Weighing double-flow type cloud optimization flotation dosing control method and system

Abstract

The invention relates to a weighing double-flow type cloud optimization flotation dosing control method and system, and belongs to the field of mineral separation automation. The invention comprises a weighing double-flow type automatic dosing system, a weighing double-flow type automatic dosing control method, a dosing system cloud optimization system and a dosing system cloud optimization method. The weighing double-flow type automatic dosing system comprises a control host system and a dosing control device; the weighing double-flow type automatic dosing control method comprises a weighing type dosing flow detection method and a coarse-fine double-flow electromagnetic valve type dosing control method. The dosing system cloud optimization system comprises a foam camera, an image processing host, a network connection device and a cloud server. The medicine adding system cloud optimization method comprises a cloud medicine adding system self-optimization method and a cloud artificial expert cooperative work method. According to the invention, the dosing amount can be accurately controlled, the system adapts to changes of chemicals and environment temperature, the dosing system can be automatically optimized according to flotation conditions, and the system plays a positive role in stabilizing and improving flotation process indexes.

Description

Technical Field

[0001] The invention relates to a weighing double-flow cloud-based optimized flotation dosing control method and system, belonging to the technical field of mineral processing automation. Background Art

[0002] Flotation is the most widely used separation method in the mineral processing industry. The accuracy of flotation dosing and the rationality of the dosing system are related to the quality of flotation production indicators. If the amount of reagent added is too high, it will not only waste reagents and increase production costs, but also worsen the flotation environment and reduce flotation indicators; if the amount of reagent added is too low, the mineral and reagent will not interact enough, which will also reduce the flotation index. Therefore, accurate dosing and a reasonable dosing system are the prerequisites for efficient operation of flotation production.

[0003] The reagents added in flotation production need to reasonably set the given values ​​of the dosage of various reagents (collectively referred to as the dosing system) according to the actual situation or the change of ore properties, and accurately add them according to the dosing system. The control requirements for dosing in the flotation production process are mainly reflected in the following two aspects: (a) accurately control the dosage according to the given value requirements; (b) reasonably modify the dosing system according to the actual flotation production situation. Traditional dosing control systems are generally solenoid valve type and metering pump type. The dosage is mainly controlled by controlling the dosing time of the solenoid valve and the speed of the metering pump. The control accuracy is easily affected by factors such as changes in the physical and chemical properties of the liquid, changes in ambient temperature, and wear of components. In addition, most of the current dosing control systems output the dosage according to the given value requirements of the dosing. When the properties of the ore change, it is often necessary to rely on on-site workers to modify the flotation dosing system. It is impossible to automatically optimize the dosing system through the Internet, cloud, etc., or to ask remote manual experts to provide timely assistance through the Internet.

[0004] The applicant of the present invention has disclosed a patent technology, named A method and device for measuring and controlling the amount of liquid medicine in the mineral processing process, with the patent number of ZL202011224939.8. Its technical solution is: measure the weight of the liquid medicine by a force sensor, and control the output of the liquid medicine by a solenoid valve switch, and its control target is mainly the precise control of a specific liquid medicine. This technology still has some shortcomings, mainly manifested in: 1) Only one solenoid valve is used for dosing at the dosing point, and the output flow of the dosing solenoid valve cannot be changed according to control needs. When the dosing is nearing the end, it cannot be changed to a small flow rate for dosing, resulting in certain errors; 2) A dosing box is used to dosing at multiple dosing points at different times, and there is not much room for adjustment. If a small solenoid valve flow rate is selected, the dosing of multiple dosing points cannot be taken into account. If a large solenoid valve flow rate is selected, the dosing amount error is large; 3) There is no function of filtering the liquid medicine and caching the filtered liquid, which is easy to cause The mixing of impurities causes the dosing solenoid valve to malfunction, and the dosing is easily stopped due to the malfunction of the previous dosing device; 4) The fluidity of the frother used in flotation is greatly affected by temperature changes, and even normal dosing cannot be achieved when the temperature is very low. The technology does not provide a heating function for the reagents that are easily affected by temperature changes; 5) The technology requires on-site workers to set and modify the dosing system, and does not provide a remote optimization function for the dosing system. It cannot automatically optimize the dosing system according to changes in the properties of the flotation ore; 6) The technology also does not provide remote manual experts to help optimize the on-site dosing system.

[0005] Flotation dosing control is a complex process, which not only requires ensuring the accuracy of dosing, but also requires timely optimization of the dosing system according to changes in ore properties, so as to ensure stable and efficient operation of flotation production. Summary of the invention

[0006] In order to overcome the shortcomings of the existing flotation dosing control methods and control systems, further improve the accuracy, rationality and adaptability of flotation dosing and realize the intelligent optimization of the dosing system, the present invention provides a weighing dual-flow cloud-based optimized flotation dosing control method and system, which has a positive effect on improving the economic indicators of the flotation production process and reducing the consumption of reagents.

[0007] The technical solution of the present invention is: a weighing double-flow cloud-based optimization flotation dosing control system, including a weighing double-flow automatic dosing system and a dosing system cloud-based optimization system;

[0008] The weighing dual-flow automatic dosing system is used to automatically adapt to the dosing requirements of various physical and chemical properties of medicines, and accurately control the dosing amount of each dosing point according to the dosing system;

[0009] The dosing system cloud optimization system is responsible for automatically optimizing the flotation dosing system dynamically according to the changes in ore properties and actual flotation conditions;

[0010] The dosing system cloud optimization system includes a foam camera, an image processing host, a network connection device and a cloud server. The foam camera is used to obtain on-site flotation foam images, and the image processing host is used to perform feature image analysis and feature image extraction, and transmit the feature images to the cloud server.

[0011] The cloud server uses the cloud-based dosing system self-optimization method to analyze and intelligently calculate the characteristic images to obtain an optimized dosing system;

[0012] Manual experts remotely analyze the characteristic foam images and review and confirm the dosing system obtained by the cloud-based manual expert collaborative working method. The dosing system that needs to be confirmed by manual experts can be transmitted to the dosing control host system of the weighing dual-flow automatic dosing system for execution; the cloud-based dosing system self-optimization method is equipped with boundary conditions for the dosing system. When the dosing system is within the boundary range, the dosing system output by the cloud-based dosing system self-optimization method is directly transmitted to the dosing control host system for execution.

[0013] As a further solution of the present invention, the weighing dual-flow automatic dosing system includes a dosing control host system and a dosing control device; the dosing control host system is used for controlling and managing the weighing dual-flow automatic dosing system; the dosing control device is used for controlling the replenishment, filtration, storage, heating, liquid level detection, weight detection, drug output, and drug delivery of the drug solution.

[0014] As a further solution of the present invention, the dosing control host system comprises a dosing control host (1) and a peripheral circuit, wherein the peripheral circuit comprises a switching power supply, a drive circuit module, a switching electrical appliance and its circuit;

[0015] The dosing control device comprises: a medicine storage box replenishing solenoid valve 2, a filter 3, a medicine storage box 4, a medicine storage box liquid level tube 5, a pressure transmitter 6, a frame 7, a dosing output stop valve 8, a rough control solenoid valve 9, a fine control solenoid valve 10, a dosing flow regulator 11, a medicine receiving funnel 12, a dosing flushing water hand valve 13, a dosing flushing water main solenoid valve 14, a force sensor 15, a signal transmitter 16, a dosing box 17, a dosing replenishing solenoid valve 18, a dosing replenishing stop valve 19, a medicine storage box electric heating plate 22, and a medicine storage box thermal resistor 23;

[0016] The dosing control device is an integrated design; the connection method of the various components of the dosing control device is:

[0017] The medicine storage box medicine replenishment solenoid valve 2 is installed above the filter 3, the filter 3 is connected to the medicine storage box 4 below, the medicine storage box electric heating plate 22 and the medicine storage box thermal resistor 23 are respectively installed below and in front of the medicine storage box 4, and the medicine storage box 4 is installed on the frame 7; the pressure transmitter 6 and the medicine storage box liquid level tube 5 are installed in front of the medicine storage box 4;

[0018] A dosing stop valve 19 is installed on the side of the bottom of the rear side of the medicine storage box 4, and the dosing stop valve 19 is connected to the dosing solenoid valve 18, and the pipeline output end of the dosing solenoid valve 18 is fixed above the dosing box 17; a dosing output stop valve 8 is installed on the pipeline at the front and bottom of the dosing box 17, and a coarse control solenoid valve 9 and a fine control solenoid valve 10 are respectively installed on the pipeline with the dosing output stop valve 8, and a dosing flow regulator 11 is installed at the outlet of the coarse control solenoid valve 9 and the fine control solenoid valve 10, and a medicine receiving funnel 12 is installed below the outlet of the dosing flow regulator 11; there are several dosing boxes 17, and a force sensor is installed below each dosing box 17 15, the force sensor 15 is connected to the signal transmitter 16; the total solenoid valve 14 for dosing and flushing water is connected to a plurality of dosing and flushing water hand valves 13, the dosing and flushing water hand valves 13 are installed on the flushing water pipeline, and the flushing water pipeline outlet is installed above the medicine receiving funnel 12; the medicine storage box replenishment solenoid valve 2, the medicine storage box liquid level pipe 5, the pressure transmitter 6, the dosing output stop valve 8, the coarse control solenoid valve 9, the fine control solenoid valve 10, the dosing and flushing water hand valve 13, the total solenoid valve 14 for dosing and flushing water, the signal transmitter 16, the dosing replenishment solenoid valve 18, the dosing replenishment stop valve 19, the medicine storage box electric heating plate 22, and the medicine storage box thermal resistor 23 are all connected to the dosing control host 1;

[0019] The purpose of each component of the dosing control device is as follows: the medicine storage box replenishment solenoid valve 2, filter 3, medicine storage box 4, medicine storage box electric heating plate 22, and medicine storage box thermal resistor 23 are respectively used for liquid replenishment, filtration, storage, heating, and temperature measurement; the medicine storage box liquid level tube 5 and pressure transmitter 6 are respectively used for indicating and detecting the liquid level of the medicine storage box 4; the dosing replenishment solenoid valve 18 and dosing replenishment stop valve 19 are used for liquid replenishment and pipeline switch of the dosing box 17; the coarse control solenoid valve 9 and fine control solenoid valve 10 are respectively used for liquid replenishment and pipeline switch of the dosing box 17; the coarse control solenoid valve 9 and fine control solenoid valve 11 are respectively used for liquid replenishment and pipeline switch of the dosing box 17; the coarse control solenoid valve 9 and fine control solenoid valve 12 are respectively used for liquid replenishment and pipeline switch of the dosing box 17; the coarse control solenoid valve 9 and fine control solenoid valve 13 ... The valve 10, the dosing output stop valve 8, the dosing flow regulator 11, and the dosing funnel 12 are respectively used for large-flow dosing, small-flow dosing, dosing flow switch, liquid medicine flow regulation and liquid medicine collection; the force sensor 15 and the signal transmitter 16 are respectively used for weight measurement of the dosing box 17 and amplification and conditioning of the force sensor signal; the dosing flushing water manual valve 13 and the dosing flushing water main solenoid valve 14 are respectively used for regulating the flushing water flow of each dosing solenoid valve and controlling the main flushing water switch.

[0020] As a further solution of the present invention, the dosing flow regulator 11 includes: a dosing flow regulator connector 24, an adjusting screw 25, an adjusting pipe seat 26, and an elastic hose 27; the dosing flow regulator 11 is connected to the output ends of the coarse control solenoid valve 9 and the fine control solenoid valve 10 respectively through the dosing flow regulator connector 24, and the other end of the dosing flow regulator connector 24 is connected to the elastic hose 27, the elastic hose 27 is installed in the adjusting pipe seat 26, and the adjusting screw 25 is installed in the middle of the adjusting pipe seat 26; the liquid output flow of the coarse control solenoid valve 9 and the fine control solenoid valve 10 is adjusted by operating the adjusting screw 25.

[0021] As a further solution of the present invention, the image processing host 20 of the cloud-based optimization system for dosing system communicates with the cloud server, the dosing control host 1 and the foam camera 21 respectively through a network connection device; the foam camera 21 collects the foam images of the rough selection, fine selection and scanning selection of the flotation process in real time and transmits the images to the image processing host 20; the image processing host 20 analyzes the foam images, extracts characteristic foam images with strong correlation with flotation, and transmits the characteristic foam images together with the dosing system data provided by the dosing control host 1 to the cloud server.

[0022] The present invention also provides a weighing dual-flow cloud-based optimized flotation dosing control method, which includes a weighing dual-flow automatic dosing control method and a dosing system cloud-based optimization method;

[0023] The weighing dual-flow automatic dosing control method can accurately control the dosing amount of each dosing point according to the dosing system;

[0024] The cloud-based optimization method for the dosing system includes a cloud-based self-optimization method for the dosing system and a cloud-based human expert collaborative working method;

[0025] The cloud-based dosing system self-optimization method automatically analyzes and optimizes the characteristic image to obtain an optimized dosing system;

[0026] The cloud-based human expert collaborative working method optimizes the on-site medication dosing system through remote assistance when the human expert is not on-site, and specifically includes the following three situations: (1) the human expert independently proposes the medication dosing system; (2) the human expert reviews and confirms the medication dosing system output by the cloud-based medication dosing system self-optimization method; (3) after the human expert sets the boundary conditions, the medication dosing system obtained by the cloud-based medication dosing system self-optimization method is directly adopted when it does not exceed the boundary conditions.

[0027] As a further solution of the present invention, the weighing dual-flow automatic dosing control method includes:

[0028] Step S101: Calibration of the weighing device of the medicine adding box; the medicine adding control host system provides a calibration function for all weighing devices of the medicine adding box. The calibration does not require the use of liquid medicine, but only requires a standard weight. By opening the calibration interface of the weighing device of the medicine adding box, the weighing device of the medicine adding box to be calibrated is selected, and the calibration of the weighing device of the medicine adding box can be completed through the operations of "zeroing", "calibration", "calculation" and "save".

[0029] Step S102: adopt periodic dosing control mode to control the dosing amount of each dosing point; all dosing points of the dosing control system are in periodic dosing control mode, dosing is performed once in each dosing control cycle, and the control system automatically performs accurate dosing according to the given dosing value. The dosing control cycle ranges from 30s to 90s and is set according to actual needs; each dosing control cycle is divided into two parts: dosing time and replenishing time. The dosing time is about three quarters of the dosing control cycle, and the replenishing time is about one quarter of the dosing control cycle.

[0030] Step S103: the diameter of the elastic hose at the outlet of the coarse control solenoid valve is changed by the dosing flow regulator, and the maximum flow of the coarse control solenoid valve is adjusted to the range required for dosing control; the diameter of the elastic hose at the outlet of the coarse control solenoid valve is changed by adjusting the adjusting screw 25 of the dosing flow regulator 11, and the maximum flow of the coarse control solenoid valve 9 at each dosing point is adjusted to the required value, and the maximum flow is the average value of the output liquid amount of the coarse control solenoid valve within one-third of the dosing cycle, which is approximately the upper limit and the lower limit of the given dosing value;

[0031] Step S104: the diameter of the elastic hose at the outlet of the precision control solenoid valve is changed by the dosing flow regulator, and the maximum flow of the precision control solenoid valve is adjusted to the range required for dosing control; the diameter of the elastic hose at the outlet of the precision control solenoid valve is changed by adjusting the adjusting screw 25 of the dosing flow regulator 11, and the maximum flow of the precision control solenoid valve 10 is adjusted to the required value, and the maximum flow is that the amount of liquid medicine output by the precision control solenoid valve within a quarter of the dosing cycle is approximately one tenth of the average value of the upper and lower limits of the given dosing value;

[0032] Step S105: when the flow set value is very small, only the precision control solenoid valve is used to control the dosage; when the flow set value at the dosing point is less than one tenth of the average value of the upper and lower limits of the dosing set value, only the precision control solenoid valve 10 is used to control the dosage;

[0033] Step S106: The weight of the dosing box is measured by the force sensor, and the dosing amount is calculated by the change of the weight of the dosing box during the dosing cycle; the weight of the dosing box 17 is measured by the force sensor 15, and the force sensor signal is amplified and conditioned by the signal transmitter 16, converted into a current or voltage signal and transmitted to the dosing control host 1, and the weight of the dosing box 17 is calculated. The dosing amount is calculated by the change of the weight of the dosing box during the dosing control cycle. The mathematical model for calculating the dosing amount is:

[0034] W=K W (N0-N)

[0035] Where W is the dosage of this dosing control cycle, K W is the weight coefficient, N0 is the weight A / D conversion value of the dosing box 17 before dosing in this dosing control cycle, and N is the weight A / D conversion value of the current dosing box 17;

[0036] Step S107: The dosage is controlled to be about 95% of the given value by switching the coarse control solenoid valve; the control system automatically opens the coarse control solenoid valve 9 to add medicine, and when the measured value of the dosage reaches about 85% of the given value, the pulse switch coarse control solenoid valve 9 is used to open the solenoid valve for a short time to add medicine, and close the solenoid valve for a short time to perform static measurement, and this is repeated until the dosage is controlled to be about 95% of the given value;

[0037] Step S108: The dosage is controlled within the required error range by switching the precision control solenoid valve. The system enters the precision control solenoid valve switching control stage, and the control system automatically opens the precision control solenoid valve 10 to add medicine. When the dosage reaches 98.5% of the given value, intermittent switching control is adopted to add medicine, the solenoid valve is opened for a short time to add medicine, and the solenoid valve is closed for a short time to perform static measurement, and this is repeated until the dosage accuracy meets the requirements.

[0038] Step S109: During the replenishment time, when the weight of the dosing box is lower than the lower limit, replenishment of the dosing box is performed; when the weight of the dosing box is higher than the upper limit, replenishment of the dosing box is stopped. During the replenishment time of each dosing control cycle, when the weight of the dosing box 17 is lower than the lower limit, replenishment of the dosing box is performed; when the weight of the dosing box 17 is higher than the upper limit, replenishment of the dosing box is stopped;

[0039] Step S110: Detect the liquid level of the medicine storage box. When the liquid level is lower than the lower limit, open the medicine storage box medicine replenishment solenoid valve. When the liquid level is higher than the upper limit, close the medicine storage box medicine replenishment solenoid valve. Use the pressure transmitter 15 to detect the liquid level of the medicine storage box 4 in real time. When the liquid level of the medicine storage box 4 is lower than the lower limit setting value, automatically open the medicine storage box medicine replenishment solenoid valve 2 to replenish medicine. When the liquid level of the medicine storage box 4 is higher than the upper limit setting value, automatically close the medicine storage box medicine replenishment solenoid valve 2 to stop replenishing medicine.

[0040] Step S111: Detecting the temperature of the medicine liquid in the medicine storage box and heating control as needed; whether to detect the temperature of the medicine liquid in the medicine storage box and heating control according to actual needs; the method for detecting the temperature of the medicine liquid in the medicine storage box and heating control is: using the medicine storage box thermal resistor 23 to detect the temperature of the medicine liquid in the medicine storage box 4, when the temperature is lower than the lower limit set value, automatically turning on the medicine storage box electric heating plate 22 for heating, when the temperature is higher than the upper limit set value, turning off the power supply of the medicine storage box electric heating plate 22, so that the temperature of the medicine liquid is kept between the upper limit and the lower limit at low temperature;

[0041] Step S112: Regulate the flow rate of the dosing flushing water and control the switch as needed; select whether flushing water is needed according to actual conditions; the method of regulating the flow rate of the dosing flushing water and controlling the switch is: adjust the flushing water flow rate required for each dosing point through the dosing flushing water manual valve 13, and then control the switch of the flushing water of each dosing point through the dosing flushing water main solenoid valve 14, automatically open the flushing water when dosing control is performed, and automatically close the flushing water when dosing control is stopped.

[0042] As a further solution of the present invention, the working steps of the cloud-based optimization method for dosing system are as follows:

[0043] Step S201: flotation foam image acquisition, characteristic foam image extraction, and cloud transmission of characteristic foam images and dosing system data; specifically, the foam camera 21 acquires the foam images of the flotation tanks of roughing, fine selection, and sweeping selection in real time, and transmits the foam images to the image processing host 20; the image processing host 20 analyzes the foam images and extracts characteristic foam images closely related to flotation; the image processing host 20 obtains the current dosing system data from the dosing control host 1; the current dosing system data together with the characteristic foam images are transmitted to the cloud server, and processed by the cloud dosing system self-optimization method and the cloud artificial expert collaborative working method;

[0044] Step S202: Optimize the dosing system through the cloud-based dosing system self-optimization method; the cloud-based dosing system self-optimization method analyzes the characteristic foam image and the dosing system data, and optimizes the dosing system through the following intelligent algorithm; the technical solution for optimizing the dosing system is:

[0045] The state space S of the system t Screening was performed to obtain the multi-dimensional characteristics of flotation foam and the composition of the sequential dosing system:

[0046] S t ={F opt (t),Q t}

[0047] In the formula, F opt (t) is the set of flotation foam features that have been screened at time t; in particular, the number of features included in the set is not necessarily fixed, depending on the screening results; Q t represents the dosing system at the current time t;

[0048] Action Space A t It indicates the change of the dosing system adjustment strategy, which includes three actions: increase the given value, decrease the given value or keep it unchanged; the action indicates the adjustment amount of the dosing system, and the adjustment amount of the dosing system is set to ΔQ. The three actions are respectively expressed as:

[0049] Q t+1 =Q t +ΔQ,Q t+1 =Q t -ΔQ,Q t+1 =Q t

[0050] According to the current state S t Select an action and obtain the dosing system adjustment amount ΔQ through cloud decision-making to calculate the dosing system, thereby optimizing the flotation dosing system;

[0051] Step S203: Review, modify, confirm or re-formulate the dosing system through the cloud-based artificial expert collaborative working method; the cloud-based artificial expert collaborative working method provides the function of reviewing, modifying, confirming or re-setting the dosing system output by the cloud-based dosing system self-optimization method and the dosing system used on site, so as to ensure the safety and efficiency of flotation production operation and promptly solve the unreasonable dosing system problems arising in the production process.

[0052] As a further solution of the present invention, the image processing host 20 analyzes the foam image and extracts the characteristic foam image closely related to flotation. The technical solution is:

[0053] 1) The basic factors for the image processing host 20 to extract the characteristic foam image are: foam size, foam quantity, foam shape roundness, foam color, foam thickness and foam texture. The characteristic vector F is constructed as:

[0054] F={F1,F2,F3,F4,F5,F6}

[0055] In the formula, F1-F6 represent foam size, foam quantity, foam shape (roundness), foam color, foam thickness, and foam texture, respectively;

[0056] 2) Using multi-scale analysis, perform wavelet transform on the foam image I(x,y) to obtain the wavelet coefficient ψ a,b (x, y), after multi-scale analysis, six key foam characteristics are extracted; the wavelet transform multi-scale analysis formula is:

[0057]

[0058] Where a is the scale factor, which is used to control the width of the wavelet; b is the translation factor, which is used to control the position of the wavelet;

[0059] The six extracted foam characteristics were evaluated, and the steps included:

[0060] Step 1: Divide the foam image into image blocks of fixed size. Each image block is mapped to dimension D through linear embedding. The position encoding PE is added to the embedded feature. The position encoding formula is as follows:

[0061]

[0062] Among them, pos represents the position of the block, and i represents the dimension index;

[0063] Step 2: By calculating self-attention to focus on the important feature areas in the image, and after processing by multiple layers of self-attention and feedforward neural networks, the output feature vector can capture the dependency between foam features. The calculation formula for self-attention is:

[0064]

[0065] For foam feature evaluation and screening, the system generates the importance score P of each foam feature by linearly mapping the output feature vector. i , set the threshold δ, when a certain importance score P i When it is less than the threshold value δ, it is considered that the contribution of this feature to the flotation result is not obvious and is eliminated. The foam feature after screening is represented by the feature vector F opt (t) is expressed as:

[0066] F opt (t) = {F i |F i ∈F,P i ≥δ}

[0067] In the formula, F opt (t) is the filtered feature vector, which contains all the features that satisfy P at the current time t. i Characteristics of ≥δ.

[0068] As a further solution of the present invention, the cloud-based artificial expert collaborative working method includes:

[0069] 1) When the on-site flotation production is abnormal due to changes in ore properties, improper operation, etc., cloud-based artificial experts can access the cloud-based artificial experts through the Internet anywhere to work together. By extracting the current characteristic foam image, dosing system data and current flotation production index data, a new dosing system is made and tracked and optimized until the flotation production returns to normal;

[0070] 2) When the dosing system provided by the cloud-based dosing system self-optimization method undergoes major changes, cloud-based experts will review, modify, and confirm it before providing on-site production execution;

[0071] 3) When the source and properties of the ore being flotated change, cloud-based human experts will provide a new dosing system;

[0072] 4) Cloud-based human experts can provide on-site assistance in optimizing the dosing system at any location and at any time.

[0073] The beneficial effects of the present invention are:

[0074] 1. The present invention overcomes the problems existing in the existing dosing control system and method, which are easily affected by factors such as pressure changes, temperature changes, impurities in the liquid medicine, scaling of the pipeline, and wear of components, and has a wider range of applications and stronger adaptability;

[0075] 2. Each dosing point of the present invention is independently controlled, which is suitable for dosing control of large flow liquid medicine, as well as small flow or tiny flow liquid medicine; it is suitable for low viscosity liquid medicine, as well as medium and high viscosity liquid medicine;

[0076] 3. The present invention automatically detects the weight of each dosing box and calculates the output of each dosing point according to the change of the weight of the dosing box. The detection component does not contact the liquid medicine. The detection and control are carried out in a relatively independent environment, with higher control accuracy and better stability.

[0077] 4. The present invention uses one standard weight to calibrate all weighing devices of the medicine adding box, without the need to use liquid medicine for calibration, and the calibration is faster, more accurate and convenient;

[0078] 5. The present invention provides a cloud-based dosing system self-optimization method and a cloud-based human expert collaborative working method, which can realize the functions of automatic optimization and manual optimization of the dosing system. Human experts can access the cloud through the Internet at any time and any place to solve the problems raised by on-site flotation production and realize the high efficiency and real-time performance of flotation production;

[0079] 6. In the present invention, when the source of the ore treated by flotation and the properties of the ore change, a new dosing system can be provided by cloud-based artificial experts. The cloud-based artificial experts can be experts of the company or experts outside the company. They can solve the problems of the on-site dosing system and provide assistance in optimizing the on-site dosing system at any place and at any time. This can greatly improve the accuracy, efficiency, real-time and speed of the on-site dosing system optimization and greatly save the cost of artificial expert assistance. BRIEF DESCRIPTION OF THE DRAWINGS

[0080] Figure 1 It is a control system structure diagram of the present invention;

[0081] Figure 2 It is a control system connection diagram of the present invention;

[0082] Figure 3 It is a dosing control flow chart of the present invention;

[0083] Figure 4 It is a front view of the dosing control device of the present invention;

[0084] Figure 5 It is a structural diagram of the dosing flow regulator of the present invention;

[0085] Figure 6 is a set of characteristic foam images of the present invention, Figure 6 (a) represents the roughly selected feature foam image, Figure 6 (b) represents the selected feature foam image, Figure 6 (c) shows the scanned feature foam image.

[0086] The numbers in the figure are: 1-dosing control host, 2-drug storage box replenishment solenoid valve, 3-filter, 4-drug storage box, 5-drug storage box liquid level tube, 6-pressure transmitter, 7-frame, 8-dosing output stop valve, 9-coarse control solenoid valve, 10-fine control solenoid valve, 11-dosing flow regulator, 12-drug receiving funnel, 13-dosing flushing water hand valve, 14-dosing flushing water main solenoid valve, 15-force sensor, 16-signal transmitter, 17-dosing box, 18-dosing replenishment solenoid valve, 19-dosing replenishment stop valve, 20-image processing host, 21-foam camera, 22-drug storage box heating plate, 23-drug storage box thermal resistor, 24-dosing flow regulator joint, 25-adjusting screw, 26-adjusting pipe seat, 27-elastic hose. DETAILED DESCRIPTION

[0087] Example 1: A chalcopyrite concentrator is a flotation production process with an ore processing capacity of 2500t / d. The added reagents are ethyl xanthate, copper sulfate, sodium sulfite, and pine oil. Ethyl xanthate, copper sulfate, and sodium sulfite need to be added with water to form a 5% concentration of liquid. Pine oil is added directly, and the temperature of the pine oil is required to be not less than 6°C. The basic dosing system is shown in Table 1 below:

[0088] Table 1 Basic dosing system

[0089]

[0090] Note: The data in Table 1 are the mass of the drug solution.

[0091] The basic dosing system provides the given values ​​of the dosage of various reagents at each dosing point under normal circumstances. Since the properties of the ore will change frequently, it is necessary to make reasonable adjustments to the dosing system in a timely manner according to the actual flotation situation to optimize the flotation reagent conditions and stabilize and improve the flotation index. The fluctuation of the dosage of each dosing point is about ±15% of the basic dosing system.

[0092] According to the ore properties, basic dosing system and control requirements of this embodiment, the present invention provides a weighing dual-flow cloud-based optimized flotation dosing control method and system, which can improve the accuracy, rationality, adaptability and intelligent control level of dosing, and play a positive role in improving the economic indicators of flotation production process and reducing reagent consumption.

[0093] like Figure 1-Figure 5 As shown, the present invention first provides a weighing dual-flow cloud-based optimized flotation dosing control system, including a weighing dual-flow automatic dosing system and a dosing system cloud-based optimization system;

[0094] The weighing dual-flow automatic dosing system is used to automatically adapt to the dosing requirements of various physical and chemical properties of medicines, and accurately control the dosing amount of each dosing point according to the dosing system;

[0095] The dosing system cloud optimization system is responsible for automatically optimizing the flotation dosing system dynamically according to the changes in ore properties and actual flotation conditions;

[0096] The dosing system cloud optimization system includes a foam camera, an image processing host, a network connection device and a cloud server. The foam camera is used to obtain on-site flotation foam images, and the image processing host is used to perform feature image analysis and feature image extraction, and transmit the feature images to the cloud server.

[0097] The cloud server uses the cloud-based dosing system self-optimization method to analyze and intelligently calculate the characteristic images to obtain an optimized dosing system;

[0098] Manual experts remotely analyze the characteristic foam images and review and confirm the dosing system obtained by the cloud-based manual expert collaborative working method. The dosing system that needs to be confirmed by manual experts can be transmitted to the dosing control host system of the weighing dual-flow automatic dosing system for execution; the cloud-based dosing system self-optimization method is equipped with boundary conditions for the dosing system. When the dosing system is within the boundary range, the dosing system output by the cloud-based dosing system self-optimization method is directly transmitted to the dosing control host system for execution.

[0099] As a further solution of the present invention, the weighing dual-flow automatic dosing system includes a dosing control host system and a dosing control device; the dosing control host system is used for the control and management of the weighing dual-flow automatic dosing system; the dosing control device is used for the control of the replenishment, filtration, storage, heating, liquid level detection, weight detection, liquid output, and liquid delivery of the liquid medicine. The fluidity of low-viscosity liquid medicine is less affected by temperature changes and does not require heating. The fluidity of medium and high-viscosity liquid medicine is greatly affected by temperature changes. When the temperature is lower than a certain value, the fluidity becomes very poor or even loses fluidity, and heating is required.

[0100] The dosing control host system includes a dosing control host 1 and peripheral circuits. The core components of the dosing control host 1 are PLC and touch screen computer. The PLC selection is: 1 CPU module 1214DC / DC / DC, 3 DO modules 16-point voltage 24VDC, 5 AI modules 16-bit resolution 4 channels, 1 AO module 12-bit resolution 8 channels; the touch screen computer selection is: CPUI5 3320M, 8G memory, 256G hard disk, 17-inch screen; the peripheral circuit includes: 6 100W24VDC switching power supplies, 5 drive circuit modules, 1 20A2P circuit breaker, 6 10A2P circuit breakers, etc. The dosing control host 1 is connected to the medicine storage box replenishment solenoid valve 2, the pressure transmitter 6, the coarse control solenoid valve 9, the fine control solenoid valve 10, the dosing flushing water main solenoid valve 14, the force sensor 15, the signal transmitter 16, the dosing replenishment solenoid valve 18, the medicine storage box electric heating plate 22, and the medicine storage box thermal resistor 23 through the peripheral circuit;

[0101] The dosing control device comprises: a medicine storage box replenishing solenoid valve 2, a filter 3, a medicine storage box 4, a medicine storage box liquid level tube 5, a pressure transmitter 6, a frame 7, a dosing output stop valve 8, a rough control solenoid valve 9, a fine control solenoid valve 10, a dosing flow regulator 11, a medicine receiving funnel 12, a dosing flushing water hand valve 13, a dosing flushing water main solenoid valve 14, a force sensor 15, a signal transmitter 16, a dosing box 17, a dosing replenishing solenoid valve 18, a dosing replenishing stop valve 19, a medicine storage box electric heating plate 22, and a medicine storage box thermal resistor 23;

[0102] The dosing control device is an integrated design; all components are mounted on the frame 7 to form a whole, and can be used in combination according to actual needs. In this embodiment, there are 4 kinds of medicines, and each medicine requires at most four dosing points, so 4 dosing control devices need to be prepared.

[0103] The dosing control device, each dosing control device has the following components, specifications and quantities: medicine storage box replenishment solenoid valve 2: caliber DN25, power supply 24VDC, diaphragm type, 1; filter 3: diameter 150mm, filter screen 80 mesh, material 304 stainless steel, 1; medicine storage box 4: length * width * height = 800mm * 500mm * 500mm, 1; pressure transmitter 5: range 0-1000ka, power supply 24VDC, output signal 4-20mA, 1; medicine storage box liquid level tube 6: height 500mm, diameter 16mm, transparent ABS material, 1; rack: length * width * height = 900mm * 520mm * 1400mm, 1; dosing output stop valve 8: caliber DN20 material 304 stainless steel, 4; coarse control solenoid valve 9: caliber DN20 material 304 stainless steel, 4; precision control solenoid valve 10: caliber DN20, material 304 stainless steel, 4; dosing flow regulator 11: 8; medicine receiving funnel 12: , 4; dosing flushing water hand valve 13: caliber DN20, material 304 stainless steel, 4; dosing flushing water main solenoid valve 14: caliber DN25, material 304 stainless steel, 1; force sensor 15: for ethyl xanthate, copper sulfate, sodium sulfite agents, the specification is 20kg, 4, for pine oil, the specification is 10kg, 4; signal transmitter 16: power supply 24VDC, output signal 4-20mA, 4; dosing box 17: width * thickness * height = 180mm * 250mm * 300mm, 4; dosing supplement solenoid 18: caliber DN25, power supply 24VDC, diaphragm type, 4; dosing supplement stop valve 19: caliber DN25, 4.

[0104] The connection mode of the components of the dosing control device is as follows:

[0105] The medicine storage box medicine replenishment solenoid valve 2 is installed above the filter 3, the filter 3 is connected to the medicine storage box 4 below, the medicine storage box electric heating plate 22 and the medicine storage box thermal resistor 23 are respectively installed below and in front of the medicine storage box 4, and the medicine storage box 4 is installed on the frame 7; the pressure transmitter 6 and the medicine storage box liquid level tube 5 are installed in front of the medicine storage box 4;

[0106] A dosing stop valve 19 is installed on the side of the bottom of the back of the medicine storage box 4, and the dosing stop valve 19 is connected to the dosing solenoid valve 18, and the pipeline output end of the dosing solenoid valve 18 is fixed above the dosing box 17; a dosing output stop valve 8 is installed on the pipeline at the front and bottom of the dosing box 17, and a coarse control solenoid valve 9 and a fine control solenoid valve 10 are respectively installed on the pipeline with the dosing output stop valve 8, and a dosing flow regulator 11 is installed at the outlet of the coarse control solenoid valve 9 and the fine control solenoid valve 10, and a medicine receiving funnel 12 is installed below the outlet of the dosing flow regulator 11; there are 16 dosing boxes 17, and a force sensor is installed below each dosing box 17 15, the force sensor 15 is connected to the signal transmitter 16; the total solenoid valve 14 for dosing and flushing water is connected to 16 dosing and flushing water hand valves 13, the dosing and flushing water hand valves 13 are installed on the flushing water pipeline, and the flushing water pipeline outlet is installed above the medicine receiving funnel 12; the medicine storage box replenishment solenoid valve 2, the medicine storage box liquid level pipe 5, the pressure transmitter 6, the dosing output stop valve 8, the rough control solenoid valve 9, the fine control solenoid valve 10, the dosing and flushing water hand valve 13, the dosing and flushing water total solenoid valve 14, the signal transmitter 16, the dosing replenishment solenoid valve 18, the dosing replenishment stop valve 19, the medicine storage box electric heating plate 22, and the medicine storage box thermal resistor 23 are all connected to the dosing control host 1;

[0107] The purpose of each component of the dosing control device is as follows: the medicine storage box replenishment solenoid valve 2, filter 3, medicine storage box 4, medicine storage box electric heating plate 22, and medicine storage box thermal resistor 23 are respectively used for liquid replenishment, filtration, storage, heating, and temperature measurement; the medicine storage box liquid level tube 5 and pressure transmitter 6 are respectively used for indicating and detecting the liquid level of the medicine storage box 4; the dosing replenishment solenoid valve 18 and dosing replenishment stop valve 19 are used for liquid replenishment and pipeline switch of the dosing box 17; the coarse control solenoid valve 9 and fine control solenoid valve 10 are respectively used for liquid replenishment and pipeline switch of the dosing box 17; the coarse control solenoid valve 9 and fine control solenoid valve 11 are respectively used for liquid replenishment and pipeline switch of the dosing box 17; the coarse control solenoid valve 9 and fine control solenoid valve 12 are respectively used for liquid replenishment and pipeline switch of the dosing box 17; the coarse control solenoid valve 9 and fine control solenoid valve 13 ... The valve 10, the dosing output stop valve 8, the dosing flow regulator 11, and the dosing funnel 12 are respectively used for large-flow dosing, small-flow dosing, dosing flow switch, liquid medicine flow regulation and liquid medicine collection; the force sensor 15 and the signal transmitter 16 are respectively used for weight measurement of the dosing box 17 and amplification and conditioning of the force sensor signal; the dosing flushing water manual valve 13 and the dosing flushing water main solenoid valve 14 are respectively used for regulating the flushing water flow of each dosing solenoid valve and controlling the main flushing water switch.

[0108] As a further solution of the present invention, the dosing flow regulator 11 includes: a dosing flow regulator connector 24, an adjusting screw 25, an adjusting pipe seat 26, and an elastic hose 27; the dosing flow regulator 11 is connected to the output ends of the rough control solenoid valve 9 and the fine control solenoid valve 10 respectively through the dosing flow regulator connector 24, and the other end of the dosing flow regulator connector 24 is connected to the elastic hose 27, the elastic hose 27 is installed in the adjusting pipe seat 26, and the adjusting screw 25 is installed in the middle of the adjusting pipe seat 26; the output flow rate of the drug solution of the rough control solenoid valve 9 and the fine control solenoid valve 10 is adjusted by operating the adjusting screw 25. The dosing flow regulator 11 is connected to the output end of the rough control solenoid valve 9 or the fine control solenoid valve 10 through the dosing flow regulator connector 24, and the maximum output flow rate of the drug solution of the rough control solenoid valve 9 or the fine control solenoid valve 10 is adjusted by operating the inlet and outlet length of the adjusting screw 25 to meet the precise control requirements of different drug solution flows. By adjusting the dosing flow regulator 11, the dosing solenoid valves of the same specification can meet different dosing flow control requirements, which can avoid the problems of a wide variety of dosing solenoid valve spare parts in actual applications, resulting in high spare parts costs, inconvenient management and maintenance, etc.

[0109] As a further solution of the present invention, the cloud-based optimization system for dosing system includes a foam camera 21, an image processing host 20, a network connection device and a cloud server; the image processing host 20 communicates with the cloud server, the dosing control host 1 and the foam camera 21 respectively through the network connection device; the foam camera 21 collects the foam images of the rough selection, fine selection and scanning selection of the flotation process in real time and transmits the images to the image processing host 20, the image processing host 20 analyzes the foam images, extracts characteristic foam images with strong correlation with flotation, and transmits the characteristic foam images together with the dosing system data provided by the dosing control host 1 to the cloud server.

[0110] The present invention also provides a weighing dual-flow cloud-based optimized flotation dosing control method, which includes a weighing dual-flow automatic dosing control method and a dosing system cloud-based optimization method;

[0111] The weighing dual-flow automatic dosing control method is responsible for accurately controlling the dosing amount of each dosing point according to the dosing system; the dosing system cloud optimization method is responsible for automatically optimizing the flotation dosing system according to the flotation situation;

[0112] The weighing dual-flow automatic dosing control method includes a weighing dosing flow detection method and a coarse-fine dual-flow electromagnetic valve dosing control method. The dosing control can be unaffected by the physical and chemical properties of the liquid medicine and the changes in the ambient temperature, and can achieve accurate dosing control under various liquid medicine conditions and environmental conditions.

[0113] The cloud-based optimization method for the medication dosing system of this embodiment includes a cloud-based medication dosing system self-optimization method and a cloud-based human expert collaborative working method; the cloud-based medication dosing system self-optimization method automatically analyzes and optimizes the feature image to obtain an optimized medication dosing system; the cloud-based human expert collaborative working method optimizes the on-site medication dosing system through remote assistance when the human expert is not on-site, and specifically includes the following three situations: (1) the human expert independently proposes the medication dosing system; (2) the human expert reviews and confirms the medication dosing system output by the cloud-based medication dosing system self-optimization method; (3) after the human expert sets the boundary conditions, when the medication dosing system obtained by the cloud-based medication dosing system self-optimization method does not exceed the boundary conditions, it is directly adopted.

[0114] As a further solution of the present invention, the weighing dual-flow automatic dosing control method includes:

[0115] Step S101: Calibration of the medicine-dosing box weighing device. The medicine-dosing control host system provides the calibration function for all medicine-dosing box weighing devices. No liquid medicine is required for calibration, only a standard weight is needed, and the mass of the standard weight is 5kg. Open the calibration interface of the medicine-dosing box weighing device, select the medicine-dosing box weighing device that needs to be calibrated, click the "zero calibration" button when there is no liquid medicine in the medicine-dosing box, and complete the zero calibration operation after a few seconds. Then place the standard weight on the medicine-dosing box, click the "calibration" button, and complete the calibration operation after a few seconds. Click the "calculate" and "save" buttons to complete the calibration of the medicine-dosing box weighing device. Use the above method to calibrate all medicine-dosing box weighing devices one by one.

[0116] Step S102: Control the dosage of each dosing point by periodic dosing control. All dosing points of the dosing control system are in periodic dosing control mode. Dosing is performed once in each dosing control cycle. The control system automatically performs accurate dosing according to the given dosing value. The dosing control cycle is set to 60s. Each dosing control cycle is divided into two parts: dosing time and replenishing time. The dosing time is 45s and the replenishing time is 15s.

[0117] Step S103: The diameter of the elastic hose at the outlet of the coarse control solenoid valve is changed by the dosing flow regulator, and the maximum flow of the coarse control solenoid valve is adjusted to the range required for dosing control. The diameter of the elastic hose at the outlet of the coarse control solenoid valve is changed by adjusting the adjusting screw 25 of the dosing flow regulator 11, and the maximum flow of the coarse control solenoid valve 9 at each dosing point is adjusted to the required value. The maximum flow is within one-third of the dosing cycle, and the amount of liquid medicine output by the dosing solenoid valve is approximately the average of the upper and lower limits of the given dosing value. The flow debugging values ​​of the coarse control solenoid valve in this embodiment are shown in Table 2:

[0118] Table 2 Rough control electromagnetic flow debugging value

[0119]

[0120] Step S104: The diameter of the elastic hose at the outlet of the precision control solenoid valve is changed by the dosing flow regulator, and the maximum flow of the precision control solenoid valve is adjusted to the range required for dosing control. The diameter of the elastic hose at the outlet of the precision control solenoid valve is changed by adjusting the adjusting screw 23 of the dosing flow regulator 11, and the maximum flow of the precision control solenoid valve 10 is adjusted to the required value. The maximum flow is within a quarter of the dosing cycle, and the amount of liquid medicine output by the dosing solenoid valve is about one tenth of the average value of the upper and lower limits of the dosing given value. The flow debugging values ​​of the precision control solenoid valve in this embodiment are shown in Table 3:

[0121] Table 3 Precision control electromagnetic flow debugging value

[0122]

[0123] Step S105: When the given flow rate is very small, only the precision control solenoid valve is used to control the dosage. When the given flow rate at the dosing point is less than the precision control solenoid flow rate debugging value in Table 3, in order to ensure control accuracy, only the precision control solenoid valve 10 is used to control the dosage.

[0124] Step S106: Measure the weight of the dosing box through the force sensor, and calculate the dosing amount through the change of the weight of the dosing box during the dosing cycle. The weight of the dosing box 17 is measured through the force sensor 15, and the force sensor signal is amplified and conditioned through the signal transmitter 16, converted into a current or voltage signal and transmitted to the dosing control host 1, and the weight of the dosing box 17 is calculated. The dosing amount is calculated through the change of the weight of the dosing box during the dosing control cycle. The mathematical model for calculating the dosing amount is:

[0125] W=K W (N0-N)

[0126] Where W is the dosage of this dosing control cycle, K W is the weight coefficient, N0 is the weight computer A / D conversion value of the dosing box 17 before dosing in this dosing control cycle, and N is the weight computer A / D conversion value of the current dosing box 17.

[0127] The A / D conversion values ​​N0 and N are determined by the actual weight of the drug solution in the drug adding box. In this embodiment, the range of N0 and N is 8000 to 60000. Weight coefficient K W It is determined by the actual conditions of the sensors, signal amplification circuits, mechanical structures, etc. of each weighing system. In this embodiment, K W The range is 0.192 to 0.557.

[0128] Step S107: The dosage is controlled to be about 95% of the given value by switching the coarse control solenoid valve. If the system selects the coarse control solenoid valve 9 for dosing, the system automatically opens the coarse control solenoid valve 9 for dosing. Since there is a large reverse force when dosing, when the dosage measurement value reaches about 85% of the given value, the pulse switch coarse control solenoid valve 9 is used instead, the solenoid valve is opened for a short time to dosing, and the solenoid valve is closed for a short time to perform static measurement, and this is repeated until the dosage is controlled to be about 95% of the given value.

[0129] Step S108: Control the dosage within the required error range by switching the precision-controlled solenoid valve. The control system enters the precision-controlled solenoid valve switch control stage, and the system automatically opens the precision-controlled solenoid valve 10 to add medicine. Since the output flow of the medicine liquid from the precision-controlled solenoid valve 10 is very small, the medicine adding box detection device performs measurements in a nearly static state, and the detection accuracy is greatly improved. When the dosage reaches 98.5% of the given value, the intermittent switch control is used to add medicine, the solenoid valve is opened for a short time to add medicine, and the solenoid valve is closed for a short time to perform static measurement, and this is repeated until the dosing accuracy meets the requirements.

[0130] Step S109: During the replenishment time, when the weight of the dosing box is lower than the lower limit, replenishment is performed in the dosing box; when the weight of the dosing box is higher than the upper limit, replenishment is stopped. During the replenishment time of each dosing control cycle, the system automatically replenishes the medicine according to the weight of the dosing box 17; when the weight of the dosing box 17 is lower than the lower limit, replenishment is performed in the dosing box; when the weight of the dosing box 17 is higher than the upper limit, replenishment is stopped. In this embodiment, the upper weight limit of the dosing box for ethyl xanthate, copper sulfate, and sodium sulfite is 14kg, and the lower weight limit is 7kg; the upper weight limit of the pine oil dosing box is 7kg, and the lower weight limit is 4kg.

[0131] Step S110: Detect the liquid level of the medicine storage box. When the liquid level is lower than the lower limit, open the medicine storage box medicine replenishment solenoid valve. When the liquid level is higher than the upper limit, close the medicine storage box medicine replenishment solenoid valve. Use the pressure transmitter 15 to detect the liquid level of the medicine storage box 4 in real time. When the liquid level of the medicine storage box 4 is lower than the lower limit setting value, automatically open the medicine storage box medicine replenishment solenoid valve 2 to replenish medicine. When the liquid level of the medicine storage box 4 is higher than the upper limit setting value, automatically close the medicine storage box medicine replenishment solenoid valve 2 to stop replenishing medicine. The upper limit of the liquid level of the medicine storage box in this embodiment is 400mm, and the lower limit of the liquid level is 200mm.

[0132] Step S111: Detect the temperature of the medicine liquid in the medicine storage box and control the heating as needed. The method for detecting the temperature of the medicine liquid in the medicine storage box and controlling the heating is as follows: use the medicine storage box thermal resistor 23 to detect the temperature of the medicine liquid in the medicine storage box 4, and when the temperature is lower than 6°C, automatically turn on the medicine storage box heating plate 22 for heating, and when the temperature is higher than 8°C, turn off the power of the medicine storage box heating plate 22, so as to keep the temperature of the medicine liquid between 6°C and 8°C under low temperature conditions.

[0133] Step S112: Regulate the flow rate of the dosing flushing water and control the switch as needed. In this embodiment, pine oil, all the liquid medicines at the 3# dosing point, and all the liquid medicines at the 4# dosing point are flushing water. The method for regulating the flow rate of the dosing flushing water and controlling the switch is: adjust the flushing water flow rate required for each dosing point through the dosing flushing water hand valve 13, and then control the switch of the flushing water at each dosing point through the dosing flushing water main electromagnetic valve 14, turn on the flushing water when dosing control is performed, and turn off the flushing water when dosing control is stopped.

[0134] The dosing system cloud optimization system includes the following equipment and their selection: image processing host 20: model NVIDIA Jetson Nano, basic configuration is CPU quad-core ARM Cortex-A57, 1.43GHz, GPU NVIDIA MaxwellGPU, memory 4GB LPDDR4, network interface is 1 Gigabit Ethernet interface; foam camera 21 model is BasleracA2440-75um, 3 units, 1 unit is installed in each of the flotation tanks for roughing, concentrating and scanning, and the configuration of foam camera 21 is 2448×2048 resolution, maximum frame rate is 75 frames per second; cloud connection network module 100-1000M adaptive 1. A technical solution of a dosing system cloud optimization system is: the flotation foam image is collected in real time by the foam camera 21 and the image is transmitted to the image processing host 20, the image processing host 20 is connected to the cloud server through the network connection device, and directly communicates with the dosing control host 1 at the same time. The foam camera 21 collects the foam images of the roughing, fine and scanning of the flotation process in real time and transmits the images to the image processing host 20. The image processing host 20 analyzes the foam images, extracts characteristic foam images with strong correlation with flotation, and transmits the characteristic foam images together with the dosing system data provided by the dosing control host 1 to the cloud server.

[0135] As a further solution of the present invention, the working steps of the cloud-based optimization method for dosing system are as follows:

[0136] Step S201: flotation foam image acquisition, characteristic foam image extraction, and cloud transmission of characteristic foam images and dosing system data; specifically, the foam camera 21 collects the foam images of the flotation tanks of the roughing, fine and sweeping selections of the flotation process in real time, and transmits the foam images to the image processing host 20; the image processing host 20 analyzes the foam images and extracts the characteristic foam images closely related to flotation; the image processing host 20 obtains the current dosing system data from the dosing control host 1; the current dosing system data together with the characteristic foam images are transmitted to the cloud server, and processed by the cloud dosing system self-optimization method and the cloud artificial expert collaborative working method; one of the characteristic foam images extracted by the image processing host 20 in this embodiment is as follows Figure 6 As shown;

[0137] Step S202: Optimize the dosing system through the cloud-based dosing system self-optimization method; the cloud-based dosing system self-optimization method analyzes the characteristic foam image and the dosing system data, and optimizes the dosing system through the following intelligent algorithm; the technical solution for optimizing the dosing system is:

[0138] The state space S of the system t Screening was performed to obtain the multi-dimensional characteristics of flotation foam and the composition of the sequential dosing system:

[0139] S t ={F opt (t),Q t}

[0140] In the formula, F opt (t) is the set of flotation foam features that have been screened at time t; in particular, the number of features included in the set is not necessarily fixed, depending on the screening results; Q t represents the dosing system at the current time t;

[0141] Action Space A t It indicates the change of the dosing system adjustment strategy, which includes three actions: increase the given value, decrease the given value or keep it unchanged; the action indicates the adjustment amount of the dosing system, and the adjustment amount of the dosing system is set to ΔQ. The three actions are respectively expressed as:

[0142] Q t+1 =Q t +ΔQ,Q t+1 =Q t -ΔQ,Q t+1 =Q t

[0143] According to the current state S t Select an action and obtain the dosing system adjustment amount ΔQ through cloud decision-making to calculate the dosing system, thereby optimizing the flotation dosing system;

[0144] The resulting dosing regime is:

[0145] Table 4 Dosing system provided by cloud-based dosing system self-optimization method

[0146]

[0147] Step S203: Review, modify, confirm or re-formulate the dosing system through the cloud-based artificial expert collaborative working method; the cloud-based artificial expert collaborative working method provides the function of reviewing, modifying, confirming or re-setting the dosing system output by the cloud-based dosing system self-optimization method and the dosing system used on site, so as to ensure the safety and efficiency of flotation production operation and promptly solve the unreasonable dosing system problems that arise in the production process. The dosing system provided by the cloud-based artificial expert collaborative working module in this embodiment is:

[0148] Table 5 Dosing system provided by the collaborative work method of cloud-based artificial experts

[0149]

[0150] As a further solution of the present invention, the image processing host 20 analyzes the foam image and extracts the characteristic foam image closely related to flotation. The technical solution is:

[0151] 1) The basic factors for the image processing host 20 to extract the characteristic foam image are: foam size, foam quantity, foam shape roundness, foam color, foam thickness and foam texture. The characteristic vector F is constructed as:

[0152] F={F1,F2,F3,F4,F5,F6}

[0153] In the formula, F1-F6 represent foam size, foam quantity, foam shape (roundness), foam color, foam thickness, and foam texture, respectively;

[0154] 2) Using multi-scale analysis, perform wavelet transform on the foam image I(x,y) to obtain the wavelet coefficient ψ a,b (x, y), after multi-scale analysis, six key foam characteristics are extracted; the wavelet transform multi-scale analysis formula is:

[0155]

[0156] Where a is the scale factor, which is used to control the width of the wavelet; b is the translation factor, which is used to control the position of the wavelet;

[0157] The six extracted foam characteristics were evaluated, and the steps included:

[0158] Step 1: Divide the foam image into image blocks of fixed size. Each image block is mapped to dimension D through linear embedding. The position encoding PE is added to the embedded feature. The position encoding formula is as follows:

[0159]

[0160] Among them, pos represents the position of the block, and i represents the dimension index;

[0161] Step 2: By calculating self-attention to focus on the important feature areas in the image, and after processing by multiple layers of self-attention and feedforward neural networks, the output feature vector can capture the dependency between foam features. The calculation formula for self-attention is:

[0162]

[0163] For foam feature evaluation and screening, the system generates the importance score P of each foam feature by linearly mapping the output feature vector. i , set the threshold δ, when a certain importance score P i When it is less than the threshold value δ, it is considered that the contribution of this feature to the flotation result is not obvious and is eliminated. The foam feature after screening is represented by the feature vector F opt (t) is expressed as:

[0164] F opt (t) = {F i |F i ∈F,P i ≥δ}

[0165] In the formula, F opt (t) is the filtered feature vector, which contains all the features that satisfy P at the current time t. i Characteristics of ≥δ.

[0166] As a further solution of the present invention, the cloud-based artificial expert collaborative working method includes:

[0167] 1) When the on-site flotation production is abnormal due to changes in ore properties, improper operation, etc., cloud-based artificial experts can access the cloud-based artificial experts through the Internet anywhere to work together. By extracting the current characteristic foam image, dosing system data and current flotation production index data, a new dosing system is made and tracked and optimized until the flotation production returns to normal;

[0168] 2) When the dosing system provided by the cloud-based dosing system self-optimization method undergoes major changes, cloud-based experts will review, modify, and confirm it before providing on-site production execution;

[0169] 3) When the source and properties of the ore being flotated change, cloud-based human experts will provide a new dosing system;

[0170] 4) Cloud-based human experts can provide on-site assistance in optimizing the dosing system at any location and at any time.

[0171] The specific implementation modes of the present invention are described in detail above in conjunction with the accompanying drawings, but the present invention is not limited to the above implementation modes, and various changes can be made within the knowledge scope of ordinary technicians in this field without departing from the purpose of the present invention.

Claims

1. A weighing dual-flow cloud-based optimized flotation dosing control system, characterized by: Including weighing dual-flow automatic dosing system and dosing system cloud optimization system; The weighing dual-flow automatic dosing system is used to automatically adapt to the dosing requirements of various physical and chemical properties of medicines, and accurately control the dosing amount of each dosing point according to the dosing system; The dosing system cloud optimization system is responsible for automatically optimizing the flotation dosing system dynamically according to the changes in ore properties and actual flotation conditions; The dosing system cloud optimization system includes a foam camera, an image processing host, a network connection device and a cloud server. The foam camera is used to obtain on-site flotation foam images, and the image processing host is used to perform feature image analysis and feature image extraction, and transmit the feature images to the cloud server. The cloud server uses the cloud-based dosing system self-optimization method to analyze and intelligently calculate the feature images to obtain an optimized dosing system; Manual experts remotely analyze the characteristic foam images and review and confirm the dosing system obtained by the cloud-based manual expert collaborative working method. The dosing system that needs to be confirmed by manual experts can be transmitted to the dosing control host system of the weighing dual-flow automatic dosing system for execution; the cloud-based dosing system self-optimization method is equipped with boundary conditions for the dosing system. When the dosing system is within the boundary range, the dosing system output by the cloud-based dosing system self-optimization method is directly transmitted to the dosing control host system for execution.

2. According to claim 1, a weighing dual-flow cloud-based optimized flotation dosing control system is characterized by: The weighing dual-flow automatic dosing system includes a dosing control host system and a dosing control device; the dosing control host system is used for controlling and managing the weighing dual-flow automatic dosing system; the dosing control device is used for controlling the replenishment, filtration, storage, heating, liquid level detection, weight detection, drug output, and drug delivery of the drug solution.

3. A weighing dual-flow cloud-based optimized flotation dosing control system according to claim 2, characterized in that: The dosing control host system comprises a dosing control host (1) and a peripheral circuit, wherein the peripheral circuit comprises a switching power supply, a drive circuit module, a switching electrical appliance and its circuit; The dosing control device comprises: a medicine storage box replenishing solenoid valve (2), a filter (3), a medicine storage box (4), a medicine storage box liquid level tube (5), a pressure transmitter (6), a frame (7), a dosing output stop valve (8), a rough control solenoid valve (9), a fine control solenoid valve (10), a dosing flow regulator (11), a medicine receiving funnel (12), a dosing flushing water hand valve (13), a dosing flushing water main solenoid valve (14), a force sensor (15), a signal transmitter (16), a dosing box (17), a dosing replenishing solenoid valve (18), a dosing replenishing stop valve (19), a medicine storage box electric heating plate (22), and a medicine storage box thermal resistor (23); The dosing control device is an integrated design; the connection method of the various components of the dosing control device is: The medicine storage box medicine replenishing solenoid valve (2) is installed above the filter (3), the filter (3) is connected to the medicine storage box (4) below, the medicine storage box electric heating plate (22) and the medicine storage box thermal resistor (23) are installed below and in front of the medicine storage box (4) respectively, and the medicine storage box (4) is installed on the frame (7); the pressure transmitter (6) and the medicine storage box liquid level tube (5) are installed in front of the medicine storage box (4); A dosing stop valve (19) is installed on the side of the bottom of the back of the medicine storage box (4), and the dosing stop valve (19) is connected to the dosing solenoid valve (18), and the pipeline output end of the dosing solenoid valve (18) is fixed above the dosing box (17); a dosing output stop valve (8) is installed on the pipeline at the front and bottom of the dosing box (17), and a rough control solenoid valve (9) and a fine control solenoid valve (10) are respectively installed on the pipeline on which the dosing output stop valve (8) is installed, and a dosing flow regulator (11) is installed at the outlet of the rough control solenoid valve (9) and the fine control solenoid valve (10), and a medicine receiving funnel (12) is installed below the outlet of the dosing flow regulator (11); there are a plurality of dosing boxes (17), and a force sensor (15) is installed below each dosing box (17). ), the force sensor (15) is connected to the signal transmitter (16); the main solenoid valve (14) for dosing and flushing water is connected to a plurality of dosing and flushing water hand valves (13), the dosing and flushing water hand valves (13) are installed on the flushing water pipeline, and the flushing water pipeline outlet is installed above the medicine receiving funnel (12); the medicine storage box replenishment solenoid valve (2), the medicine storage box liquid level pipe (5), the pressure transmitter (6), the dosing output stop valve (8), the rough control solenoid valve (9), the fine control solenoid valve (10), the dosing and flushing water hand valve (13), the dosing and flushing water main solenoid valve (14), the signal transmitter (16), the dosing replenishment solenoid valve (18), the dosing replenishment stop valve (19), the medicine storage box electric heating plate (22), and the medicine storage box thermal resistor (23) are all connected to the dosing control host (1); The components of the medicine adding control device are used as follows: the medicine storage box medicine replenishment solenoid valve (2), the filter (3), the medicine storage box (4), the medicine storage box electric heating plate (22), and the medicine storage box thermal resistor (23) are respectively used for medicine liquid replenishment, filtering, storage, heating, and temperature measurement; the medicine storage box liquid level tube (5) and the pressure transmitter (6) are respectively used for indicating and detecting the liquid level of the medicine storage box (4); the medicine adding replenishment solenoid valve (18) and the medicine adding replenishment stop valve (19) are used for medicine liquid replenishment and pipeline switch of the medicine adding box (17); the coarse control solenoid valve (9) and the fine control solenoid valve (10) are respectively used for controlling the medicine liquid level of the medicine storage box (17); and the fine control solenoid valve (11) and the fine control solenoid valve (12) are respectively used for controlling the medicine liquid level of the medicine storage box (17); and the fine control solenoid valve (13) and the fine control solenoid valve (14) are respectively used for controlling the medicine liquid level of the medicine storage box (17); and the fine control solenoid valve (15) and the fine control solenoid valve (16) are respectively used for controlling the medicine liquid level of the medicine storage box (17); and the fine control solenoid valve (17) and the fine control solenoid valve (18) are respectively used for controlling the medicine liquid level of the medicine storage box (17); and the fine control solenoid valve (15) and the fine control solenoid valve (16 ...4); and the fine control solenoid valve (15) and the fine control solenoid valve (16) are respectively used for controlling the medicine liquid level of the medicine storage box (4); and the fine control solenoid valve (15) and the fine control solen The control solenoid valve (10), the dosing output stop valve (8), the dosing flow regulator (11), and the dosing funnel (12) are respectively used for large-flow dosing, small-flow dosing, dosing flow switch, liquid medicine flow regulation, and liquid medicine collection; the force sensor (15) and the signal transmitter (16) are respectively used for weight measurement of the dosing box (17) and amplification and conditioning of the force sensor signal; the dosing flushing water manual valve (13) and the dosing flushing water main solenoid valve (14) are respectively used for regulating the flushing water flow of each dosing solenoid valve and controlling the flushing water main switch.

4. A weighing dual-flow cloud-based optimized flotation dosing control system according to claim 3, characterized in that: The dosing flow regulator (11) comprises: a dosing flow regulator connector (24), an adjusting screw (25), an adjusting pipe seat (26), and an elastic hose (27); the dosing flow regulator (11) is respectively connected to the output ends of the coarse control solenoid valve (9) and the fine control solenoid valve (10) through the dosing flow regulator connector (24); the other end of the dosing flow regulator connector (24) is connected to the elastic hose (27); the elastic hose (27) is installed in the adjusting pipe seat (26); the adjusting screw (25) is installed in the middle of the adjusting pipe seat (26); the output flow of the drug solution of the coarse control solenoid valve (9) and the fine control solenoid valve (10) is adjusted by operating the adjusting screw (25).

5. According to claim 1, a weighing dual-flow cloud-based optimized flotation dosing control system is characterized by: The image processing host (20) of the dosing system cloud optimization system communicates with the cloud server, the dosing control host (1) and the foam camera (21) respectively through a network connection device; the foam images of rough selection, fine selection and scanning selection of the flotation process are collected in real time by the foam camera (21) and the images are transmitted to the image processing host (20); the foam images are analyzed by the image processing host (20), and characteristic foam images with strong correlation with flotation are extracted; the characteristic foam images are transmitted to the cloud server together with the dosing system data provided by the dosing control host (1).

6. A weighing dual-flow cloud-based optimized flotation dosing control method, characterized in that: The method includes a weighing dual-flow automatic dosing control method and a dosing system cloud optimization method; The weighing dual-flow automatic dosing control method can accurately control the dosing amount of each dosing point according to the dosing system; The cloud-based optimization method for the dosing system includes a cloud-based self-optimization method for the dosing system and a cloud-based human expert collaborative working method; The cloud-based medication dosing system self-optimization method automatically analyzes and optimizes the characteristic image to obtain an optimized medication dosing system; The cloud-based human expert collaborative working method optimizes the on-site medication dosing system through remote assistance when the human expert is not on-site, and specifically includes the following three situations: (1) the human expert independently proposes the medication dosing system; (2) the human expert reviews and confirms the medication dosing system output by the cloud-based medication dosing system self-optimization method; (3) after the human expert sets the boundary conditions, the medication dosing system obtained by the cloud-based medication dosing system self-optimization method is directly adopted when it does not exceed the boundary conditions.

7. A weighing dual-flow cloud-based optimized flotation dosing control method according to claim 6, characterized in that: The weighing dual-flow automatic dosing control method comprises: Step S101: Calibration of the weighing device of the medicine adding box; the medicine adding control host system provides the calibration function for all weighing devices of the medicine adding box. The calibration does not require the use of liquid medicine, but only requires a standard weight. By opening the calibration interface of the weighing device of the medicine adding box, the weighing device of the medicine adding box to be calibrated is selected, and the calibration of the weighing device of the medicine adding box can be completed through the operations of "zeroing", "calibration", "calculation" and "save". Step S102: adopt periodic dosing control mode to control the dosing amount of each dosing point; all dosing points of the dosing control system are in periodic dosing control mode, dosing is performed once in each dosing control cycle, and the control system automatically performs accurate dosing according to the given dosing value. The dosing control cycle ranges from 30s to 90s and is set according to actual needs; each dosing control cycle is divided into two parts: dosing time and replenishing time. The dosing time is about three quarters of the dosing control cycle, and the replenishing time is about one quarter of the dosing control cycle. Step S103: by changing the caliber of the elastic hose at the outlet of the coarse control solenoid valve through the dosing flow regulator, the maximum flow of the coarse control solenoid valve is adjusted to the range required for dosing control; by adjusting the adjusting screw (25) of the dosing flow regulator (11), the caliber of the elastic hose at the outlet of the coarse control solenoid valve is changed, and the maximum flow of the coarse control solenoid valve (9) at each dosing point is adjusted to a required value, wherein the maximum flow is the average value of the output amount of liquid medicine of the coarse control solenoid valve within one third of the dosing cycle and the upper and lower limits of the given dosing value; Step S104: the diameter of the elastic hose at the outlet of the precision control solenoid valve is changed by the dosing flow regulator, and the maximum flow of the precision control solenoid valve is adjusted to the range required for dosing control; the diameter of the elastic hose at the outlet of the precision control solenoid valve is changed by adjusting the adjusting screw (25) of the dosing flow regulator (11), and the maximum flow of the precision control solenoid valve (10) is adjusted to the required value, wherein the maximum flow is such that the amount of liquid medicine output by the precision control solenoid valve within one quarter of the dosing cycle is approximately one tenth of the average value of the upper and lower limits of the given dosing value; Step S105: when the flow rate set value is very small, only the precision control solenoid valve is used to control the dosage; when the flow rate set value at the dosing point is less than one tenth of the average value of the upper and lower limits of the dosing set value, only the precision control solenoid valve (10) is used to control the dosage; Step S106: The weight of the medicine adding box is measured by the force sensor, and the dosage is calculated by the change of the weight of the medicine adding box during the medicine adding cycle; the weight of the medicine adding box (17) is measured by the force sensor (15), and the force sensor signal is amplified and conditioned by the signal transmitter (16), converted into a current or voltage signal and transmitted to the medicine adding control host (1), and the weight of the medicine adding box (17) is calculated. The dosage is calculated by the change of the weight of the medicine adding box during the medicine adding control cycle. The mathematical model for calculating the dosage is: W=K W (N0-N) Where W is the dosage of this dosing control cycle, K W is the weight coefficient, N0 is the weight A / D conversion value of the dosing box (17) before dosing in this dosing control cycle, and N is the weight A / D conversion value of the current dosing box (17); Step S107: The dosage is controlled to be about 95% of the given value by switching the coarse control solenoid valve; the control system automatically opens the coarse control solenoid valve (9) to add the drug, and when the measured value of the dosage reaches about 85% of the given value, the pulse-type switching coarse control solenoid valve (9) is used to open the solenoid valve for a short time to add the drug, and close the solenoid valve for a short time to perform static measurement, and this is repeated until the dosage is controlled to be about 95% of the given value; Step S108: The dosage is controlled within the required error range by switching the precision control solenoid valve. The system enters the precision control solenoid valve switching control stage, and the control system automatically opens the precision control solenoid valve (10) to add medicine. When the dosage reaches 98.5% of the given value, intermittent switching control is adopted to add medicine, the solenoid valve is opened for a short time to add medicine, and the solenoid valve is closed for a short time to perform static measurement, and this is repeated until the dosage accuracy meets the requirements. Step S109: During the replenishment time, when the weight of the medicine adding box is lower than the lower limit, replenishment of the medicine adding box is performed; when the weight of the medicine adding box is higher than the upper limit, replenishment of the medicine adding box is stopped. During the replenishment time of each medicine adding control cycle, when the weight of the medicine adding box (17) is lower than the lower limit, replenishment of the medicine adding box is performed; when the weight of the medicine adding box (17) is higher than the upper limit, replenishment of the medicine adding box is stopped; Step S110: detecting the liquid level of the medicine storage box; when the liquid level is lower than the lower limit, opening the medicine storage box medicine replenishment solenoid valve; when the liquid level is higher than the upper limit, closing the medicine storage box medicine replenishment solenoid valve; using a pressure transmitter (15) to detect the liquid level of the medicine storage box (4) in real time; when the liquid level of the medicine storage box (4) is lower than the lower limit setting value, automatically opening the medicine storage box medicine replenishment solenoid valve (2) to replenish medicine; when the liquid level of the medicine storage box (4) is higher than the upper limit setting value, automatically closing the medicine storage box medicine replenishment solenoid valve (2) to stop replenishing medicine; Step S111: Detecting the temperature of the medicine liquid in the medicine storage box and controlling the heating as required; selecting whether to detect the temperature of the medicine liquid in the medicine storage box and controlling the heating as required; the method for detecting the temperature of the medicine liquid in the medicine storage box and controlling the heating as required is as follows: using a medicine storage box thermal resistor (23) to detect the temperature of the medicine liquid in the medicine storage box (4); when the temperature is lower than a lower limit set value, automatically turning on the medicine storage box electric heating plate (22) for heating; when the temperature is higher than an upper limit set value, turning off the power supply of the medicine storage box electric heating plate (22), thereby keeping the temperature of the medicine liquid between the upper limit and the lower limit at low temperatures; Step S112: Regulate the flow rate of the dosing flushing water and control the switch as needed; select whether flushing water is needed according to the actual situation; the method of regulating the flow rate of the dosing flushing water and controlling the switch is: adjust the flushing water flow rate required for each dosing point through the dosing flushing water manual valve (13), and then control the switch of the flushing water of each dosing point through the dosing flushing water main electromagnetic valve (14), automatically turn on the flushing water when dosing control is performed, and automatically turn off the flushing water when dosing control is stopped.

8. The weighing dual-flow cloud-based optimized flotation dosing control method according to claim 6 is characterized by: The working steps of the cloud-based optimization method for dosing system are as follows: Step S201: flotation foam image acquisition, characteristic foam image extraction, and cloud transmission of characteristic foam images and dosing system data; specifically, the steps include: using a foam camera (21) to collect foam images of flotation tanks of roughing, fine selection, and sweeping selection in real time, and transmitting the foam images to an image processing host (20); the image processing host (20) analyzes the foam images and extracts characteristic foam images that are closely related to flotation; the image processing host (20) obtains current dosing system data from a dosing control host (1); the current dosing system data together with the characteristic foam images are transmitted to a cloud server, and processed using a cloud dosing system self-optimization method and a cloud artificial expert collaborative working method; Step S202: Optimize the dosing system through the cloud-based dosing system self-optimization method; the cloud-based dosing system self-optimization method analyzes the characteristic foam image and the dosing system data, and optimizes the dosing system through the following intelligent algorithm; the technical solution for optimizing the dosing system is: The state space S of the system t Screening was performed to obtain the multi-dimensional characteristics of flotation foam and the composition of the sequential dosing system: S t ={F opt (t),Q t } In the formula, F opt (t) is the set of flotation foam features that have been screened at time t; in particular, the number of features included in the set is not necessarily fixed, depending on the screening results; Q t represents the dosing system at the current time t; Action Space A t It indicates the change of the dosing system adjustment strategy, which includes three actions: increase the given value, decrease the given value or keep it unchanged; the action indicates the adjustment amount of the dosing system, and the adjustment amount of the dosing system is set to ΔQ. The three actions are respectively expressed as: Q t+1 =Q t +ΔQ,Q t+1 =Q t -ΔQ,Q t+1 =Q t According to the current state S t Select an action and obtain the dosing system adjustment amount ΔQ through cloud decision-making to calculate the dosing system, thereby optimizing the flotation dosing system; Step S203: Review, modify, confirm or re-formulate the dosing system through the cloud-based artificial expert collaborative working method; the cloud-based artificial expert collaborative working method provides the function of reviewing, modifying, confirming or re-setting the dosing system output by the cloud-based dosing system self-optimization method and the dosing system used on site, so as to ensure the safety and efficiency of flotation production operation and promptly solve the unreasonable dosing system problems arising in the production process.

9. A weighing dual-flow cloud-based optimized flotation dosing control method according to claim 8, characterized in that: The image processing host (20) analyzes the foam image and extracts the characteristic foam image closely related to flotation. The technical solution is: 1) The basic factors for extracting characteristic foam images by the image processing host (20) are: foam size, foam quantity, foam shape (roundness), foam color, foam thickness and foam texture. The characteristic vector F is constructed as: F={F1,F2,F3,F4,F5,F6} In the formula, F1-F6 represent foam size, foam quantity, foam shape (roundness), foam color, foam thickness, and foam texture, respectively; 2) Using multi-scale analysis, perform wavelet transform on the foam image I(x,y) to obtain the wavelet coefficient ψ a,b (x, y), after multi-scale analysis, six key foam characteristics are extracted; the wavelet transform multi-scale analysis formula is: Where a is the scale factor, which is used to control the width of the wavelet; b is the translation factor, which is used to control the position of the wavelet; The six extracted foam characteristics were evaluated, and the steps included: Step 1: Divide the foam image into image blocks of fixed size. Each image block is mapped to dimension D through linear embedding. The position encoding PE is added to the embedded feature. The position encoding formula is as follows: and Among them, pos represents the position of the block, and i represents the dimension index; Step 2: By calculating self-attention to focus on the important feature areas in the image, and after processing by multiple layers of self-attention and feedforward neural networks, the output feature vector can capture the dependency between foam features. The calculation formula for self-attention is: For foam feature evaluation and screening, the system generates the importance score P of each foam feature by linearly mapping the output feature vector. i , set the threshold δ, when a certain importance score P i When it is less than the threshold value δ, it is considered that the contribution of this feature to the flotation result is not obvious and is eliminated. The foam feature after screening is represented by the feature vector F opt (t) is expressed as: F opt (t)={F i |F i ∈F,P i ≥δ} In the formula, F opt (t) is the filtered feature vector, containing all the features that satisfy P at the current time t. i Characteristics of ≥δ.

10. A weighing dual-flow cloud-based optimized flotation dosing control method according to claim 8 and claim 9, characterized in that: Cloud-based human expert collaboration methods include: 1) When the on-site flotation production is abnormal due to changes in ore properties, improper operation, etc., cloud-based artificial experts can access the cloud-based artificial experts through the Internet anywhere to work together. By extracting the current characteristic foam image, dosing system data and current flotation production index data, a new dosing system is made and tracked and optimized until the flotation production returns to normal; 2) When the dosing system provided by the cloud-based dosing system self-optimization method undergoes major changes, cloud-based experts will review, modify, and confirm it before providing on-site production execution; 3) When the source and properties of the ore being flotated change, cloud-based human experts will provide a new dosing system; 4) Cloud-based human experts can provide on-site assistance in optimizing the dosing system at any location and at any time.

Images