CN112666047A - Liquid viscosity detection method - Google Patents
Liquid viscosity detection method Download PDFInfo
- Publication number
- CN112666047A CN112666047A CN202110046283.3A CN202110046283A CN112666047A CN 112666047 A CN112666047 A CN 112666047A CN 202110046283 A CN202110046283 A CN 202110046283A CN 112666047 A CN112666047 A CN 112666047A
- Authority
- CN
- China
- Prior art keywords
- viscosity
- liquid
- image
- solution
- vector machine
- 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.)
- Granted
Links
Images
Landscapes
- Investigating Or Analysing Biological Materials (AREA)
- Automatic Analysis And Handling Materials Therefor (AREA)
Abstract
A liquid viscosity detection method, the invention is that LED backlight light source, bubble generating device, solution collecting tank, bracket, CCD high-speed camera are fixed and installed on the adjustable horizontal base platform from left to right in turn, the bracket is installed with normal pressure liquid supply device; a solution collecting tank is fixedly arranged on a platform below the normal-pressure liquid supply device; the controller is arranged above the CCD high-speed camera, and the normal-pressure liquid supply device and the bubble generation groove are connected with the controller through communication cables; the upper PC is arranged beside the whole device, and the CCD high-speed camera and the controller are connected with the upper PC through communication cables. The detection device is controlled to carry out video acquisition on the generation process of liquid drops of the solution to be detected and bubbles in the solution, the characteristic data of the solution to be detected is obtained through video image processing and characteristic extraction, the viscosity of the liquid to be detected is identified by adopting a soft measurement method, and the accuracy and the real-time performance of online viscosity detection in an industrial scene are improved.
Description
Technical Field
The invention relates to the field of detection equipment, in particular to a liquid viscosity detection method.
Background
Viscosity is a macroscopic expression of the acting force between fluid molecules, is an important physical parameter of the fluid, and is often used as a key index for evaluating the quality of the fluid. Viscosity measurement is widely used in industrial process control, cardiovascular disease analysis, food and beverage quality detection, and metal particle detection. The traditional viscosity measurement method mainly comprises a fluid method and a motion method according to different detection principles, wherein measurement devices of the fluid method comprise an outflow cup viscometer, a capillary viscometer, a falling ball viscometer and the like; the measuring device of the kinematic method includes a vibration viscometer, a rotary viscometer, and the like. These viscosity measurement devices are often very bulky and complex, have low detection accuracy, and have high requirements on external environments. With the continuous improvement of the production process level, the requirements on the measurement precision, the measurement efficiency and the measurement automation degree in the industrial production process are higher and higher. These traditional detection techniques often can not satisfy the accurate simple and convenient efficient requirement under the on-line detection scene.
Disclosure of Invention
The invention provides a liquid viscosity detection method, aiming at the problems that the traditional viscosity measurement equipment is huge, complex and difficult to deploy on site, the detection precision is not high, the requirement on the external environment is high and the like, the contour characteristic data of liquid drops with different viscosities in the growth process under the same condition is obtained by building an experiment platform, the kinematic viscosity of target liquid is trained and recognized by combining a soft measurement technology and utilizing an improved support vector machine regression method, so that the production process is detected and controlled by utilizing the machine vision technology, the online monitoring of the viscosity is realized, the non-contact detection is carried out on the liquid viscosity by utilizing the machine vision technology, the equipment is convenient to install on an industrial site, the recognition precision is high in a certain viscosity range, and the requirements on accuracy, simplicity, convenience, rapidness and high efficiency under an online detection scene are.
The invention relates to a liquid viscosity detection device which comprises an adjustable horizontal base 1, an LED backlight light source 2, a support 3, a normal-pressure liquid supply device 4, a bubble generation device 5, a solution collecting tank 6, a controller 7, a CCD high-speed camera 8 and an upper PC (personal computer) 9.
An LED backlight light source 2, a bubble generating device 5, a solution collecting tank 6, a bracket 3 and a CCD high-speed camera 8 are fixedly arranged on a platform of the adjustable horizontal base 1 from left to right in sequence, and a normal-pressure liquid supply device 4 is arranged on the bracket 3; a solution collecting tank 6 is fixedly arranged on a platform below the normal-pressure liquid supply device 4; the controller 7 is arranged above the CCD high-speed camera 8, and the normal-pressure liquid supply device 4 and the bubble generation tank 5 are connected with the controller 7 through communication cables; the upper PC 9 is arranged beside the whole device, and the CCD high-speed camera 8 and the controller 7 are connected with the device through communication cables.
The adjustable horizontal base 1 comprises an adjustable foot pad 111, a platform 112 and a level meter 113; the adjustable foot pad 111 is arranged at the bottom of the platform 112, and the level meter 113 is arranged on the upper surface of the platform 112.
The normal-pressure liquid supply device 4 comprises a liquid storage tank 411, a liquid supply tank 412, a liquid supply port 413, a regulating valve 414, an overflow port 415, a detachable capillary drip port 416, a pressurizing liquid discharge valve 417 and a liquid level sensor 418; the liquid supply tank 412 is connected to the lower left of the liquid storage tank 411 through a pipeline; the regulating valve 414 is installed in the pipeline of the liquid supply port 413, the liquid supply port 413 and the overflow port 415 are respectively arranged at a certain height position at two sides of the groove wall of the liquid supply groove 412, and the overflow port 415 and the groove wall form a mechanical angle inclined downwards by 45 degrees; the removable capillary drop 416 is vertically mounted at the bottom of the feed tank 412; the pressurized drain valve 417 is positioned right above the liquid supply tank 412 and is fixedly arranged outside the liquid storage tank 411; the controller 7 is connected to the regulating valve 414, the level sensor 418 and the pressurized drain valve motor via communication cables.
The bubble generation device 5 comprises a transparent liquid storage tank 511, a micro-flow pump 512 and a bubble generation tube 513; the bubble generating tube is arranged in the transparent liquid storage tank 511; the micro-flow pump 512 is arranged at the top of the transparent liquid storage tank 511 and is connected with the bubble generating pipe 513 through a hose, and the micro-flow pump 512 is connected with the controller 7 through a communication cable.
The invention relates to a liquid viscosity detection method, which comprises the following steps:
(1) collecting viscosity data of different solutions, and collecting videos of the growth process of liquid drops of the solutions with different viscosities and the growth process of bubbles in the solutions by using the viscosity detection device;
(2) segmenting the target areas of the liquid drop growth video and the bubble growth video to obtain an image sequence of a bubble growth process and an image sequence of a liquid drop growth process;
(3) converting the image sequence from RGB color space to single-channel gray image, and carrying out image preprocessing;
(4) carrying out binarization processing on the gray level image by using a self-adaptive threshold value, and converting the image into a binary image;
(5) carrying out edge detection and contour extraction on the binary image, and calculating and extracting relevant contour features; the contour features include: width W, length H, perimeter L, area a, circularity C, rectangularity R, elongation Q, etc.;
(6) and judging the growth period of the liquid drop according to the profile characteristics, recording the video frame number F of one liquid drop growth period, and sampling to the maximum liquid drop according to a certain frequency, namely the profile characteristics of the image of the previous frame before the liquid drop grows to break. The n characteristic sequences after storage are combined with the video frame number F of the droplet growth cycle and are marked as S = [ F, T ]1，T2，…，Tn]Wherein T isi=[Wi， Hi， Li， Ai， Ci， Qi]；
(7) Taking the combination of the characteristic sequence T and the corresponding solution viscosity V as a sample point, and marking as Di={(Si，Vi) }; all the collected videos were subjected to feature sequence determination, and combined with the corresponding solution viscosity to determine all the sample sets D = { (S)1，V1)，（S2，V2），…，（SN，VN）}；
(8) Normalizing all the characteristics of the sample set D;
(9) selecting features from the sample set D by using a random forest feature selection RFFS, sorting the features by using variable importance measurement of a random forest algorithm, then removing one least important feature from the feature set each time by using a sequence backward search method, successively iterating, calculating the accuracy, and finally obtaining a feature subset set with the least variable number and the highest accuracy as a feature selection result;
(10) establishing a support vector machine regression SVR model, and optimizing a target by using a quadratic programming algorithm (SMO); selecting the feature subset as an input variable; the solution viscosity as the target variable is the expected output of the support vector machine regression SVR model, denoted as yo(ii) a The actual output of the support vector machine regression SVR model is Op(ii) a The solution viscosity utilizes a particle swarm algorithm to optimize the hyperparameter of the regression SVR model of the support vector machine, the mean square error between the minimized actual output and the expected output is taken as an optimization target, and the fitness function is as follows:
(11) and detecting the test sample, testing the trained support vector machine regression SVR model by using the test sample, and outputting the support vector machine regression SVR model, namely the identification result of the solution viscosity.
The invention has the beneficial effects that: the method comprises the steps of controlling the generation process of liquid drops and bubbles in a solution by building a detection device, collecting video images of the liquid drops and the bubbles in the solution, extracting outline characteristics of the liquid drops and the bubbles in the solution by using a video image processing method, determining a soft measurement model and an optimization target by using the outline of the solution and the real solution viscosity, and optimizing the target by using a quadratic programming algorithm (SMO) of a support vector machine regression (SVR) model; wherein the soft measurement model is trained and established according to the existing data. Firstly, training historical data through a vector machine regression SVR model, and optimizing hyper-parameters of the support vector machine regression SVR model by utilizing a particle swarm algorithm, so that the modeling accuracy is further improved. And finally, performing solution viscosity fitting by using the optimized soft measurement model, and verifying the effectiveness of the design of the invention through error analysis of the real viscosity. The method solves the technical problems that the traditional viscosity detection device is large in size and poor in real-time performance, and online measurement is difficult to realize, improves the accuracy and real-time performance of online viscosity detection in an industrial scene, and enables the processing, detection and control processes to be better combined together.
Drawings
FIG. 1 is a system diagram of a liquid viscosity detection device according to the present invention.
FIG. 2 is a schematic view of a normal pressure liquid supply device of the liquid viscosity detection device of the present invention.
FIG. 3 is a schematic view of a bubble generating device of the liquid viscosity detecting device according to the present invention.
FIG. 4 is a flow chart of a method for detecting liquid viscosity according to the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all embodiments, and all other embodiments obtained by a person of ordinary skill in the art without creative efforts based on the embodiments of the present invention belong to the protection scope of the present invention.
In the description of the present invention, it should be noted that the terms "top", "bottom", "one side", "the other side", "front", "back", "middle part", "inside", "top", "bottom", etc. indicate orientations or positional relationships based on the orientations or positional relationships shown in the drawings, and are only for convenience of describing the present invention and simplifying the description, but do not indicate or imply that the referred device or element must have a specific orientation, be constructed in a specific orientation, and be operated, and thus, should not be construed as limiting the present invention; the terms "first," "second," and "third" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance; furthermore, unless expressly stated or limited otherwise, the terms "mounted," "connected," and "connected" are to be construed broadly, as they may be fixedly connected, detachably connected, or integrally connected, for example; can be mechanically or electrically connected; they may be connected directly or indirectly through intervening media, or they may be interconnected between two elements. The specific meanings of the above terms in the present invention can be understood in specific cases to those skilled in the art.
The liquid viscosity detection device shown in fig. 1 comprises an adjustable horizontal base 1, an LED backlight light source 2, a support 3, a normal-pressure liquid supply device 4, a bubble generation device 5, a solution collection tank 6, a controller 7, a CCD high-speed camera 8 and an upper PC 9; an LED backlight light source 2, a bubble generating device 5, a solution collecting tank 6, a bracket 3 and a CCD high-speed camera 8 are fixedly arranged on the platform of the adjustable horizontal base 1 from left to right in sequence, and the height of three adjustable foot pads 111 is adjusted by observing a level gauge 113 on the adjustable horizontal base 1 so as to ensure that the platform 112 keeps a horizontal state; light rays emitted by LED light beads of the LED backlight light source 2 are changed into a surface light source with uniform light emission through the rectangular diffuse reflection plate, so that the quality of image collection is improved; the light rays are imaged on a CCD image sensor of a CCD high-speed camera 8 through a bubble generating device 5 and a normal-pressure liquid supply device 4 in sequence, and video images of the growth process of liquid drops of solutions with different viscosities and the growth process of bubbles in the solutions are obtained by adjusting the focal length of the CCD high-speed camera 8; the controller 7 is arranged on a CCD high-speed camera 8, the normal-pressure liquid supply device 4 and the bubble generation tank 5 are connected with the controller 7 through communication cables so as to control the video data acquisition process, and the controller 7 is a programmable logic controller; the upper PC 9 is arranged beside the whole device, the CCD high-speed camera 8 and the controller 7 are connected with the upper PC through communication cables, the whole system is subjected to master control, and the acquired video image data is processed, analyzed and the solution viscosity is identified.
The normal pressure liquid supply device of the liquid drop video image acquisition device shown in fig. 2, wherein the normal pressure liquid supply device 4 comprises a liquid storage tank 411, a liquid supply tank 412, a liquid supply port 413, a regulating valve 414, an overflow port 415, a detachable capillary drop port 416, a pressurized liquid discharge valve 417 and a liquid level sensor 418; the liquid supply tank 412 is connected to the left lower part of the liquid storage tank 411 through a pipeline, the solution in the liquid storage tank 411 continuously supplies liquid to the liquid supply tank 412 through the liquid supply port 413 under the action of gravity, and the redundant liquid overflows from the overflow port 415; the liquid supply port 413 and the overflow port 415 are respectively arranged at a certain height at two sides of the groove wall of the liquid supply groove 412, the overflow port 415 and the groove wall form a mechanical angle of 45 degrees of downward inclination, the generation frequency of liquid drops is related to the height from the overflow port 415 to the detachable capillary drop port 416, and the change rule of the profile characteristic of the liquid drops can be analyzed by extracting the image characteristic that the shape changes along with time in the process from the formation to the fracture of the axisymmetric liquid drops formed by vertically downward capillary drop ports under the same liquid level height by using solutions with different viscosities; the detachable capillary drip opening 416 is vertically arranged at the bottom of the liquid supply tank 412, and the viscosity measurement range can be changed by replacing capillary drip openings with different calibers; the liquid level sensor 418 is arranged in the cavity of the liquid supply tank 412, and the opening of the regulating valve 414 is adjusted by detecting the liquid level in the liquid supply tank 412 so as to control the liquid supply flow of the liquid supply port 413; the regulating valve 414 is arranged in a pipeline of the liquid supply port 413, when the detection is started, the regulating valve 414 is in a full-open state, the liquid level in the liquid supply tank 412 is rapidly raised, the opening degree of the regulating valve 414 is gradually reduced according to the rising speed of the liquid level monitored by the liquid level sensor 418, and when the liquid level reaches the height of the overflow port 415, the liquid supply port 413 supplies liquid to the liquid supply tank 412 at a speed slightly larger than the flow of the detachable capillary drop port 416, so that the loss of the solution to be detected is reduced; the pressurizing liquid discharge valve 414 is positioned right above the liquid supply tank 412 and is fixedly installed outside the liquid storage tank 411, when the video acquisition is finished, the adjusting valve 414 is closed after the liquid storage tank 411 is emptied of the solution by the maximum opening degree, the motor of the pressurizing liquid discharge valve 417 is started to assist in emptying the residual solution in the liquid supply tank 412, so that the next measurement can be started quickly.
The bubble generation device of the liquid drop video image acquisition device shown in fig. 3, wherein the bubble generation device 5 comprises a transparent liquid storage tank 511, a micro-flow pump 512 and a bubble generation tube 513; the micro-flow pump 512 is arranged at the top of the transparent liquid storage tank 511 and is connected with the bubble generating tube 513 through a hose, and the micro-flow pump 512 injects air into the transparent liquid storage tank filled with the solution to be detected at a constant speed to enable the tail end of the bubble generating tube to slowly and stably generate bubbles; the bubble generating tube 513 is vertically installed in the transparent liquid storage tank 511, and the relationship between the profile characteristics and the solution viscosity is analyzed and learned by extracting the image characteristics of the change of the bubble shape along with the time in the process that the bubble generating tube 513 generates bubbles with the same aperture in the solutions with different viscosities by using the constant flow air.
Fig. 4 is a flowchart of a method for detecting liquid viscosity according to an embodiment of the present invention, where the method includes the following steps:
(1) collecting videos of the growth process of the liquid drops of the solutions with different viscosities and the growth process of bubbles in the solutions by using a detection device; collecting viscosity data of solutions with different concentrations, wherein experimental data come from ASHRAE handbook, and collecting and arranging 100 groups of data;
(2) segmenting the target areas of the liquid drop growth video and the bubble growth video to obtain an image sequence of a bubble growth process and an image sequence of a liquid drop growth process;
(3) converting the image sequence from an RGB color space into a single-channel gray image;
(4) using a threshold value method to carry out binarization processing on the gray level image, converting the image into a binary image, recording the binary image as A, and after a plurality of tests, considering that the threshold value is the best effect when the gray level number is 200, and the expression is
(ii) a B, performing morphological closing operation on the binary image, filling fine holes in the image, and keeping detailed information such as necking lines, satellite droplets and the like when the droplets in the image are broken while playing a role of image noise reduction so as to extract complete profile information of the droplets; the morphological closing operation is to perform expansion processing on the binary image and then perform corrosion processing, and the expression is as follows:
(5) carrying out edge detection and contour extraction on the processed binary image, sequencing all contour positions in the image, selecting a liquid drop contour at a liquid drop opening at the top, namely a target contour, and calculating and extracting relevant contour features; the physical and chemical characteristics of different solution samples are different due to different composition and concentration of different solution samples, so that the conditions are constantNext, the growth profile characteristics of the liquid drops passing through the same-caliber liquid drop drippers are also different, and representative profile characteristics of the liquid drop images are extracted through image processing; the contour features include: width W, length H, perimeter L, area a, circularity C, squareness R, elongation Q, etc., wherein circularity is used to delineate the degree to which a target is close to a circle, and circularity is maximum when the target is circular (maximum)C=1），C=(4π*A)/L2. The degree of rectangularity R reflects the degree of fullness of the image to the circumscribed rectangle, i.e., the degree of fullness of the droplet, the larger the R value, the fuller the droplet is R = (a/W × H). The elongation reflects the stretching degree of the target in the vertical direction, and the more slender the target is, the greater the elongation is;
(6) and judging the growth period of the liquid drop according to the profile characteristics, judging the growth period of the liquid drop by the sudden change of the area A of the liquid drop when the liquid drop is broken, recording the video frame number F of one growth period of the liquid drop, and sampling according to a certain frequency until the maximum liquid drop, namely the profile characteristics of the image of the liquid drop before the liquid drop grows to the broken frame. The n characteristic sequences after storage are combined with the video frame number F of the droplet growth cycle and are marked as S = [ F, T ]1，T2，…，Tn]Wherein T isi=[Wi， Hi， Li， Ai， Ci， Qi]；
(7) Combining the characteristic sequence T and the corresponding solution viscosity V as a sample point, and marking as Di={(Si，Vi) }; all the collected videos were subjected to feature sequence determination, and combined with the corresponding solution viscosity to determine all the sample sets D = { (S)1，V1)，（S2，V2），…，（SN，VN）}；
(8) All the characteristics of the sample set D are subjected to standardization processing, all characteristic data are converted into a state with a mean value of 0 and a variance of 1, the influence of different data scales on the model performance is reduced, the identification precision is improved, and a data standardization formula is as follows:
；
(9) selecting features from the sample set D by using a random forest feature selection RFFS, sorting the features by using variable importance measurement of a random forest algorithm, then removing one least important feature from the feature set each time by using a sequence backward search method, successively iterating, calculating the accuracy, and finally obtaining a feature subset with the least variable number and the highest accuracy as a feature selection result, wherein the feature selection method can be used for screening out features irrelevant to or redundant to a target to improve the model performance;
(10) establishing a support vector machine regression (SVR) model, selecting the feature subset as an input variable and recording the input variable as xi(ii) a The solution viscosity index as the target variable is the expected output of the support vector machine regression SVR model, denoted as yi(ii) a The support vector machine regression SVR model is a characteristic value matrix x of the characteristic subsetiAnd its corresponding viscosity index yiAnd is mapped to the high-dimensional feature space by a non-linear mapping function ϕ (x). x is the number ofiThe nonlinear relationship between the input data and the f (x) output data is:
(ii) a The support vector machine regression SVR model focuses on obtaining the optimal hyperplane and minimizing the error between the training samples and the loss function, and then minimizing the overall error. Therefore, the SVR can be written as the following objective function to be optimized:
wherein C is an adjustment parameter for the first and second terms of the formula; xii *And xiiIs a relaxation factor. The first term in the above equation is used to adjust the weight magnitude to maintain regression function flatness and penalize large weights, and to adjust them by using the idea of maximizing the distance between two training samples separated from each other; the second term is used to penalize f (x) and y by employing insensitive penaltiesiThe constraint conditions of the error between the two are as follows:
the optimization problem can be converted into a lagrangian function by introducing lagrangian factors, and the following is obtained by deducting:
substituting the original formula to obtain a new SVR regression formula as follows:
wherein beta isi *、βiIs a lagrange multiplier; k (x)i，xj) For the kernel function, the model trained by the RBF kernel function in the model for identifying the viscosity of the target solution has better overall performance compared with the models trained by other kernel functions, and the RBF (radial basis function) kernel function is expressed as:
(11) optimizing hyper-parameters of the support vector machine regression SVR model by using particle swarm optimization
By using the algorithm, the optimal hyper-parameter of the vector machine regression SVR model can be quickly and accurately found in the tensor space, the performance of the vector machine regression SVR model is enhanced, and the identification precision of viscosity is improved. The following is the implementation of the algorithm:
a) initializing relevant parameters;
b) and taking the minimum mean square error as an optimization target, and evaluating the initial adaptive value of each particle by using a fitness function, wherein the fitness function is as follows:
c) taking the initial adaptive value as the optimal value of each current particle and recording;
d) the current position is taken as a local optimal position (Pbest);
e) taking the optimal initial adaptation value as a current global optimum value, and recording the current position (Gbest);
f) the velocity and position are calculated according to the following formula, and the maximum velocity amplitude is limited,
wherein w is an inertia factor; c. C1And c2Is the acceleration constant; r is1And r2Is a random number within [0, 1 ]; a is a constraint factor and controls the speed weight;
(12) and detecting the test sample. And testing the trained support vector machine regression SVR model by using the test sample, wherein the output of the support vector machine regression SVR model is the identification result of the viscosity of the solution.
Claims (1)
1. The liquid viscosity detection method comprises the following steps:
(1) collecting the viscosity data of solutions with different concentrations, and collecting videos of the growth process of liquid drops of the solutions with different viscosities and the growth process of bubbles in the solutions by using the viscosity detection device;
(2) segmenting the target areas of the liquid drop growth video and the bubble growth video to obtain an image sequence of a bubble growth process and an image sequence of a liquid drop growth process;
(3) converting the image sequence from RGB color space to single-channel gray image, and carrying out image preprocessing;
(4) carrying out binarization processing on the gray level image by using a self-adaptive threshold value, and converting the image into a binary image;
(5) carrying out edge detection and contour extraction on the binary image, and calculating and extracting relevant contour features; the contour features include: width W, length H, perimeter L, area a, circularity C, rectangularity R, elongation Q, etc.;
(6) judging the growth period of the liquid drop according to the profile characteristics, recording the video frame number F of one liquid drop growth period, sampling to the maximum liquid drop according to a certain frequency, namely the profile characteristics of the image frame before the liquid drop grows to break, storing the n characteristic sequences, combining the n characteristic sequences with the video frame number F of the liquid drop growth period, and marking as S = [ F, T ]1，T2，…，Tn]Wherein T isi=[Wi， Hi， Li， Ai， Ci， Qi]；
(7) Combining the characteristic sequence T and the corresponding solution viscosity V as a sample point, and marking as Di={(Si，Vi) }; all the collected videos were subjected to feature sequence determination, and combined with the corresponding solution viscosity to determine all the sample sets D = { (S)1，V1)，（S2，V2），…，（SN，VN）}；
(8) Normalizing all the characteristics of the sample set D;
(9) selecting features from the sample set D by using a random forest feature selection RFFS, sorting the features by using variable importance measurement of a random forest algorithm, then removing one least important feature from the feature set each time by using a sequence backward search method, successively iterating, calculating the accuracy, and finally obtaining a feature subset set with the least variable number and the highest accuracy as a feature selection result;
(10) establishing a support vector machine regression SVR model, and optimizing a target by using a quadratic programming algorithm (SMO); selecting the feature subset as an input variable; the solution viscosity as the target variable is the expected output of the support vector machine regression SVR model, denoted as yo(ii) a The actual output of the support vector machine regression SVR model is Op(ii) a Optimizing the hyperparameters of the support vector machine regression SVR model by using a particle swarm algorithm, taking the minimum mean square error as an optimization target, wherein the fitness function is as follows:
(11) detecting the test sample; and testing the trained support vector machine regression SVR model by using the test sample, wherein the output of the support vector machine regression SVR model is the identification result of the viscosity of the solution.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110046283.3A CN112666047B (en) | 2021-01-14 | 2021-01-14 | Liquid viscosity detection method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110046283.3A CN112666047B (en) | 2021-01-14 | 2021-01-14 | Liquid viscosity detection method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112666047A true CN112666047A (en) | 2021-04-16 |
| CN112666047B CN112666047B (en) | 2022-04-29 |
Family
ID=75415078
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110046283.3A Active CN112666047B (en) | 2021-01-14 | 2021-01-14 | Liquid viscosity detection method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112666047B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116474157A (en) * | 2023-04-28 | 2023-07-25 | 广州迅合医疗科技有限公司 | Preparation method and system of hard tissue biological adhesive for continuous adhesion in wet environment |
| CN116589652A (en) * | 2023-04-28 | 2023-08-15 | 广州迅合医疗科技有限公司 | Preparation method and system of soft tissue biological adhesive based on polyurethane |
| CN117606980A (en) * | 2023-09-22 | 2024-02-27 | 中煤科工开采研究院有限公司 | Method for measuring the flow properties of a liquid and device for observing liquid drops |
Citations (26)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| GB1185445A (en) * | 1967-04-24 | 1970-03-25 | Rhodiaceta | Apparatus for Automatically Measuring Viscosity. |
| RU2059221C1 (en) * | 1992-07-15 | 1996-04-27 | Уфимский государственный нефтяной технический университет | Process of examination of rheological properties of liquids and gases |
| CN2903971Y (en) * | 2006-05-18 | 2007-05-23 | 宁波迅高智能科技有限公司 | Measurement analyzer for liquid surface interface dynamic characters based on high-speed image process |
| CN101156705A (en) * | 2007-11-09 | 2008-04-09 | 新疆大学 | A red date, hop zymolysis beverage and method for making same |
| JP2009292021A (en) * | 2008-06-04 | 2009-12-17 | Ricoh Co Ltd | Image formation device |
| CN102301269A (en) * | 2008-12-05 | 2011-12-28 | 优尼森索股份公司 | Optical Sectioning Of A Sample And Detection Of Particles In A Sample |
| CN102298706A (en) * | 2011-08-12 | 2011-12-28 | 河海大学 | Inland waterway ship large-scale prediction method in restricted conditions |
| CN102998316A (en) * | 2012-12-20 | 2013-03-27 | 山东大学 | Transparent liquid impurity detection system and detection method thereof |
| AU2012250496A1 (en) * | 2011-05-03 | 2013-11-28 | Commonwealth Scientific And Industrial Research Organisation | Method for detection of a neurological disease |
| JP2014083723A (en) * | 2012-10-22 | 2014-05-12 | Fuji Xerox Co Ltd | Image processing device and image processing program |
| CN103797131A (en) * | 2011-06-16 | 2014-05-14 | 卡里斯生命科学卢森堡控股有限责任公司 | Biomarker compositions and methods |
| DE102013217149A1 (en) * | 2013-08-28 | 2015-03-05 | Kuchenmeister Gmbh | Method and device for process control of a plant for the continuous production of foams |
| US20170166333A1 (en) * | 2015-12-10 | 2017-06-15 | Nova Chemicals (International) S.A. | Hot Fill Process With Closures Made From Bimodal Polyethylene Compositions |
| CN108475328A (en) * | 2015-12-30 | 2018-08-31 | 文塔纳医疗系统公司 | System and method for chemically examining monitoring in real time |
| CN108507905A (en) * | 2018-06-29 | 2018-09-07 | 华中科技大学 | A kind of micro fluid viscosity measurement technology sprayed using drop |
| CN109668858A (en) * | 2019-02-14 | 2019-04-23 | 大连理工大学 | Method based near infrared spectrum detection fermentation process biomass and concentration of component |
| WO2019103981A1 (en) * | 2017-11-22 | 2019-05-31 | Geltor Inc. | Methods and systems for engineering collagen |
| CN110599901A (en) * | 2018-06-12 | 2019-12-20 | 三星显示有限公司 | Window and display device including the same |
| CN111089632A (en) * | 2019-12-09 | 2020-05-01 | 中北大学 | Method and device for detecting liquid level of resin solution tank |
| CN111565632A (en) * | 2017-10-13 | 2020-08-21 | 奥特美医疗有限责任公司 | System and microtubule conductivity for characterizing, diagnosing and treating a health condition of a patient and methods of use thereof |
| CN111598305A (en) * | 2020-04-21 | 2020-08-28 | 汉谷云智(武汉)科技有限公司 | Method for optimizing and predicting running state of light hydrocarbon separation device |
| CN111738482A (en) * | 2020-04-20 | 2020-10-02 | 东华大学 | Method for adjusting technological parameters in polyester fiber polymerization process |
| CN112036090A (en) * | 2020-09-10 | 2020-12-04 | 浙江工业大学 | Linear hydraulic pressure polishing waviness prediction optimization method |
| CN112098272A (en) * | 2020-09-04 | 2020-12-18 | 浙江大学 | Method and device for simultaneously measuring surface tension and viscosity of liquid on line |
| CN112611680A (en) * | 2021-01-14 | 2021-04-06 | 新疆大学 | Liquid viscosity detection device |
| CN214472543U (en) * | 2021-01-14 | 2021-10-22 | 新疆大学 | Liquid viscosity detecting system |
-
2021
- 2021-01-14 CN CN202110046283.3A patent/CN112666047B/en active Active
Patent Citations (27)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| GB1185445A (en) * | 1967-04-24 | 1970-03-25 | Rhodiaceta | Apparatus for Automatically Measuring Viscosity. |
| RU2059221C1 (en) * | 1992-07-15 | 1996-04-27 | Уфимский государственный нефтяной технический университет | Process of examination of rheological properties of liquids and gases |
| CN2903971Y (en) * | 2006-05-18 | 2007-05-23 | 宁波迅高智能科技有限公司 | Measurement analyzer for liquid surface interface dynamic characters based on high-speed image process |
| CN101156705A (en) * | 2007-11-09 | 2008-04-09 | 新疆大学 | A red date, hop zymolysis beverage and method for making same |
| JP2009292021A (en) * | 2008-06-04 | 2009-12-17 | Ricoh Co Ltd | Image formation device |
| CN102301269A (en) * | 2008-12-05 | 2011-12-28 | 优尼森索股份公司 | Optical Sectioning Of A Sample And Detection Of Particles In A Sample |
| AU2012250496A1 (en) * | 2011-05-03 | 2013-11-28 | Commonwealth Scientific And Industrial Research Organisation | Method for detection of a neurological disease |
| CN103797131A (en) * | 2011-06-16 | 2014-05-14 | 卡里斯生命科学卢森堡控股有限责任公司 | Biomarker compositions and methods |
| CN102298706A (en) * | 2011-08-12 | 2011-12-28 | 河海大学 | Inland waterway ship large-scale prediction method in restricted conditions |
| JP2014083723A (en) * | 2012-10-22 | 2014-05-12 | Fuji Xerox Co Ltd | Image processing device and image processing program |
| CN102998316A (en) * | 2012-12-20 | 2013-03-27 | 山东大学 | Transparent liquid impurity detection system and detection method thereof |
| DE102013217149A1 (en) * | 2013-08-28 | 2015-03-05 | Kuchenmeister Gmbh | Method and device for process control of a plant for the continuous production of foams |
| US20170166333A1 (en) * | 2015-12-10 | 2017-06-15 | Nova Chemicals (International) S.A. | Hot Fill Process With Closures Made From Bimodal Polyethylene Compositions |
| CN108475328A (en) * | 2015-12-30 | 2018-08-31 | 文塔纳医疗系统公司 | System and method for chemically examining monitoring in real time |
| US20200297286A1 (en) * | 2017-10-13 | 2020-09-24 | Autem Medical, Llc | System for characterization, diagnosis, and treatment of a health condition of a patient and microtubule conductivity, and methods of using same |
| CN111565632A (en) * | 2017-10-13 | 2020-08-21 | 奥特美医疗有限责任公司 | System and microtubule conductivity for characterizing, diagnosing and treating a health condition of a patient and methods of use thereof |
| WO2019103981A1 (en) * | 2017-11-22 | 2019-05-31 | Geltor Inc. | Methods and systems for engineering collagen |
| CN110599901A (en) * | 2018-06-12 | 2019-12-20 | 三星显示有限公司 | Window and display device including the same |
| CN108507905A (en) * | 2018-06-29 | 2018-09-07 | 华中科技大学 | A kind of micro fluid viscosity measurement technology sprayed using drop |
| CN109668858A (en) * | 2019-02-14 | 2019-04-23 | 大连理工大学 | Method based near infrared spectrum detection fermentation process biomass and concentration of component |
| CN111089632A (en) * | 2019-12-09 | 2020-05-01 | 中北大学 | Method and device for detecting liquid level of resin solution tank |
| CN111738482A (en) * | 2020-04-20 | 2020-10-02 | 东华大学 | Method for adjusting technological parameters in polyester fiber polymerization process |
| CN111598305A (en) * | 2020-04-21 | 2020-08-28 | 汉谷云智(武汉)科技有限公司 | Method for optimizing and predicting running state of light hydrocarbon separation device |
| CN112098272A (en) * | 2020-09-04 | 2020-12-18 | 浙江大学 | Method and device for simultaneously measuring surface tension and viscosity of liquid on line |
| CN112036090A (en) * | 2020-09-10 | 2020-12-04 | 浙江工业大学 | Linear hydraulic pressure polishing waviness prediction optimization method |
| CN112611680A (en) * | 2021-01-14 | 2021-04-06 | 新疆大学 | Liquid viscosity detection device |
| CN214472543U (en) * | 2021-01-14 | 2021-10-22 | 新疆大学 | Liquid viscosity detecting system |
Non-Patent Citations (2)
| Title |
|---|
| LI GY 等: "An integrated smoothed particle hydrodynamics method for numerical simulation of the droplet impacting with heat transfer", 《ENGINEERING ANALYSIS WITH BOUNDARY ELEMENTS》 * |
| 郑桂波: "插入式阵列电导传感器两相流测量方法研究", 《中国博士学位论文全文数据库工程科技II辑》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116474157A (en) * | 2023-04-28 | 2023-07-25 | 广州迅合医疗科技有限公司 | Preparation method and system of hard tissue biological adhesive for continuous adhesion in wet environment |
| CN116589652A (en) * | 2023-04-28 | 2023-08-15 | 广州迅合医疗科技有限公司 | Preparation method and system of soft tissue biological adhesive based on polyurethane |
| CN116474157B (en) * | 2023-04-28 | 2023-09-15 | 广州迅合医疗科技有限公司 | Preparation method and system of hard tissue biological adhesive for continuous adhesion in wet environment |
| CN116589652B (en) * | 2023-04-28 | 2023-10-27 | 广州迅合医疗科技有限公司 | Preparation method and system of soft tissue biological adhesive based on polyurethane |
| CN117606980A (en) * | 2023-09-22 | 2024-02-27 | 中煤科工开采研究院有限公司 | Method for measuring the flow properties of a liquid and device for observing liquid drops |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112666047B (en) | 2022-04-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112666047B (en) | Liquid viscosity detection method | |
| CN110779937B (en) | Casting product internal defect intelligent detection device | |
| CN214472543U (en) | Liquid viscosity detecting system | |
| CN101713776B (en) | Neural network-based method for identifying and classifying visible components in urine | |
| CN109977780A (en) | A kind of detection and recognition methods of the diatom based on deep learning algorithm | |
| CN111753912A (en) | Coal slime flotation clean coal ash content prediction method based on deep learning | |
| CN1120163A (en) | Dynamic fluid level and bubble inspection for quality and process control | |
| CN109829882A (en) | A kind of stages of DR prediction technique | |
| CN108663367A (en) | A kind of egg quality lossless detection method based on egg unit weight | |
| CN108318494B (en) | The red online vision detection and classification devices and methods therefor for proposing fruit powder | |
| CN114383985A (en) | Flotation concentrate grade detection method based on multi-parameter feature fusion | |
| CN109115777A (en) | A kind of Fresh Grade Breast degree of lignification hierarchy system and stage division based on image deformation feature | |
| CN109657718A (en) | SPI defect classification intelligent identification Method on a kind of SMT production line of data-driven | |
| CN110490099A (en) | A kind of subway common location stream of people's analysis method based on machine vision | |
| CN107564000A (en) | Hericium erinaceus Non-Destructive Testing stage division based on computer vision | |
| CN111931827B (en) | Hydraulic pump health status detection system based on multi-sensor information fusion | |
| CN116595881A (en) | Method for predicting temperature field of crude oil storage tank and evaluating oil storage safety | |
| CN111553085A (en) | Flow field bubble image pressure recognition algorithm based on HOG + SVM | |
| JPS63500116A (en) | Method and apparatus for measuring particle size distribution of particles in suspension | |
| CN108230313A (en) | Based on the adaptively selected SAR image object detection method with discrimination model of component | |
| CN106599902B (en) | A kind of sugar crystal Classification and Identification and crystalline quality control method based on image | |
| CN104330336B (en) | Gas-liquid two-phase flow pattern recognition methods based on ICA and SVM | |
| CN116205969B (en) | Sample liquid volume identification method, system and medium | |
| CN112611680A (en) | Liquid viscosity detection device | |
| CN110675363A (en) | Automatic calculation method of DNA index for cervical cells |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |