CN102636624A - Method for soft measurement of alumina concentration in electrolyzer during aluminum electrolysis process - Google Patents
Method for soft measurement of alumina concentration in electrolyzer during aluminum electrolysis process Download PDFInfo
- Publication number
- CN102636624A CN102636624A CN2012101316017A CN201210131601A CN102636624A CN 102636624 A CN102636624 A CN 102636624A CN 2012101316017 A CN2012101316017 A CN 2012101316017A CN 201210131601 A CN201210131601 A CN 201210131601A CN 102636624 A CN102636624 A CN 102636624A
- Authority
- CN
- China
- Prior art keywords
- parameter
- value
- colony
- alumina concentration
- max
- 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
Images
Abstract
The invention discloses a method for soft measurement of alumina concentration in an electrolyzer during an aluminum electrolysis process, comprising the following steps: (1) collecting process production parameter data; (2) preprocessing the process production parameter data; (3) establishing samples of process production parameter data; (4) establishing a multi-alumina concentration soft measurement model based on different electrolyzer statuses and obtaining model parameters; (5) employing the model to estimate alumina concentration. The method disclosed by the invention can ensure fast, accurate and real-time detection of changes in the alumina concentration so as to achieve optimal control of the production process.
Description
Technical field
The present invention relates to the measuring technique of aluminium electrolysis process manufacturing parameter, relate in particular to a kind of flexible measurement method that aluminium electrolysis process is difficult to measure with the physical sensors online in real time alumina concentration in the electrolytic tank that is used for solving.
Background technology
In aluminium electrolytic industry; Alumina concentration is the important state parameter of reflection aluminium cell production run process; Also be an important control parameter keeping the aluminium cell material balance simultaneously, the process schedule requirement according to the modern aluminum electrolytic industry is produced " three low height " must be controlled at a lower concentration range with alumina concentration; But because at present to economy, technical reasons such as high, the easy damages of the direct measured sensor cost of alumina concentration; Cause lacking the measurement means of online direct measurement alumina concentration, this has become restriction aluminium electroloysis industry and has further improved control efficiency, a bottleneck that cuts down the consumption of energy always.At present; Soft-measuring technique provides valid approach for solving this type of problem; And be considered to fruitful and the most attractive new method; Add under the condition of investment not increasing or reduce, soft-measuring technique will be used widely, thereby process control and detection system are produced tremendous influence.Soft measurement is exactly (to be called auxiliary variable according to the process variable that easy survey can be surveyed; Like speed, pressure, temperature etc.) (be called leading variable with the variable to be measured that is difficult to directly detect; Like material component, product quality etc.) mathematical relation; According to certain optiaml ciriterion, adopt various computing method, realize measurement or estimation with soft-sensing model to variable to be measured.
At present, to the soft measurement of alumina concentration, mainly still be cell resistance slope meter algorithm, promptly through can extrapolate the concentration of aluminium oxide to the analysis of cell resistance size, its relation curve is as shown in Figure 1.Because cell resistance R calculates gained by tank voltage and potline current intensity; Can relation shown in Figure 1 be converted into the relation between tank voltage and the alumina concentration; Being about to the alumina concentration variation is transformed in the less time interval scope; In this scope,, adopt the method for linear theories such as linear regression model (LRM), set up the soft-sensing model of alumina concentration according to tank voltage or cell resistance slope value and corresponding alumina concentration value thereof; This method can only effectively detect alumina concentration and change in very little perform region; And can not thoroughly solve the soft problems of measurement of alumina concentration of whole perform region, this be because alumina concentration except change with current tank voltage or cell resistance have outside the Pass, also with the current or residing electrolytic tank state of electrolytic tank in early time; The action of the groove that adopted, relevant with the fluctuation of other manufacturing parameters etc.In addition; Based on artificial neural network (Artificial Neural Network; Brief note is for ANN) the soft measurement of alumina concentration also the someone adopted, can be under the condition of the priori that does not possess object, based on the soft measurement of artificial neural network according to the direct modeling of the input/output relation of object; The on-line correction ability of model is strong, and can be applicable to highly non-linear and serious time-dependent system.But; The selection of quality and quantity, learning algorithm, topology of networks and the type etc. of the learning training sample of neural network all has significant impact to the soft-sensing model that constitutes; And the selection of training sample proper vector short of reliable technological basis often, so adopt the alumina concentration soft-sensing model of neural network that its limitation is also arranged.In sum; Because aluminium cell is a non-homogeneous unsteady Model; Each item factor that the alumina concentration that relates to changes intercouples; So combined process is chosen the characteristic input of suitable alumina concentration model, adopting a kind of accurate, reliable non-linear soft-sensing model is vital to setting up the alumina concentration soft-sensing model.
Summary of the invention
In view of this, the flexible measurement method that the purpose of this invention is to provide alumina concentration in a kind of accurate, reliable aluminium electrolysis process electrolytic tank.
The objective of the invention is to realize through following technical scheme:
The flexible measurement method of alumina concentration in this aluminium electrolysis process electrolytic tank may further comprise the steps:
Step 1: gatherer process manufacturing parameter data comprise that the process of gathering tank voltage, potline current, reinforced (NB), aluminum yield at interval produces parameter, and set up the historical data base of each parameter values of storage through hardware device;
Step 2: set a time for reading parameter and time for reading section parameter; Every at a distance from one section time for reading parameter; Just read the time for reading section parameter data in the past in this moment of historical data base; The process manufacturing parameter is carried out data processing, set up training sample set and test sample book set, the form of sample set is { x
i, y
i, wherein
x
i={ F (k), H (k), R (k), S (k), T (k), NB (k), L (k), V
s(k) }, y
i={ C (k) }, x
i∈ R
8, y
i∈ R, wherein, the implication of each parameter is following in the formula:
K representes residing certain sampling instant of sample i, constantly the k-moment (k-1)=time for reading parameter;
F (k) is the cell voltage fluctuation number of times, to the time for reading section parameter that collects with all interior tank voltage data-signal V
k, then the signal of (8,0) frequency range is carried out reconstruct and obtain the tank voltage signal V of time for reading section parameter with eight layers of DB8 wavelet decomposition signals to the according to tree construction with interior (8,0) frequency range
80k, statistics time for reading section parameter is with the tank voltage signal wave crest trough number of interior (8,0) frequency range, and a Wave crest and wave trough is designated as 1 secondary undulation, and crest or trough are 0.5 secondary undulation;
H (k) is a high-frequency energy number percent, is designated as V with collecting time for reading section parameter with interior tank voltage
k, calculate tank voltage signal gross energy
Calculate (8,0) frequency range tank voltage signal energy
High-frequency energy number percent then
R (k) is a cell resistance, handles the tank voltage V that obtains (8,0) frequency range constantly through k
80kPotline current I (k) divided by this moment obtains; And utilize the exceptional value in the cell resistance sequential value of ten thousand formulas deletion sampling rate of following smoothing algorithm, wherein, e is the cell resistance value and the deviation of level and smooth calculated value; ω represents shift time; ω=5s, R (k-ω) are the cell resistance value of k 5s before the moment, and R (k-2 ω) is the cell resistance value of k 10s before the moment.
