CN111709942B - Zinc flotation dosing amount prediction control method based on texture degree optimization - Google Patents

Zinc flotation dosing amount prediction control method based on texture degree optimization Download PDF

Info

Publication number
CN111709942B
CN111709942B CN202010602623.1A CN202010602623A CN111709942B CN 111709942 B CN111709942 B CN 111709942B CN 202010602623 A CN202010602623 A CN 202010602623A CN 111709942 B CN111709942 B CN 111709942B
Authority
CN
China
Prior art keywords
foam
texture
image
value
neural network
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN202010602623.1A
Other languages
Chinese (zh)
Other versions
CN111709942A (en
Inventor
唐朝晖
李涛
张虎
张国勇
罗金
李耀国
袁鹤
戴智恩
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Central South University
Original Assignee
Central South University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Central South University filed Critical Central South University
Priority to CN202010602623.1A priority Critical patent/CN111709942B/en
Publication of CN111709942A publication Critical patent/CN111709942A/en
Application granted granted Critical
Publication of CN111709942B publication Critical patent/CN111709942B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/0002—Inspection of images, e.g. flaw detection
    • G06T7/0004—Industrial image inspection
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B03—SEPARATION OF SOLID MATERIALS USING LIQUIDS OR USING PNEUMATIC TABLES OR JIGS; MAGNETIC OR ELECTROSTATIC SEPARATION OF SOLID MATERIALS FROM SOLID MATERIALS OR FLUIDS; SEPARATION BY HIGH-VOLTAGE ELECTRIC FIELDS
    • B03D—FLOTATION; DIFFERENTIAL SEDIMENTATION
    • B03D1/00—Flotation
    • B03D1/14—Flotation machines
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/11—Region-based segmentation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/40—Analysis of texture
    • G06T7/41—Analysis of texture based on statistical description of texture
    • G06T7/45—Analysis of texture based on statistical description of texture using co-occurrence matrix computation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/60—Analysis of geometric attributes
    • G06T7/62—Analysis of geometric attributes of area, perimeter, diameter or volume
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/90—Determination of colour characteristics
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20081—Training; Learning
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20084—Artificial neural networks [ANN]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30108—Industrial image inspection

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • General Physics & Mathematics (AREA)
  • Mathematical Physics (AREA)
  • Probability & Statistics with Applications (AREA)
  • Computing Systems (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Biotechnology (AREA)
  • Geometry (AREA)
  • Quality & Reliability (AREA)
  • Image Processing (AREA)
  • Image Analysis (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The invention discloses a zinc flotation dosing prediction control method based on texture degree optimization, which comprises the following steps of: firstly, extracting foam image characteristics including entropy, energy, inverse difference moment, foam size and foam color characteristics of a rough selection foam image and texture characteristics of a fine selection foam image, and defining texture complexity according to texture parameters of the fine selection foam image and a concentrate grade relation; and secondly, constructing a prediction model of the complexity of the selected foam texture by using the extracted characteristic parameters and a neural network, and finally calculating a medicament dosage adjustment value by using an optimization method according to the square of the difference value between the predicted value of the complexity of the selected foam texture and the expected optimal value as a target function to finish medicament adding control. The invention combines the foam state characteristics of fine selection and rough selection, so that the control result is more excellent, the recovery efficiency of minerals is improved, and the consumption of medicaments is reduced.

Description

Zinc flotation dosing amount prediction control method based on texture degree optimization
Technical Field
The invention belongs to the technical field of froth flotation, and particularly relates to a method for predicatively controlling the dosage in a zinc flotation process.
Background
The froth flotation is a process for separating minerals from useless gangue by utilizing the difference of the hydrophilicity and the hydrophobicity of the minerals, and the dosing amount in the zinc flotation process is mainly adjusted by judging the production state by observing the states of the color, the size, the texture and the like of the froth on the surface of the flotation cell by an experienced operator with naked eyes. However, the manual observation has strong subjectivity and randomness of operation, so that the dosing operation error is large and the efficiency is low in the flotation process, the serious waste of chemicals and mineral resources can be caused due to inaccurate judgment, the texture characteristics of the foam in the concentration tank are closely related to the concentrate grade of the foam, and the finally obtained concentrate grade can be well controlled to reach the standard by controlling the texture complexity in the concentration tank within a certain range.
The existing flotation process dosing control method mainly aims at the size characteristics of flotation foam, a Hammerstein-Wiener model is established to predict a foam size distribution curve of a roughing tank after dosing for 6 minutes by researching a size distribution probability function (PDF) of the foam size, whether the flotation is good or not is judged by analyzing the foam size, a dosing adjustment calculation model is established by utilizing the difference between the foam size distribution curve predicted by the model and an expected optimal size distribution curve (a foam bath size distribution curve in the roughing production process with high concentrate grade is obtained), and reasonable dosing prediction control is realized. However, the flotation process is a complex multi-process flow and is completed through the roughing, concentrating and scavenging processes, the method only analyzes and judges the foam state in the roughing tank, and the foam state of a plurality of flotation tanks is not combined, so that the problem that the production condition cannot be comprehensively monitored and adjusted in time exists in the practical industrial application.
The invention mainly combines two process flows of rough concentration and fine concentration to research the dosage prediction control of the zinc flotation process, and a neural network prediction model is constructed to predict the foam texture complexity of a fine concentration tank by combining the characteristic information of the color, the size, the entropy value, the energy, the inverse difference moment and the like of foam in a rough concentration tank. And taking the square of the difference between the predicted selected foam texture degree and the optimal texture complexity as a control target, establishing an expert experience fuzzy rule base, solving by adopting an optimization method, and calculating the value of the medicament quantity required to be modified to reach the control target. Through characteristic prediction of the fine selection groove, future working conditions are judged, the dosage of the medicament can be adjusted in advance, and the production process is adjusted in time.
Disclosure of Invention
The invention aims to provide a texture-based method for optimizing the dosage prediction control in the zinc flotation process.
The technical scheme adopted by the invention comprises the following specific steps:
s1, extracting foam image characteristics in a zinc flotation roughing tank, including entropy, energy, inverse difference moment, foam size and foam color of a flotation foam image: extracting features of n rough-selected foam images by adopting a space gray level co-occurrence matrix method, namely an SGLCM method, and extracting entropy, energy and inverse difference moment of the foam images to obtain specific data sets respectively Er＝[E1,E2,E3,E4...En]、Ar＝[A1,A2,A3,A4,...An]、Cr＝[C1,C2,C3,C4,...Cn]And (3) segmenting and extracting foam size characteristic data S ═ S1, S2, S3, S4.. Sn by adopting a classical watershed algorithm]The foam color characteristics were calculated via the HSV color channel to yield data Co ═ Co1, Co2, Co3,Co4...Con]forming a rough feature data set U by the five feature data;
s2, selecting selected foam images corresponding to the n selected foam images, and extracting texture characteristic parameters of the n selected foam images by adopting an SGLCM method:
entropy value
Figure GDA0003365879750000021
(Energy)
Figure GDA0003365879750000022
Moment of adverse difference
Figure GDA0003365879750000023
By analyzing the relationship between the three characteristic parameters and the foam image texture and the actual concentrate grade, the energy of the selected foam image corresponding to the high concentrate grade is low, the entropy value and the inverse difference moment are high, so that the complexity of the groove texture of the selected foam image is defined:
Figure GDA0003365879750000024
calculating texture complexity TC for the n selected foam images respectively to form a selected groove texture complexity data set T [ TC ]1,TC2,TC3,TC4...TCn]Where i, j is the pixel value of the bubble image, L is the quantization level of the image, pd(i, j) is the element in the ith row and the jth column of the spatial gray level co-occurrence matrix.
S3, constructing a BP neural network prediction model to predict the texture complexity of the selected foam image:
a. entropy E of the extracted rough foam imager＝[E1,E2,E3,E4...En]Energy Ar＝[A1,A2,A3,A4,...An]Negative moment of deviation Cr＝[C1,C2,C3,C4,...Cn]And a foam size S ═ S1, S2, S3, S4.. Sn]And foam color Co ═ Co1, Co2, Co3, Co4]5 input variables as input layers; selecting flute texture complexity TC as spiritOutput variables via the network.
b. Carrying out confidence degree distribution on 5 input variables according to the characteristic importance degree, and obtaining weight values of entropy, energy, inverse difference moment, foam size and foam color as w1, w2, w3, w4 and w5 according to the confidence degree;
c. the hidden layer node determination of the neural network is according to an empirical formula:
Figure GDA0003365879750000025
wherein m is the number of hidden layer nodes, g is the number of input layer nodes, o is the number of output layer nodes, and u is a constant; because the number of input variables is 5 and the number of output variables is 1, the number m of nodes of the hidden layer is determined to be between 2 and 13.
d. Inputting the input variable and the output variable with the determined weight into a neural network to train a selected groove texture complexity prediction model: and selecting M pieces of corresponding data from the feature data set U obtained in the step S1 and the texture data set T obtained in the step S2 to form a training set of the neural network, inputting the training set into the constructed BP neural network, and selecting a prediction error less than delta as a training end condition to obtain a prediction model.
e. Testing and modifying a predictive model using a test set
S4, counting the corresponding optimal texture T under the condition that the concentrate grade obtained in the flotation process is 52% -56% by adopting a statistical methodbThe selected foam texture degree predicted value T calculated by the BP neural network prediction modelpOptimum texture T with selected foambThe square of the difference is used as an objective function, an expert experience fuzzy rule base is established, an optimization method is adopted for solving, and the medicine adding amount x is calculated:
f(x)＝min{(Tp-Tb)2}
the fuzzy rule is defined as:
if E＝e0,A＝a0,C＝c0,S＝s0,Co＝Co0,Tp＝y0；thenx＝x0；
if E＝e1,A＝a1,C＝c1,S＝s1,Co＝Co1,Tp＝y1；thenx＝x1；
……
if E＝en,A＝an,C＝cn,S＝sn,Co＝Con,Tp＝yn；thenx＝xn；
according to the actual calculation process, wherein e0,e1…e n，a0,a1,…an，c0,c1,…cn，s0,s1,…sn，Co0,Co1,…ConRespectively the entropy of the extracted rough foam image, and the value range is [2.0,3.0 ]]Energy value range [0.2,0.4 ]]The inverse difference moment interval is [0.6,0.8 ]]The size range is [0.5,0.8 ]]The value interval of the color parameter sequence is in [0.4,0.7 ]]，y0,y1,…ynFor the prediction value of the selected foam texture degree, the value falls in [0.8,1.0 ]]，x0,x1,…xnAdjust the amount for the calculated dose [2000,4500]，Tb＝0.85。
In step a of S3, the weight values corresponding to the features of the roughly selected foam image are w 1-0.1, w 2-0.3, w 3-0.2, w 4-0.3, and w 5-0.1.
In step d of S3, the prediction model error Δ is 0.01.
In step c of S3, u is 0 < u < 10, and the number m of hidden layer nodes is 7.
Compared with the prior art, the method has the advantages that compared with the traditional flotation process agent amount control method, the method can consider the production state among multiple production processes, analyze and research the foam state between the fine concentration tank and the rough concentration tank, predict the actual production process and timely and effectively calculate the amount control of the chemicals added in the flotation process. The flotation process has long process flow and serious multi-process coupling condition, and the traditional medicament quantity control method only relates to the research on the influence and the effect of a single process or a single foam characteristic on the medicament adding quantity in the flotation process, so that the problems of misjudgment of production state and insufficient or excessive medicament quantity are easily increased, the waste of ore resources is caused, and the economic benefit of enterprises is reduced. Therefore, the invention considers the importance of a plurality of characteristics, establishes a neural network prediction model and utilizes a medicament quantity calculation method, effectively combines two processes of selection and scavenging, and realizes reasonable and efficient medicament quantity prediction control.
Drawings
FIG. 1 is a schematic flow diagram of a method for predictive control of dosing in a zinc flotation process according to the present invention;
fig. 2 is a block diagram of the present invention.
Detailed Description
In order to more specifically describe the technical solutions and advantages of the present invention, the following further describes the present invention in detail with reference to the accompanying drawings and embodiments.
S1, extracting foam image characteristics in a zinc flotation roughing tank, including entropy, energy, inverse difference moment, foam size and foam color of a flotation foam image: extracting features of n rough-selected foam images by adopting a space gray level co-occurrence matrix method, namely an SGLCM method, and extracting entropy, energy and inverse difference moment of the foam images to obtain specific data sets respectively Er＝[E1,E2,E3,E4...En]、Ar＝[A1,A2,A3,A4,...An]、Cr＝[C1,C2,C3,C4,...Cn]And (3) segmenting and extracting foam size characteristic data S ═ S1, S2, S3, S4.. Sn by adopting a classical watershed algorithm]The foam color characteristics were calculated via HSV color channel to obtain data Co ═ Co1, Co2, Co3, Co4.. Con]Forming a rough feature data set U by the five feature data;
s2, selecting selected foam images corresponding to the n selected foam images, and extracting texture characteristic parameters of the n selected foam images by adopting an SGLCM method:
entropy value
Figure GDA0003365879750000041
(Energy)
Figure GDA0003365879750000042
Moment of adverse difference
Figure GDA0003365879750000043
By analyzing the relationship between the three characteristic parameters and the foam image texture and the actual concentrate grade, the energy of the selected foam image corresponding to the high concentrate grade is low, the entropy value and the inverse difference moment are high, so that the complexity of the groove texture of the selected foam image is defined:
Figure GDA0003365879750000044
calculating texture complexity TC for the n selected foam images respectively to form a selected groove texture complexity data set T [ TC ]1,TT2,TC3,TC4...TCn]Where i, j is the pixel value of the bubble image, L is the quantization level of the image, pd(i, j) is the element of the ith row and the jth column in the space gray level co-occurrence matrix;
s3: constructing a BP neural network prediction model to predict the texture complexity of a selected foam image, and specifically comprising the following steps of:
step 1, extracting the entropy value E of the rough foam imager＝[E1,E2,E3,E4...En]Energy Ar＝[A1,A2,A3,A4,...An]Negative moment of deviation Cr＝[C1,C2,C3,C4,...Cn]And a foam size S ═ S1, S2, S3, S4.. Sn]And foam color Co ═ Co1, Co2, Co3, Co4]5 input variables as input layers; selecting groove texture complexity TC as an output variable of a neural network;
step 2, carrying out confidence degree distribution on the 5 input variables according to the characteristic importance degree, and obtaining weighted values of an entropy value, energy, an adverse moment, a foam size and a foam color as w1, w2, w3, w4 and w5 according to the confidence degree;
step 3, determining hidden layer nodes of the neural network according to an empirical formula:
Figure GDA0003365879750000051
wherein m is the number of hidden layer nodes, g is the number of input layer nodes, and o is the number of output layer nodes; because the number of input variables is 5 and the number of output variables is 1, the number m of nodes of the hidden layer is determined to be 2-13, and the calculation speed is increased and the calculation is reducedAnd calculating the error, and selecting m as 7.
And 4, step 4: inputting the input variable and the output variable with the determined weight into a neural network to train a selected groove texture complexity prediction model: selecting M pieces of corresponding data from the feature data set U obtained in the step S1 and the texture data set T obtained in the step S2 to form a training set of the neural network, inputting the training set into the constructed BP neural network, and selecting a prediction error smaller than delta as a training end condition to obtain a prediction model;
and 5: and (4) according to the complexity of the texture of the actual selected foam image and the error of the predicted value, checking and correcting the prediction model by using the test set.
S4: counting the corresponding optimal texture T under the condition that the concentrate grade obtained in the flotation process is about 54 percent by adopting a statistical methodbThe selected foam texture degree predicted value T calculated by the BP neural network prediction modelpOptimum texture T with selected foambThe square of the difference is used as an objective function, an expert experience fuzzy rule base is established, an optimization method is adopted for solving, and the medicine adding amount x is calculated:
f(x)=min{(Tp-Tb)2}
the fuzzy rule is defined as:
if E＝e0,A＝a0,C＝c0,S＝s0,Co＝Co0,Tb＝y0；thenx＝x0；
if E＝e1,A＝a1,C＝c1,S＝s1,Co＝Co1,Tb＝y1；thenx＝x1；
……
if E＝en,A＝an,C＝cn,S＝sn,Co＝Con,Tb＝yn；thenx＝xn
according to the actual calculation process, wherein e0,e1…e n，a0,a1,…an，c0,c1,…cn，s0,s1,…sn，Co0,Co1,…ConRespectively the entropy of the extracted rough foam image, and the value range is [2.0,3.0 ]]Energy value range [0.2,0.4 ]]The inverse difference moment interval is [0.6,0.8 ]]The size range is [0.5,0.8 ]]The value interval of the color parameter sequence is in [0.4,0.7 ]]，y0,y1,…ynFor the prediction value of the selected foam texture degree, the value falls in [0.8,1.0 ]]，x0,x1,…xnAdjust the amount for the calculated dose [2000,4500]，Tb＝0.85。

Claims (6)

1. A zinc flotation dosing quantity prediction control method based on texture degree optimization is characterized by comprising the following steps:
s1, extracting foam image characteristics in a zinc flotation roughing tank, including entropy, energy, inverse difference moment, foam size and foam color of a flotation foam image: extracting features of n rough selected foam images by adopting a space gray level co-occurrence matrix method, namely an SGLCM method, extracting entropy values, energy and inverse difference moments of the foam images to obtain specific data sets respectively Er ═ E1, E2, E3 and E4.]、Ar＝[A1，A2，A3，A4，...An]、Cr＝[C1，C2，C3，C4，...Cn]And (3) segmenting and extracting foam size characteristic data S ═ S1, S2, S3, S4.. Sn by adopting a classical watershed algorithm]The foam color characteristics were calculated via HSV color channel to obtain data Co ═ Co1, Co2, Co3, Co4.. Con]Forming a rough feature data set U by the five feature data;
s2, selecting selected foam images corresponding to the n selected foam images, and extracting texture characteristic parameters of the n selected foam images by adopting an SGLCM method:
entropy value
Figure FDA0003365879740000011
(Energy)
Figure FDA0003365879740000012
Moment of adverse difference
Figure FDA0003365879740000013
By analyzing the relationship between the three characteristic parameters and the foam image texture and the actual concentrate grade, the energy of the selected foam image corresponding to the high concentrate grade is low, the entropy value and the inverse difference moment are high, so that the complexity of the groove texture of the selected foam image is defined:
Figure FDA0003365879740000014
calculating texture complexity TC for the n selected foam images respectively to form a selected groove texture complexity data set T [ TC ]1，TC2，TC3，TC4...TCn]Where i, j is the pixel value of the bubble image, L is the quantization level of the image, pd(i, j) is the element of the ith row and the jth column in the space gray level co-occurrence matrix;
s3, constructing a BP neural network prediction model to predict the texture complexity of the selected foam image:
a. entropy values Er of the extracted rough foam images are [ E1, E2, E3, E4... En ═ E1, E2, E3, E E4... En]Energy Ar＝[A1，A2，A3，A4，...An]Negative moment of deviation Cr＝[C1，C2，C3，C4，...Cn]And a foam size S ═ S1, S2, S3, S4.. Sn]And foam color Co ═ Co1, Co2, Co3, Co4]5 input variables as input layers; selecting groove texture complexity TC as an output variable of a neural network;
b. carrying out confidence degree distribution on 5 input variables according to the characteristic importance degree, and obtaining weight values of entropy, energy, inverse difference moment, foam size and foam color as w1, w2, w3, w4 and w5 according to the confidence degree;
c. the hidden layer node determination of the neural network is according to an empirical formula:
Figure FDA0003365879740000015
wherein m is the number of hidden layer nodes, g is the number of input layer nodes, and o is the number of output layer nodes; the number of input variables is 5, the number of output variables is 1, and the number m of nodes of the selected hidden layer is 2-13;
d. inputting the input variable and the output variable of the determined weight into a neural network to train a selected groove texture complexity prediction model: selecting M pieces of corresponding data from the feature data set U obtained in the step S1 and the texture data set T obtained in the step S2 to form a training set of the neural network, inputting the training set into the constructed BP neural network, and selecting a prediction error smaller than delta as a training end condition to obtain a prediction model;
e. testing and modifying a predictive model using a test set
S4, counting the corresponding optimal texture T under the condition that the concentrate grade obtained in the flotation process is 52% -56% by adopting a statistical methodbThe selected foam texture degree predicted value T calculated by the BP neural network prediction modelpOptimum texture T with selected foambThe square of the difference is used as an objective function, an expert experience fuzzy rule base is established, an optimization method is adopted for solving, and the medicine adding amount x is calculated:
f(x)＝min{(Tp-Tb)2}
the fuzzy rule is defined as:
if E＝e0，A＝a0，C＝c0，S＝s0，Co＝Co0，Tp＝y0；then x＝x0；
if E＝e1，A＝a1，C＝c1，S＝s1，Co＝Co1，Tp＝y1；then x＝x1；
……
if E＝en，A＝an，C＝cn，S＝sn，Co＝Con，Tp＝yn；then x＝xn；
according to the actual calculation process, wherein e0，e1...en，a0，a1，...an，c0，c1，...cn，s0，s1，...sn，Co0，Co1，...ConRespectively is the entropy of the extracted rough foam image, and the value range is [2 ]2.0，3.0]Energy value range [0.2,0.4 ]]The inverse difference moment interval is [0.6,0.8 ]]The size range is [0.5,0.8 ]]The value interval of the color parameter sequence is in [0.4,0.7 ]]，y0，y1，...ynFor the prediction value of the selected foam texture degree, the value falls in [0.8,1.0 ]]，x0，x1，...xnAdjust the amount for the calculated agent [2000,4500 ]]，Tb＝0.85。
2. The zinc flotation dosing prediction control method based on texture degree optimization as claimed in claim 1, wherein the total number n of roughed froth images is 3000; the number of training set data pieces M was 2000.
3. The texture-based optimization zinc flotation dosing prediction control method is characterized in that the weight value of each feature of the rough foam image is w 1-0.1, w 2-0.3, w 3-0.2, w 4-0.3 and w 5-0.1.
4. The zinc flotation dosing prediction control method based on texture optimization as claimed in claim 1, wherein the prediction model error Δ is 0.01.
5. The zinc flotation dosing prediction control method based on texture optimization as claimed in claim 1, wherein in step c of S3, u is 0 < u < 10.
6. The zinc flotation dosing prediction control method based on texture degree optimization as claimed in claim 1, wherein the number m of hidden layer nodes is 7.
CN202010602623.1A 2020-06-29 2020-06-29 Zinc flotation dosing amount prediction control method based on texture degree optimization Active CN111709942B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010602623.1A CN111709942B (en) 2020-06-29 2020-06-29 Zinc flotation dosing amount prediction control method based on texture degree optimization

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010602623.1A CN111709942B (en) 2020-06-29 2020-06-29 Zinc flotation dosing amount prediction control method based on texture degree optimization

Publications (2)

Publication Number Publication Date
CN111709942A CN111709942A (en) 2020-09-25
CN111709942B true CN111709942B (en) 2022-04-12

Family

ID=72544266

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010602623.1A Active CN111709942B (en) 2020-06-29 2020-06-29 Zinc flotation dosing amount prediction control method based on texture degree optimization

Country Status (1)

Country Link
CN (1) CN111709942B (en)

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113112466A (en) * 2021-03-30 2021-07-13 湖州霍里思特智能科技有限公司 Ore identification method, detection mechanism and mineral product sorting machine
CN113128126B (en) * 2021-04-26 2022-06-10 湖南理工学院 Modeling method of flotation dosing process based on generation of countermeasure network
CN113713968A (en) * 2021-08-30 2021-11-30 东北大学 In-situ online characterization method for structural characteristics of micro-fine particle iron mineral aggregate

Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2005293840A (en) * 2005-05-25 2005-10-20 Hoya Corp Glass disk substrate, magnetic disk, method for manufacturing glass disk substrate for magnetic disk, and method for manufacturing magnetic disk
CN102855492A (en) * 2012-07-27 2013-01-02 中南大学 Classification method based on mineral flotation foam image
CN103839057A (en) * 2014-03-28 2014-06-04 中南大学 Antimony floatation working condition recognition method and system
CN104331714A (en) * 2014-11-28 2015-02-04 福州大学 Image data extraction and neural network modeling-based platinum flotation grade estimation method
CN105787562A (en) * 2016-03-24 2016-07-20 上海工程技术大学 Yarn dyed fabric texture type recognition method
CN106530296A (en) * 2016-11-07 2017-03-22 首都医科大学 Lung detection method and device based on PET/CT image features
CN110310286A (en) * 2019-07-02 2019-10-08 中南大学 Layering operating condition sensing method of discrimination based on froth images multiple features
CN110738271A (en) * 2019-07-13 2020-01-31 中南大学 Concentrate grade prediction method in zinc flotation processes
CN110918266A (en) * 2019-12-10 2020-03-27 中南大学 Control device and method for improving flotation froth quality

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP2309738A1 (en) * 2005-05-23 2011-04-13 Thomas S. Gilley Distributed scalable media environment
CN101334366B (en) * 2008-07-18 2011-05-25 中南大学 Flotation recovery rate prediction method based on image characteristic analysis
US10896535B2 (en) * 2018-08-13 2021-01-19 Pinscreen, Inc. Real-time avatars using dynamic textures
CN111324989B (en) * 2020-03-19 2024-01-30 重庆大学 Gear contact fatigue life prediction method based on GA-BP neural network

Patent Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2005293840A (en) * 2005-05-25 2005-10-20 Hoya Corp Glass disk substrate, magnetic disk, method for manufacturing glass disk substrate for magnetic disk, and method for manufacturing magnetic disk
CN102855492A (en) * 2012-07-27 2013-01-02 中南大学 Classification method based on mineral flotation foam image
CN103839057A (en) * 2014-03-28 2014-06-04 中南大学 Antimony floatation working condition recognition method and system
CN104331714A (en) * 2014-11-28 2015-02-04 福州大学 Image data extraction and neural network modeling-based platinum flotation grade estimation method
CN105787562A (en) * 2016-03-24 2016-07-20 上海工程技术大学 Yarn dyed fabric texture type recognition method
CN106530296A (en) * 2016-11-07 2017-03-22 首都医科大学 Lung detection method and device based on PET/CT image features
CN110310286A (en) * 2019-07-02 2019-10-08 中南大学 Layering operating condition sensing method of discrimination based on froth images multiple features
CN110738271A (en) * 2019-07-13 2020-01-31 中南大学 Concentrate grade prediction method in zinc flotation processes
CN110918266A (en) * 2019-12-10 2020-03-27 中南大学 Control device and method for improving flotation froth quality

Non-Patent Citations (4)

* Cited by examiner, † Cited by third party
Title
Color co-occurrence matrix based froth image texture extraction for mineral flotation;Weihua Gui 等;《Minerals Engineering》;20130630;第46-47卷;第60-67页 *
Machine Vision Based Production Condition Classification and Recognition for Mineral Flotation Process Monitoring;Liu Jinpng 等;《International Journal of Computational Interlligence Systems》;20130930;第6卷(第5期);第969-986页 *
基于CCHS的浮选泡沫图像纹理特征提取;陈宁 等;《中南大学学报（自然科学版）》;20131130;第44卷(第11期);第4506-4512页 *
浮选泡沫图像识别系统的设计与实现;黄胜国 等;《工业控制计算机》;20061231;第19卷(第6期);第62-63、68页 *

Also Published As

Publication number Publication date
CN111709942A (en) 2020-09-25

Similar Documents

Publication Publication Date Title
CN111709942B (en) Zinc flotation dosing amount prediction control method based on texture degree optimization
CN110378799B (en) Alumina comprehensive production index decision method based on multi-scale deep convolution network
Wu et al. Regional water resource carrying capacity evaluation based on multi-dimensional precondition cloud and risk matrix coupling model
CN108846526A (en) A kind of CO2 emissions prediction technique
CN103632203A (en) Distribution network power supply area division method based on comprehensive evaluation
CN110175416A (en) Three Gorges Reservoir water temperature prediction method based on principal component analysis and neural network
CN105404926A (en) Aluminum electrolytic production technology optimization method based on BP neural network and MBFO algorithm
CN104375478A (en) Method and device for online predicting and optimizing product quality in steel rolling production process
CN111260117A (en) CA-NARX water quality prediction method based on meteorological factors
CN103440525A (en) Urban lake and reservoir water bloom emergency treatment multiple-target multiple-layer decision-making method based on Vague value similarity measurement improved algorithm
CN114777192B (en) Secondary network heat supply autonomous optimization regulation and control method based on data association and deep learning
CN110728329B (en) Concentrate grade prediction method based on feedback compensation mechanism optimization in zinc flotation process
Xu et al. Evaluation of island tourism sustainable development in the context of smart tourism
Ai et al. Shape-weighted bubble size distribution based reagent predictive control for the antimony flotation process
CN106874568B (en) A kind of material particular diameter distribution forecasting method of mechanical milling process
CN115271437A (en) Water resource configuration method and system based on multi-decision-making main body
Zhang et al. Generative adversarial network-based image-level optimal setpoint calculation for flotation reagents control
CN101702172A (en) Data discretization method based on category-attribute relation dependency
CN109359320A (en) Blast furnace index prediction technique based on multi-sampling rate ARDL model
CN114202086A (en) Multi-objective optimization method for mining scheme
CN109636021B (en) Task reliability oriented manufacturing system selective maintenance decision method
Zhao et al. Simulation and evaluation on the eco-industrial system of Changchun economic and technological development zone, China
CN116503378A (en) Concentrate grade prediction method for bimodal CNN secondary transfer learning
Liu et al. Improved design of risk assessment model for PPP project under the development of marine architecture
CN114169699A (en) Water resource system sustainable development evaluation method considering combination of uncertainty and subjective and objective weights

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant