CN109885012A - A kind of gold hydrometallurgy whole process real-time optimization compensation method - Google Patents
A kind of gold hydrometallurgy whole process real-time optimization compensation method Download PDFInfo
- Publication number
- CN109885012A CN109885012A CN201910236152.4A CN201910236152A CN109885012A CN 109885012 A CN109885012 A CN 109885012A CN 201910236152 A CN201910236152 A CN 201910236152A CN 109885012 A CN109885012 A CN 109885012A
- Authority
- CN
- China
- Prior art keywords
- variable
- data
- whole process
- evaluation
- optimization
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/02—Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]
Landscapes
- Feedback Control In General (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present invention relates to a kind of golden hydrometallurgy whole process real-time optimization compensation methodes；It include: that S1 application process evaluation of running status method carries out on-line analysis acquisition evaluation result to golden hydrometallurgy whole process real-time optimization result；S2 selects matched compensation method to be handled for evaluation result；S21 is directed to the case where evaluation result is suboptimum, using the compensation method of self-optimizing control；S22 is non-optimum situation for evaluation result, using the operating quantity optimal setting compensation method based on data；S23 obtains Optimum Operation by such preceding floor data using the method for golden hydrometallurgy whole process re-optimization for the case where being can not find in historical data base with current working set of metadata of similar data；The present invention avoids the problem that production process can not establish mechanism model there are uncertain disturbances or uncertain variables and can not acquire Optimum Operation, is of great significance for improving production efficiency, improving Business Economic Benefit by establishing compensation model and solving.
Description
Technical field
The invention belongs to golden hydrometallurgy process real-time optimization field more particularly to a kind of golden hydrometallurgy whole process are real-time
Optimization Compensation method.
Background technique
Hydrometallurgical processes are gradually decreased with high grade ore, have begun the great attention by countries in the world.
Compared with traditional pyrometallurgy, hydrometallurgical technology has efficient, cleaning, is suitable for low-grade complex metallic mineral resources time
The advantages such as receipts.It is more especially for Mineral Resources in China lean ore, complicated symbiosis, the high feature of impurity content, hydrometallurgical processes
The comprehensive utilization ratio for improving mineral resources is industrialized, solid waste yield is reduced, reduces environmental pollution, have great
Meaning.Hydrometallurgical processes progress was rapid in recent years.But since hydrometallurgy reaction mechanism is complicated, process conditions are severe,
Such as high temperature, high pressure, deep-etching, process flow is long, device type multiplicity, therefore big rule are only continuously improved in hydrometallurgy enterprise
Mould industrializes intelligent control level, just can guarantee production safety, stabilization, continuous operation, and then guarantee the quality and production of product
Amount.The research of China's hydrometallurgical processes technology is in international most advanced level, even in leading position in terms of some of them, but
Due to hydrometallurgical processes type is more, process conditions difference is big, scale is relatively small etc., automatization level is relatively low.
And the single optimization control of each process has been far from satisfying its industrial needs.In order to improve production to the maximum extent
Amount, metal recovery rate, mineral resources the technical-economic indexes such as comprehensive utilization ratio, reduce operating cost and solid waste produce
Amount, reduces environmental pollution, reaches the purpose of high yield and high quality, energy-saving, urgently need excellent in real time to hydrometallurgy whole process
Change is studied, final to realize the target for improving the economic benefit of enterprise.
Golden hydrometallurgy whole process include ore grinding, flotation, dehydration size mixing, Cyanide Leaching, pressure filtration washing, the wet processes smelting such as displacement
The process flow of refining.Tcrude ore obtains certain ore pulp, Zhi Houtong after the preprocessing process such as ore grinding and separating flotation first
It crosses dense pressure-filtering process to be detached from the medicament carried in preprocessing process from ore, obtains the filter cake with few quantity of fluid.
Then filter cake and water of sizing mixing are stirred to get by certain density ore pulp by the process of sizing mixing, mixed ore pulp is squeezed by ore discharge pump
Into the subsequent leaching tanks for leaching process.Leach process be made of two-stage leaching process, in concentrate indissoluble gold by with leaching
Agent (NaCN), which reacts, out generates water-soluble ion, and the ore pulp after two-stage leaching generates after being passed through filter press pressure filtration washing
Your liquid obtains metallic gold eventually by displacement process.Gold hydrometallurgy master operation is as shown in Figure 1.
Reasonable hydrometallurgical processes process is to ensure that Gold in Ores effective recycling, enterprise obtain high yield return
Basic premise.Currently, both at home and abroad focusing mostly in each subprocess the modeling of hydrometallurgy whole process, optimization and base control (as soaked
Out, pressure filtration washing/dense washing, extraction/displacement) level on, there are no answering for related hydrometallurgy whole process real-time optimization
With and research, many researchs or precision it is not high or lack to subprocess each in whole process and mutual physics
The considerations of characteristic, therefore they cannot reflect hydrometallurgy whole process, limit the actual application ability of model.Hydrometallurgy is complete
Process generally has many characteristics, such as more process, close coupling, large time delay, non-linear.Therefore, the model established should be able to also embody
Above-mentioned complexity, to realize that hydrometallurgy whole process real-time optimization establishes solid foundation.In addition, in many actual industrial productions
In the process, there is error related with working condition, measured deviation, raw material characteristics etc. or uncertainties.And it is right both at home and abroad
It focuses mostly in the research of model foundation in simple mechanism model or simple data model, since uncertain factor exists,
Causing the model of the certain local links of production process can not obtain, at this time also just can not Kernel-based methods model progress overall process optimization
Control.Therefore the mixed model that reasonable establishment process qualitative model and quantitative model coexist for improve enterprises production efficiency and
Economic benefit, adjustment easy to produce have important practical significance.
Currently, studying seldom the full-range real-time optimization of hydrometallurgy both at home and abroad, automatization level is not also high, and theory is ground
Study carefully and also only rests in the optimization level to each process.Since hydrometallurgy whole process is answered by what a series of typical process formed
Miscellaneous process, with the continuous development of industry, the optimization of single process has been unable to meet the full-range production requirement of hydrometallurgy.Needle
To such challenge, extensive concern is had been obtained in hydrometallurgy overall process optimization, and becomes one of minerals processing industry
Important development target.However, hydrometallurgy whole process scale is excessive, process and variable are excessive, process model be may be subjected to not
Determining random perturbation.These features of hydrometallurgy process production, may make the process model established and actual production process
There are mismatch, the optimum results for causing the optimization method based on model to obtain are not practical mistake when being applied in real process
Journey optimal solution.The above feature makes hydrometallurgy whole process real-time optimization problem increasingly complex, and there is an urgent need to complete to hydrometallurgy
Process real-time optimization method is studied.Therefore, it is necessary to seek whole process modeling method appropriate and real-time optimization method, this hair
The bright method provided suitable for hydrometallurgy whole process real-time optimization.
Summary of the invention
(1) technical problems to be solved
In order to solve the model and actual production process mismatch that existing golden hydrometallurgical is established, and when the application model
The technical issues of optimal solution cannot be obtained, the present invention provide a kind of golden hydrometallurgy whole process real-time optimization compensation method.
(2) technical solution
In order to achieve the above object, the main technical schemes that the present invention uses include:
On-line evaluation and operation compensation part, on-line evaluation and operation compensation section point the following steps are included:
S1, application process evaluation of running status method divide golden hydrometallurgy whole process real-time optimization result online
Analysis obtains evaluation result；
S2, it is directed to evaluation result, selection is handled with the matched compensation method of the evaluation result, specifically included:
S21, it is directed to the case where evaluation result is suboptimum, using the compensation method of self-optimizing control, selects whole process output
The linear combination of variable realizes hydrometallurgy whole process mistake as controlled variable, by controlling the controlled variable tracking fixed valure
Self-optimizing control of the journey under uncertain disturbances；
S22, for evaluation result be non-optimum situation, using the operating quantity optimal setting compensation method based on data, base
Operation is compensated to current operation amount in timely study thoughts and deflected secondary air, and then realizes process operation from excellent
Change；
S23, for being can not find in historical data base with current working set of metadata of similar data the case where, by such preceding floor data
Using the method for golden hydrometallurgy whole process re-optimization, and then obtain Optimum Operation under current working.
Optionally, including off-line modeling part；
The off-line modeling part specifically includes: golden hydrometallurgy overall process optimization model foundation, process operation state are commented
Valence model foundation, Optimum Operation case library is established and self-optimizing control model foundation step.
Optionally, golden hydrometallurgy overall process optimization model foundation includes:
With the minimum optimization aim of economic cost of golden hydrometallurgy whole process production process, with cyanidation-leaching process twice
In Cymag additive amount Qji,cn, j=1,2 be two-stage leaching process, and i=1,2,3 be leaching tanks number in a leaching process
With zinc powder additive amount Q in replacement processznFor performance variable, constitutes golden hydrometallurgy whole process and operate vector u=[Q11,cn,...,
Q23,cn,Qzn]T；
It is final to establish golden wet process by meeting the quality index of each sub-process and the actual requirement range constraint of operating quantity
Metallurgical overall process optimization model is as follows:
Wherein, J is golden hydrometallurgy whole process economic cost function, Pcn,Pzn,Pcnd, PAuFor the unit valence of every kind of material
Lattice, EC are production total energy consumption；fpIndicate the mechanism model of p-th of sub-process；xtpIt (p=1,2,3) is the leaching of a leaching process
The replacement rate of extracting rate, the leaching rate of secondary leaching process and replacement process,It is leached for a leaching leaching rate, two leachings
Rate and the minimum of replacement rate require constraint；upFor the operation vector of p-th of process, z is working condition, including ore flow, Gu
Gold grade, ore partial size etc.；CwFor pulp density, random perturbation variable represent.
Optionally, process operation state evaluation model foundation includes:
According to the data characteristic of whole process industrial property and each sub-process, hydrometallurgy whole process evaluation model is established,
Specific implementation step is as follows:
A1, the quantitative data and qualitative data of online acquisition in metallurgy are pre-processed, is obtained to be analyzed
Quantitative data and qualitative data, steps are as follows for data prediction:
A11, select the economic cost in whole process production process for evaluation index, and choosing being capable of influence process operation shape
The process variable of state；
A12, for qualitative variable, sequentially indicate different conditions grade with a series of positive integers；For quantitative variable, carry out
Simple smoothing processing chooses the sliding window of certain length, and the information of entire window is characterized with window internal variable mean value；
A2, according to the quantitative data and qualitative data in the metallurgy in historical time section, establish for evaluating
The evaluation of running status model of full-range operating status grade, specific implementation step are as follows:
A21, whole process grade determine；The modeling data is enabled to beH is number of samples, and J is variable number, includes
Sizing variable and quantitative variable；According to whole process evaluation index economic cost, process operation state is divided into several grades, such as
Excellent, suboptimum and non-optimum is denoted as X1,X2,X3, wherein X includes 6 sub-block datas, process of respectively sizing mixing sub-block, cyaniding leaching
Process sub-block out, a pressure filtration washing process sub-block, secondary cyanidation-leaching process sub-block, secondary pressure filtration washing process sub-block and sets
Change process sub-block；It is excellent, suboptimum and non-optimum that footmark 1,2,3, which respectively indicates performance rate,；
A22, whole process evaluation model are established；Obtain each level data XlIt is coarse using fuzzy probability after (l=1,2,3)
Collection is to whole process model foundation decision table；
Wherein, conditional attribute XlEach variable in (l=1,2,3), decision attribute are whole process operating status grade
L, domain include Xl(l=1,2,3) all elements in.
Optionally, Optimum Operation case library, which is established, includes:
Based on the process operation data { x that history is excellentn}N=1, K, N, N is that the process operation state grade retained in history is excellent
When data amount check, establish Optimum Operation case libraryWherein znFor working condition, unTo be in working condition
znWhen Optimum Operation variable, NcFor data amount check in optimal case library.
Optionally, self-optimizing control model foundation includes:
Using the economic costs of golden hydrometallurgy whole process process operation, energy efficiency as cost function, select whole process defeated
The linear combination of variable establishes the full-range self-optimizing control model of golden hydrometallurgy, specific self-optimizing as controlled variable out
It is as follows to control off-line modeling process:
B1, to the distribution space C of uncertain disturbance variable pulp densityw∈ [32%, 40%], using Monte Carlo
Sampling generates NsGroup sequence
B2, to each group of pulp density situation d(n)Offline optimization solution is carried out using overall process optimization method, is calculated
Optimal input uopt=[Q11,cn,...,Q23,cn,Qzn]T；
B3, according to front solving result, record corresponding optimal output variable under each disturbance situationWherein export
VariableIncluding optimal input variable uopt, twice
Solid gold grade C in leaching processji,sWith cyanogen root particle concentration Cji,cn, cyanogen cinder grade C in replacement processcnd；Measurement noise takes
Gaussian noise, standard deviation are the 5% of nominal value；The optimal output variable acquired under all disturbance situations is constituted into matrixWhereinIt is the covariance matrix for measuring noise；
B4, nominal operation point C is takenw=36%, the gain matrix G of input variable value is calculated with finite difference methody,refWith
Hessian matrix J of the loss function J relative to performance variable uuu,ref；
B5, optimum combination matrix H in combination C=Hy is calculated, so that average loss is minimum, i.e.,
Combinatorial matrix H can be obtained by solving following formula:Base
Optimal output variable under nominal operation pointControlled variable setting value C is acquired using formula C=Hys。
Optionally, on-line evaluation is specifically included with operation compensation part: process operation state on-line evaluation, self-optimizing control
Online compensation and operating quantity optimal setting compensation model establishment step based on data, it is specific as described below:
The Optimum Operation variable under nominal operation point is acquired by overall process optimization, performance variable is brought into reality
In golden hydrometallurgy production process, wherein there are uncertain disturbances for real process；
The first step, Kernel-based methods evaluation of running status method judge active procedure runnability online；
Second step, for different performance rates, self-optimizing control method is mended with the operating quantity optimal setting based on data
The corresponding compensation policy of compensation method is implemented respectively；
Compensated performance variable is brought into actual production process, and then obtains Optimum Operation by third step；It comments online
Specific step is as follows for valence and operation compensation:
101, the performance variable for acquiring nominal operation pointIt is brought into practical gold hydrometallurgy production process, obtained
Cheng Bianliang xtWith output variable yt；
102, process operation state on-line evaluation；It is complete using what is established offline for active procedure variable and output variable
Process flow operation state evaluation model, handles active procedure variable, determines whole process operating status in metallurgy
Grade
Specific step is as follows for on-line evaluation:
C1, online data x is obtainedt, and data in the decision table established offline carry out according to fuzzy probability rough set
Match, obtains xtFuzzy equivalence relation class λ cut setWherein Table
Registration is according to xtConditional attribute set, λ is given threshold value；
C2, calculating belong to the probability of first of grade are as follows:
Wherein dlIndicate first of grade, XlIt (l=1,2,3) is that operating status grade is determined according to historical data；Judgement is complete
The operating status grade of process are as follows:
103, according to different evaluation results, corresponding compensation policy is taken, specific compensating form is divided into as follows:
Situation 1: when evaluation result is excellent, showing that current operation variable is Optimum Operation, without compensating operation, until
New process disturbance occurs；
Situation 2: when evaluation result is suboptimum, i.e., process operation state deviates Optimum Operation, but deviation is little；For the feelings
Condition is operated using self-optimizing control online compensation, realizes that uncertainty is disturbed with variation by control controlled variable tracking fixed valure
Dynamic optimize on economic goal influences minimum, and then compensates to current operation, realizes hydrometallurgy process in uncertainty
Self-optimizing control under disturbance；
Specific compensation process is as follows:
Output variable y under D1, the connection matrix H and current working that are obtained based on off-line calculationt, utilize formula C=Hyt
Acquire controlled variable Ct；
D2, it is based on following formula, calculates current operation variableOffset Δ u1.k；
Δ y=Gy·Δu
H'=BH, B=(HGy)-1
C'=BC=H'y
B Δ C=H' Δ y=(HGy)-1·H·GyΔ u=I Δ u
Wherein, offset Δ u is Δ u1.k, Δ C is controlled variable setting value CsWith current required controlled variable CtIt is inclined
Difference；The general solution of H' expression connection matrix H；
D3, by new performance variableIt is brought into actual production process, executes next stage；
Situation 3: when evaluation result is non-optimum, i.e., process operation state deviates Optimum Operation, and deviation is very big；For the feelings
Condition makes production process be returned to optimal or suboptimum operating status using the operating quantity optimal setting compensation method based on data；I.e.
Using based on timely study thoughts, lookup and the most similar several groups of data of current working in history optimal data library, using inclined
Least square method establishes the data model between working condition and operating quantity offset, compensates to current operation, in turn
Realize process operation self-optimizing；
Specific compensation process is as follows:
E1, it is directed to active procedure variable xt, found and current working condition z in the excellent case library of the history established offlinet
Similar case data zi, it is based on following formula, calculates Sample Similarity Sz；
Sz=ρ exp (- Di)+(1-ρ)cos(ωi)
Di=| | zt-zi||2,
Wherein Sz∈ [0,1] is bigger, and Sample Similarity is stronger；DiWith cos (ωi) respectively indicate ziWith ztBetween Euclidean away from
From the cosine value with angle；ρ (0≤ρ≤1) is the index of equilibrium distance and angle information weight；
E2, Sample Similarity threshold epsilon (0 < ε < 1) is defined, judgement sample similarity SzWhether threshold epsilon is greater than；
First situation: if Sz>=ε shows there are data similar with current working in the excellent case library of history；If phase
Like sample number >=L, operation compensation is carried out using based on timely study thoughts, otherwise, carries out the second situation operation；
For the first situation specific steps are as follows:
F1, it is based on Sample Similarity calculation formula, L modeling numbers of the similar sample as local regression model is calculated
According to；
F2, L similar sample { z are based oni,ui}I=1,2, L, calculating operation variable ui, (i=1,2, K, L) and current operation become
AmountBetween difference, be denoted as Δ ui, (i=1,2, K, L)；Construct modeling data collection { zi,Δui}I=1,2, L；
F3, local regression model Δ u is established based on deflected secondary airi=fr(zi, θ), wherein θ is undated parameter, with
The difference of modeling data and change；For current operation working condition zt, it is based on local regression model fr, calculate current operation
The offset Δ u of variable2,k；
F4, by new performance variableIt is brought into actual production process, executes next stage；
Second situation: if Sz< ε or Sz>=ε but similar sample number<L, show to can not find in the excellent case library of history largely with
The similar data of current working, in this way, using traditional real-time optimization method, that is, identification disturbance variable, re-optimization whole process
Optimized model, and then obtain Optimum Operation.
Optionally, evaluation result be it is non-optimum when, using be based on timely study thoughts, in history optimal data library search with
The most similar several groups of data of current working are divided into two kinds of situations according to the number of set of metadata of similar data:
Situation one: if set of metadata of similar data number be greater than a certain threshold value, show in the excellent case library of history there are it is enough with work as
The similar data of preceding operating condition；Data model between working condition and operating quantity offset is established using deflected secondary air,
Current operation is compensated, and then realizes process operation self-optimizing；
Situation two: if set of metadata of similar data number be less than a certain threshold value, show to can not find in the excellent case library of history largely with work as
The similar data of preceding operating condition；Using traditional real-time optimization method, i.e. identification disturbance variable, re-optimization overall process optimization mould
Type, and then obtain Optimum Operation.
It optionally, further include process detection system, process detection system includes PLC controller, Concentration Testing, pressure detecting
And flow detection；
PLC controller has Profibus DP mouthfuls of connection distributed I/O using the CPU 414-2 of 400 series of Simens；
It is equipped with ethernet communication module for PLC, accesses plc data for host computer；PLC controller and ethernet communication module are placed on
In PLC rack in central control room；
Host computer selects i7 thinking computer, using WINDOWS XP operating system；
Golden hydrometallurgy whole process real-time optimization system is on i7 thinking computer, and using C#2008 programming software, gold is wet
Method metallurgy whole process real-time optimization algorithm uses Matlab 2016a programming software；
The signal of PLC and whole process real-time optimization transmission software is using C#2008 programming software；
Instrument is detected in hydrometallurgy process in-site installation, detection instrument passes the signal of acquisition by Profibus-DP
It is sent in PLC, PLC timing sends acquisition signal to host computer by Ethernet, and host computer is transmitted to the data of receiving golden wet
Method metallurgy whole process real-time optimization system carries out full-range real-time optimization compensating operation, and provides Operating Guideline suggestion.
(3) beneficial effect
The beneficial effects of the present invention are: firstly, this forwarding method is built using the mechanism of reliable expertise and flow process
Vertical whole process process model provides safeguard to obtain accurately and reliably Optimized model；Secondly, using can online quantitative measurment or fixed
Property estimation variable information, Real-Time Evaluation is made to process operation state, keeps production process more efficient, it is ensured that business economic
Benefit；Finally, influencing for different process disturbances, self-optimizing control compensation is compensated with the operating quantity optimal setting based on data
Rationally reliable Operating Guideline opinion is provided for operative employee, improves the optimization performance of control system, and then improve whole process mistake
Total productivity effect of journey.
Detailed description of the invention
Fig. 1 is the golden hydrometallurgy that provides of one embodiment of the invention mainly for answering flow chart；
Fig. 2 is the golden hydrometallurgy whole process real-time optimal control schematic diagram that one embodiment of the invention provides；
Fig. 3 is that the golden hydrometallurgy that provides of one embodiment of the invention is sized mixing process pulp density variable change schematic diagram；
Fig. 4 is the golden hydrometallurgy whole process on-line evaluation that one embodiment of the invention provides and operation compensation process signal
Figure；
Fig. 5 a is the Optimum Operation variable Q under the different pulp densities disturbance that one embodiment of the invention provides11,cnSchematic diagram；
Fig. 5 b is the Optimum Operation variable Q under the different pulp densities disturbance that one embodiment of the invention provides12,cnSchematic diagram；
Fig. 5 c is the Optimum Operation variable Q under the different pulp densities disturbance that one embodiment of the invention provides13,cnSchematic diagram；
Fig. 5 d is the Optimum Operation variable Q under the different pulp densities disturbance that one embodiment of the invention provides21,cnSchematic diagram；
Fig. 5 e is the Optimum Operation variable Q under the different pulp densities disturbance that one embodiment of the invention provides22,cnSchematic diagram；
Fig. 5 f is the Optimum Operation variable Q under the different pulp densities disturbance that one embodiment of the invention provides23,cnSchematic diagram；
Fig. 5 g is the Optimum Operation variable Q under the different pulp densities disturbance that one embodiment of the invention providesznSchematic diagram；
Fig. 6 is overall process optimization analog result schematic diagram under the different pulp densities that one embodiment of the invention provides；
Fig. 7 is the process operation state evaluation result schematic diagram that one embodiment of the invention provides；
Fig. 8 is evaluation result and economic loss schematic diagram under 65 kinds of random perturbations that one implementation of the present invention provides；
Fig. 9 is the controlled volume deviation schematic diagram under 65 kinds of random perturbations that one embodiment of the invention provides；
Figure 10 is the simulation result schematic diagram being directed under three kinds of different operations of economic loss that one embodiment of the invention provides.
Specific embodiment
In order to preferably explain the present invention, in order to understand, with reference to the accompanying drawing, by specific embodiment, to this hair
It is bright to be described in detail.
Hydrometallurgy full-flow process process is as shown in Figure 1, tcrude ore passes through the preprocessing process such as ore grinding and separating flotation
Certain ore pulp is obtained afterwards, and the medicament carried in preprocessing process is detached from by dense pressure-filtering process from ore later, is obtained
To the filter cake for having few quantity of fluid；And then filter cake and water of sizing mixing are stirred to get by certain density ore pulp by the process of sizing mixing,
Ore pulp after sizing mixing enters Cyanide Leaching process, and the ore pulp after leaching is by pressure filtration washing, and the filter cake after washing is by sizing mixing
Enter replacement process afterwards；Your liquid carries out zinc dust precipitation after purifying deoxidation, generates gold mud.Hydrometallurgy process detection system master
It to be made of Concentration Testing, pressure detecting, flow detection.
A kind of golden hydrometallurgy whole process real-time optimization compensation method is present embodiments provided, wherein PLC controller uses
The CPU 414-2 of 400 series of Simens, has Profibus DP mouthfuls of connection distributed I/O；Ethernet communication mould is equipped with for PLC
Block accesses plc data for host computer, and PLC controller and ethernet communication module are placed on the PLC rack in central control room
In.
PH value is to carry out pH value on-line checking by the BPHM-II type acidometer that Beijing Mine and Metallurgy General Inst develops, will be molten
The variation of liquid pH value is converted into the variation of mV signal.Glass electrode PH measuring system is by the glass of a glass-film for pH sensitivity
Glass tube end blows out blister, and casing pack has the 3mol/l KCL buffer solution of the AgCl containing saturation, pH value 7.It is present in glass
The potential difference Ag/AgCl conducting system of the reflection pH value in two face of film, exports potential difference, is then changed mA number with mA acquisition instrument
PH value is counted as to show.
Pressure is that the DSIII type pressure detecting instrument produced by SIEMENS company carries out pressure on-line checking, medium pressure
Power directly acts on sensitive diaphragm, is distributed in the Wheatstone bridge of the resistance composition on sensitive diaphragm, utilizes piezoresistive effect reality
The millivolt signal that sensing element generates is enlarged into industrial standard electric current letter by electronic circuit by existing conversion of the pressure to electric signal
Number.
Dissolved oxygen concentration is the inpro6870+M400 type oxygen content measurement sensor produced by Mei Teletuo benefit company
Carry out on-line checking.Oxygen content measurement sensor is made of the counterelectrode of cathode and belt current, currentless reference electrode, electrode
In the electrolyte, sensor has diaphragm covering for submergence, and electrode and electrolyte and measured liquid are separated, only dissolved by overlay film
Gas-permeable overlay film, therefore sensor is protected, it can prevent electrolyte from escaping and preventing invading people and leading for foreign substance
It causes pollution and poisons.Current signal is admitted to transmitter, using between the oxygen content and partial pressure of oxygen, temperature stored in sensor
Relation curve calculates oxygen content, is then converted into standard signal output.
Host computer selects i7 thinking computer, using WINDOW XP operating system.
Whole process process real-time optimization system is on i7 thinking computer, using C#2008 programming software, whole process process
Real-time optimization algorithm uses Matlab 2016a programming software.
PLC and the signal of process real-time optimization system transmission software are using C#2008 programming software.
Instrument is detected in hydrometallurgy process in-site installation, detection instrument passes the signal of acquisition by Profibus-DP
It is sent in PLC, PLC timing sends acquisition signal to host computer by Ethernet, and the data of receiving are transmitted to process by host computer
Real-time optimization system carries out real-time optimization, and provides Operating Guideline suggestion.
Whole process process real-time optimization block diagram of the present invention by taking the high copper mine of golden hydrometallurgy as an example is as shown in Figure 2.Golden wet process
The specific implementation steps are as follows for metallurgical (high copper mine) whole process process real-time optimization method:
Step 1: in conjunction with expertise and site operation personnel's experience, taking sizing mixing in a period of time to obtain in the process practical
Pulp concentration value.
Step 2: being passed through as shown in table 1 based on the nominal data in practical golden hydrometallurgy whole process production process
The Monte Carlo method of sampling generates 500 groups of samples to pulp density variable at random, is optimized using overall process optimization model
It solves, obtains the Optimum Operation variable under different pulp densities.
Specifically for example optimum results are as shown in Figure 5；As can be seen that different pulp densities from Fig. 5 a to Fig. 5 g
The lower Optimum Operation variable Q acquired using overall process optimization method11,cn、Q12,cn、Q13,cn、Q21,cn、Q22,cn、Q23,cnAnd QznIt is
Variation, therefore, pulp density is studied golden hydrometallurgy whole process real-time optimization side as a kind of random perturbation by the present invention
The validity of method.
1 gold medal hydrometallurgy whole process process variable nominal value of table
Step 3: in order to further explain the variation meeting of pulp density so that overall process optimization result is changed asks
Topic, literary invention study optimum results by the pulp density variation specifically enumerated under three kinds of different operating conditions.Emulation experiment step
As follows.Choosing pulp density respectively is Cw=0.32, nominal value Cw=0.36 and Cw=0.40, utilize overall process optimization side
Method solves respectively obtains corresponding Optimum Operation variable u=[Q11,cn,...,Q23,cn,Qzn]T.Later, respectively by pulp density Cw
=0.36 and Cw=0.40 optimum results are brought into simulation model, and wherein the pulp density in simulation model is Cw=0.40,
Optimum results are as shown in Fig. 6 and table 2.
Overall process optimization result under the different pulp densities of table 2
| Variable | Optimum results Cw=0.32 (g/g) | Optimum results Cw=0.36 (g/g) | Optimum results Cw=0.40 (g/g) |
| xt1(%) | 0.9529 | 0.9549 | 0.9573 |
| xt2(%) | 0.8732 | 0.8734 | 0.8737 |
| xt3(%) | 0.9997 | 0.9997 | 0.9997 |
| Q11,cn(kg/h) | 11.9835 | 12.0058 | 12.0495 |
| Q12,cn(kg/h) | 0.1001 | 0.1001 | 0.3980 |
| Q13,cn(kg/h) | 0.0553 | 0.0059 | 0.0587 |
| Q21,cn(kg/h) | 24.6387 | 22.5764 | 22.8302 |
| Q22,cn(kg/h) | 10.4865 | 11.4495 | 11.9523 |
| Q23,cn(kg/h) | 0.1001 | 1.9160 | 3.3507 |
| Qzn(kg/h) | 0.2779 | 0.2781 | 0.2784 |
| xc1(member/when) | 155.3776 | 155.0312 | 160.0790 |
| xc2(member/when) | 450.8838 | 460.0551 | 488.1045 |
| J (member/when) | 2977.2293 | 3047.0352 | 3119.2983 |
Whole process economic cost J that as can be seen from Figure 5 different pulp densities acquires (member/when) be different.Mine
Slurry concentration is higher, and leaching rate is also bigger, however in order to meet the requirement of leaching rate, material consumption is also bigger, therefore economic cost
Also bigger.By pulp density Cw=0.36 and Cw=0.40 optimum results are brought into simulation model respectively as can be seen that practical
Economic cost be 3119.2983 (members/when).However, nominal operation point Cw=0.36 obtained Optimum Operation variable is not
Actual optimum operation, being brought into economic cost obtained in simulation model is 3175.7859 (members/when), hence it is evident that is greater than practical warp
Ji cost.
From table 3 more as can be seen that when the performance variable that nominal operation point acquires is brought into real process, leaching rate is small
Leaching rate required by real process is not able to satisfy actual production requirement.Therefore, the variation of pulp density should be by as one kind
Random perturbation considers.
Leaching rate under 3 two kinds of different operations of table compares
| Variable | Total leaching rate | Leaching rate | Secondary leaching rate |
| Cw=0.36-0.40 | 0.9945 | 0.9569 | 0.8730 |
| Cw=0.40-0.40 | 0.9946 | 0.9573 | 0.8737 |
Step 4: implementation being compensated to the performance variable under suboptimum operating status using self-optimizing control method.It is sharp first
450 samples of stochastic production in space are disturbed in pulp density with Monte Carlo method.It is carried out with overall process optimization method excellent
Change and solve, obtains Optimum Operation variable u=[Q11,cn,...,Q23,cn,Qzn]TAnd process variable.The present invention chooses whole process production
Output variable is y=[Q in the process11,cn,...,Q23,cn,Qzn,C11,s,...,C23,s,C11,cn,...,C23,cn,Ccnd]T, measurement
Noise is that mean value is 0, and standard deviation is the 2% of nominal value.Controlled variable is acquired using self-optimizing control off-line modeling method to set
Value is Cs=Hy=[0.0083-0.00390.03100.012920.0064-0.00910.0005]T。
Emulation experiment carries out verification experimental verification: method one using two methods, is implemented using nominal operation point.Optimum Operation becomes
Amount chooses the Optimum Operation that nominal operation point acquires, and actual production process is changed by pulp density random perturbation, i.e. d=[Cw]=
[0.320.40]；Method two is implemented using self-optimizing control method.Performance variable changes in real time according to self-optimizing control method,
Actual production process is changed d=[C by pulp density random perturbationw]=[0.320.40].
Random simulation of the present invention is tested 65 times, firstly, before the economic loss result of relatively two methods, by running shape
State evaluation method assesses this 65 kinds of process performance results, and performance rate is divided into 1,2 and 3, respectively indicates excellent grade, suboptimum
Grade and non-optimum grade.Evaluation result is as shown in Figure 7.Since the present invention is that 65 disturbances are randomly generated to emulate, three
Kind performance rate is likely to exist.Operating status grade illustrates pulp density disturbance variable close to nominal value C for excellentw=
0.36, due to being worked using nominal operation point, so obtained process performance is excellent.
And for suboptimum and non-optimum performance, illustrate that pulp density is deviated considerably from nominal value, non-optimum degree is bigger, partially
It is bigger from also.It is very big for can be seen that from Fig. 8 a economic loss using the economic loss that nominal operation point obtains, maximum damage
Lose for 34.81 (member/when), average loss be 9.98 (member/when).And operation is compensated by self-optimizing control compensation method
Afterwards, economic loss is as shown in Figure 8 b, and maximum loss is 0.699 (member/when), average loss be 0.0130 (member/when).It will be apparent that
Economic loss has obtained huge improvement.
As shown in figure 9, controlled variable deviation schematic diagram can be seen that and obtain when being operated using nominal operation point
Controlled variable deviation is very big, maximum deviation 2.0601, average deviation 0.9881.And it is obtained by self-optimizing control compensation method
The controlled variable deviation maximum value arrived is 0.3849, average deviation 0.1397.This is also further illustrated when nominal operation point
When being applied in actual production process, performance variable and process variable are not actual optimums, therefore reduce productivity effect, are brought
Huge economic losses.
Step 5: utilizing the operating quantity optimal setting compensation method based on data, be returned to non-optimum production process optimal
Or suboptimum operating status；The present invention carries out 5 groups of emulation experiments at random, for every group of experiment, is obtained under this group of working condition in advance
Optimum Operation variable and corresponding economic loss, and every group of experiment is all made of nominal operation point and carries out experiment comparison.
Firstly, being directed to every group of operating condition, using based on study just-in-time thought (hereinafter referred to as JIT) in time, going through
It is searched and the most similar several groups of data of current working in history optimal data library；
Secondly, seeking the difference between the performance variable under similar sample data and current working, modeling data collection is obtained
{zi,Δui}I=1,2, K, 7, wherein ziFor the working condition of similar sample, Δ uiFor the difference of performance variable.Using minimum two partially
Multiply the data model Δ u that PLS method is established between working condition and operating quantity offseti=fr(zi,θ)；
Finally, being directed to each group of experiment, working condition is brought into established data model, is based on regression model Δ
u2,k=fr(zt, θ) current operation is compensated, obtain offset Δ u2,k, and then obtain new performance variableNew performance variable is brought into actual production process, the economic cost J of every group of experiment is obtained.In order to
Experiment comparison is preferably carried out, the present invention takes three kinds of methods to carry out simulating, verifying, is respectively as follows: method one, nominal operation point behaviour
Make；Method two, the operating quantity optimal setting compensation method operation based on data；Method three, the Optimum Operation of real process；Experiment
The results are shown in Figure 10.
Since nominal operation point is not actual optimum operation, the economic cost obtained when being brought into actual production process
Be not it is optimal, there are biggish deviations.And after using the operating quantity optimal setting compensation method operation based on data, it is compensated
Performance variable is operated close to actual optimum, therefore while being brought into actual production process obtained economic cost and actual optimum pass through
Ji costvariation very little.Table 4 lists the root-mean-square error (hereinafter referred to as RMSE) of economic loss after distinct methods operate and puts down
Equal absolute error (hereinafter referred to as MAE).As can be seen from the table, compare for the experiment of five groups of random simulations, by using being based on
Economic loss after the operating quantity optimal setting compensation method operation of data substantially reduces, and RMSE and MAE are respectively less than and are based on nominally
The error that operating point operates.
Application condition under 4 two kinds of different operations of table
By above example, the validity of the golden hydrometallurgy whole process real-time optimization method of the present invention-- is shown,
It realizes from different evaluation result angles, real-time optimization compensating operation, solution is carried out to hydrometallurgy whole process from suboptimum to non-optimum
It has determined because there are uncertain factors to lead to not to establish quantitative model and optimization knot based on mechanism model for actual production process
The problem of fruit is not actual optimum provides effective ways to solve complex industrial process real-time optimization, has wide application
Prospect.
Finally, it should be noted that above-described embodiments are merely to illustrate the technical scheme, rather than to it
Limitation；Although the present invention is described in detail referring to the foregoing embodiments, those skilled in the art should understand that:
It can still modify to technical solution documented by previous embodiment, or to part of or all technical features into
Row equivalent replacement；And these modifications or substitutions, it does not separate the essence of the corresponding technical solution various embodiments of the present invention technical side
The range of case.
Claims (9)
1. a kind of gold hydrometallurgy whole process real-time optimization compensation method, golden hydrometallurgy whole process include filter-press dehydration size mixing,
Leaching, washing and replacing process, which is characterized in that including on-line evaluation and operation compensation part, on-line evaluation is compensated with operation
Part the following steps are included:
S1, application process evaluation of running status method carry out on-line analysis to golden hydrometallurgy whole process real-time optimization result, obtain
Obtain evaluation result；
S2, it is directed to evaluation result, selection is handled with the matched compensation method of the evaluation result, specifically included:
S21, it is directed to the case where evaluation result is suboptimum, using the compensation method of self-optimizing control, selects whole process output variable
Linear combination as controlled variable, realize that hydrometallurgy whole process process exists by controlling the controlled variable tracking fixed valure
Self-optimizing control under uncertain disturbances；
S22, for evaluation result be non-optimum situation, using the operating quantity optimal setting compensation method based on data, be based on and
When study thoughts and deflected secondary air operation compensated to current operation amount, and then realize process operation self-optimizing；
S23, for being can not find in historical data base with current working set of metadata of similar data the case where, such preceding floor data is used
The method of golden hydrometallurgy whole process re-optimization, and then obtain Optimum Operation under current working.
2. the method according to claim 1, wherein including off-line modeling part；
The off-line modeling part specifically includes: golden hydrometallurgy overall process optimization model foundation, process operation state evaluation mould
Type is established, Optimum Operation case library is established and self-optimizing control model foundation step.
3. according to the method described in claim 2, it is characterized in that, golden hydrometallurgy overall process optimization model foundation includes:
With the minimum optimization aim of economic cost of golden hydrometallurgy whole process production process, in cyanidation-leaching process twice
Cymag additive amount Qji,cn, j=1,2 be two-stage leaching process, i=1, and 2,3 for leaching tanks number in leaching process and set
Zinc powder additive amount Q during changingznFor performance variable, constitutes golden hydrometallurgy whole process and operate vector u=[Q11,cn,...,
Q23,cn,Qzn]T；
It is final to establish golden hydrometallurgy by meeting the quality index of each sub-process and the actual requirement range constraint of operating quantity
Overall process optimization model is as follows:
Wherein, J is golden hydrometallurgy whole process economic cost function, Pcn,Pzn,Pcnd, PAuFor the unit price of every kind of material, EC
To produce total energy consumption；fpIndicate the mechanism model of p-th of sub-process；xtpBe (p=1,2,3) leaching rate of a leaching process,
The leaching rate of secondary leaching process and the replacement rate of replacement process,For a leaching leaching rate, two leaching leaching rates and set
It changes the minimum of rate and requires constraint；upFor the operation vector of p-th of process, z is working condition, including ore flow, Gu gold grade,
Ore partial size etc.；CwFor pulp density, random perturbation variable represent.
4. according to the method described in claim 3, it is characterized in that, process operation state evaluation model foundation includes:
According to the data characteristic of whole process industrial property and each sub-process, hydrometallurgy whole process evaluation model is established, specifically
Implementation steps are as follows:
A1, the quantitative data and qualitative data of online acquisition in metallurgy are pre-processed, obtains and to be analyzed determines
Data and qualitative data are measured, steps are as follows for data prediction:
A11, it selects the economic cost in whole process production process for evaluation index, and chooses and be capable of influence process operating status
Process variable；
A12, for qualitative variable, sequentially indicate different conditions grade with a series of positive integers；For quantitative variable, carry out simple
Smoothing processing, choose the sliding window of certain length, the information of entire window characterized with window internal variable mean value；
A2, according to the quantitative data and qualitative data in the metallurgy in historical time section, establish for evaluating full stream
The evaluation of running status model of the operating status grade of journey, specific implementation step are as follows:
A21, whole process grade determine；The modeling data is enabled to beH is number of samples, and J is variable number, includes sizing
Variable and quantitative variable；According to whole process evaluation index economic cost, process operation state is divided into several grades, such as excellent,
Suboptimum and non-optimum, is denoted as X1,X2,X3, wherein X includes 6 sub-block datas, process of respectively sizing mixing sub-block, a Cyanide Leaching
Process sub-block, a pressure filtration washing process sub-block, secondary cyanidation-leaching process sub-block, secondary pressure filtration washing process sub-block and displacement
Process sub-block；It is excellent, suboptimum and non-optimum that footmark 1,2,3, which respectively indicates performance rate,；
A22, whole process evaluation model are established；Obtain each level data XlAfter (l=1,2,3), using fuzzy probability rough set to complete
Procedural model establishes decision table；
Wherein, conditional attribute XlEach variable in (l=1,2,3), decision attribute are whole process operating status grade l, domain
Include Xl(l=1,2,3) all elements in.
5. according to the method described in claim 4, it is characterized in that, the foundation of Optimum Operation case library includes:
Based on the process operation data { x that history is excellentn}N=1, K, N, N is the process operation state grade retained in history when being excellent
Data amount check establishes Optimum Operation case libraryWherein znFor working condition, unTo be z in working conditionnWhen
Optimum Operation variable, NcFor data amount check in optimal case library.
6. according to the method described in claim 5, it is characterized in that, self-optimizing control model foundation includes:
Using the economic costs of golden hydrometallurgy whole process process operation, energy efficiency as cost function, whole process output is selected to become
The linear combination of amount establishes the full-range self-optimizing control model of golden hydrometallurgy, specific self-optimizing control as controlled variable
Off-line modeling process is as follows:
B1, to the distribution space C of uncertain disturbance variable pulp densityw∈ [32%, 40%] is sampled using Monte Carlo and is produced
Raw NsGroup sequence
B2, to each group of pulp density situation d(n)Offline optimization solution is carried out using overall process optimization method, is calculated optimal
Input uopt=[Q11,cn,...,Q23,cn,Qzn]T；
B3, according to front solving result, record corresponding optimal output variable under each disturbance situationWherein output variableIncluding optimal input variable uopt, two-stage leaching
Solid gold grade C in the processji,sWith cyanogen root particle concentration Cji,cn, cyanogen cinder grade C in replacement processcnd；Measurement noise takes Gauss
Noise, standard deviation are the 5% of nominal value；The optimal output variable acquired under all disturbance situations is constituted into matrixWhereinIt is the covariance matrix for measuring noise；
B4, nominal operation point C is takenw=36%, the gain matrix G of input variable value is calculated with finite difference methody,refAnd loss
Hessian matrix J of the function J relative to performance variable uuu,ref；
B5, optimum combination matrix H in combination C=Hy is calculated, so that average loss is minimum, i.e.,Combination
Matrix H can be obtained by solving following formula:Based on mark
Claim the optimal output variable under operating pointControlled variable setting value C is acquired using formula C=Hys。
7. according to the method described in claim 6, it is characterized in that, on-line evaluation is specifically included with operation compensation part: process
Operating status on-line evaluation, self-optimizing control online compensation and the operating quantity optimal setting compensation model based on data establish step
Suddenly, specific as described below:
The Optimum Operation variable under nominal operation point is acquired by overall process optimization, it is wet that performance variable is brought into practical gold
In method metallurgical production process, wherein there are uncertain disturbances for real process；
The first step, Kernel-based methods evaluation of running status method judge active procedure runnability online；
Second step, for different performance rates, self-optimizing control method and the operating quantity optimal setting compensation side based on data
The corresponding compensation policy of method is implemented respectively；
Compensated performance variable is brought into actual production process, and then obtains Optimum Operation by third step；On-line evaluation with
Specific step is as follows for operation compensation:
101, the performance variable for acquiring nominal operation pointIt is brought into practical gold hydrometallurgy production process, obtains process and become
Measure xtWith output variable yt；
102, process operation state on-line evaluation；For active procedure variable and output variable, using the whole process established offline
Evaluation of running status model handles active procedure variable, determines whole process operating status grade in metallurgy
Specific step is as follows for on-line evaluation:
C1, online data x is obtainedt, with data in the decision table established offline, matched, obtained according to fuzzy probability rough set
xtFuzzy equivalence relation class λ cut setWherein Indicate data xt
Conditional attribute set, λ is given threshold value；
C2, calculating belong to the probability of first of grade are as follows:
Wherein dlIndicate first of grade, XlIt (l=1,2,3) is that operating status grade is determined according to historical data；Judge whole process
Operating status grade are as follows:
103, according to different evaluation results, corresponding compensation policy is taken, specific compensating form is divided into as follows:
Situation 1: when evaluation result is excellent, show that current operation variable is Optimum Operation, without compensating operation, until new
Process disturbance occurs；
Situation 2: when evaluation result is suboptimum, i.e., process operation state deviates Optimum Operation, but deviation is little；For the situation,
It is operated using self-optimizing control online compensation, uncertain and variation disturbance pair is realized by control controlled variable tracking fixed valure
Economic goal, which optimizes, influences minimum, and then compensates to current operation, realizes hydrometallurgy process in uncertain disturbances
Under self-optimizing control；
Specific compensation process is as follows:
Output variable y under D1, the connection matrix H and current working that are obtained based on off-line calculationt, utilize formula C=HytIt acquires
Controlled variable Ct；
D2, it is based on following formula, calculates current operation variableOffset Δ u1.k；
Δ y=Gy·Δu
H'=BH, B=(HGy)-1
C'=BC=H'y
B Δ C=H' Δ y=(HGy)-1·H·GyΔ u=I Δ u
Wherein, offset Δ u is Δ u1.k, Δ C is controlled variable setting value CsWith current required controlled variable CtDeviation；H'
Indicate the general solution of connection matrix H；
D3, by new performance variableIt is brought into actual production process, executes next stage；
Situation 3: when evaluation result is non-optimum, i.e., process operation state deviates Optimum Operation, and deviation is very big；For the situation,
Using the operating quantity optimal setting compensation method based on data, production process is made to be returned to optimal or suboptimum operating status；Adopt
With based on timely study thoughts, lookup and the most similar several groups of data of current working in history optimal data library, using partially most
Small least square method establishes the data model between working condition and operating quantity offset, compensates to current operation, Jin Ershi
Existing process operation self-optimizing；
Specific compensation process is as follows:
E1, it is directed to active procedure variable xt, found and current working condition z in the excellent case library of the history established offlinetIt is close
Case data zi, it is based on following formula, calculates Sample Similarity Sz；
Sz=ρ exp (- Di)+(1-ρ)cos(ωi)
Wherein Sz∈ [0,1] is bigger, and Sample Similarity is stronger；DiWith cos (ωi) respectively indicate ziWith ztBetween Euclidean distance and
The cosine value of angle；ρ (0≤ρ≤1) is the index of equilibrium distance and angle information weight；
E2, Sample Similarity threshold epsilon (0 < ε < 1) is defined, judgement sample similarity SzWhether threshold epsilon is greater than；
First situation: if Sz>=ε shows there are data similar with current working in the excellent case library of history；If similar sample
Number >=L carries out operation compensation using based on timely study thoughts, otherwise, carries out the second situation operation；
For the first situation specific steps are as follows:
F1, it is based on Sample Similarity calculation formula, L modeling datas of the similar sample as local regression model is calculated；
F2, L similar sample { z are based oni,ui}I=1,2, L, calculating operation variable ui, (i=1,2, K, L) and current operation variable
Between difference, be denoted as Δ ui, (i=1,2, K, L)；Construct modeling data collection { zi,Δui}I=1,2, L；
F3, local regression model Δ u is established based on deflected secondary airi=fr(zi, θ), wherein θ is undated parameter, with building
The difference of modulus evidence and change；For current operation working condition zt, it is based on local regression model fr, calculate current operation variable
Offset Δ u2,k；
F4, by new performance variableIt is brought into actual production process, executes next stage；
Second situation: if Sz< ε or Sz>=ε but similar sample number<L, show to can not find in the excellent case library of history it is a large amount of with it is current
The similar data of operating condition, in this way, using traditional real-time optimization method, that is, identification disturbance variable, re-optimization overall process optimization
Model, and then obtain Optimum Operation.
8. the method according to claim 1, wherein thinking when evaluation result is non-optimum using based on study in time
Think, is searched in history optimal data library and be divided into two according to the number of set of metadata of similar data with the most similar several groups of data of current working
Kind situation:
Situation one: if set of metadata of similar data number is greater than a certain threshold value, show that there are enough and current works in the excellent case library of history
The similar data of condition；Data model between working condition and operating quantity offset is established using deflected secondary air, to working as
Preceding operation compensates, and then realizes process operation self-optimizing；
Situation two: if set of metadata of similar data number is less than a certain threshold value, show to can not find a large amount of and current work in the excellent case library of history
The similar data of condition；Using traditional real-time optimization method, i.e. identification disturbance variable, re-optimization overall process optimization model, into
And obtain Optimum Operation.
9. method according to any one of claims 1 to 8, which is characterized in that further include process detection system, process inspection
Examining system includes PLC controller, Concentration Testing, pressure detecting and flow detection；
PLC controller has Profibus DP mouthfuls of connection distributed I/O using the CPU 414-2 of 400 series of Simens；For
PLC is equipped with ethernet communication module, accesses plc data for host computer；During PLC controller and ethernet communication module are placed on
It entreats in the PLC rack in control room；
Host computer selects i7 thinking computer, using WINDOWS XP operating system；
Golden hydrometallurgy whole process real-time optimization system is on i7 thinking computer, using C#2008 programming software, golden wet process smelting
Golden whole process real-time optimization algorithm uses Matlab 2016a programming software；
The signal of PLC and whole process real-time optimization transmission software is using C#2008 programming software；
Instrument is detected in hydrometallurgy process in-site installation, the signal of acquisition is transmitted to by detection instrument by Profibus-DP
In PLC, PLC timing sends acquisition signal to host computer by Ethernet, and the data of receiving are transmitted to golden wet process smelting by host computer
Golden whole process real-time optimization system carries out full-range real-time optimization compensating operation, and provides Operating Guideline suggestion.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910236152.4A CN109885012B (en) | 2019-04-12 | 2019-04-12 | Real-time optimization compensation method for gold hydrometallurgy full flow |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910236152.4A CN109885012B (en) | 2019-04-12 | 2019-04-12 | Real-time optimization compensation method for gold hydrometallurgy full flow |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109885012A true CN109885012A (en) | 2019-06-14 |
| CN109885012B CN109885012B (en) | 2020-12-25 |
Family
ID=66934633
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910236152.4A Active CN109885012B (en) | 2019-04-12 | 2019-04-12 | Real-time optimization compensation method for gold hydrometallurgy full flow |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109885012B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110918266A (en) * | 2019-12-10 | 2020-03-27 | 中南大学 | Control device and method for improving flotation froth quality |
| CN111142488A (en) * | 2019-12-30 | 2020-05-12 | 浙江中控技术股份有限公司 | Process industrial optimization method based on working condition comparison driving |
| CN111340269A (en) * | 2019-12-31 | 2020-06-26 | 浙江中控技术股份有限公司 | Real-time optimization method for process industrial process |
| CN116880427A (en) * | 2023-09-06 | 2023-10-13 | 中南大学 | Intelligent control method and system based on feeding condition estimation and comprehensive analysis of different working conditions |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101526814A (en) * | 2009-02-03 | 2009-09-09 | 东北大学 | Leaching rate prediction and optimization operation method in wet metallurgical leaching process |
| CN104062953A (en) * | 2014-06-12 | 2014-09-24 | 东北大学 | Gold hydrometallurgy process running state online evaluation method |
| CN104597755B (en) * | 2014-06-12 | 2017-02-15 | 东北大学 | Hydrometallurgical gold cyanide leaching optimization method |
| CN106886154A (en) * | 2017-04-26 | 2017-06-23 | 东北大学 | Hydrometallurgy whole process modeling method and optimization method based on interval number |
| CN106950946A (en) * | 2017-05-16 | 2017-07-14 | 东北大学 | A kind of hydrometallurgy exception control method based on optimization principles |
| CN108921352A (en) * | 2018-07-06 | 2018-11-30 | 东北大学 | A kind of wet metallurgical leaching process optimization method with bounded-but-unknown uncertainty |
-
2019
- 2019-04-12 CN CN201910236152.4A patent/CN109885012B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101526814A (en) * | 2009-02-03 | 2009-09-09 | 东北大学 | Leaching rate prediction and optimization operation method in wet metallurgical leaching process |
| CN104062953A (en) * | 2014-06-12 | 2014-09-24 | 东北大学 | Gold hydrometallurgy process running state online evaluation method |
| CN104597755B (en) * | 2014-06-12 | 2017-02-15 | 东北大学 | Hydrometallurgical gold cyanide leaching optimization method |
| CN106886154A (en) * | 2017-04-26 | 2017-06-23 | 东北大学 | Hydrometallurgy whole process modeling method and optimization method based on interval number |
| CN106950946A (en) * | 2017-05-16 | 2017-07-14 | 东北大学 | A kind of hydrometallurgy exception control method based on optimization principles |
| CN108921352A (en) * | 2018-07-06 | 2018-11-30 | 东北大学 | A kind of wet metallurgical leaching process optimization method with bounded-but-unknown uncertainty |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110918266A (en) * | 2019-12-10 | 2020-03-27 | 中南大学 | Control device and method for improving flotation froth quality |
| CN110918266B (en) * | 2019-12-10 | 2020-09-29 | 中南大学 | Control device and method for improving flotation froth quality |
| CN111142488A (en) * | 2019-12-30 | 2020-05-12 | 浙江中控技术股份有限公司 | Process industrial optimization method based on working condition comparison driving |
| CN111142488B (en) * | 2019-12-30 | 2021-05-04 | 浙江中控技术股份有限公司 | Process industrial optimization method based on working condition comparison driving |
| CN111340269A (en) * | 2019-12-31 | 2020-06-26 | 浙江中控技术股份有限公司 | Real-time optimization method for process industrial process |
| CN116880427A (en) * | 2023-09-06 | 2023-10-13 | 中南大学 | Intelligent control method and system based on feeding condition estimation and comprehensive analysis of different working conditions |
| CN116880427B (en) * | 2023-09-06 | 2023-11-14 | 中南大学 | Intelligent control method and system based on feeding condition estimation and working condition analysis |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109885012B (en) | 2020-12-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109885012A (en) | A kind of gold hydrometallurgy whole process real-time optimization compensation method | |
| CN104062953B (en) | Golden hydrometallurgy process running status on-line evaluation method | |
| CN101813932B (en) | Method for component content prediction and optimization operation in wet-process metallurgic extraction process | |
| CN101526814B (en) | Leaching rate prediction and optimization operation method in wet metallurgical leaching process | |
| CN104597755B (en) | Hydrometallurgical gold cyanide leaching optimization method | |
| CN108037747B (en) | Golden hydrometallurgy whole process hierarchical block operating status on-line evaluation method | |
| CN105278520A (en) | Complex industrial process running state evaluation method and application based on T-KPRM | |
| CN104199389B (en) | The optimization method of golden hydrometallurgy replacement process | |
| CN108921352A (en) | A kind of wet metallurgical leaching process optimization method with bounded-but-unknown uncertainty | |
| CN104298187B (en) | Golden hydrometallurgy whole process three-decker process monitoring method | |
| CN108762238A (en) | A kind of wet metallurgical leaching process method for diagnosing faults based on DCD | |
| CN107563656B (en) | Method for evaluating running state of gold hydrometallurgy cyaniding leaching process | |
| CN105975800B (en) | Multi-parameter optimizing method and device for chemical heavy metal waste water treatment process | |
| CN100370453C (en) | Rare-earth cascade extraction separation component content soft measuring method | |
| CN107121490B (en) | The evaluation method of operating status in a kind of hydrometallurgy process | |
| CN106886154B (en) | Hydrometallurgy whole process modeling method and optimization method based on interval number | |
| Liu et al. | Temporal hypergraph attention network for silicon content prediction in blast furnace | |
| CN106950946B (en) | A kind of hydrometallurgy exception control method based on optimization principles | |
| Chang et al. | Multimode process mode identification with coexistence of quantitative information and qualitative information | |
| CN110276128B (en) | Underflow concentration prediction method of thickener based on DAJYPLS algorithm | |
| Jun et al. | Gold recovery modeling based on interval prediction for a gold cyanidation leaching plant | |
| Liu et al. | Real-time optimization compensation method based on a novel two-level multi-block hybrid model for the hydrometallurgy process | |
| CN110532699A (en) | The dense washing process method for diagnosing faults of hydrometallurgy based on fuzzy DCD | |
| Wang et al. | Identification of abnormal conditions for gold flotation process based on multivariate information fusion and double‐channel convolutional neural network | |
| CN109840722A (en) | A kind of gold hydrometallurgy whole process operating status on-line evaluation method and system |
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 |