R(k)＝0.3×R(k-ω)-0.1×R(k-2ω)+0.2×(e(k)+2×e(k-ω)+e(k-2ω))；
S (k) is the cell resistance slope, and the increment that is defined as the filter resistance in the nearest time for reading parameter is a rate of change; T (k) is the accumulation slope; Be defined as the accumulation increment of the filter resistance in the in the recent period nearest time for reading parameter; Use the difference filter slope calculations, and use recursion formula calculating cumulative slope, specific algorithm is to be the cycle LPF resistance to be sampled with 30s; By following two formulas difference calculated resistance slope S and accumulation slope T
S(k)＝(R(k-6ω)-R(k-18ω)+2×(R(k)-R(k-24ω)))/5，
T(k)＝(15/16)×T(k-1)+S(k)/4；
NB (k) be alumina blanking at interval, aluminum electrolysis control system actual set blanking spacing value, expression with respect to the benchmark blanking at interval multiple, NB (k) is between 0.8-1.3, when NB (k)=1, expression is with benchmark blanking blanking at interval;
L (k) is an aluminum yield, the quality of the metallic aluminium that the expression aluminium cell extracts when k carries out out the aluminium operation constantly the time, in this method with L (k) as k average aluminum yield constantly, units, promptly
V
s(k) be setting voltage, the expression aluminum electrolytic cell control system is the tank voltage that electrolytic tank is set at moment k, the V of unit;
C (k) is an alumina concentration, and expression is dissolved into the mass percent of the aluminium oxide in the electrolyte, and general oxidation concentration all is between 0%-12%, C (k) value part of only peeking here, and promptly 2% alumina concentration is expressed as C (k)=2;
Step 3: adopt following algorithm that sample set is carried out normalization and handle, zoom to attribute between [0,1]:
In the following formula || || get the 2-norm;
Step 4:, set up alumina concentration soft-sensing model corresponding under the different electrolytic tank states, solving model, computation model parameter according to all electrolytic tank states of storing in the control system historical data base;
Step 5: according to current electrolytic cell state and its corresponding alumina concentration soft-sensing model; The test sample book combination is input in the soft-sensing model; Calculate algorithm that its corresponding alumina concentration discreet value
utilizes following formula with the anti-alumina concentration value that is normalized to of alumina concentration discreet value
; Estimate out the corresponding alumina concentration of current test sample book set
Further; In step 4; The groove state comprises health, health, inferior health, critical health, slight morbid state, ill, the serious morbid state of moderate totally seven kinds of states basically, collects sample set to the training the different slots state under, adopts least square method supporting vector machine (LSSVM) training to obtain the alumina concentration soft-sensing model of different slots state correspondence; Totally 7 kinds, model tormulation is following:
In the following formula, N is input number of samples, x
iBe i input sample, x is a certain input variable, α
i≠ 0 is pairing input sample x
iBe support vector, α
iBe the SVMs coefficient,
Be model pre-estimating output, K (x, x
i) (LSSVM) be kernel function;
Further, kernel function is RBF (RBF);
Further, in step 4, to the electrolytic tank under a certain groove state, the algorithm of alumina concentration soft-sensing model parameter is following:
Step 1:
The model parameter initialization; (comprise the quantity n of initialization colony, maximum iteration time k
Max, study factor c
1, study factor c
2, penalty coefficient pace of change lower bound Vc
Min, penalty coefficient pace of change upper bound Vc
Max, the kernel function spread factor changes lower bound V σ
Min, the kernel function spread factor changes lower bound V σ
Max, kernel function spread factor lower bound σ
Min, kernel function spread factor upper bound σ
Max, penalty coefficient lower bound c
Min, penalty coefficient upper bound c
Max, inertia weight ω);
Step 2:
The corresponding algorithm parameter c of all particles in the initialization colony and σ, i.e. initialization particle population, initial population is shone upon to produce through Logistic and is obtained, and concrete grammar is: definition ξ
11, ξ
12Be the random number in (0,1), and ξ
11≠ ξ
12, with ξ
11, ξ
12Substitution Logistic mapping, that is:
ξ
i+1，1＝4ξ
i1(1-ξ
i1)
ξ
i+1，2＝4ξ
i2(1-ξ
i2)，i＝1，...，m-1，
The following formula iteration is obtained m group Chaos Variable
for m-1 time
Again Chaos Variable is mapped in the feasible zone, even:
c
i＝c
min+ξ
i1(c
max-c
min)
σ
i＝σ
min+ξ
i2(σ
max-σ
min)，，
The definition initial population
X wherein
i=(x
I1, x
I2)=(c
i, σ
i) be i particle position;
Calculate the corresponding optimal adaptation value of all particles in the colony, the individual particles optimal location in the definition colony is:
M is colony's quantity; Definition colony optimal location: the optimal location P of colony
gInitial value be minimum that of adaptive value in the individual optimal location; With m P
iIn the radially basic kernel function substitution formula:
Use least square method and obtain m group model parameter alpha
iAnd b, then with this m group model parameter substitution successively
With
In, the f that tries to achieve is the adaptive value of m all particles correspondences, and the adaptive value of trying to achieve this moment is initialized as the corresponding optimal adaptation value of particle this moment, the optimal adaptation value is meant the difference of two squares of estimating output and actual output of adaptive value under parameter current;
Step 3:
Search colony's optimal adaptation value, i.e. the corresponding optimal adaptation value of all particles in the colony relatively, finding minimum optimal adaptation value is corresponding colony optimal adaptation value, and with particle corresponding algorithm parameter c and σ as optimum model parameter;
Step 4:
Value x and the pace of change v thereof of all particle algorithm parameter c and σ in the renewal colony;
All particle algorithm parameter c and σ are in the feasible zone scope, that is: in judgement and the change colony
If v
I1＞vc
Max, then make v
I1=vc
Max
If v
I1＜vc
Min, then make v
I1=vc
Min
If v
I2＞v σ
Max, then make v
I2=v σ
Max
If v
I2＜v σ
Min, then make v
I2=v σ
Min
If x
I1＜c
Min, then make x
I1=c
Min
If x
I1＞c
Max, then make x
I1=c
Max
If x
I2＜σ
Min, then make x
I2=σ
Min
If x
I2＞σ
Max, then make x
I2=σ
Max, i=1 ..., colony's quantity
Least square method is calculated the model parameter α of each particle under its algorithm parameter c and σ
iAnd b;
Calculate the optimal adaptation value of all particles in the colony, the optimal adaptation value value that all particles are corresponding is the minimum value between the optimal adaptation value that calculates of adaptive value that time iterative computation goes out and last iteration;
Search colony's optimal adaptation value, i.e. the corresponding optimal adaptation value of all particles in the colony relatively, finding minimum optimal adaptation value is corresponding colony optimal adaptation value, and with particle corresponding algorithm parameter c and σ as optimum model parameter;
Step 5:
Judge whether iterations reaches maximal value, if do not reach then return step 4, otherwise execution in step 6;
Step 6:
The algorithm parameter that algorithm parameter c that calculates after the last iteration and σ are decided to be the aluminium oxide soft-sensing model;
Step 7:
The model parameter α that utilization least square method computational algorithm parameter c and σ are corresponding
iAnd b;
Further, in step 1, the time for reading parameter is 10min, and time for reading section parameter is 2h.
The invention has the beneficial effects as follows:
(1),, the optimal control that realizes the electrolytic aluminium process lays the foundation for moving with optimization for the automatic control of aluminium electrolysis process realization provides monitor data;
(2) according to different electrolytic tank running statuses, training obtains the polyoxy aluminum concentration soft-sensing model structure under the different conditions, makes the adaptive faculty of soft measurement stronger;
(3) detect alumina concentration in real time and change, realize production process monitoring, improve electrolytic aluminium output;
(4) replace artificial assay, reach the purpose that promptly and accurately detects production status, further optimize technology.
Other advantages of the present invention, target and characteristic will be set forth in instructions subsequently to a certain extent; And to a certain extent; Based on being conspicuous to those skilled in the art, perhaps can from practice of the present invention, obtain instruction to investigating of hereinafter.Target of the present invention and other advantages can realize and obtain through following instructions and claims.
Description of drawings
In order to make the object of the invention, technical scheme and advantage clearer, will combine accompanying drawing that the present invention is made further detailed description below, wherein:
Fig. 1 is alumina concentration-cell resistance relation curve;
Fig. 2 is the implementation framework of soft-sensing model of the present invention;
Fig. 3 is the hardware platform structural representation that the present invention adopts;
Fig. 4 is the WAVELET PACKET DECOMPOSITION reorganization tree construction that sample data of the present invention is set up in the process to be adopted;
Fig. 5 is an electrolytic bath change in voltage curve.
Embodiment
Below will carry out detailed description to the preferred embodiments of the present invention with reference to accompanying drawing.Should be appreciated that preferred embodiment has been merely explanation the present invention, rather than in order to limit protection scope of the present invention.
As shown in Figure 2, the flexible measurement method of alumina concentration in the aluminium electrolysis process electrolytic tank of the present invention may further comprise the steps:
Step 1: gatherer process manufacturing parameter data comprise that the process of gathering tank voltage, potline current, reinforced (NB), aluminum yield at interval produces parameter, and set up the historical data base of each parameter values of storage through hardware device;
Step 2: set a time for reading parameter and time for reading section parameter; Every at a distance from one section time for reading parameter; Just read the time for reading section parameter data in the past in this moment of historical data base; The process manufacturing parameter is carried out data processing, set up training sample set and test sample book set, the form of sample set is { x
i, y
i, in the present embodiment, the time for reading parameter is 10min, time for reading section parameter is got 2h, and promptly every data before 10min reads this moment of historical data base 2h, wherein
x
i={ F (k), H (k), R (k), S (k), T (k), NB (k), L (k), V
s(k) }, y
i={ C (k) }, x
i∈ R
8, y
i∈ R, wherein, the implication of each parameter is following in the formula:
K representes residing certain sampling instant of sample i, constantly the k-moment (k-1)=time for reading parameter;
F (k) is the cell voltage fluctuation number of times, to the time for reading section parameter that collects with all interior tank voltage data-signal V
k, then the signal of (8,0) frequency range is carried out reconstruct and obtain the tank voltage signal V of time for reading section parameter with eight layers of DB8 wavelet decomposition signals to the according to tree construction with interior (8,0) frequency range
80k, statistics time for reading section parameter is with the tank voltage signal wave crest trough number of interior (8,0) frequency range, and a Wave crest and wave trough is designated as 1 secondary undulation, and crest or trough are 0.5 secondary undulation;
H (k) is a high-frequency energy number percent, is designated as V with collecting time for reading section parameter with interior tank voltage
k, calculate tank voltage signal gross energy
Calculate (8,0) frequency range tank voltage signal energy
High-frequency energy number percent then
R (k) is a cell resistance, handles the tank voltage V that obtains (8,0) frequency range constantly through k
80kPotline current I (k) divided by this moment obtains; And the mode of utilizing following smoothing algorithm deletes the exceptional value in the cell resistance sequential value of sampling rate, and wherein, e is the cell resistance value and the deviation of level and smooth calculated value; ω represents shift time; ω=5s, R (k-ω) are the cell resistance value of k 5s before the moment, and R (k-2 ω) is the cell resistance value of k 10s before the moment.
R(k)＝0.3×R(k-ω)-0.1×R(k-2ω)+0.2×(e(k)+2×e(k-ω)+e(k-2ω))；
S (k) is the cell resistance slope, and the increment that is defined as the filter resistance in the nearest time for reading parameter is a rate of change; T (k) is the accumulation slope; Be defined as the accumulation increment of the filter resistance in the in the recent period nearest time for reading parameter; Use the difference filter slope calculations, and use recursion formula calculating cumulative slope, specific algorithm is to be the cycle LPF resistance to be sampled with 30s; By following two formulas difference calculated resistance slope S and accumulation slope T
S(k)＝(R(k-6ω)-R(k-18ω)+2×(R(k)-R(k-24ω)))/5，
T(k)＝(15/16)×T(k-1)+S(k)/4；
NB (k) be alumina blanking at interval, aluminum electrolysis control system actual set blanking spacing value, expression with respect to the benchmark blanking at interval multiple, NB (k) is between 0.8-1.3, when NB (k)=1, expression is with benchmark blanking blanking at interval;
L (k) is an aluminum yield; The quality of the metallic aluminium that the expression aluminium cell extracts when k carries out out the aluminium operation during moment; Because L (k) sampling period with respect to preceding several model parameters is much bigger, thus with L (k) as k average aluminum yield constantly, in this method with L (k) as k average aluminum yield constantly; Units, promptly
V
s(k) be setting voltage, the expression aluminum electrolytic cell control system is the tank voltage that electrolytic tank is set at moment k, the V of unit;
C (k) is an alumina concentration, and expression is dissolved into the mass percent of the aluminium oxide in the electrolyte, and general oxidation concentration all is between 0%-12%, C (k) value part of only peeking here, and promptly 2% alumina concentration is expressed as C (k)=2;
Step 3: adopt following algorithm that sample set is carried out normalization and handle, zoom to attribute between [0,1]:
In the following formula || || get the 2-norm;
Step 4:, set up alumina concentration soft-sensing model corresponding under the different electrolytic tank states, solving model, computation model parameter according to all electrolytic tank states of storing in the control system historical data base; The groove state is defined as electrolysis ability and the important indicator of moving steady in a long-term that electrolytic tank has; Be how many decisions drops into electric weight to electrolytic tank; The reference frame that keeps efficient output, the electrolytic tank healthy for the groove state need drop into more energy, to improve output; On the contrary, just need to reduce the input of energy, take corresponding control measures to make the groove condition improvement for the electrolytic tank of groove state morbid state; Increase the input of energy then; Along with the carrying out that produces, the physical characteristics of electrolytic tank itself will change, and the electrolysis ability of groove also changes thereupon.
In the present embodiment; Groove state definition comprises health, health, inferior health, critical health, slight ill, ill, the serious morbid state of moderate totally seven kinds of states basically; Training under the different slots state collects sample set; Adopt least square method supporting vector machine (LSSVM) training to obtain the corresponding alumina concentration soft-sensing model of different slots state, totally 7 kinds, model tormulation is following:
In the following formula, N is input number of samples, x
iBe i input sample, x is a certain input variable, α
i≠ 0 is pairing input sample x
iBe support vector, α
iBe the SVMs coefficient,
Be model pre-estimating output, K (x, x
i) (LSSVM) be kernel function; The kernel function that this patent adopts is RBF (RBF); The parametric solution process of least square method supporting vector machine (LSSVM) adopts population (PSO) algorithm dynamic optimization; Ask for obtaining algorithm parameter penalty coefficient c and kernel function spread factor σ, c and σ can be obtained optimum model parameter α among the substitution LSSVM again
iAnd b.
In step 4, to the electrolytic tank under a certain groove state, the algorithm of alumina concentration soft-sensing model parameter is following:
Step 1:
The model parameter initialization; (comprise the quantity n of initialization colony, maximum iteration time k
Max, study factor c
1, study factor c
2, penalty coefficient pace of change lower bound Vc
Min, penalty coefficient pace of change upper bound Vc
Max, the kernel function spread factor changes lower bound V σ
Min, the kernel function spread factor changes lower bound V σ
Max, kernel function spread factor lower bound σ
Min, the kernel function spread factor upper bound
σ max, penalty coefficient lower bound c
Min, penalty coefficient upper bound c
Max, inertia weight ω);
Step 2:
The corresponding algorithm parameter c of all particles in the initialization colony and σ, i.e. initialization particle population, initial population is shone upon to produce through Logistic and is obtained, and concrete grammar is: definition ξ
11, ξ
12Be the random number in (0,1), as get ξ
11, ξ
12=0.25,0.5,0.75 and ξ
11≠ ξ
12With ξ
11, ξ
12Substitution Logistic mapping, that is:
ξ
i+1，1＝4ξ
i1(1-ξ
i1)
ξ
i+1，2＝4ξ
i2(1-ξ
i2)，i＝1，...，m-1，
The following formula iteration is obtained m group Chaos Variable
for m-1 time
Further, Chaos Variable is mapped in the feasible zone, even:
c
i＝c
min+ξ
i1(c
max-c
min)
σ
i＝σ
min+ξ
i2(σ
max-σ
min)，
The definition initial population
X wherein
i=(x
I1, x
I2)=(c
i, σ
i) be i particle position.
(explain: c is a penalty coefficient, and representative model is to the attention degree of outlier in the sample data, and we pay attention to the outlier in the sample data more the big more representative of penalty coefficient c; And the model training error can dullness descend along with the increase of penalty factor c simultaneously, but after c increased to certain value, this range of decrease degree can become very little; Even go to zero, when the slack variable of all outlier with one regularly, fixed c is big more; Loss to objective function is also big more, is just hinting and is unwilling to abandon these outlier this moment, and opposite extreme situations is to be decided to be infinity to c; Like this as long as there is a point to peel off slightly; The value of objective function becomes infinity at once, lets problem become nothing at once and separates, thereby be degenerated to hard interval problem; Kernel function spread factor σ: when spread factor σ is very little, the contact between the support vector loose (distance is less than just there being contact between the support vector of σ), study machine relative complex, extensive popularization ability is relatively poor; Otherwise σ is too big, and the influence between support vector is strong excessively, and regression model is difficult to the precision that reaches enough, is easy to generate and owes match)
Calculate the corresponding optimal adaptation value of all particles in the colony, the individual particles optimal location in the definition colony is:
m is colony's quantity.Definition colony optimal location: the optimal location P of colony
gInitial value be minimum that of adaptive value in the individual optimal location.With m P
iIn the radially basic kernel function substitution formula:
Use least square method and obtain m group model parameter alpha
iAnd b, then with this m group model parameter substitution successively
With
In, the f that tries to achieve is the adaptive value of m all particles correspondences, and the adaptive value of trying to achieve this moment is initialized as the corresponding optimal adaptation value of particle this moment, the optimal adaptation value is meant the difference of two squares of estimating output and actual output of adaptive value under parameter current;
Step 3:
Search colony's optimal adaptation value, i.e. the corresponding optimal adaptation value of all particles in the colony relatively, finding minimum optimal adaptation value is corresponding colony optimal adaptation value, and with particle corresponding algorithm parameter c and σ as optimum model parameter;
Step 4:
Value x and the pace of change v thereof of all particle algorithm parameter c and σ in the renewal colony;
All particle algorithm parameter c and σ are in the feasible zone scope, that is: in judgement and the change colony
If v
I1＞vc
Max, then make v
I1=vc
Max
If v
I1＜vc
Min, then make v
I1=vc
Min
If v
I2＞v σ
Max, then make v
I2=v σ
Max
If v
I2＜v σ
Min, then make v
I2=v σ
Min
If x
I1＜c
Min, then make x
I1=c
Min
If x
I1＞c
Max, then make x
I1=c
Max
If x
I2＜σ
Min, then make x
I2=σ
Min
If x
I2＞σ
Max, then make x
I2=σ
Max, i=1 ..., colony's quantity
Least square method is calculated the model parameter α of each particle under its algorithm parameter c and σ
iAnd b;
Calculate the optimal adaptation value of all particles in the colony, the optimal adaptation value value that all particles are corresponding is the minimum value between the optimal adaptation value that calculates of adaptive value that time iterative computation goes out and last iteration;
Search colony's optimal adaptation value, i.e. the corresponding optimal adaptation value of all particles in the colony relatively, finding minimum optimal adaptation value is corresponding colony optimal adaptation value, and with particle corresponding algorithm parameter c and σ as optimum model parameter;
Step 5:
Judge whether iterations reaches maximal value, if do not reach then return step 4, otherwise execution in step 6;
Step 6:
The algorithm parameter that algorithm parameter c that calculates after the last iteration and σ are decided to be the aluminium oxide soft-sensing model;
Step 7:
The model parameter α that utilization least square method computational algorithm parameter c and σ are corresponding
iAnd b;
Step 5: according to current electrolytic cell state and its corresponding alumina concentration soft-sensing model; The test sample book combination is input in the soft-sensing model; Calculate algorithm that its corresponding alumina concentration discreet value
utilizes following formula with the anti-alumina concentration value that is normalized to of alumina concentration discreet value
; Estimate out the corresponding alumina concentration of current test sample book set
Alumina concentration concentration on-line detecting system is made up of hardware support platform and soft Survey Software; Slot control machine collects a large amount of real-time process manufacturing parameter data in the process control level of hardware platform (as shown in Figure 2); Through the CAN bus these real time datas are write in the database, database offers data processing server with data, in data processing server, accomplishes the processing of data; Computation processes such as the soft measurement diagnosis of alumina concentration; And result of calculation offered carry out man-machine interaction in the management and monitoring machine and show that the management and monitoring machine can offer the decision-making management level with soft measurement result simultaneously, so that decision references is provided for the managerial personnel of higher level.Soft Survey Software realizes flexible measurement method proposed by the invention according to the process manufacturing parameter data that hardware platform collects, and target is exactly to detect the variation of alumina concentration, to realize the optimal control to production run.
Practical implementation is given an example:
Step 1: gather sample
Record of production according to certain aluminium manufacturer workshop; The electrolytic tank operation manufacturing parameter of No. 3326 electrolytic tank 2011-8-1300:00:00-2011-8-1401:00:00 of continuous acquisition electrolysis 3 series, collection period is 10min, totally 150 pairs of samples; Wherein preceding 100 pairs are used for modeling, and back 50 pairs are used for prediction.Electrolytic tank is as shown in Figure 5 at the tank voltage curve of this section in the time; Judge current groove state according to cell parameters curve and combination experience and be slight morbid state, then the aluminium oxide soft-sensing model that trains of the sample data of this moment is the electrolytic tank of this series electrolytic tank correspondence under slight ill groove state.
Set up sample set wherein for
x
i＝{F(k)，H(k)，R(k)，S(k)，T(k)，NB(k)，L(k)，V
s(k)}
y
i＝{C(k)}
x
i∈ R
2, y
i∈ R, k represent residing certain sampling instant of sample i.
The part sample data is as shown in table 1:
Table 1
Step 2: sample data pre-service
Owing to adopt the Euclidean distance of sample data to calculate in the LSSVM algorithm; For avoiding the data domination lesser amt range data of larger amt scope; Adopt following algorithm all to carry out the normalization processing to input data and output data; Zoom to attribute between [0,1], the pretreated sample data of part is as shown in table 2 below:
Table 2
Step 3: determine best parameter
Selected radially basic kernel function
is as the kernel function of aluminium oxide soft-sensing model; After the selected kernel function, adopt alumina concentration soft-sensing model parameter to ask for the algorithm computation model parameter.Initiation parameter is provided with as shown in table 3 below before the algorithm operation:
Table 3
Asking for the optimal parameter that algorithm determines through alumina concentration soft-sensing model parameter is:
Penalty coefficient c=90; Kernel function spread factor σ=0.33.
The alumina concentration soft-sensing model coefficient of this series electrolytic tank that training obtains correspondence under slight ill groove state is as shown in table 4 below:
Table 4
Step 5: model pre-estimating meter
Back 50 pairs of sample datas that are used for predicting are input to the alumina concentration soft-sensing model, calculate its corresponding alumina concentration discreet value
Step 6: the anti-normalization of discreet value
Algorithm below utilizing is with the alumina concentration discreet value
The anti-alumina concentration value that is normalized to, and with the alumina concentration value y of reality
iCompare, the result in the table 5 representes that this alumina concentration soft-sensing model can estimate alumina concentration preferably.
Table 5
Explanation is at last; Above embodiment is only unrestricted in order to technical scheme of the present invention to be described; Although with reference to preferred embodiment the present invention is specified, those of ordinary skill in the art should be appreciated that and can make amendment or be equal to replacement technical scheme of the present invention; And not breaking away from the aim and the scope of present technique scheme, it all should be encompassed in the middle of the claim scope of the present invention.
Claims (5)
1. the flexible measurement method of alumina concentration in the aluminium electrolysis process electrolytic tank, it is characterized in that: said flexible measurement method may further comprise the steps:
Step 1: gatherer process manufacturing parameter data comprise that the process of gathering tank voltage, potline current, reinforced (NB), aluminum yield at interval produces parameter, and set up the historical data base of each parameter values of storage through hardware device;
Step 2: set a time for reading parameter and time for reading section parameter; Every at a distance from one section time for reading parameter; Just read the time for reading section parameter data in the past in this moment of historical data base; The process manufacturing parameter is carried out data processing, set up training sample set and test sample book set, the form of sample set is { x
i, y
i, wherein
x
i={ F (k), H (k), R (k), S (k), T (k), NB (k), L (k), V
s(k) }, y
i={ C (k) }, x
i∈ R
8, y
i∈ R, wherein, the implication of each parameter is following in the formula:
K representes residing certain sampling instant of sample i, constantly the k-moment (k-1)=time for reading parameter;
F (k) is the cell voltage fluctuation number of times, to the time for reading section parameter that collects with all interior tank voltage data-signal V
k, then the signal of (8,0) frequency range is carried out reconstruct and obtain the tank voltage signal V of time for reading section parameter with eight layers of DB8 wavelet decomposition signals to the according to tree construction with interior (8,0) frequency range
80k, statistics time for reading section parameter is with the tank voltage signal wave crest trough number of interior (8,0) frequency range, and a Wave crest and wave trough is designated as 1 secondary undulation, and crest or trough are 0.5 secondary undulation;
H (k) is a high-frequency energy number percent, is designated as V with collecting time for reading section parameter with interior tank voltage
k, calculate tank voltage signal gross energy
Calculate (8,0) frequency range tank voltage signal energy
High-frequency energy number percent then
R (k) is a cell resistance, handles the tank voltage V that obtains (8,0) frequency range constantly through k
80kPotline current I (k) divided by this moment obtains; And the mode of utilizing following smoothing algorithm deletes the exceptional value in the cell resistance sequential value of sampling rate, and wherein, e is the cell resistance value and the deviation of level and smooth calculated value; ω represents shift time; ω=5s, R (k-ω) are the cell resistance value of k 5s before the moment, and R (k-2 ω) is the cell resistance value of k 10s before the moment.
R(k)＝0.3×R(k-ω)-0.1×R(k-2ω)+0.2×(e(k)+2×e(k-ω)+e(k-2ω))；
S (k) is the cell resistance slope, and the increment that is defined as the filter resistance in the nearest time for reading parameter is a rate of change; T (k) is the accumulation slope; Be defined as the accumulation increment of the filter resistance in the in the recent period nearest time for reading parameter; Use the difference filter slope calculations, and use recursion formula calculating cumulative slope, specific algorithm is to be the cycle LPF resistance to be sampled with 30s; By following two formulas difference calculated resistance slope S and accumulation slope T
S(k)＝(R(k-6ω)-R(k-18ω)+2×(R(k)-R(k-24ω)))/5，
T(k)＝(15/16)×T(k-1)+S(k)/4；
NB (k) be alumina blanking at interval, aluminum electrolysis control system actual set blanking spacing value, expression with respect to the benchmark blanking at interval multiple, NB (k) is between 0.8-1.3, when NB (k)=1, expression is with benchmark blanking blanking at interval;
L (k) is an aluminum yield, the quality of the metallic aluminium that the expression aluminium cell extracts when k carries out out the aluminium operation constantly the time, in this method with L (k) as k average aluminum yield constantly, units, promptly
V
s(k) be setting voltage, the expression aluminum electrolytic cell control system is the tank voltage that electrolytic tank is set at moment k, the V of unit;
C (k) is an alumina concentration, and expression is dissolved into the mass percent of the aluminium oxide in the electrolyte;
Step 3: adopt following algorithm that sample set is carried out normalization and handle, zoom to attribute between [0,1]:
In the following formula || || get the 2-norm;
Step 4:, set up alumina concentration soft-sensing model corresponding under the different electrolytic tank states, solving model, computation model parameter according to all electrolytic tank states of storing in the control system historical data base;
Step 5: according to current electrolytic cell state and its corresponding alumina concentration soft-sensing model; The test sample book combination is input in the soft-sensing model; Calculate algorithm that its corresponding alumina concentration discreet value
utilizes following formula with the anti-alumina concentration value that is normalized to of alumina concentration discreet value
; Estimate out the corresponding alumina concentration of current test sample book set
2. the flexible measurement method of alumina concentration in the aluminium electrolysis process electrolytic tank according to claim 1; It is characterized in that: in step 4; The groove state comprises health, health, inferior health, critical health, slight morbid state, ill, the serious morbid state of moderate totally seven kinds of states basically; Training under the different slots state collects sample set; Adopt least square method supporting vector machine (LSSVM) training to obtain the corresponding alumina concentration soft-sensing model of different slots state, totally 7 kinds, the optimal function of employing is following:
In the following formula, ω
T, ω is used for the complexity of controlling models, c is the error penalty coefficient, representative function smoothness and permissible error are greater than the compromise between the numerical value of ε; R
EmpBe empiric risk, promptly about the insensitive loss function of ε; Least square method supporting vector machine selects for use quadratic loss function as the optimization problem loss function, and quadratic loss function is expressed as the quadratic sum of error ξ i, and it is used for measuring the loss that the special parameter selection brings; The definition kernel function
Be the inner product operation of high-dimensional feature space, K (x
i, x
j) satisfy the symmetric function of Mercer condition, optimization problem is converted into to find the solution linear model following:
In the following formula, N is input number of samples, x
iBe i input sample, x is a certain input variable, α
i≠ 0 is pairing input sample x
iBe support vector, α
iBe the SVMs coefficient,
Be model pre-estimating output, K (x, x
i) (LSSVM) be kernel function.
3. the flexible measurement method of alumina concentration in the aluminium electrolysis process electrolytic tank according to claim 2; It is characterized in that: kernel function is RBF (RBF); Being expressed as
σ is the kernel function spread factor, and being needs definite algorithm parameter.
4. the flexible measurement method of alumina concentration in the aluminium electrolysis process electrolytic tank according to claim 1 is characterized in that: in step 4, to the electrolytic tank under a certain groove state, the algorithm of alumina concentration soft-sensing model parameter is following:
Step 1:
The model parameter initialization; Comprise the quantity n of initialization colony, maximum iteration time k
Max, study factor c
1, study factor c
2, penalty coefficient pace of change lower bound Vc
Min, penalty coefficient pace of change upper bound Vc
Max, the kernel function spread factor changes lower bound V σ
Min, the kernel function spread factor changes lower bound V σ
Max, kernel function spread factor lower bound σ
Min, kernel function spread factor upper bound σ
Max, penalty coefficient lower bound c
Min, penalty coefficient upper bound c
Max, inertia weight ω;
Step 2: the corresponding algorithm parameter c of all particles in the initialization colony and σ, i.e. initialization particle population, initial population is shone upon to produce through Logistic and is obtained, and concrete grammar is: definition ξ
11, ξ
12Be the random number in (0,1), and ξ
11≠ ξ
12, with ξ
11, ξ
12Substitution Logistic mapping, that is:
ξ
i+1，1＝4ξ
i1(1-ξ
i1)
ξ
i+1，2＝4ξ
i2(1-ξ
i2)，i＝1，...，m-1，
The following formula iteration is obtained m group Chaos Variable
for m-1 time
Again Chaos Variable is mapped in the feasible zone, even:
c
i＝c
min+ξ
i1(c
max-c
min)
σ
i＝σ
min+ξ
i2(σ
max-σ
min)，，
The definition initial population
X wherein
i=(x
I1, x
I2)=(c
i, σ
i) be i particle position;
Calculate the corresponding optimal adaptation value of all particles in the colony, the individual particles optimal location in the definition colony is:
M is colony's quantity; Definition colony optimal location: the optimal location P of colony
gInitial value be minimum that of adaptive value in the individual optimal location; With m P
iIn the radially basic kernel function substitution formula:
Use least square method and obtain m group model parameter alpha
iAnd b, then with this m group model parameter substitution successively
With
In; The f that tries to achieve is m the adaptive value that all particles are corresponding; The adaptive value of trying to achieve this moment is initialized as the corresponding optimal adaptation value of particle this moment, and the optimal adaptation value value that all particles are corresponding is the minimum value between the optimal adaptation value that calculates of adaptive value that time iterative computation goes out and last iteration;
Step 3:
Search colony's optimal adaptation value, i.e. the corresponding optimal adaptation value of all particles in the colony relatively, finding minimum optimal adaptation value is corresponding colony optimal adaptation value, and with particle corresponding algorithm parameter c and σ as optimum model parameter;
Step 4:
Value x and the pace of change v thereof of all particle algorithm parameter c and σ in the renewal colony;
All particle algorithm parameter c and σ are in the feasible zone scope, that is: in judgement and the change colony
If v
I1＞vc
Max, then make v
I1=vc
Max
If v
I1＜vc
Min, then make v
I1=vc
Min
If v
I2＞v σ
Max, then make v
I2=v σ
Max
If v
I2＜v σ
Min, then make v
I2=v σ
Min
If x
I1＜c
Min, then make x
I1=c
Min
If x
I1＞c
Max, then make x
I1=c
Max
If x
I2＜σ
Min, then make x
I2=σ
Min
If x
I2＞σ
Max, then make x
I2=σ
Max, i=1 ..., colony's quantity,
Utilize least square method to calculate the model parameter α of each particle under its algorithm parameter c and σ
iAnd b;
Calculate the corresponding optimal adaptation value of all particles in the colony, the optimal adaptation value value that all particles are corresponding is the minimum value between the optimal adaptation value that calculates of adaptive value that time iterative computation goes out and last iteration;
Search colony's optimal adaptation value, i.e. the corresponding optimal adaptation value of all particles in the colony relatively, finding minimum optimal adaptation value is corresponding colony optimal adaptation value, and with particle corresponding algorithm parameter c and σ as optimum model parameter;
Step 5:
Judge whether iterations reaches maximal value, if do not reach then return step 4, otherwise execution in step 6;
Step 6:
The algorithm parameter that algorithm parameter c that calculates after the last iteration and σ are decided to be the aluminium oxide soft-sensing model;
Step 7:
The model parameter α that utilization least square method computational algorithm parameter c and σ are corresponding
iAnd b.
5. the flexible measurement method of alumina concentration in the aluminium electrolysis process electrolytic tank according to claim 1 is characterized in that: in step 1, the time for reading parameter is 10min, and time for reading section parameter is 2h.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210131601.7A CN102636624B (en) | 2012-04-28 | 2012-04-28 | Method for soft measurement of alumina concentration in electrolyzer during aluminum electrolysis process |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210131601.7A CN102636624B (en) | 2012-04-28 | 2012-04-28 | Method for soft measurement of alumina concentration in electrolyzer during aluminum electrolysis process |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102636624A true CN102636624A (en) | 2012-08-15 |
| CN102636624B CN102636624B (en) | 2014-07-02 |
Family
ID=46621094
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210131601.7A Expired - Fee Related CN102636624B (en) | 2012-04-28 | 2012-04-28 | Method for soft measurement of alumina concentration in electrolyzer during aluminum electrolysis process |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102636624B (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105301985A (en) * | 2015-11-23 | 2016-02-03 | 浙江中控软件技术有限公司 | Method and system for measuring length of calcium carbide furnace electrode |
| CN107992645A (en) * | 2017-10-30 | 2018-05-04 | 嘉兴学院 | Sewage disposal process soft-measuring modeling method based on chaos-fireworks hybrid algorithm |
| CN108360021A (en) * | 2018-04-10 | 2018-08-03 | 中南大学 | Aluminium cell process control parameters acquisition methods and equipment based on data and knowledge |
| CN108446508A (en) * | 2018-03-27 | 2018-08-24 | 昆明理工大学 | A kind of electrolytic cell load substitutional resistance computational methods based on least square method |
| CN108595458A (en) * | 2017-12-30 | 2018-09-28 | 贵阳铝镁设计研究院有限公司 | Data integration platform based on aluminium electroloysis factory |
| CN108959847A (en) * | 2018-07-20 | 2018-12-07 | 东北大学 | A method of monitoring alumina dissolution rate is calculated by calorimetry |
| CN109933031A (en) * | 2019-03-26 | 2019-06-25 | 沈阳铝镁设计研究院有限公司 | A kind of system and method automatically correcting soft measuring instrument according to analysis data |
| CN110129832A (en) * | 2019-06-21 | 2019-08-16 | 广西大学 | A kind of Multipurpose Optimal Method of aluminium electrolysis process tank voltage |
| CN110282645A (en) * | 2019-06-21 | 2019-09-27 | 广西大学 | A kind of aluminium oxide calcining process operating parameters optimization method |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101275249A (en) * | 2007-12-20 | 2008-10-01 | 中国铝业股份有限公司 | Method for real-time detection of concentration of aluminum oxide in aluminum cell |
| CN101275248A (en) * | 2007-12-20 | 2008-10-01 | 中国铝业股份有限公司 | Method for controlling materials balance of electrolytic cell |
| CN102230192A (en) * | 2011-06-20 | 2011-11-02 | 东北大学 | On-line identification method for concentration signal of aluminum oxide in electrolyte in aluminum electrolysis cell |
| CN102234818A (en) * | 2011-06-21 | 2011-11-09 | 中国铝业股份有限公司 | Constant value control method for alumina concentration of aluminum cell |
-
2012
- 2012-04-28 CN CN201210131601.7A patent/CN102636624B/en not_active Expired - Fee Related
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101275249A (en) * | 2007-12-20 | 2008-10-01 | 中国铝业股份有限公司 | Method for real-time detection of concentration of aluminum oxide in aluminum cell |
| CN101275248A (en) * | 2007-12-20 | 2008-10-01 | 中国铝业股份有限公司 | Method for controlling materials balance of electrolytic cell |
| CN102230192A (en) * | 2011-06-20 | 2011-11-02 | 东北大学 | On-line identification method for concentration signal of aluminum oxide in electrolyte in aluminum electrolysis cell |
| CN102234818A (en) * | 2011-06-21 | 2011-11-09 | 中国铝业股份有限公司 | Constant value control method for alumina concentration of aluminum cell |
Non-Patent Citations (1)
| Title |
|---|
| 阎钢等: "《铝电解槽参数的最小二乘支持向量机软测量模型》", 《小型微型计算机系统》 * |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105301985A (en) * | 2015-11-23 | 2016-02-03 | 浙江中控软件技术有限公司 | Method and system for measuring length of calcium carbide furnace electrode |
| CN107992645A (en) * | 2017-10-30 | 2018-05-04 | 嘉兴学院 | Sewage disposal process soft-measuring modeling method based on chaos-fireworks hybrid algorithm |
| CN108595458A (en) * | 2017-12-30 | 2018-09-28 | 贵阳铝镁设计研究院有限公司 | Data integration platform based on aluminium electroloysis factory |
| CN108446508A (en) * | 2018-03-27 | 2018-08-24 | 昆明理工大学 | A kind of electrolytic cell load substitutional resistance computational methods based on least square method |
| CN108360021A (en) * | 2018-04-10 | 2018-08-03 | 中南大学 | Aluminium cell process control parameters acquisition methods and equipment based on data and knowledge |
| CN108959847A (en) * | 2018-07-20 | 2018-12-07 | 东北大学 | A method of monitoring alumina dissolution rate is calculated by calorimetry |
| CN108959847B (en) * | 2018-07-20 | 2021-08-13 | 东北大学 | Method for calculating and monitoring dissolution rate of alumina by calorimetry |
| CN109933031A (en) * | 2019-03-26 | 2019-06-25 | 沈阳铝镁设计研究院有限公司 | A kind of system and method automatically correcting soft measuring instrument according to analysis data |
| CN109933031B (en) * | 2019-03-26 | 2021-08-31 | 沈阳铝镁设计研究院有限公司 | System and method for automatically correcting soft measuring instrument according to assay data |
| CN110129832A (en) * | 2019-06-21 | 2019-08-16 | 广西大学 | A kind of Multipurpose Optimal Method of aluminium electrolysis process tank voltage |
| CN110282645A (en) * | 2019-06-21 | 2019-09-27 | 广西大学 | A kind of aluminium oxide calcining process operating parameters optimization method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102636624B (en) | 2014-07-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102636624B (en) | Method for soft measurement of alumina concentration in electrolyzer during aluminum electrolysis process | |
| Dong et al. | Wind power day-ahead prediction with cluster analysis of NWP | |
| CN105117602B (en) | A kind of metering device running status method for early warning | |
| CN105300692B (en) | A kind of bearing failure diagnosis and Forecasting Methodology based on expanded Kalman filtration algorithm | |
| CN103135009B (en) | Electric appliance detection method and system based on user feedback information | |
| CN110263866A (en) | A kind of power consumer load setting prediction technique based on deep learning | |
| CN108388962B (en) | Wind power prediction system and method | |
| CN102999791A (en) | Power load forecasting method based on customer segmentation in power industry | |
| Zhang et al. | An anomaly identification model for wind turbine state parameters | |
| CN103679288B (en) | Method for short-term prediction of generated output of runoff small hydropower station cluster and prediction system | |
| CN112857658B (en) | Pressure detection system based on cloud platform | |
| CN106405280B (en) | A kind of intelligent substation on-line monitoring parameter trend method for early warning | |
| Yu et al. | Pump remaining useful life prediction based on multi-source fusion and monotonicity-constrained particle filtering | |
| CN104932898B (en) | A kind of component selection method to be increased based on improvement multi-objective particle | |
| CN106443164A (en) | Transient state event and steady state event monitoring method during electric energy classification metering process | |
| CN110033181B (en) | Power generation equipment state evaluation method based on self-encoder | |
| CN104008433A (en) | Method for predicting medium-and-long-term power loads on basis of Bayes dynamic model | |
| CN111525548B (en) | Method and system for evaluating static voltage stability margin of power grid | |
| Kutschenreiter-Praszkiewicz | Application of artificial neural network for determination of standard time in machining | |
| CN107634544A (en) | The dynamic power control method and system of fired power generating unit | |
| CN104680400B (en) | The short-term and long-range forecast method of enterprise's electricity sales amount variation based on grey correlation | |
| CN107633271B (en) | Method for calculating inaccurate probability of steady-state availability of power system | |
| CN113505909B (en) | Error compensation method for short-term wind power trend prediction | |
| CN111898871B (en) | Method, device and system for evaluating data quality of power grid power supply end | |
| CN115310648A (en) | Medium-and-long-term wind power combination prediction method based on multi-meteorological variable model identification |
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 |
Granted publication date: 20140702 Termination date: 20180428 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |