CN103970016B - Multipolar magnesium electrolysis bath state recognition control system and recognition control method thereof - Google Patents
Multipolar magnesium electrolysis bath state recognition control system and recognition control method thereof Download PDFInfo
- Publication number
- CN103970016B CN103970016B CN201410219060.2A CN201410219060A CN103970016B CN 103970016 B CN103970016 B CN 103970016B CN 201410219060 A CN201410219060 A CN 201410219060A CN 103970016 B CN103970016 B CN 103970016B
- Authority
- CN
- China
- Prior art keywords
- module
- electrolysis
- electrolysis bath
- slot status
- status
- 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.)
- Expired - Fee Related
Links
Landscapes
- Electrolytic Production Of Metals (AREA)
Abstract
The invention relates to a magnesium electrolysis technology. A multipolar magnesium electrolysis bath state recognition control system is provided for solving the problem that only a single control strategy exists for different electrolysis bath states in the prior art. The multipolar magnesium electrolysis bath state recognition control system comprises an electrolysis bath, and further comprises a data acquisition module, a database module, a state recognition module, a study module, a fuzzy reasoning module and a control module. The data acquisition module is connected with the study module and the state recognition module respectively, the state recognition module is connected with the database module and the fuzzy reasoning module respectively, the fuzzy reasoning module is connected with the study module, the database module and the control module respectively, and the control module is connected with the electrolysis bath. The drawback that only a single control strategy exists for different electrolysis bath states before is changed, and the electrolysis process is kept between energy balance and material balance; the multipolar magnesium electrolysis bath state recognition control system and the recognition control method of the multipolar magnesium electrolysis bath state play a good effect in saving electric energy, reducing equipment faults, prolonging the service life of the electrolysis bath, improving work environment and other aspects, and the method is suitable for the multipolar magnesium electrolysis bath control system.
Description
Technical field
The present invention relates to magnesium eletrolysis technology, particularly to multipole groove magnesium eletrolysis technology.
Background technology
Multipole groove magnesium eletrolysis technology is adopted by increasing enterprise as process matched therewith in titanium sponge production, how to save
Can consumption reduction, extension slot service life, raising single cavity production capacity always enterprise's problem in the urgent need to address.Multipole magnesium eletrolysis
Can groove obtain good technical-economic index, depend primarily on and realize two balances, i.e. thing in electrolysis bath running
Material balance and the energy balance.Material balance substantially can be divided into charging, go out the aspects such as magnesium, electrolyte level control, chlorine output,
The energy balance can be divided into the spontaneous hot systems of electrolysis, exchange heating system, fan radiating system.In prior art, not one
Plant efficient control method, go preferably to control the setting electric current of feeding interval and electrolysis bath.
Content of the invention
The technical problem to be solved it is simply that provide a kind of multipole closing down magnesium electrolysis bath state recognition control system and its
Identification control method, thus preferably controlling the setting electric current of feeding interval and electrolysis bath, reduces tank electromagnetism valve events under liquid
Number, so that electrolytic process maintains between the energy balance and material balance, reaches energy-saving purpose.
The present invention solves described technical problem, employed technical scheme comprise that multipole closing down magnesium electrolysis bath state recognition control system,
Including electrolysis bath, also include data acquisition module, DBM, state recognition module, study module, fuzzy reasoning module and
Control module, described data acquisition module is connected with study module and state recognition module respectively, state recognition module respectively with
DBM and fuzzy reasoning module connect, fuzzy reasoning module respectively with study module, DBM and control module
Connect, control module is connected with electrolysis bath；
Described data acquisition module, the parameter information for collection electrolysis slot status and the real-time process data producing, pass
It is defeated by study module；
Described study module, for the parameter information of the electrolysis slot status according to collection, draws and can affect to be electrolysed channel-shaped
The parameter information of state, and can affect to be electrolysed the relation information with electrolysis slot status for the parameter information of slot status, described relation letter
Breath is designated as historical data and drives rule, and is transmitted to state recognition module, and control data acquisition module is only possible to simultaneously
The parameter information of impact electrolysis slot status and the real-time process data producing, are transferred to state recognition module；
Described DBM, for collecting other decision rules and the electrolysis bath state classification of impact electrolysis slot status,
It is additionally operable to collect the control method corresponding with electrolysis slot status；
Described state recognition module, for according to data collecting module collected to can affect be electrolysed slot status parameter
Information and the real-time process data producing, drive rule and other decision rules of impact electrolysis slot status according to historical data,
Judge electrolysis slot status, and be transmitted to fuzzy reasoning module；
Described fuzzy reasoning module, for driving rule, impact to be electrolysed slot status according to electrolysis slot status, historical data
Other decision rules and the electrolysis corresponding control method of slot status draw corresponding Optimal Control Strategy, and are transmitted to control
Molding block；
Described control module, for carrying out corresponding to control to multipole closing down magnesium electrolysis bath according to Optimal Control Strategy.
Specifically, described electrolysis bath state classification includes health status, basic health status, sub-health state, critical strong
Health state, slight deterioration state, moderate degradation mode and/or severe exacerbation state.
Specifically, described study module drawn according to automatic cluster algorithm can affect be electrolysed slot status parameter information and
Historical data drives rule.
Further, described historical data drives that rule can be in condition discrimination function, cluster centre and threshold value
Plant or multiple.
Further, the optimisation strategy that in described fuzzy reasoning module, electrolysis bath controls is manually set by technical staff, and
And it is corresponding with electrolysis slot status.
Multipole closing down magnesium electrolysis bath state recognition control method, comprises the following steps:
Step 1, system real-time monitoring electrolysis bath, the parameter information of collection electrolysis bath；
Step 2, system, according to the parameter information of electrolysis bath, carry out cluster analyses, draw and can affect to be electrolysed slot status
Parameter information and can affect to be electrolysed the relation information with electrolysis slot status for the parameter information of slot status, described relation information is designated as
Historical data drives rule；
Step 3, system can affect to be electrolysed the parameter information of slot status and real-time production according to what acquisition module collected
Process data, according to historical data drive rule and data base, judge electrolysis slot status, in described data base at least provided with
Other decision rules of impact electrolysis slot status, electrolysis bath state classification and the control method corresponding with electrolysis slot status；
Step 4, system, according to electrolysis bath status information, draw excellent accordingly in conjunction with historical data driving rule and data base
Change control strategy；
Step 5, system are controlled accordingly to electrolysis bath according to Optimal Control Strategy control.
Specifically, in described step 2, system is learnt to electrolysis bath running status and control method using learner,
Thus clustered analysis, draw the parameter information that can affect to be electrolysed slot status, and can affect to be electrolysed the parameter letter of slot status
Breath and the relation information being electrolysed slot status, described relation information is designated as historical data and drives rule.
Specifically, in described step 2, it can be condition discrimination function, cluster centre and threshold value that historical data drives rule
One or more of.
Specifically, in described step 3, impact electrolysis slot status other decision rules and with electrolysis slot status corresponding
Control method can be manually set.
Specifically, in described step 3, electrolysis bath state classification includes health status, basic health status, subhealth state shape
State, critical health status, slight deterioration state, moderate degradation mode and/or severe exacerbation state.
The invention has the beneficial effects as follows, come with respect to the multipole magnesium eletrolysis production control system being presently using or promote
Say, add electrolysis bath condition diagnosing, based on the real time data of electrolysis bath, using clustering algorithm, by study module, right
Substantial amounts of unordered, strong noise electrolysis bath service data is learnt, thus from substantial amounts of unordered, strong noise electrolysis bath
The relation rule that can reflect electrolysis slot status is excavated out in service data.By study module, reaction electrolysis slot status
Characteristic signal carries out pretreatment, extraction and identification, through study after a while and training, as the discrimination standard of slot status.
Artificially it is provided with the diagnosis basis of combined process expertise research electrolysis slot status system and various state in DBM,
And based on this state system, infer electrolysis bath control method under different conditions.And these diagnosis basis can according to
Family demand is manually set.DBM is divided into healthy shape according to the parameter rule of impact electrolysis slot status by being electrolysed slot status
State, basic health status, sub-health state, critical health status, slight deterioration state, moderate degradation mode and severe exacerbation shape
State, every kind of state all corresponds to a kind of control strategy of optimization.The present invention adds fuzzy reasoning on the basis of original control strategy,
In conjunction with Fuzzy Inference, set up electrolysis bath Based Intelligent Control mechanism to realize the Based Intelligent Control to electric current and feeding interval.Change
The drawbacks of conventional difference slot status only single control strategy, is so that electrolytic process maintains the energy balance and material balance
Between, in saves energy, reduce equipment fault, improve bath life, improve the aspects such as working environment, all serving good
Effect.
Brief description
Fig. 1 is the system structure of multipole closing down magnesium electrolysis bath state recognition control system of the present invention and its control method embodiment
Figure；
Fig. 2 is the state recognition stream of multipole closing down magnesium electrolysis bath state recognition control system of the present invention and its control method embodiment
Cheng Tu；
Fig. 3 is the fuzzy reasoning stream of multipole closing down magnesium electrolysis bath state recognition control system of the present invention and its control method embodiment
Cheng Tu.
Specific embodiment
Below in conjunction with the accompanying drawings and embodiment describe in detail technical scheme:
The present invention is directed to the problem that different electrolysis slot status in prior art only have single control strategy, provides one kind many
Pole closing down magnesium electrolysis bath state recognition control system, including electrolysis bath, also includes data acquisition module, DBM, state recognition
Module, study module, fuzzy reasoning module and control module, described data acquisition module respectively with study module and state recognition
Module connects, and state recognition module is connected with DBM and fuzzy reasoning module respectively, fuzzy reasoning module learns respectively
Module, DBM and control module connect, and control module is connected with electrolysis bath；Described data acquisition module, for gathering
The parameter information of electrolysis slot status and the real-time process data producing, are transferred to study module；Described study module, for basis
The parameter information of the electrolysis slot status of collection, draws the parameter information that can affect to be electrolysed slot status, and can affect electrolysis bath
The parameter information of state with electrolysis slot status relation information, described relation information be designated as historical data drive rule, and by its
It is transferred to state recognition module, control data acquisition module is only possible to affect the parameter information of electrolysis slot status and in real time simultaneously
The process data producing, is transferred to state recognition module；Described DBM, affects to be electrolysed other of slot status for collecting
Decision rule and electrolysis bath state classification, are additionally operable to collect the control method corresponding with electrolysis slot status；Described state recognition
Module, for according to data collecting module collected to can affect be electrolysed the parameter information of slot status and the real-time technique producing
Data, according to historical data drive rule and impact electrolysis slot status other decision rules, judge electrolysis slot status, and by its
It is transferred to fuzzy reasoning module；Described fuzzy reasoning module, for driving rule, impact according to electrolysis slot status, historical data
Other decision rules of electrolysis slot status and the electrolysis corresponding control method of slot status draw corresponding Optimal Control Strategy, and
It is transmitted to control module；Described control module, for carrying out corresponding to control to multipole closing down magnesium electrolysis bath according to Optimal Control Strategy
System.Multipole closing down magnesium electrolysis bath state recognition control method, first, system real-time monitoring electrolysis bath, the parameter information of collection electrolysis bath；
Then, system, according to the parameter information of electrolysis bath, carries out cluster analyses, draw can affect be electrolysed slot status parameter information and
Can affect to be electrolysed the relation information with electrolysis slot status for the parameter information of slot status, described relation information is designated as historical data and drives
Dynamic rule；Secondly, system can affect to be electrolysed the parameter information of slot status and real-time produces according to what acquisition module collected
Process data, drives rule and data base according to historical data, judges electrolysis slot status, at least provided with shadow in described data base
Ring other decision rules of electrolysis slot status and the control method corresponding with electrolysis slot status；Thirdly, system is according to electrolysis
Slot status information, drives rule and data base to draw corresponding Optimal Control Strategy in conjunction with historical data；Finally, system is according to excellent
Change control strategy control electrolysis bath is controlled accordingly.Produce with respect to the multipole magnesium eletrolysis being presently using or promote and control
For system, add electrolysis bath condition diagnosing, based on the real time data of electrolysis bath, using clustering algorithm, by study
Module, learns to substantial amounts of unordered, strong noise electrolysis bath service data, thus from substantial amounts of unordered, strong noise
Electrolysis bath service data in excavate out can reflect electrolysis slot status relation rule.By study module, reaction electrolysis
The characteristic signal of slot status carries out pretreatment, extraction and identification, through study after a while and training, sentencing as slot status
Other standard.Artificially it is provided with combined process expertise research electrolysis slot status system and the examining of various state in DBM
Disconnected foundation, and based on this state system, infer electrolysis bath control method under different conditions.And these diagnosis basis are permissible
It is manually set according to user's request.DBM is divided into being electrolysed slot status according to the parameter rule of impact electrolysis slot status
Health status, basic health status, sub-health state, critical health status, slight deterioration state, moderate degradation mode and serious
Degradation mode, every kind of state all corresponds to a kind of control strategy of optimization.The present invention adds mould on the basis of original control strategy
Paste reasoning, in conjunction with Fuzzy Inference, sets up electrolysis bath Based Intelligent Control mechanism to realize the intelligence control to electric current and feeding interval
System.The drawbacks of change the only single control strategy of different slot status in the past so that electrolytic process maintain the energy balance and
Between material balance, in saves energy, reduce equipment fault, improve bath life, improve the aspects such as working environment, all playing
Good effect.
Embodiment
The multipole closing down magnesium electrolysis bath state recognition control system of this example, as shown in figure 1, including electrolysis bath, also including data and adopting
Collection module, DBM, state recognition module, study module, fuzzy reasoning module and control module, described data acquisition module
Block is connected with study module and state recognition module respectively, state recognition module respectively with DBM and fuzzy reasoning module
Connect, study module, DBM and control module connect fuzzy reasoning module respectively；
Specifically, described data acquisition module, the parameter information for collection electrolysis slot status and the real-time technique producing
Data, is transferred to study module；Described study module, for the parameter information of the electrolysis slot status according to collection, drawing can
The parameter information of impact electrolysis slot status, and can affect to be electrolysed the relation letter with electrolysis slot status for the parameter information of slot status
Breath, described relation information is designated as historical data and drives rule, and is transmitted to state recognition module, simultaneously control data collection
Module is only possible to affect the parameter information of electrolysis slot status and the real-time process data producing, and is transferred to state recognition module；
Described DBM, for collecting other decision rules and the electrolysis bath state classification of impact electrolysis slot status, is additionally operable to receive
The collection control method corresponding with electrolysis slot status；Described state recognition module, for arrive according to data collecting module collected
Can affect to be electrolysed the parameter information of slot status and the real-time process data producing, drive rule and impact electricity according to historical data
Other decision rules of solution slot status, judge electrolysis slot status, and are transmitted to fuzzy reasoning module；Described fuzzy reasoning mould
Block, for driving rule, other decision rules of impact electrolysis slot status and electrolysis channel-shaped according to electrolysis slot status, historical data
The corresponding control method of state draws corresponding Optimal Control Strategy, and is transmitted to control module；Described control module, uses
In carrying out corresponding to control to multipole closing down magnesium electrolysis bath according to Optimal Control Strategy.Described Optimal Control Strategy is system according to electrolysis bath
State, historical data drive rule, other decision rules of impact electrolysis slot status and are electrolysed the corresponding controlling party of slot status
Method analyzes the control strategy thus drawing with reference to what art electrolysis cells data of lower a moment and density of magnesium chloride were made, to ensure to be electrolysed
Groove real-time ensuring is in normal operating condition.
Preferably, the rule of described DBM and control method can be manually set, user can according to experience and
Knowledge, arranges the relation information of corresponding cell parameters and state in DBM.
Preferably, described control module is plc control module.
The multipole closing down magnesium electrolysis bath state recognition control system of this example (is gone out by analysis feeding interval, tank voltage, special operational
Magnesium, change coil, change anode etc.) etc. the process data that produces in real time and density of magnesium chloride relation rule.Analysis electrolysis bath event
Barrier reason, carries out slot status identification, predicts electrolysis bath state change in advance, by live existing plc system, increases study mould
Block and state recognition module, the current state of Real time identification electrolysis bath, in conjunction with next stage density of magnesium chloride, in conjunction with fuzzy reasoning
Technology, thus realize the optimal control to multipole closing down magnesium electrolysis bath.With respect to the multipole magnesium eletrolysis life being presently using or promote
For producing control system, add electrolysis bath condition diagnosing, based on the real time data of electrolysis bath, using clustering algorithm, lead to
Cross study module, substantial amounts of unordered, strong noise electrolysis bath service data learnt, thus from substantial amounts of unordered,
The relation rule that can reflect electrolysis slot status is excavated out in the electrolysis bath service data of strong noise.Artificially set in DBM
It is equipped with the diagnosis basis of combined process expertise research electrolysis slot status system and various state, and with this state system as base
Plinth, infers electrolysis bath under different conditions and controls.And what these diagnosis basis can be manually set according to user's request.
Embodiment 2
The multipole closing down magnesium electrolysis bath state recognition control method of this example is:
First, system real-time monitoring electrolysis bath, the parameter information of collection electrolysis bath.
Then, system, according to the parameter information of electrolysis bath, carries out cluster analyses, draws the ginseng that can affect to be electrolysed slot status
Number information and can affect to be electrolysed the relation information with electrolysis slot status for the parameter information of slot status, described relation information is designated as going through
History data driven rule.Specifically, system is learnt to electrolysis bath running status and control method using learner, thus warp
Cluster analyses, show that the parameter information that can affect to be electrolysed slot status and historical data drive rule.Preferably, historical data is driven
Dynamic rule can be one or more of condition discrimination function, cluster centre and threshold value.
Secondly, system can affect to be electrolysed the parameter information of slot status and real-time produces according to what acquisition module collected
Process data, drives rule and data base according to historical data, judges electrolysis slot status, at least provided with shadow in described data base
Ring other decision rules of electrolysis slot status and the control method corresponding with electrolysis slot status.Specifically, impact electrolysis channel-shaped
Other decision rules of state and the control method corresponding with electrolysis slot status can be manually set.Preferably, electrolysis slot status is divided
For health status, basic health status, sub-health state, critical health status, slight deterioration state, moderate degradation mode and tight
Weight degradation mode.
Thirdly, system, according to electrolysis bath status information, drives rule and data base to draw accordingly in conjunction with historical data
Optimal Control Strategy.Finally, system controls according to Optimal Control Strategy and electrolysis bath is controlled accordingly.
Specifically, the multipole closing down magnesium electrolysis bath state recognition flow process of this example, as shown in Fig. 2 comprise feature selection, pretreatment,
Feature set composition, feature extraction, the study of state identifier and training and state recognition.
Detailed, first, the real-time production data that system collects to acquisition module carries out feature selection and pretreatment shape
Become initial condition feature set, the importance due to constituting each parameter in state feature set is different, and may deposit between these parameters
In dependency.Study module learns to these huge data, judges the parameter letter that can affect to be electrolysed slot status
Cease and can affect to be electrolysed the relation information with electrolysis slot status for the parameter information of slot status.
Preferably, study module is by drawing the parameter information that can affect to be electrolysed slot status according to automatic cluster algorithm
And can affect to be electrolysed the parameter information of slot status and the relation information of electrolysis slot status.
Thus system is according to the judged result of study module, primitive character collection is adopted suitable conversion extract optimum ginseng
Amount, reduces dimension, forms low-dimensional feature set.The step for be not only convenient for the process of computer, improve arithmetic speed, be conducive to reality
When state recognition, but also the increase of the state fuzziness causing due to the interference of secondary parameter can be eliminated, impact identification
Accuracy.Original thick feature set is namely converted into smart feature set by this process, that is, the process discarded the dross and selected the essential.
Hypothesis x is n Wei Yuanshitezhengji, y smart feature set after processed for m dimension (m < n), then linear transformation relation is y
=ax, in formula, a is m*n transformation matrix.If transformation matrix a takes orthogonal transform matrix, transformed after smart feature set y have
Following features:
1. due to adopting orthogonal transform matrix, so that each dimension parameter is separate in feature set y, reduce feature set
Fuzziness, highlights its separability, beneficial to state classification, identification.
2. construct transformation matrices a under conditions of least mean-square error so as to be m*n matrix, and m < n, then make feature set y
Dimension be less than feature set x dimension.Achieve the purpose of data compression, condensed information.
In practical operation, totally taken a walk matrix by feature set x, according to least mean-square error condition by totally taking a walk
The characteristic value of matrix and eigenvector tectonic transition matrix a.
Study module, the parameter information of the electrolysis slot status according to collection, draw the parameter that can affect to be electrolysed slot status
Information and can affect to be electrolysed the parameter information of slot status and the relation information of electrolysis slot status, and is transmitted to state recognition
Module, the parameter information that control data acquisition module is only possible to affect to be electrolysed slot status simultaneously is transferred to state recognition module；
The design of state recognition module is according to the smart feature set obtaining after feature extraction: { yi, i=1,2 ..., n warp
Cross training or the study of computer, seek condition discrimination function or cluster centre and threshold value, then state recognition module is according to this
State recognition is carried out to feature set.
Can be used as being electrolysed slot status with the study module of training and state recognition module through study after a while
Discrimination standard, just may recognize that the equipment of its reaction in the case of the characteristic parameter obtaining any group of description electrolysis slot status
Status categories.
The multipole closing down magnesium electrolysis bath state recognition control system of this example is chosen tank voltage and is analyzed for cluster data, is set by monitoring
Determine voltage and sampled voltage by voltage curve parametrization, obtain being electrolysed the parameter needed for slot status.
Detailed, the state of electrolysis bath is mainly made up of groove condition and operating mode, and the operating mode of electrolysis bath is to represent that electrolysis bath is current
The index of the matching relationship of the electrolysis ability being occurred and input energy, electrolysis bath is in different operating modes, and its electrolysis ability is just
Difference, this index is independent of groove condition.The judgement basic thought of operating mode is: in a period of time, according to the whole ripple of voltage curve
The lifting situation of shape, the fluctuating margin of waveform, the situation of waveform mutation are determining the poles such as operating mode.
The groove condition of electrolysis bath is the index of the electrolysis ability representing that electrolysis bath has, with the carrying out producing, electrolysis bath
The physical characteristics of itself will change, and the electrolysis ability of groove also changes therewith.The change of groove condition be one slow
Process, the operating mode in a period of time may indicate that the variation tendency of groove condition.Electrolysis bath is in different groove conditions, real-time for producing
Control should be treated with a certain discrimination.The thinking of groove condition analysis: by analyzing the tank voltage ripple in each electrolysis bath a period of time recently
Move value and operating mode situation to determine.
DBM is divided into health status, substantially strong according to the parameter rule of impact electrolysis slot status by being electrolysed slot status
Health state, sub-health state, critical health status, slight deterioration state, moderate degradation mode and severe exacerbation state.
The parameter of system monitor in real time reflection electrolysis slot status, judges the state of electrolysis bath by state identifier, in conjunction with
Density of magnesium chloride is predicted, draws control strategy by fuzzy reasoning module.Again by plc control module according to control strategy to electrolysis
Groove makes corresponding control.As shown in figure 3, system combines DBM according to electrolysis bath instantaneous operating conditions and density of magnesium chloride
Fuzzy reasoning is carried out to it, ambiguity solution obtains setting electric current and feeding interval is controlled to electrolysis bath.
Specific fuzzy reasoning module, is according to electrolysis bath combinations of states technological parameter, design, this fuzzy reasoning module
Optimisation strategy and fuzzy control rule that middle electrolysis bath controls, can be manually set by technical staff, and with electrolysis slot status phase
Corresponding.
In sum, with respect to be presently using or the multipole magnesium eletrolysis production control system promoted for, add
Electrolysis bath condition diagnosing, based on the real time data of electrolysis bath, using clustering algorithm, by study module, to substantial amounts of nothing
Sequence, strong noise electrolysis bath service data is learnt, thus from substantial amounts of unordered, strong noise electrolysis bath service data
Middle excavation out can reflect the relation rule being electrolysed slot status.Artificially it is provided with combined process expertise in DBM to grind
Study carefully the diagnosis basis of electrolysis slot status system and various state, and based on this state system, infer electricity under different conditions
Solution groove controls.And what these diagnosis basis can be manually set according to user's request.DBM is according to impact electrolysis channel-shaped
The parameter rule of state is divided into health status, basic health status, sub-health state, critical health status, light by being electrolysed slot status
Micro- degradation mode, moderate degradation mode and severe exacerbation state, every kind of state all corresponds to a kind of control strategy of optimization.The present invention
Add fuzzy reasoning on the basis of original control strategy, in conjunction with Fuzzy Inference, set up electrolysis bath Based Intelligent Control mechanism
Realize the Based Intelligent Control to electric current and feeding interval.Change the drawbacks of different slot status only have single control strategy in the past,
Electrolytic process is maintained between the energy balance and material balance, in saves energy, reduces equipment fault, improve the electrolysis bath longevity
Order, improve the aspects such as working environment, all serve good effect.
Claims (10)
1. multipole closing down magnesium electrolysis bath state recognition control system, including electrolysis bath it is characterised in that also including data acquisition module,
DBM, state recognition module, study module, fuzzy reasoning module and control module, described data acquisition module is respectively
It is connected with study module and state recognition module, state recognition module is connected with DBM and fuzzy reasoning module respectively,
Fuzzy reasoning module is connected with study module, DBM and control module respectively, and control module is connected with electrolysis bath；
Described data acquisition module, the parameter information for collection electrolysis slot status and the real-time process data producing, are transferred to
Study module；
Described study module, the parameter information for the electrolysis slot status according to collection carries out feature selection, pretreatment, feature set
Draw the parameter information that can affect to be electrolysed slot status after composition and feature extraction, and can affect to be electrolysed the parameter letter of slot status
Breath and the relation information being electrolysed slot status, described relation information is designated as historical data and drives rule, and the state that is transmitted to is known
Other module, simultaneously control data acquisition module be only possible to affect the parameter information of electrolysis slot status and the real-time technique number producing
According to being transferred to state recognition module；
Described DBM, for collecting other decision rules and the electrolysis bath state classification of impact electrolysis slot status, also uses
In the collection control method corresponding with electrolysis slot status；
Described state recognition module, for according to data collecting module collected to can affect be electrolysed slot status parameter information
And the process data of production in real time, drive rule according to historical data and affect to be electrolysed other decision rules of slot status, judge
Electrolysis slot status, and it is transmitted to fuzzy reasoning module；
Described fuzzy reasoning module, for driving rule, affecting to be electrolysed other of slot status according to electrolysis slot status, historical data
Decision rule and the electrolysis corresponding control method of slot status draw corresponding Optimal Control Strategy, and are transmitted to control mould
Block；
Described control module, for carrying out corresponding to control to multipole closing down magnesium electrolysis bath according to Optimal Control Strategy.
2. multipole closing down magnesium electrolysis bath state recognition control system according to claim 1 is it is characterised in that described electrolysis channel-shaped
State classification includes health status, basic health status, sub-health state, critical health status, slight deterioration state, moderate deterioration
State and/or severe exacerbation state.
3. multipole closing down magnesium electrolysis bath state recognition control system according to claim 1 is it is characterised in that described study module
Show that according to automatic cluster algorithm the parameter information that can affect to be electrolysed slot status and historical data drive rule.
4. multipole closing down magnesium electrolysis bath state recognition control system according to claim 3 is it is characterised in that described historical data
Drive rule can be one or more of condition discrimination function, cluster centre and threshold value.
5. multipole closing down magnesium electrolysis bath state recognition control system according to claim 4 is it is characterised in that described fuzzy reasoning
The optimisation strategy that in module, electrolysis bath controls is manually set by technical staff, and corresponding with electrolysis slot status.
6. multipole closing down magnesium electrolysis bath state recognition control method is it is characterised in that comprise the following steps:
Step 1, system real-time monitoring electrolysis bath, the parameter information of collection electrolysis bath；
Step 2, system according to the parameter information of electrolysis bath, carry out cluster analyses, feature selection, pretreatment, feature set constitute and
Feature extraction, parameter information and the electrolysis drawing the parameter information that can affect to be electrolysed slot status and can affecting to be electrolysed slot status
The relation information of slot status, described relation information is designated as historical data and drives rule；
Step 3, system can affect to be electrolysed the parameter information of slot status and the real-time work producing according to what acquisition module collected
Skill data, drives rule and data base according to historical data, judges electrolysis slot status, at least provided with having an impact in described data base
Other decision rules of electrolysis slot status, electrolysis bath state classification and the control method corresponding with electrolysis slot status；
Step 4, system, according to electrolysis bath status information, drive rule and data base to show that corresponding optimization is controlled in conjunction with historical data
System strategy；
Step 5, system are controlled accordingly to electrolysis bath according to Optimal Control Strategy control.
7. multipole closing down magnesium electrolysis bath state recognition control method according to claim 6 is it is characterised in that in described step 2,
System is learnt to electrolysis bath running status and control method using learner, thus clustered analysis, draws and can affect
The parameter information of electrolysis slot status, and can affect to be electrolysed the relation information with electrolysis slot status for the parameter information of slot status, institute
State relation information and be designated as historical data driving rule.
8. multipole closing down magnesium electrolysis bath state recognition control method according to claim 6 is it is characterised in that in described step 2,
It can be one or more of condition discrimination function, cluster centre and threshold value that historical data drives rule.
9. multipole closing down magnesium electrolysis bath state recognition control method according to claim 6 is it is characterised in that in described step 3,
Other decision rules of impact electrolysis slot status and the control method corresponding with electrolysis slot status can be manually set.
10. multipole closing down magnesium electrolysis bath state recognition control method according to claim 6 is it is characterised in that described step 3
In, electrolysis bath state classification includes health status, basic health status, sub-health state, critical health status, slight deterioration shape
State, moderate degradation mode and/or severe exacerbation state.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410219060.2A CN103970016B (en) | 2014-05-22 | 2014-05-22 | Multipolar magnesium electrolysis bath state recognition control system and recognition control method thereof |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410219060.2A CN103970016B (en) | 2014-05-22 | 2014-05-22 | Multipolar magnesium electrolysis bath state recognition control system and recognition control method thereof |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103970016A CN103970016A (en) | 2014-08-06 |
| CN103970016B true CN103970016B (en) | 2017-01-25 |
Family
ID=51239659
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410219060.2A Expired - Fee Related CN103970016B (en) | 2014-05-22 | 2014-05-22 | Multipolar magnesium electrolysis bath state recognition control system and recognition control method thereof |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103970016B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109100598B (en) * | 2018-10-23 | 2021-01-26 | 攀钢集团攀枝花钢铁研究院有限公司 | Method for judging single-cell running state of assembly line type magnesium electrolytic cell |
| CN116288532B (en) * | 2023-03-13 | 2024-09-10 | 赛富能科技(深圳)有限公司 | Method and equipment for monitoring electrolytic tank |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101624713A (en) * | 2009-07-31 | 2010-01-13 | 深圳市库马克新技术股份有限公司 | Aluminum electrolysis visual production information system |
-
2014
- 2014-05-22 CN CN201410219060.2A patent/CN103970016B/en not_active Expired - Fee Related
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101624713A (en) * | 2009-07-31 | 2010-01-13 | 深圳市库马克新技术股份有限公司 | Aluminum electrolysis visual production information system |
Non-Patent Citations (4)
| Title |
|---|
| 孙乐彬.基于数据驱动和预测原理的电解槽优化控制系统设计.《中国优秀硕士学位论文全文数据库 工程科技Ⅰ辑》.2011,(第04期),第B023-33页. * |
| 林学会等.铝电解智能控制技术的研究与应用.《自动化技术与应用》.2008,第27卷(第6期),第109-112页. * |
| 汪成亮等.基于工况槽况的电解生产控制研究及实现.《轻金属》.2009,(第4期),第24-28页. * |
| 满朝元等.基于数据挖掘的电解槽优化控制.《重庆理工大学学报(自然科学)》.2010,第24卷(第6期),第89-93页. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103970016A (en) | 2014-08-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107578288B (en) | Non-invasive load decomposition method considering user power consumption mode difference | |
| CN108876163B (en) | Transient state power angle stability rapid evaluation method integrating causal analysis and machine learning | |
| CN111563827B (en) | Load decomposition method based on physical characteristics of electrical appliance and electricity utilization behaviors of residents | |
| CN105823948A (en) | Non-invasive resident load identification method | |
| CN114676742A (en) | Power grid abnormal electricity utilization detection method based on attention mechanism and residual error network | |
| CN109558897A (en) | Power grid planning multi-scene extraction method based on data mining | |
| CN112686456A (en) | Power load prediction system and method combining edge calculation and energy consumption identification | |
| CN103970016B (en) | Multipolar magnesium electrolysis bath state recognition control system and recognition control method thereof | |
| CN117254505B (en) | Energy storage power station optimal operation mode decision method and system based on data processing | |
| CN112670999A (en) | Low-voltage distribution network real-time voltage control method based on user-side flexible resources | |
| CN118133203A (en) | Fault diagnosis method for electric energy metering detection information | |
| CN107974696A (en) | Aluminium cell production process Anodic effect forecast method | |
| CN114880939A (en) | Intelligent prediction method and device for service life of power battery | |
| CN113379005B (en) | Intelligent energy management system and method for power grid power equipment | |
| Zhou et al. | Non-intrusive extraction and forecasting of residential electric vehicle charging load | |
| CN116298576A (en) | Time-segmentation-considered extensible non-invasive load monitoring method | |
| CN116205345A (en) | Optimal scheduling method and system considering wind-solar prediction and maintenance plan | |
| CN109492802A (en) | A kind of year-end level of multi-year regulating storage reservoir for timed position preferred method based on expert system | |
| CN114819362A (en) | Power grid power load balancing method for wind-solar power system | |
| CN111695599A (en) | Elastic identification method for user electrical load time | |
| CN111598145A (en) | Non-invasive load monitoring method based on mixed probability label time-varying constraint distribution | |
| CN109657735A (en) | Service life of aluminum electrolysis cell prediction technique based on Method Using Relevance Vector Machine | |
| CN218404436U (en) | Electrolytic hydrogen production control system based on artificial intelligence | |
| CN115186900B (en) | Dynamic blast furnace gas generation amount prediction method and system suitable for multiple working condition categories | |
| CN115169381A (en) | Low-frequency signal non-invasive load state identification algorithm |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20170125 Termination date: 20210522 |