Monitoring system and method for operation process of microbubble flotation machine
Through high-speed industrial cameras, the microbubble images are collected and the image semantic feature extraction and confidence reset are solved, and the traditional microbubble flotation machine monitoring method is accurately controlled and the flotation efficiency is improved.
Patent Information
- Application Number
- CN202410900559.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-05
- Publication Date
- 2024-12-06
- Estimated Expiration
- 2044-07-05
AI Technical Summary
The traditional microbubble flotation machine monitoring method relies on manual observation and empirical judgment, is inefficient and difficult to achieve precise control, affecting the flotation efficiency and selectivity.
By obtaining the time series of microbubble images collected by high-speed industrial cameras, image semantic feature extraction, confidence reset and prior information transmission optimization are carried out, and control instructions for water feeding are generated to realize the timing expression of microbubble features and water feeding Adaptive control.
The monitoring accuracy and control accuracy of the operation process of the microbubble flotation machine are improved, the flotation efficiency is optimized, energy consumption and cost are reduced, and the flotation process is automated and intelligent.
Smart Images

Figure CN118437508B_ABST
Abstract
Description
Technical Field
[0001] The present disclosure relates to the technical field of microbubble flotation machines, and in particular, to a monitoring system and method for the operation process of a microbubble flotation machine. Background Art
[0002] In the field of mineral processing, flotation is a widely used technology for separating valuable minerals from mineral mixtures. The efficiency and selectivity of the flotation process depend largely on the performance of the flotation machine. As a high-efficiency flotation equipment, the monitoring and control of the operating status of the microbubble flotation machine is crucial to improving the flotation efficiency.
[0003] Microbubble flotation machines capture mineral particles by generating tiny bubbles, and the interaction between these bubbles and mineral particles determines the efficiency of flotation. However, the behavior of microbubbles and foam characteristics during flotation are affected by many factors, including chemical environment, water supply, etc. Traditional flotation machine monitoring methods often rely on manual observation and empirical judgment, which is not only inefficient but also difficult to achieve precise control.
[0004] Therefore, an optimized monitoring system and method for the operation process of a microbubble flotation machine is desired. Summary of the invention
[0005] This summary is provided to introduce concepts in a brief form that will be described in detail in the detailed description below. This summary is not intended to identify key features or essential features of the claimed technical solution, nor is it intended to limit the scope of the claimed technical solution.
[0006] In a first aspect, the present disclosure provides a method for monitoring the operation process of a microbubble flotation machine, the method comprising:
[0007] Acquire a time series of microbubble images acquired by a camera module;
[0008] Extracting image semantic features from the time series of the microbubble images to obtain a sequence of microbubble semantic feature vectors;
[0009] Resetting the confidence and optimizing the prior information transfer of the sequence of microbubble semantic feature vectors to obtain a microbubble semantic temporal dynamic feature vector;
[0010] The control instruction of water supply amount is determined based on the microbubble semantic temporal dynamic feature vector.
[0011] Optionally, the camera module is a high-speed industrial camera.
[0012] Optionally, image semantic feature extraction is performed on the time series of microbubble images to obtain a sequence of microbubble semantic feature vectors, including: image preprocessing is performed on the time series of microbubble images to obtain a sequence of grayscale microbubble images; and each grayscale microbubble image in the sequence of grayscale microbubble images is passed through a microbubble feature extractor to obtain a sequence of microbubble semantic feature vectors.
[0013] Optionally, performing image preprocessing on the time series of microbubble images to obtain a sequence of grayscale microbubble images includes: performing grayscale processing on each microbubble image in the time series of microbubble images to obtain the sequence of grayscale microbubble images.
[0014] Optionally, the microbubble feature extractor is CNN, AlexNet or DenseNet.
[0015] Optionally, the sequence of microbubble semantic feature vectors is subjected to confidence reset and prior information transfer optimization to obtain a microbubble semantic temporal dynamic feature vector, including: passing the sequence of microbubble semantic feature vectors through a feature autocorrelation confidence reset network to obtain a sequence of importance reset microbubble semantic feature vectors; and passing the sequence of importance reset microbubble semantic feature vectors through an information transfer module based on prior distribution to obtain the microbubble semantic temporal dynamic feature vector.
[0016] Optionally, passing the sequence of microbubble semantic feature vectors through a feature autocorrelation confidence reset network to obtain a sequence of importance reset microbubble semantic feature vectors, comprising: processing the sequence of microbubble semantic feature vectors using the following feature autocorrelation confidence reset formula to obtain the sequence of importance reset microbubble semantic feature vectors; wherein the feature autocorrelation confidence reset formula is:
[0017]
[0018] in, For the The attention score coefficient, For the A weight coefficient matrix, For the A vector of weight coefficients, For the A bias vector, represents the Selu activation function, is the first in the sequence of microbubble semantic feature vectors microbubble semantic feature vectors, and are weighting coefficients with different values, represents the value input into the Selu activation function, represents the softmax activation function, represents dot product, The first in the sequence of microbubble semantic feature vectors is reset for the importance The importance of resetting the microbubble semantic feature vector.
[0019] Optionally, the sequence of the importance-reset microbubble semantic feature vectors is passed through an information transfer module based on a prior distribution to obtain the microbubble semantic temporal dynamic feature vector, comprising: processing the sequence of the importance-reset microbubble semantic feature vectors with the following information transfer formula to obtain the microbubble semantic temporal dynamic feature vector; wherein the information transfer formula is:
[0020]
[0021] in, and are the first weight coefficient and the second weight coefficient respectively, and are the first random coefficient and the second random coefficient randomly sampled from the standard normal distribution, resetting the mean feature vector of the sequence of microbubble semantic feature vectors for said importance, The first in the sequence of microbubble semantic feature vectors is reset for the importance importance reset microbubble semantic feature vector, the number of sequences of microbubble semantic feature vectors reset for said importance, randomly selecting an importance-reset microbubble semantic feature vector from the sequence of importance-reset microbubble semantic feature vectors, is the microbubble semantic temporal dynamic feature vector.
[0022] Optionally, determining a water supply control instruction based on the microbubble semantic temporal dynamic feature vector includes: passing the microbubble semantic temporal dynamic feature vector through a classifier-based water supply controller to obtain the control instruction, wherein the control instruction is used to indicate whether to increase the water supply.
[0023] Optionally, the monitoring method of the operation process of the microbubble flotation machine also includes a training phase for training the microbubble feature extractor, the feature autocorrelation confidence reset network, the information transfer module based on the prior distribution and the water supply controller based on the classifier; wherein the training phase includes: acquiring a time series of training microbubble images; performing image preprocessing on the time series of training microbubble images to obtain a sequence of training grayscale microbubble images; passing each training grayscale microbubble image in the sequence of training grayscale microbubble images through a microbubble feature extractor to obtain a sequence of training microbubble semantic feature vectors; passing the sequence of training microbubble semantic feature vectors through a feature autocorrelation confidence reset network to obtain a training importance weight vector. The invention discloses a method for resetting the sequence of microbubble semantic feature vectors; passing the sequence of the training importance-reset microbubble semantic feature vectors through an information transfer module based on a prior distribution to obtain a training microbubble semantic temporal dynamic feature vector; passing the training microbubble semantic temporal dynamic feature vector through a water supply controller based on a classifier to obtain a classification loss function value; based on the classification loss function value, the microbubble feature extractor, the feature autocorrelation confidence reset network, the information transfer module based on a prior distribution and the water supply controller based on the classifier are trained, wherein each time the training microbubble semantic temporal dynamic feature vector is classified and iterated through the water supply controller based on the classifier, the training microbubble semantic temporal dynamic feature vector is corrected.
[0024] In a second aspect, the present disclosure provides a monitoring system for the operation process of a microbubble flotation machine, the system comprising:
[0025] A microbubble image acquisition module, used to acquire a time series of microbubble images collected by a camera module;
[0026] An image semantic feature extraction module, used for performing image semantic feature extraction on the time series of the microbubble images to obtain a sequence of microbubble semantic feature vectors;
[0027] A confidence reset and prior information transfer optimization module, used for resetting the confidence and optimizing the prior information transfer of the sequence of microbubble semantic feature vectors to obtain a microbubble semantic temporal dynamic feature vector;
[0028] A water supply control instruction determination module is used to determine the water supply control instruction based on the microbubble semantic temporal dynamic feature vector.
[0029] The above technical solution is adopted, by obtaining the time series of microbubble images collected by the camera module; extracting image semantic features from the time series of the microbubble images to obtain a sequence of microbubble semantic feature vectors; resetting the confidence and optimizing the prior information transfer of the sequence of microbubble semantic feature vectors to obtain a microbubble semantic time series dynamic feature vector; and determining the control instruction of the water supply amount based on the microbubble semantic time series dynamic feature vector. In this way, the dynamic time series characteristics of microbubbles can be captured from the time series data of the microbubble images, and further feature transfer can be performed on them with the prior distribution as the guiding information, thereby optimizing the time series expression of the microbubble characteristics, and realizing adaptive control of the water supply amount based on the dynamic time series expression information of the microbubble characteristics.
[0030] Other features and advantages of the present disclosure will be described in detail in the following detailed description. BRIEF DESCRIPTION OF THE DRAWINGS
[0031] The above and other features, advantages and aspects of the embodiments of the present disclosure will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. Throughout the drawings, the same or similar reference numerals represent the same or similar elements. It should be understood that the drawings are schematic and the originals and elements are not necessarily drawn to scale. In the drawings:
[0032] Figure 1 It is a schematic structural diagram of a microbubble flotation machine according to an exemplary embodiment.
[0033] Figure 2 The present invention is a flow chart of a method for monitoring the operation process of a microbubble flotation machine according to an exemplary embodiment.
[0034] Figure 3 The invention is a block diagram of a monitoring system for the operation process of a microbubble flotation machine according to an exemplary embodiment.
[0035] Figure 4 It is a block diagram of an electronic device according to an exemplary embodiment.
[0036] Figure 5 It is an application scenario diagram of a method for monitoring the operation process of a microbubble flotation machine according to an exemplary embodiment. DETAILED DESCRIPTION
[0037] Embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although certain embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be construed as being limited to the embodiments described herein, which are instead provided for a more thorough and complete understanding of the present disclosure. It should be understood that the drawings and embodiments of the present disclosure are only for exemplary purposes and are not intended to limit the scope of protection of the present disclosure.
[0038] It should be understood that the various steps described in the method embodiments of the present disclosure may be performed in different orders and / or in parallel. In addition, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present disclosure is not limited in this respect.
[0039] The term "including" and its variations used herein are open inclusions, i.e., "including but not limited to". The term "based on" means "based at least in part on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments". The relevant definitions of other terms will be given in the following description.
[0040] It should be noted that the concepts such as "first" and "second" mentioned in the present disclosure are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence of the functions performed by these devices, modules or units.
[0041] It should be noted that the modifications of "one" and "plurality" mentioned in the present disclosure are illustrative rather than restrictive, and those skilled in the art should understand that unless otherwise clearly indicated in the context, it should be understood as "one or more".
[0042] The names of the messages or information exchanged between multiple devices in the embodiments of the present disclosure are only used for illustrative purposes and are not used to limit the scope of these messages or information.
[0043] Figure 1 FIG. 1 is a schematic diagram of the structure of a microbubble flotation machine according to an exemplary embodiment. Figure 1 As shown, in one embodiment of the present disclosure, a microbubble flotation machine is provided, which includes: a feed circulation pump 1, a concentrate outlet 2, a cleaning water regulating valve 3, a downpipe regulating valve 4, a feed pressure gauge 5, a tailings outlet 6, an air flow meter 7, a gas volume regulating valve 8, a tailings control valve 9, and a vacuum pressure gauge 10.
[0044] When the micro-bubble flotation machine is running, the feed slurry is pumped into the high-pressure feed pipeline by the high-pressure pump, and the pressure provided by the high-pressure pump forms a high-speed jet at the outlet. The high-speed downward impact slurry jet is mixed with the air provided by the self-priming inflation valve in the downpipe, and the mineralization process of the mineral is completed at the same time. The mineralized bubbles carrying minerals are squeezed out at the bottom of the downpipe, and the bubbles carrying the concentrate automatically rise, and the tailings slurry is discharged downward. The entire flotation process is completed in the downpipe, and the separation of concentrate and tailings is completed in the flotation machine cylinder outside the downpipe. Each working area is independent of each other and does not interfere with each other. The conditions can be optimized separately to ensure that the flotation efficiency in the downpipe is optimal.
[0045] Each working area of the microbubble flotation machine, including the bubble generation area, the pulp and bubble mixing area, the concentrate recovery area and the tailings discharge area, is relatively independent, which provides a certain feasibility for optimizing the working conditions of each working area.
[0046] The average diameter of mineralized bubbles generated by the slurry mixed with air after passing through the jet in the micro-bubble flotation machine is about 0.3mm, which is about one tenth of the bubble diameter of conventional flotation machines and flotation columns. When flotating minerals, the number of mineralized bubbles generated by the former is more than 50 times that of the latter under the same conditions. This is one of the key factors that enable the micro-bubble flotation machine to simultaneously improve the recovery rate and concentrate grade of the ore dressing.
[0047] The operation process includes:
[0048] 1. Turn on the power of the electric control cabinet and check whether the display of each instrument is normal.
[0049] 2. Set the target liquid level (this value is obtained during the debugging process, reference value: 1200).
[0050] 3. The feeding pump starts feeding.
[0051] 4. Feed material to the current liquid level display value and start the circulation pump at a low frequency (reference value: 28Hz).
[0052] 5. When the current liquid level display is consistent with the target liquid level display, adjust the circulation pump frequency.
[0053] 6. Adjust the frequency of the circulating pump to the target feed pressure.
[0054] 7. Adjust the target liquid level according to the flotation foam phenomenon.
[0055] 8. Adjust the air intake to an appropriate vacuum level of approx.
[0056] 9. Adjust the amount of cleaning water.
[0057] 10. Repeat the above steps 6, 7, 8, and 9 to achieve a stable output of target concentrate and tailings.
[0058] 11. Record various parameters.
[0059] Factors that affect production indicators include:
[0060] When the concentrate grade is low, the adjustment methods adopted are: a. lower the target liquid level; b. increase the amount of cleaning water; c. reduce the air intake; d. increase the ore feed.
[0061] The tailings grade is high and the recovery rate is low. The adjustment methods used are: a. increase the target liquid level; b. increase the feed pressure; c. increase the air intake; d. reduce the amount of washing water; e. reduce the feed amount.
[0062] Simple fault judgment and processing are shown in Table 1 below:
[0063] Table 1
[0064]
[0065] In order to solve the above problems, the present disclosure provides a monitoring system and method for the operation process of a microbubble flotation machine, by acquiring a time series of microbubble images collected by a camera module; extracting image semantic features from the time series of the microbubble images to obtain a sequence of microbubble semantic feature vectors; resetting the confidence and optimizing the prior information transfer of the sequence of microbubble semantic feature vectors to obtain a microbubble semantic time series dynamic feature vector; and determining the control instructions for the water supply based on the microbubble semantic time series dynamic feature vector. In this way, the dynamic time series characteristics of microbubbles can be captured from the time series data of the microbubble images, and further feature transfer can be performed on them with the prior distribution as the guiding information, thereby optimizing the time series expression of the microbubble characteristics, and realizing adaptive control of the water supply based on the dynamic time series expression information of the microbubble characteristics.
[0066] The specific embodiments of the present disclosure are described in detail below with reference to the accompanying drawings.
[0067] Figure 2 is a flow chart of a method for monitoring the operation process of a microbubble flotation machine according to an exemplary embodiment. Figure 2 As shown, the method includes:
[0068] Step 101, obtaining a time series of microbubble images collected by a camera module;
[0069] Step 102, performing image semantic feature extraction on the time series of the microbubble images to obtain a sequence of microbubble semantic feature vectors;
[0070] Step 103, resetting the confidence and optimizing the prior information transfer of the sequence of microbubble semantic feature vectors to obtain a microbubble semantic temporal dynamic feature vector; and
[0071] Step 104: Determine a water supply control instruction based on the microbubble semantic temporal dynamic feature vector.
[0072] Wherein, the camera module is a high-speed industrial camera.
[0073] It should be understood that in the actual operation of the microbubble flotation machine, controlling the water supply plays an extremely important role in the microbubble flotation process. Specifically, changes in the water supply will directly affect the concentration of the slurry. In actual application scenarios, too high a slurry concentration may lead to reduced contact between bubbles and mineral particles, while too low a concentration may increase energy consumption and reduce the flotation rate. At the same time, the adjustment of the water supply will affect the chemical environment in the flotation liquid, including pH value, reagent concentration, etc. These factors have a significant effect on the wettability of mineral particles, the stability of bubbles, and the adhesion between minerals and bubbles. In summary, the control of water supply is a key link in the operation of the microbubble flotation machine, which is of great significance to improving flotation efficiency, reducing costs, protecting the environment, and realizing the automation and intelligence of the flotation process. However, the existing control methods usually rely on manual observation and empirical judgment, which is not only inefficient, but also difficult to achieve precise control. Therefore, an optimized solution is expected.
[0074] In response to the above technical problems, the technical concept of this application is to use computer vision technology and deep learning algorithms to build an intelligent operation monitoring system for microbubble flotation machines based on automatic image acquisition and processing. More specifically, it is expected to collect time series data of microbubble images through high-speed industrial cameras, capture the dynamic time series characteristics of microbubbles, and further transmit the characteristics with the prior distribution as the guiding information, so as to optimize the time series expression of microbubble characteristics, and realize adaptive control of water supply based on the dynamic time series expression information of microbubble characteristics.
[0075] Based on this, in the technical solution of the present application, first, a time series of microbubble images collected by a high-speed industrial camera is obtained. In a specific example of the present application, the high-speed industrial camera is installed inside a microbubble flotation machine, which can collect microbubble images in situ in a harsh slurry solution. In particular, the high-speed industrial camera has a high frame rate and can capture the rapid movement of microbubbles. In another specific example of the present application, considering that the slurry solution may be corrosive, a protective cover is deployed outside the high-speed industrial camera to protect the high-speed industrial camera. Wherein, the protective cover is made of corrosion-resistant material, such as stainless steel or special plastic.
[0076] Because the color features in the microbubble image are not key clue information for controlling the water supply. Specifically, the color information of the microbubble is not important for judging the behavior of the microbubble and the characteristics of the foam. Even in the process of shooting the microbubble image, it may be affected by the illumination, resulting in uneven colors between local areas. If the focus of feature extraction is placed on color changes, it will undoubtedly have an adverse effect on the target task of controlling the water supply. Therefore, in order to reduce the excessive attention to color information, in the technical solution of the present application, each microbubble image in the time series of the microbubble image is gray-scaled to obtain a sequence of grayscale microbubble images. Here, each grayscale microbubble image processed by grayscale does not contain color information, which helps to highlight the characteristic information such as the outline and texture of the microbubble, so as to facilitate the capture of the scale, shape and motion information of the microbubble. It is worth mentioning that although each grayscale microbubble image contains brightness information, it does not directly show the illumination changes in the image. The grayscale microbubble image only reflects the brightness difference of each area in the image, and does not show the specific changes or illumination intensity of the illumination. Therefore, grayscale images cannot directly express the lighting changes in the image, but instead focus on showing the brightness differences in different areas of the image, namely edge information, contour information, and texture information.
[0077] In one embodiment of the present disclosure, image semantic feature extraction is performed on the time series of microbubble images to obtain a sequence of microbubble semantic feature vectors, including: image preprocessing is performed on the time series of microbubble images to obtain a sequence of grayscale microbubble images; and each grayscale microbubble image in the sequence of grayscale microbubble images is passed through a microbubble feature extractor to obtain a sequence of microbubble semantic feature vectors.
[0078] Further, performing image preprocessing on the time series of microbubble images to obtain a sequence of grayscale microbubble images includes: performing grayscale processing on each microbubble image in the time series of microbubble images to obtain the sequence of grayscale microbubble images.
[0079] Then, each grayscale microbubble image in the sequence of grayscale microbubble images is passed through a microbubble feature extractor to obtain a sequence of microbubble semantic feature vectors. That is, the microbubble feature extractor is used to capture the local implicit image semantic features in each of the grayscale microbubble images. Subsequently, the sequence of microbubble semantic feature vectors is passed through a feature autocorrelation confidence reset network to obtain a sequence of importance reset microbubble semantic feature vectors. That is, the full time domain autocorrelation in the sequence of microbubble semantic feature vectors is obtained by resetting the feature autocorrelation confidence network, capturing and strengthening the global dependencies and time dynamic features in the data. Among them, the essence of the feature autocorrelation confidence reset network is a weighted probability distribution mechanism, that is, a greater confidence is assigned to important content. Such a mechanism is more focused on finding key information with unique contribution in the input data, and discovering the autocorrelation between each microbubble semantic feature vector. Here, the feature autocorrelation confidence reset network can learn the temporal dependencies between microbubble behaviors in different local time spans in the sequence data of the microbubble semantic feature vector, and comprehensively understand the dynamic temporal operation mode and foam characteristics of microbubbles by modeling the autocorrelation of the entire sequence.
[0080] Wherein, the microbubble feature extractor is CNN, AlexNet or DenseNet.
[0081] In one embodiment of the present disclosure, the sequence of microbubble semantic feature vectors is subjected to confidence reset and prior information transfer optimization to obtain a microbubble semantic temporal dynamic feature vector, including: passing the sequence of microbubble semantic feature vectors through a feature autocorrelation confidence reset network to obtain a sequence of importance reset microbubble semantic feature vectors; and passing the sequence of importance reset microbubble semantic feature vectors through an information transfer module based on prior distribution to obtain the microbubble semantic temporal dynamic feature vector.
[0082] Further, the sequence of the importance reset microbubble semantic feature vector is passed through the information transfer module based on the prior distribution to obtain the microbubble semantic time series dynamic feature vector. Here, the sequence of the importance reset microbubble semantic feature vector is passed through the information transfer module based on the prior distribution to use the prior information as a guide to optimize the feature distribution presentation and expression of the sequence of the importance reset microbubble semantic feature vector. More specifically, in an embodiment of the present application, the information transfer module based on the prior distribution increases the diversity of the model search space and the possibility of jumping out of the local extreme value through the Gaussian difference mutation strategy. Unlike the traditional differential mutation strategy using a uniform distribution function, since the uniform distribution function is equally likely in intervals of the same length interval, it is sometimes impossible to generate a large disturbance near the variant data, so sometimes it is impossible to jump out of the local extreme value. However, the Gaussian distribution function has a small peak at the origin and a relatively long distribution at both ends. In other words, the Gaussian difference mutation strategy generates disturbances through the Gaussian distribution function, so that a larger disturbance is generated near the current data, which helps to better explore the search space and avoid falling into a local optimal solution. In this way, the information transfer module based on prior distribution combined with the Gaussian difference mutation strategy can use prior information to guide and optimize the dynamic timing pattern of microbubbles, thereby performing feature constraint expression based on search space randomization enhancement.
[0083] In one embodiment of the present disclosure, the sequence of microbubble semantic feature vectors is passed through a feature autocorrelation confidence reset network to obtain a sequence of importance reset microbubble semantic feature vectors, including: processing the sequence of microbubble semantic feature vectors with the following feature autocorrelation confidence reset formula to obtain the sequence of importance reset microbubble semantic feature vectors; wherein the feature autocorrelation confidence reset formula is:
[0084]
[0085] in, For the The attention score coefficient, For the A weight coefficient matrix, For the A vector of weight coefficients, For the A bias vector, represents the Selu activation function, is the first in the sequence of microbubble semantic feature vectors microbubble semantic feature vectors, and are weighting coefficients with different values, represents the value input into the Selu activation function, represents the softmax activation function, represents dot product, The first in the sequence of microbubble semantic feature vectors is reset for the importance The importance of resetting the microbubble semantic feature vector.
[0086] Further, the sequence of the importance-reset microbubble semantic feature vectors is passed through an information transfer module based on a prior distribution to obtain the microbubble semantic temporal dynamic feature vector, including: processing the sequence of the importance-reset microbubble semantic feature vectors with the following information transfer formula to obtain the microbubble semantic temporal dynamic feature vector; wherein the information transfer formula is:
[0087]
[0088] in, and are the first weight coefficient and the second weight coefficient respectively, and are the first random coefficient and the second random coefficient randomly sampled from the standard normal distribution, resetting the mean feature vector of the sequence of microbubble semantic feature vectors for said importance, The first in the sequence of microbubble semantic feature vectors is reset for the importance importance reset microbubble semantic feature vector, the number of sequences of microbubble semantic feature vectors reset for said importance, randomly selecting an importance-reset microbubble semantic feature vector from the sequence of importance-reset microbubble semantic feature vectors, is the microbubble semantic temporal dynamic feature vector.
[0089] Then, the microbubble semantic temporal dynamic feature vector is passed through a water supply controller based on a classifier to obtain a control instruction, wherein the control instruction is used to indicate whether to increase the water supply.
[0090] In one embodiment of the present disclosure, a control instruction for water supply is determined based on the microbubble semantic timing dynamic feature vector, including: passing the microbubble semantic timing dynamic feature vector through a classifier-based water supply controller to obtain the control instruction, wherein the control instruction is used to indicate whether to increase the water supply.
[0091] In one embodiment of the present disclosure, the monitoring method of the operation process of the microbubble flotation machine also includes a training phase for training the microbubble feature extractor, the feature autocorrelation confidence reset network, the information transfer module based on the prior distribution and the water supply controller based on the classifier; wherein the training phase includes: acquiring a time series of training microbubble images; performing image preprocessing on the time series of training microbubble images to obtain a sequence of training grayscale microbubble images; passing each training grayscale microbubble image in the sequence of training grayscale microbubble images through a microbubble feature extractor to obtain a sequence of training microbubble semantic feature vectors; passing the sequence of training microbubble semantic feature vectors through a feature autocorrelation confidence reset network to obtain a training The invention relates to a method for resetting the importance of microbubble semantic feature vectors; passing the sequence of training importance-reset microbubble semantic feature vectors through an information transfer module based on a prior distribution to obtain a training microbubble semantic temporal dynamic feature vector; passing the training microbubble semantic temporal dynamic feature vector through a water supply controller based on a classifier to obtain a classification loss function value; based on the classification loss function value, the microbubble feature extractor, the feature autocorrelation confidence reset network, the information transfer module based on a prior distribution and the water supply controller based on the classifier are trained, wherein each time the training microbubble semantic temporal dynamic feature vector is classified and iterated through the water supply controller based on the classifier, the training microbubble semantic temporal dynamic feature vector is corrected.
[0092] In the technical solution described above, each training importance reset microbubble semantic feature vector in the sequence of training importance reset microbubble semantic feature vectors represents the confidence image semantic feature based on the autocorrelation of the local time domain image semantic feature of the corresponding training microbubble image in the time series. Therefore, when the sequence of the training importance reset microbubble semantic feature vectors passes through the information transmission module based on the prior distribution, it is expected that the training microbubble semantic time series dynamic feature vectors will maintain the confidence image semantic feature distribution distinguishability of each training importance reset microbubble semantic feature vector for the local time domain while improving the predetermined category harmony of each training importance reset microbubble semantic feature vector under the local time domain distribution distinguishability to the global time domain image semantic feature domain that transmits the prior image semantic feature distribution information to the local time domain based on the global time domain, thereby improving the global time domain image semantic feature expression effect of the training microbubble semantic time series dynamic feature vector for the predetermined class probability of the classifier.
[0093] Based on this, the present application corrects the training microbubble semantic time series dynamic feature vector each time the training microbubble semantic time series dynamic feature vector is classified and iterated by the water supply controller based on the classifier, specifically comprising: firstly calculating the autocorrelation matrix of the training microbubble semantic time series dynamic feature vector and its own transposition, and calculating the first and The inner product of the row vectors is used as the first The matrix value of the position is obtained to obtain a weight matrix, and then the weight matrix is matrix-multiplied with the autocorrelation matrix, and then further matrix-vector multiplied with the training microbubble semantic temporal dynamic feature vector to obtain a correction vector, and finally the correction vector is dot-multiplied with the training microbubble semantic temporal dynamic feature vector to obtain the corrected training microbubble semantic temporal dynamic feature vector.
[0094] Specifically, each time the training microbubble semantic temporal dynamic feature vector is classified and iterated by the water supply controller based on the classifier, the training microbubble semantic temporal dynamic feature vector is corrected by the following correction formula; wherein the correction formula is:
[0095]
[0096] in, represents the training microbubble semantic temporal dynamic feature vector, represents the transpose of a vector, represents the autocorrelation matrix, and The autocorrelation matrix is represented by and row vector, represents the weight matrix, represents the weight matrix The matrix value of the position, represents the height of the autocorrelation matrix, represents the set of real numbers, represents matrix multiplication, represents dot product, Represents the corrected semantic temporal dynamic feature vector of trained microbubbles.
[0097] In this way, based on the self-correlation dimension of the training microbubble semantic temporal dynamic feature vector as the object to be modulated, the correlation expansion is performed based on the spatial sub-dimensional complexity of the high-dimensional feature space of the feature distribution, thereby introducing the decomposable correlation dimension offset of the training microbubble semantic temporal dynamic feature vector into the heterogeneous correlation embedding space, thereby enhancing its correlation self-consistency through the joint fine-tuning of the decomposable dimension set represented by the heterogeneous correlation embedding space, so as to improve the predetermined category harmony of the feature set of the training microbubble semantic temporal dynamic feature vector in the associated target classification domain, thereby improving the global time domain image semantic feature expression effect of the training microbubble semantic temporal dynamic feature vector for the predetermined class probability of the classifier, and improving the accuracy of the classification result.
[0098] In summary, the above scheme is adopted to build an intelligent operation monitoring system of microbubble flotation machine based on automatic image acquisition and processing by using computer vision technology and deep learning algorithm. More specifically, it is expected to collect time series data of microbubble images through high-speed industrial cameras, capture the dynamic time series characteristics of microbubbles, and further transfer the characteristics with the prior distribution as the guiding information, so as to optimize the time series expression of microbubble characteristics, and realize adaptive control of water supply based on the dynamic time series expression information of microbubble characteristics.
[0099] Figure 3 FIG. 1 is a block diagram of a monitoring system for the operation process of a microbubble flotation machine according to an exemplary embodiment. Figure 3 As shown, the system 200 includes:
[0100] A microbubble image acquisition module 201 is used to acquire a time series of microbubble images collected by a camera module;
[0101] An image semantic feature extraction module 202 is used to extract image semantic features from the time series of the microbubble images to obtain a sequence of microbubble semantic feature vectors;
[0102] A confidence resetting and prior information transfer optimization module 203, configured to perform confidence resetting and prior information transfer optimization on the sequence of microbubble semantic feature vectors to obtain a microbubble semantic temporal dynamic feature vector; and
[0103] The water supply control instruction determination module 204 is used to determine the water supply control instruction based on the microbubble semantic temporal dynamic feature vector.
[0104] Reference below Figure 4 , which shows a schematic diagram of the structure of an electronic device 600 suitable for implementing the embodiment of the present disclosure. The terminal device in the embodiment of the present disclosure may include but is not limited to mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), vehicle terminals (such as vehicle navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 4 The electronic device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present disclosure.
[0105] like Figure 4As shown, the electronic device 600 may include a processing device (e.g., a central processing unit, a graphics processing unit, etc.) 601, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 602 or a program loaded from a storage device 608 into a random access memory (RAM) 603. In the RAM 603, various programs and data required for the operation of the electronic device 600 are also stored. The processing device 601, the ROM 602, and the RAM 603 are connected to each other via a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.
[0106] Typically, the following devices may be connected to the I / O interface 605: an input device 606 including, for example, a touch screen, a touch pad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; an output device 607 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 608 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 609. The communication device 609 may allow the electronic device 600 to communicate with other devices wirelessly or by wire to exchange data. Although Figure 4 The electronic device 600 is shown with various devices, but it should be understood that it is not required to implement or possess all the devices shown. More or fewer devices may be implemented or possessed instead.
[0107] In particular, according to an embodiment of the present disclosure, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present disclosure includes a computer program product, which includes a computer program carried on a non-transitory computer-readable medium, and the computer program contains program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through the communication device 609, or installed from the storage device 608, or installed from the ROM 602. When the computer program is executed by the processing device 601, the above-mentioned functions defined in the method of the embodiment of the present disclosure are executed.
[0108] It should be noted that the computer-readable medium disclosed above may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that may be used by or in combination with an instruction execution system, device or device. In the present disclosure, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, in which a computer-readable program code is carried. This propagated data signal may take a variety of forms, including but not limited to an electromagnetic signal, an optical signal, or any suitable combination of the above. The computer readable signal medium may also be any computer readable medium other than a computer readable storage medium, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. The program code contained on the computer readable medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.
[0109] In some embodiments, the client and the server may communicate using any currently known or future developed network protocol such as HTTP (HyperText Transfer Protocol), and may be interconnected with any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network ("LAN"), a wide area network ("WAN"), an internet (e.g., the Internet), and a peer-to-peer network (e.g., an ad hoc peer-to-peer network), as well as any currently known or future developed network.
[0110] The computer-readable medium may be included in the electronic device, or may exist independently without being incorporated into the electronic device.
[0111] Computer program code for performing the operations of the present disclosure may be written in one or more programming languages or a combination thereof, including, but not limited to, object-oriented programming languages, such as Java, Smalltalk, C++, and conventional procedural programming languages, such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).
[0112] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present disclosure. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some implementations as replacements, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.
[0113] The modules involved in the embodiments described in the present disclosure may be implemented by software or hardware. The name of a module does not limit the module itself in some cases. For example, a test parameter acquisition module may also be described as a "module for acquiring device test parameters corresponding to a target device".
[0114] The functions described above herein may be performed at least in part by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that may be used include: field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), and the like.
[0115] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.
[0116] Figure 5 FIG. 1 is an application scenario diagram of a method for monitoring the operation process of a microbubble flotation machine according to an exemplary embodiment. Figure 5 As shown, in this application scenario, first, a time series of microbubble images collected by a camera module is obtained (for example, Figure 5 Then, the time series of the acquired microbubble images is input to a server (for example, Figure 5 In the S) shown in FIG. 1 , the server is capable of processing the time series of the microbubble images based on a monitoring algorithm of the operation process of the microbubble flotation machine to determine a control instruction for the water supply amount.
[0117] The above description is only a preferred embodiment of the present disclosure and an explanation of the technical principles used. Those skilled in the art should understand that the scope of disclosure involved in the present disclosure is not limited to the technical solutions formed by a specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the above disclosed concept. For example, the above features are replaced with the technical features with similar functions disclosed in the present disclosure (but not limited to) by each other to form a technical solution.
[0118] In addition, although each operation is described in a specific order, this should not be understood as requiring these operations to be performed in the specific order shown or in a sequential order. Under certain circumstances, multitasking and parallel processing may be advantageous. Similarly, although some specific implementation details are included in the above discussion, these should not be interpreted as limiting the scope of the present disclosure. Some features described in the context of a separate embodiment can also be implemented in a single embodiment in combination. On the contrary, the various features described in the context of a single embodiment can also be implemented in multiple embodiments individually or in any suitable sub-combination mode.
[0119] Although the subject matter has been described in language specific to structural features and / or method logic actions, it should be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or actions described above. On the contrary, the specific features and actions described above are merely example forms of implementing the claims. Regarding the device in the above embodiment, the specific manner in which each module performs the operation has been described in detail in the embodiment related to the method, and will not be elaborated here.
Claims
1. A method for monitoring the operation process of a microbubble flotation machine, characterized in that: include: Acquire a time series of microbubble images acquired by a camera module; Extracting image semantic features from the time series of the microbubble images to obtain a sequence of microbubble semantic feature vectors; Resetting the confidence and optimizing the prior information transfer of the sequence of microbubble semantic feature vectors to obtain a microbubble semantic temporal dynamic feature vector; Determine the control instruction of water supply amount based on the microbubble semantic time series dynamic feature vector; The process of resetting the confidence level and optimizing the prior information transmission of the sequence of microbubble semantic feature vectors to obtain a microbubble semantic temporal dynamic feature vector comprises: Passing the sequence of microbubble semantic feature vectors through a feature autocorrelation confidence resetting network to obtain a sequence of importance-reset microbubble semantic feature vectors; Passing the sequence of importance-reset microbubble semantic feature vectors through an information transfer module based on prior distribution to obtain the microbubble semantic temporal dynamic feature vector; The process of passing the sequence of microbubble semantic feature vectors through a feature autocorrelation confidence resetting network to obtain a sequence of importance-reset microbubble semantic feature vectors comprises: The sequence of microbubble semantic feature vectors is processed using the following feature autocorrelation confidence reset formula to obtain the sequence of importance reset microbubble semantic feature vectors; wherein the feature autocorrelation confidence reset formula is: in, For the The attention score coefficient, For the A weight coefficient matrix, For the A vector of weight coefficients, For the A bias vector, represents the Selu activation function, is the first in the sequence of microbubble semantic feature vectors microbubble semantic feature vectors, and are weighting coefficients with different values, represents the value input into the Selu activation function, represents the softmax activation function, represents dot product, The first in the sequence of microbubble semantic feature vectors is reset for the importance The importance of resetting the microbubble semantic feature vector.
2. The method for monitoring the operation process of a microbubble flotation machine according to claim 1, characterized in that: The camera module is a high-speed industrial camera.
3. The method for monitoring the operation process of a microbubble flotation machine according to claim 2, characterized in that: Extracting image semantic features from the time series of the microbubble images to obtain a sequence of microbubble semantic feature vectors includes: performing image preprocessing on the time series of microbubble images to obtain a sequence of grayscale microbubble images; Each grayscale microbubble image in the sequence of grayscale microbubble images is passed through a microbubble feature extractor to obtain a sequence of microbubble semantic feature vectors.
4. The method for monitoring the operation process of a microbubble flotation machine according to claim 3, characterized in that: Performing image preprocessing on the time series of microbubble images to obtain a sequence of grayscale microbubble images includes: Grayscale processing is performed on each microbubble image in the time series of microbubble images to obtain a sequence of grayscale microbubble images.
5. The method for monitoring the operation process of a microbubble flotation machine according to claim 4, characterized in that: The sequence of the importance-reset microbubble semantic feature vectors is passed through an information transfer module based on a priori distribution to obtain the microbubble semantic temporal dynamic feature vector, including: The sequence of the importance-reset microbubble semantic feature vectors is processed using the following information transfer formula to obtain the microbubble semantic temporal dynamic feature vector; wherein the information transfer formula is: in, and are the first weight coefficient and the second weight coefficient respectively, and are the first random coefficient and the second random coefficient randomly sampled from the standard normal distribution, resetting the mean feature vector of the sequence of microbubble semantic feature vectors for said importance, The first in the sequence of microbubble semantic feature vectors is reset for the importance importance reset microbubble semantic feature vector, the number of sequences of microbubble semantic feature vectors reset for said importance, randomly selecting an importance-reset microbubble semantic feature vector from the sequence of importance-reset microbubble semantic feature vectors, is the microbubble semantic temporal dynamic feature vector.
6. The method for monitoring the operation process of a microbubble flotation machine according to claim 5, characterized in that: The control instruction of water supply amount is determined based on the microbubble semantic time series dynamic feature vector, including: The microbubble semantic temporal dynamic feature vector is passed through a water supply controller based on a classifier to obtain the control instruction, and the control instruction is used to indicate whether to increase the water supply.
7. The method for monitoring the operation process of a microbubble flotation machine according to claim 6, characterized in that: Also included is a training phase for training the microbubble feature extractor, the feature autocorrelation confidence reset network, the information transfer module based on the prior distribution, and the water supply controller based on the classifier; The training phase includes: Acquire a time series of training microbubble images; Performing image preprocessing on the time series of the training microbubble images to obtain a sequence of training grayscale microbubble images; Passing each training grayscale microbubble image in the sequence of training grayscale microbubble images through a microbubble feature extractor to obtain a sequence of training microbubble semantic feature vectors; Passing the sequence of training microbubble semantic feature vectors through a feature autocorrelation confidence reset network to obtain a sequence of training importance reset microbubble semantic feature vectors; Passing the sequence of the training importance reset microbubble semantic feature vectors through an information transfer module based on a prior distribution to obtain a training microbubble semantic temporal dynamic feature vector; Passing the training microbubble semantic temporal dynamic feature vector through a water supply controller based on a classifier to obtain a classification loss function value; The microbubble feature extractor, the feature autocorrelation confidence reset network, the information transfer module based on the prior distribution and the classifier-based water supply controller are trained based on the classification loss function value, wherein the training microbubble semantic temporal dynamic feature vector is corrected each time the training microbubble semantic temporal dynamic feature vector is classified iteratively through the classifier-based water supply controller.
8. A monitoring system for the operation process of a microbubble flotation machine, characterized in that: include: A microbubble image acquisition module, used to acquire a time series of microbubble images collected by a camera module; An image semantic feature extraction module, used for performing image semantic feature extraction on the time series of the microbubble images to obtain a sequence of microbubble semantic feature vectors; A confidence reset and prior information transfer optimization module, used for resetting the confidence and optimizing the prior information transfer of the sequence of microbubble semantic feature vectors to obtain a microbubble semantic temporal dynamic feature vector; A water supply control instruction determination module, used to determine the water supply control instruction based on the microbubble semantic time series dynamic feature vector; The process of resetting the confidence level and optimizing the prior information transmission of the sequence of microbubble semantic feature vectors to obtain a microbubble semantic temporal dynamic feature vector comprises: Passing the sequence of microbubble semantic feature vectors through a feature autocorrelation confidence resetting network to obtain a sequence of importance-reset microbubble semantic feature vectors; Passing the sequence of importance-reset microbubble semantic feature vectors through an information transfer module based on prior distribution to obtain the microbubble semantic temporal dynamic feature vector; The process of passing the sequence of microbubble semantic feature vectors through a feature autocorrelation confidence resetting network to obtain a sequence of importance-reset microbubble semantic feature vectors comprises: The sequence of microbubble semantic feature vectors is processed using the following feature autocorrelation confidence reset formula to obtain the sequence of importance reset microbubble semantic feature vectors; wherein the feature autocorrelation confidence reset formula is: in, For the The attention score coefficient, For the A weight coefficient matrix, For the A vector of weight coefficients, For the A bias vector, represents the Selu activation function, is the first in the sequence of microbubble semantic feature vectors microbubble semantic feature vectors, and are weighting coefficients with different values, represents the value input into the Selu activation function, represents the softmax activation function, represents dot product, The first in the sequence of microbubble semantic feature vectors is reset for the importance The importance of resetting the microbubble semantic feature vector.
Citation Information
Patent Citations
Real-time monitoring and early warning system and method for microbubble flotation machine
CN117943213A
