (Gx) and the vertical direction convolution kernel (Gy) are respectively:
Gx＝[[-1,0,1],
[-2,0,2],
[-1,0,1]]
Gy＝[[-1,-2,-1],
[0,0,0],
[1,2,1]]
performing two-dimensional convolution operation on the gray level image by using Gx and Gy respectively to obtain gradient approximate values in the horizontal direction and the vertical direction;
calculating the gradient amplitude in the horizontal and vertical directions for each pixel point, i.e
gradient_magnitude＝sqrt(Gx^2+Gy^2)
And gradient direction, i.e.
gradient_direction＝arctan2(Gy,Gx)
Wherein arctan2 is an arctangent function, the correct gradient direction can be obtained from the symbols of Gy and Gx.
5. The method for mass analysis of low-carbon lubricating oil according to claim 1, wherein the method for establishing the reference model is as follows:
collecting a plurality of representative lubricating oil image data, including a lubricating oil image in a normal state and an image under abnormal color and uneven particles;
preprocessing the acquired lubricating oil image data, including denoising, color correction and size standardization operation, so as to ensure the quality and consistency of the data;
extracting characteristic parameters including color distribution characteristics, particle shape characteristics and oil uniformity characteristics from the preprocessed image data by utilizing image processing and computer vision technologies;
labeling the extracted characteristic parameters, marking the lubricating oil image in a normal state as normal, and marking the image in an abnormal state as a corresponding abnormal type;
training by using the marked characteristic parameter data and using a deep learning neural network model, and establishing a reference model of the lubricating oil quality.
6. The method for analyzing the quality of low-carbon lubricant according to claim 5, wherein the method for extracting the characteristic parameters from the preprocessed image data comprises the steps of:
extracting color information in the image using the color histogram;
analyzing texture information of the image by using the gray level co-occurrence matrix;
describing shape features of objects in the image using edge detection;
selecting and dimension-reducing the extracted features by using principal component analysis;
the extracted feature parameters are represented as a computer-processable feature matrix.
7. The method for analyzing the quality of low-carbon lubricating oil according to claim 1, wherein the method for detecting and classifying the abnormality of the lubricating oil image comprises:
the extracted characteristic parameters are input into a reference model, the model is utilized to carry out anomaly detection, whether the anomaly exists or not is judged by comparing the difference degree between the characteristic parameters of the image to be detected and normal characteristic distribution, and the image is divided into a color anomaly image and a particle non-uniformity image according to the anomaly type.
8. The method for analyzing the quality of low-carbon lubricating oil according to claim 1, wherein the method for combining the results of the image processing analysis with the chemical analysis data is:
sampling the lubricating oil with the abnormal image, and analyzing chemical components of the sampled lubricating oil to obtain detection data of viscosity, acid value and base number indexes of the lubricating oil;
comparing the detection data of the lubricating oil with the standard data, judging whether the detection data is in the standard data interval, and if not, determining that the index of the lubricating oil is abnormal.
9. A mass analysis system for low-carbon lubricating oil, comprising:
the image acquisition module is used for acquiring the real-time image of the lubricating oil on the production line and comprises high-definition camera equipment;
the image processing module is used for preprocessing the acquired image and comprises a denoising and contrast enhancement module;
the feature extraction module is used for extracting feature parameters such as color distribution, particle shape and the like of the lubricating oil image;
the reference model building module is used for taking the image characteristics of the normal lubricating oil as a standard to build a reference model;
computer vision module for detecting and classifying the abnormality of lubricating oil image by computer vision technique to find out the image characteristics different from reference model including abnormal color and uneven particles
The image processing analysis module is used for combining the result of the image processing analysis with the chemical analysis data to further confirm the abnormal image;
and the feedback module is used for feeding back the abnormal image and the analysis result to production management personnel.
10. The mass analysis system of low-carbon lubricating oil according to claim 9, wherein: and the system also comprises an alarm device which gives an alarm to an operator or a monitoring system when the abnormal quality of the lubricating oil is verified.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311715868.5A CN117723739A (en) | 2023-12-13 | 2023-12-13 | Quality analysis method and system for low-carbon lubricating oil |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311715868.5A CN117723739A (en) | 2023-12-13 | 2023-12-13 | Quality analysis method and system for low-carbon lubricating oil |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117723739A true CN117723739A (en) | 2024-03-19 |
Family
ID=90210066
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311715868.5A Pending CN117723739A (en) | 2023-12-13 | 2023-12-13 | Quality analysis method and system for low-carbon lubricating oil |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117723739A (en) |
Citations (20)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101196510A (en) * | 2007-12-25 | 2008-06-11 | 深圳市亚泰光电技术有限公司 | Method and device for detecting pollution degree of lubricating oil |
| JP2011122968A (en) * | 2009-12-11 | 2011-06-23 | Jx Nippon Oil & Energy Corp | Lubricating oil deterioration monitoring device |
| JP2012068048A (en) * | 2010-09-21 | 2012-04-05 | Jx Nippon Oil & Energy Corp | Device for monitoring lubricating oil deterioration |
| CN103278431A (en) * | 2013-04-22 | 2013-09-04 | 西安交通大学 | On-line monitoring and analyzing method of lubrication oil wear particles based on video acquisition |
| DE102013110011A1 (en) * | 2013-09-12 | 2015-03-12 | Gert Horstmeyer | Method and device for the analysis of oils and technical operating fluids and for the qualified assessment of the operating states of aggregates |
| CN104764489A (en) * | 2015-03-27 | 2015-07-08 | 西安交通大学 | Online monitoring method for lubricating oil of wind-power transmission |
| CN108389216A (en) * | 2018-02-06 | 2018-08-10 | 西安交通大学 | Local auto-adaptive threshold segmentation method towards on-line ferrograph image wear Particles Recognition |
| WO2020028313A1 (en) * | 2018-07-31 | 2020-02-06 | The Regents Of The University Of Colorado A Body Corporate | Systems and methods for applying machine learning to analyze microcopy images in high-throughput systems |
| JPWO2020203995A1 (en) * | 2019-03-29 | 2020-10-08 | ||
| CN112288000A (en) * | 2020-10-27 | 2021-01-29 | 西安科技大学 | Intelligent ferrographic image identification method based on support vector machine |
| WO2022082010A1 (en) * | 2020-10-16 | 2022-04-21 | Saudi Arabian Oil Company | Detecting equipment defects using lubricant analysis |
| CN114663821A (en) * | 2022-05-18 | 2022-06-24 | 武汉大学 | Real-time nondestructive detection method for product quality based on video hyperspectral imaging technology |
| CN115115621A (en) * | 2022-08-24 | 2022-09-27 | 聊城市泓润能源科技有限公司 | Lubricating oil pollution degree detection method based on image processing |
| CN115496692A (en) * | 2022-11-17 | 2022-12-20 | 卡松科技股份有限公司 | Lubricating oil abrasive particle image enhancement method |
| CN115619708A (en) * | 2022-08-26 | 2023-01-17 | 华能新能源股份有限公司山西分公司 | Method for judging fault based on image recognition of oil quality change of main shaft bearing lubricating oil |
| CN115797347A (en) * | 2023-02-06 | 2023-03-14 | 临沂农业科技职业学院(筹) | Automatic production line abnormity monitoring method based on computer vision |
| CN115908945A (en) * | 2022-12-20 | 2023-04-04 | 内蒙古农业大学 | Method for detecting water content of lubricating oil based on shape characteristic parameters of distorted object image |
| CN116402810A (en) * | 2023-06-05 | 2023-07-07 | 山东天力润滑油有限公司 | Image processing-based lubricating oil anti-abrasive particle quality detection method |
| CN117058414A (en) * | 2023-08-18 | 2023-11-14 | 怀来欧洛普过滤器制造有限公司 | Online monitoring and early warning method and system for oil particles |
| CN117115494A (en) * | 2023-10-23 | 2023-11-24 | 卡松科技股份有限公司 | Lubricating oil impurity pollution detection method and device based on artificial intelligence |
-
2023
- 2023-12-13 CN CN202311715868.5A patent/CN117723739A/en active Pending
Patent Citations (20)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101196510A (en) * | 2007-12-25 | 2008-06-11 | 深圳市亚泰光电技术有限公司 | Method and device for detecting pollution degree of lubricating oil |
| JP2011122968A (en) * | 2009-12-11 | 2011-06-23 | Jx Nippon Oil & Energy Corp | Lubricating oil deterioration monitoring device |
| JP2012068048A (en) * | 2010-09-21 | 2012-04-05 | Jx Nippon Oil & Energy Corp | Device for monitoring lubricating oil deterioration |
| CN103278431A (en) * | 2013-04-22 | 2013-09-04 | 西安交通大学 | On-line monitoring and analyzing method of lubrication oil wear particles based on video acquisition |
| DE102013110011A1 (en) * | 2013-09-12 | 2015-03-12 | Gert Horstmeyer | Method and device for the analysis of oils and technical operating fluids and for the qualified assessment of the operating states of aggregates |
| CN104764489A (en) * | 2015-03-27 | 2015-07-08 | 西安交通大学 | Online monitoring method for lubricating oil of wind-power transmission |
| CN108389216A (en) * | 2018-02-06 | 2018-08-10 | 西安交通大学 | Local auto-adaptive threshold segmentation method towards on-line ferrograph image wear Particles Recognition |
| WO2020028313A1 (en) * | 2018-07-31 | 2020-02-06 | The Regents Of The University Of Colorado A Body Corporate | Systems and methods for applying machine learning to analyze microcopy images in high-throughput systems |
| JPWO2020203995A1 (en) * | 2019-03-29 | 2020-10-08 | ||
| WO2022082010A1 (en) * | 2020-10-16 | 2022-04-21 | Saudi Arabian Oil Company | Detecting equipment defects using lubricant analysis |
| CN112288000A (en) * | 2020-10-27 | 2021-01-29 | 西安科技大学 | Intelligent ferrographic image identification method based on support vector machine |
| CN114663821A (en) * | 2022-05-18 | 2022-06-24 | 武汉大学 | Real-time nondestructive detection method for product quality based on video hyperspectral imaging technology |
| CN115115621A (en) * | 2022-08-24 | 2022-09-27 | 聊城市泓润能源科技有限公司 | Lubricating oil pollution degree detection method based on image processing |
| CN115619708A (en) * | 2022-08-26 | 2023-01-17 | 华能新能源股份有限公司山西分公司 | Method for judging fault based on image recognition of oil quality change of main shaft bearing lubricating oil |
| CN115496692A (en) * | 2022-11-17 | 2022-12-20 | 卡松科技股份有限公司 | Lubricating oil abrasive particle image enhancement method |
| CN115908945A (en) * | 2022-12-20 | 2023-04-04 | 内蒙古农业大学 | Method for detecting water content of lubricating oil based on shape characteristic parameters of distorted object image |
| CN115797347A (en) * | 2023-02-06 | 2023-03-14 | 临沂农业科技职业学院(筹) | Automatic production line abnormity monitoring method based on computer vision |
| CN116402810A (en) * | 2023-06-05 | 2023-07-07 | 山东天力润滑油有限公司 | Image processing-based lubricating oil anti-abrasive particle quality detection method |
| CN117058414A (en) * | 2023-08-18 | 2023-11-14 | 怀来欧洛普过滤器制造有限公司 | Online monitoring and early warning method and system for oil particles |
| CN117115494A (en) * | 2023-10-23 | 2023-11-24 | 卡松科技股份有限公司 | Lubricating oil impurity pollution detection method and device based on artificial intelligence |
Non-Patent Citations (2)
| Title |
|---|
| FOROUGHI, B, 等: "Detection of Gasoline Adulteration Using Modified Distillation Curves and Artificial Neural Network", CHEMICAL ENGINEERING & TECHNOLOGY, vol. 44, no. 03, 31 March 2021 (2021-03-31), pages 527 - 534 * |
| 郝延龙;何红坤;常青;严志军;潘新祥;朱新河;程东;马春生: "基于显微图像识别的在线润滑油中磨粒分析方法", 基于显微图像识别的在线润滑油中磨粒分析方法, vol. 41, no. 05, 15 May 2016 (2016-05-15), pages 59 - 64 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108037133B (en) | Intelligent electric power equipment defect identification method and system based on unmanned aerial vehicle inspection image | |
| CN111242123B (en) | Power equipment fault diagnosis method based on infrared image | |
| CN107085846B (en) | Workpiece surface defect image identification method | |
| CN112419299A (en) | Bolt loss detection method, device, equipment and storage medium | |
| CN112173636B (en) | Method for detecting faults of belt conveyor carrier roller by inspection robot | |
| CN111126196A (en) | Equipment oil leakage detection method | |
| CN115802027B (en) | On-line monitoring equipment operation abnormity detection method based on image processing | |
| CN111563896A (en) | Image processing method for catenary anomaly detection | |
| CN111754737A (en) | Online identification and evaluation device and method for installation acceptance of metering device | |
| CN116231504A (en) | Remote intelligent inspection method, device and system for booster station | |
| CN114694050A (en) | Power equipment running state detection method based on infrared image | |
| CN117664990A (en) | Intelligent PCBA appearance defect detection method | |
| CN113128555B (en) | Method for detecting abnormality of train brake pad part | |
| CN117252840B (en) | Photovoltaic array defect elimination evaluation method and device and computer equipment | |
| CN111626104B (en) | Cable hidden trouble point detection method and device based on unmanned aerial vehicle infrared thermal image | |
| CN117689646A (en) | High-precision defect detection method, system and medium for positive and negative sample fusion | |
| CN117556211A (en) | Method for constructing power transmission line defect and hidden danger identification algorithm model | |
| CN117114420A (en) | Image recognition-based industrial and trade safety accident risk management and control system and method | |
| CN116840240A (en) | Visual detection system of power supply distributor | |
| CN117723739A (en) | Quality analysis method and system for low-carbon lubricating oil | |
| CN116862712A (en) | Electric power construction potential safety risk detection method and system based on thunder fusion | |
| CN116630809A (en) | Geological radar data automatic identification method and system based on intelligent image analysis | |
| CN109272480B (en) | Detection method for boundary of immiscible solution | |
| CN115619763A (en) | Power transmission line small part image target detection method based on multistage reasoning detection | |
| Wang et al. | Visual defect detection for substation equipment based on joint inspection data of camera and robot |
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 |