CN108549757A - A kind of reciprocating mixing pump discharge flow rate prediction technique that model selects certainly - Google Patents
A kind of reciprocating mixing pump discharge flow rate prediction technique that model selects certainly Download PDFInfo
- Publication number
- CN108549757A CN108549757A CN201810288855.7A CN201810288855A CN108549757A CN 108549757 A CN108549757 A CN 108549757A CN 201810288855 A CN201810288855 A CN 201810288855A CN 108549757 A CN108549757 A CN 108549757A
- Authority
- CN
- China
- Prior art keywords
- gpr
- prediction
- model
- test sample
- flow rate
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
- G06F18/24155—Bayesian classification
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Probability & Statistics with Applications (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- Control Of Positive-Displacement Pumps (AREA)
Abstract
The invention discloses a kind of models from the reciprocating mixing pump discharge flow rate prediction technique of selection, it includes the following steps：(1) it determines the variable that outputs and inputs of prediction model, collects modeling sample；(2) the local GPR models of reciprocating mixing pump discharge flow rate are established；(3) the weighting GPR models of reciprocating mixing pump discharge flow rate are established；(4) the instant GPR models of reciprocating mixing pump discharge flow rate are established；(5) it is based on prediction probability information, it is automatic to select suitable prediction model for each new input sample point；(6) step (2) to (5) is repeated, it can be from local GPR, weighting GPR and instant GPR models, most suitable prediction model is found for each input sample point under new operating mode, then it obtains under new operating mode, the discharge flow rate curve of reciprocating mixing pump, the present invention is based on limited modeling sample, the modeling and prediction to mixing reciprocating mixing pump discharge flow rate under defeated operating mode are realized, implementation, accuracy height are easy in engineering.
Description
Technical field
The present invention relates to the modeling of reciprocating mixing pump design phase important parameter and prediction techniques, more particularly to a kind of suitable
It closes complicated frequency to become under oil and gas multiphase flow operating mode, reciprocating mixing pump discharge flow rate prediction technique of the model from selection.
Background technology
Reciprocating mixing pump has both the function of pump and compressor, can be effectively increased oil and gas during oil exploitation
Yield.When reciprocating pump conveys incompressible medium, the periodic motion of piston and terminal valve so that pump chamber discharge flow rate
Equal to its rate of volumetric change, sinusoidal variations are presented in the pump chamber discharge flow rate curve of a stroke.Reciprocating pump conveys oil gas etc. can
When compression and incompressible mixing medium, pumps the interior heterogeneous stream frequency generated and become shock loading, holding for terminal valve can be induced
It opens, close hysteresis.Since gas is compressible and the incompressible characteristic of liquid, the unlatching lag of outlet valve will lead to pump chamber
Flow rate is climbed to peak value, and then fluctuation restores to sinusoidal variations, and mutation section is presented on discharge flow rate curve；Outlet valve
Closing lag will produce of short duration and smaller reflux, reflux section is presented on discharge flow rate curve.As it can be seen that oil and gas multiphase flow work
Under condition, zero stream section, mutation section, sinusoidal section and reflux section four is presented in the pump chamber discharge flow rate curve of one stroke of reciprocating pump
Stage.Jumping phenomenon in reciprocating pump flow discharge process is aggravation flow pulsation, induces noise, vibration, reduces pump efficiency
One of the main reasons.Therefore, it studies and interacts between reciprocating pump discharge flow rate characteristic and oil and gas multiphase flow operating mode, it is mixed to instructing
Defeated pump engineering design, it is ensured that its stable operation is of great significance.
Domestic and foreign scholars are based on mechanism model and Fluid Mechanics Computation (CFD) emulation technology, mainly to screw, blade
Formula, the discharge characteristic of gear type mixing pump are studied.As a kind of recently used mixing pump type, for reciprocating oil
The research that gas mixing pump discharge characteristic is carried out is seldom.Only Zhang Shengchang etc. does not consider any volumetric loss by CFD modeling tools,
Several instantaneous flows for mixing and being pumped under defeated operating mode have been obtained, and its flow pulsation characteristic has been studied.Generally speaking, above-mentioned
Do not consider leakage and the energy loss of pump mostly for the research of various mixing pumps, and flow pattern in pump is simplified as homogeneous flow, no
It is enough to describe the flow behavior that frequency complicates in mixing pump.Meanwhile the mechanism model for describing various mixing pumps is complex, mould
The parameters such as valve clearance instantaneous pressure, temperature, discharge coefficient are difficult to obtain from experiment in type, limit its engineer application.It removes
Except this, the division of dynamic grid, the selection of multiphase flow and turbulence model, the CFD modeling process such as User-Defined Functions, all
Height relies on research and the experience of designer.Therefore, it is necessary to establish, a kind of versatility is stronger, and accuracy is higher reciprocating
Mixing pump discharge flow rate model is to adapt to the oil and gas multiphase flow operating mode that complicated frequency becomes.
In recent years, data-driven empirical model, need not be excessive not need the complicated internal phenomena of substantive understanding
Dependence designer experience, have been used for predicting there is the variable of larger measurement delay in nonlinear industrial processes.These advantages
Above-mentioned mechanism and CFD modeling problems can be solved simultaneously, and a kind of new side is provided for the modeling and prediction of multiphase mixing transmission pump discharge flow rate
Method.
Difference is mixed under defeated operating mode, and different process characteristics is presented in discharge flow rate curve；It is same to mix under defeated operating mode, discharge stream
Different local characteristics are also presented in zero stream section, mutation section, sinusoidal section and the reflux section of rate curve, and are mutated the discharge flow rate of section
The characteristics such as quick nonlinear change are presented.In addition, on the market multi-phase flowmeter product few in number exist involve great expense,
It is extremely difficult to improve estimated performance by a large amount of reliable modeling datas are obtained from experiment for the limitations such as measurement delay.Consider this
A little factors are returned the prediction Uncertainty information that (GPR) model provides using Gaussian process, are modeled using ADAPTIVE MIXED
Method, entire complex process characteristic is more completely described.However, finding by literature search, GPR models are reciprocal for predicting
Formula mixing pump discharge flow rate does not have but.
Invention content
The present invention will overcome the disadvantages mentioned above of the prior art, provide a kind of suitable complicated frequency and become under oil and gas multiphase flow operating mode
Reciprocating mixing pump discharge flow rate prediction technique of the model from selection.
The present invention is for insufficient and defect existing for mechanism and CFD modeling process, it is proposed that a kind of to be predicted not using GPR
The adaptive modeling and prediction technique of degree of certainty can be based on limited sample, select suitable prediction automatically for each test point
Model improves the prediction accuracy and the degree of reliability of discharge flow rate.
A kind of model includes the following steps from the reciprocating mixing pump discharge flow rate prediction technique of selection：
(1) input variable and output variable of prediction model are determined, collects modeling sample collection S, and carry out to sample set S
Classification, obtains M sample set, i.e. S=(S1,…,Sm)T, m=1 ..., M；M-th of sample set is expressed asWherein NmIndicate m-th of sample set sample number；
(2) the local GPR models of reciprocating mixing pump discharge flow rate are established
Learning training is individually carried out to above-mentioned each sample set, establishes the local GPR prediction models of discharge flow rate；
It is defined according to GPR, m-th of partial model GPRmOutput be expressed as：
In formula, CmIndicate that covariance matrix, the i-th row jth column element are expressed as：
In formula, xm,idIndicate xm,iD-th of element；I=j, then δm,ij=1, otherwise δm,ij=0；D is training sample point
xm,iInput dimension；θm=[am,0,am,1,vm,0,wm,1,…,wm,d,bm]TIt is model coefficient；
Using formula (1) and formula (2), M part GPR model completes off-line modeling, is defined as GPRm, m=
1,…,M；
To new input test sample setWherein T indicates new input test sample set number, Nt
Indicate the number of samples of t-th of new input test sample set, xt,iIndicate XtI-th of sample point；GPRmTo xt,iPrediction outputAnd varianceIt indicates respectively as follows：
In formula,Indicate new input test sample and training sample
This covariance；km,ti=C (xt,i,xt,i) be new input test sample covariance；
The input test sample point new to one calculates from formula (3) and formula (4) and obtains M based on part GPR moulds
Type on-line prediction information；
(3) the weighting GPR models of reciprocating mixing pump discharge flow rate are established；
Based on Bayesian inference, probability P of offering (GPRm|xt,i), to GPRmEach sample of model and new input set
Point xt,iBetween relationship assessed；P(GPRm|xt,i) calculate it is as follows：
In formula, P (GPRm) and P (xt,i|GPRm) it is prior probability and conditional probability respectively；When there is no process priori
When, P (GPRm|xt,i) be expressed as：
Based on probability analysis method, to new input test sample point xT, iFor, P (GPRm|xT, i) bigger, then GPRmMould
Type is more suitable to predict it；
Merge above-mentioned M part GPR models to xt,iThe probabilistic information of prediction weights the predicted value of GPR modelsAnd its
VarianceIt is expressed as follows：
(4) the instant GPR models of reciprocating mixing pump discharge flow rate are established；
(4.1) from sample set S={ xn, yn }, in n=1 ..., N (N be sample set total sample number), surveyed for new input
Try sample point xt,iSelect suitable Similar moulding sample；Definition
ηt,ni=exp (- | | xn-xt,i| |), n=1 ..., N (9)
Descriptive modelling sample point xnThe x between new input test sample pointt,iSimilarity relation；ηt,niBetween 0 and 1,
Its value is bigger, illustrates that relationship between the two is more similar；Therefore, by the way that suitable threshold value λ is arranged, pass through formula
ηt,ni＞ λ (10)
For new input test sample point xt,iSelect suitable Similar moulding sample set；
(4.2) it is based on the selected Similar moulding sample set of formula (10), using formula (1) and formula (2), is established instant
GPR models；Instant GPR models are calculated to sample point x from formula (3) and formula (4)t,iPredicted valueAnd its variance
(5) it is based on prediction probability information, it is automatic to select suitable prediction model for each new input test sample point；
(5.1) it is each new input test sample point xt,iThe suitable part GPR prediction models of selection；
Based on the probabilistic information that formula (6) provides, there is the model of maximal condition probability (MCP), i.e.,
MCPt,i=maxP (GPRm|xt,i), m=1 ..., M (11)
For new input test sample point xt,iMost suitable part GPR prediction models, corresponding predicted value and its variance difference
It is denoted asWith
(5.2) from part, the instant GPR models of weighted sum, for each new input test sample point xt,iSelection is suitable
Prediction model；
Prediction variance can be used for describing input test sample point xt,iWith the uncertainty of its prediction model；If one is not
Suitable model is to new input test sample xt,iIt is predicted, then corresponding variance yields is with regard to big；It, can be from based on thisWithThe middle model for selecting to have minimum variance (MV), i.e.,
As new input test sample point xt,iMost suitable prediction model；
(6) step (2) to (5) is repeated, is every under new operating mode from local GPR, weighting GPR and instant GPR models
A input test sample point finds most suitable prediction model, then obtains under new operating mode, the discharge stream of reciprocating mixing pump
Rate curve.
A kind of reciprocating mixing pump discharge flow rate prediction technique of the model from selection, it is characterised in that step (5)
In, with the maximal condition probability and prediction variance of proposition, each input test sample point x is describedt,iNot with its prediction model
Degree of certainty, to select suitable model to be predicted for it.
By using above-mentioned technology, compared with prior art, beneficial effects of the present invention are as follows：Mould proposed by the present invention
For type from the reciprocating mixing pump discharge flow rate prediction technique of selection, it is based on limited modeling sample, realizes past under defeated operating mode to mixing
The modeling and prediction of compound mixing pump discharge flow rate are set compared to the complexity of mechanism model modeling process, CFD modeling process pair
The dependence etc. of meter person's experience level, method of the invention are easy implementation, accuracy height in engineering.
Description of the drawings
Fig. 1 is the flow chart of the method for the present invention；
Fig. 2 a~Fig. 2 d are the experimental result and selection model prediction knot certainly of the new input sample collection of one of present invention
The comparison of fruit, wherein Fig. 2 a are new input sample collection part GPR prediction result figures, and Fig. 2 b are new input sample collection weightings
GPR prediction result figures, Fig. 2 c are the instant GPR prediction results figures of new input sample collection, and input sample collection model new Fig. 2 d is certainly
The prediction result figure of selection.
Specific implementation mode
With reference to the accompanying drawings of the specification and the technical solution that further illustrates the present invention of embodiment.
Embodiment
As shown in Figure 1, a kind of model includes the following steps from the reciprocating mixing pump flow rate prediction technique of selection：
(1) input variable and output variable of prediction model are determined, modeling sample is collected；
In the present embodiment, the three cylinder Double-action reciprocatings of influence factor and experiment of reciprocating mixing pump flow rate are considered
The design parameter of formula mixing pump model machine selects inlet pressure Ps(0.2~0.4 MPa), outlet pressure Pd(1.0~3.0MPa), contain
Gas rate β (0.1~0.8), crank angle θ (180 °~discharge process terminates) are input variable；In view of the stream of conventional reciprocating pump
Rate can be obtained by the flow rate of a pump chamber through theoretical formula method, and frequently with different number of pump in practical implementation
Cylinder, without loss of generality, it is output variable to select the flow rate Q of one pump chamber of model machine；
From the experimental data obtained in experimental system under above-mentioned different operating modes；By a stroke, that is, there is identical inlet-pressure
Power, outlet pressure, void fraction, revolution speed, clearance volume, the sample under different crank angles are classified as a sample set.It obtains
15 sample sets be expressed as S=(S1,…,S15)T, wherein 9 are used for being training sample set i.e. (S1,…,S9), it is left 6
It is a to be used for being test sample collection i.e. (S10,…,S15), corresponding operating mode is respectively Ps=0.3,0.35,0.25,0.2,0.4,
0.4MPa, outlet pressure Pd=3.0,3.0,2.5,3.0,1.0,1.0MPa), void fraction β (0.5,0.3,0.4,0.1,0.7,
0.8), crank angle θ (180 °~discharge process terminates)；In order to illustrate the advantage of the method for the invention, 6 test samples
It concentrates, preceding 3 operating modes and training operating mode are more similar, and rear three operating modes are then different；
(2) the local GPR models of reciprocating mixing pump discharge flow rate are established
Based on formula (1) and (2), 9 part GPR models are established offline；
Based on formula (3) and (4), 9 part GPR model on-line prediction information can be obtained respectively, to test sample collection S10
A sample point predicted value and variance；
(3) the weighting GPR models of reciprocating mixing pump discharge flow rate are established
Based on formula (5) to (8), weighting GPR models are established, and obtain S10A sample point predicted value and variance；
(4) the instant GPR models of reciprocating mixing pump discharge flow rate are established.
It is S based on formula (9) and (10)10A sample point select suitable Similar moulding sample set；Based on formula
(1) and formula (2) instant GPR models, are established；Instant GPR models are calculated to the sample using formula (3) and formula (4)
The predicted value and variance of point；
(5) it is based on prediction probability information, it is automatic to select suitable prediction model for each new input sample point；
It is S based on formula (11)10A sample point selection it is most suitable, that is, correspond to the part of maximal condition probability
GPR models；It is S from part, the instant GPR models of weighted sum based on formula (12)10A sample point selection it is most suitable,
The model of i.e. corresponding minimum prediction variance；
(6) step (2) to (5) is repeated, can be test sample collection from local GPR, weighting GPR and instant GPR models
S10Each sample point finds most suitable prediction model, then obtains S10Discharge flow rate curve；
(7) step (2) to (6) is repeated, can obtain the discharge flow rate curve of remaining 5 test sample collection.
6 test sample collection prediction results that this method obtains are compared with experimental result, select mean square error
(abbreviation RMSE) this common index is as evaluation criterion.To the t test sample collection, the definition of RMSE evaluation criterions is such as
Under：
RMSE indexs are nonnegative numbers, and numerical value is smaller, illustrate that prediction effect is better.Comparison result is as shown in table 1.
1 the method for the present invention of table is directed to the estimated performance of test sample collection
As can be known from Table 1, limited modeling sample, the method for the present invention (prediction side selected certainly based on model are based on
Method) can capture well preceding 3 test samples (to training operating mode it is more similar) characteristic information；Also 3 after capable of preferably capturing
The main feature information of test sample (different from training operating mode)；Compared with directly with single part GPR, weighting GPR and instant
GPR prediction models, the method for the present invention can preferably capture the characteristic information of each test sample collection, obtain preferably predictive
Energy.
From test sample collection S10Detailed prediction result (such as attached drawing 2a~2d) is it is found that part GPR prediction models are being mutated
The ascent stage prediction effect of section is preferable, and weighting GPR prediction models prediction effect near the peak value of mutation section is preferable, instant GPR
Prediction model is preferable in zero stream section and sinusoidal section prediction effect, and model preferably combines three kinds of prediction moulds from selection method
The advantage of type, prediction effect are best.Therefore, the method for the present invention utilizes the probabilistic information that GPR is provided, and is automatically each test specimens
This point selects suitable prediction model, can preferably extract the characteristic information in sample, improves whole precision of prediction.
Finally, the modeling data of the 15 kinds of operating modes provided based on experiment completes the on-line prediction of 6 test sample collections only
A few minutes are only used.Under identical calculations resources supplIes, traditional CFD modelings link usually will take two weeks or more, and
The CFD model established might not be accurate, also not necessarily suitable for the test set under new operating mode.
It is therefore proposed that model from select prediction technique can be provided for the discharge flow rate of reciprocating mixing pump relatively accurately
Model and prediction.In addition, its simple and reliable implementation can reduce design complexities, design cost is reduced, saves modeling
Time provides a kind of effective supplementary means for current reciprocating mixing pump design.
Content described in this specification embodiment is only enumerating to the way of realization of inventive concept, guarantor of the invention
Shield range is not construed as being only limitted to the concrete form that embodiment is stated, protection scope of the present invention is also and in this field
Technical staff according to present inventive concept it is conceivable that equivalent technologies mean.
Claims (2)
1. a kind of model includes the following steps from the reciprocating mixing pump discharge flow rate prediction technique of selection：
(1) input variable and output variable of prediction model are determined, modeling sample collection S is collected, and classify to sample set S, obtains
To M sample set, i.e. S=(S1,…,Sm)T, m=1, M；M-th of sample set is expressed as
Wherein NmIndicate m-th of sample set sample number；
(2) the local GPR models of reciprocating mixing pump discharge flow rate are established
Learning training is individually carried out to above-mentioned each sample set, establishes the local GPR prediction models of discharge flow rate；
It is defined according to GPR, m-th of partial model GPRmOutput be expressed as：
In formula, CmIndicate that covariance matrix, the i-th row jth column element are expressed as：
In formula, xm,idIndicate xm,iD-th of element；I=j, then δm,ij=1, otherwise δm,ij=0；D is training sample point xm,i's
Input dimension；θm=[am,0,am,1,vm,0,wm,1,…,wm,d,bm]TIt is model coefficient；
Using formula (1) and formula (2), M part GPR model completes off-line modeling, is defined as GPRm, m=1 ..., M；
To new input test sample setWherein T indicates new input test sample set number, NtIndicate the
The number of samples of t new input test sample sets, xt,iIndicate XtI-th of sample point；GPRmTo xt,iPrediction outputThe side and
DifferenceIt indicates respectively as follows：
In formula,It indicates between new input test sample and training sample
Covariance；km,ti=C (xt,i,xt,i) be new input test sample covariance；
It is online based on part GPR models to calculate acquisition M from formula (3) and formula (4) for the input test sample point new to one
Predictive information；
(3) the weighting GPR models of reciprocating mixing pump discharge flow rate are established；
Based on Bayesian inference, probability P of offering (GPRm|xt,i), to GPRmEach sample point x of model and new input sett,i
Between relationship assessed；P(GPRm|xt,i) calculate it is as follows：
In formula, P (GPRm) and P (xt,i|GPRm) it is prior probability and conditional probability respectively；When there is no process priori, P
(GPRm|xt,i) be expressed as：
Based on probability analysis method, to new input test sample point xt,iFor, P (GPRm|xt,i) bigger, then GPRmModel is got over
Properly it is predicted；
Merge above-mentioned M part GPR models to xt,iThe probabilistic information of prediction weights the predicted value of GPR modelsAnd its varianceIt is expressed as follows：
(4) the instant GPR models of reciprocating mixing pump discharge flow rate are established；
(4.1) from sample set S={ xn,yn, it is new input test sample in n=1 ..., N (N be sample set total sample number)
Point xt,iSelect suitable Similar moulding sample；Definition
ηt,ni=exp (- | | xn-xt,i| |), n=1 ..., N (9)
Descriptive modelling sample point xnThe x between new input test sample pointt,iSimilarity relation；ηt,niBetween 0 and 1, value
It is bigger, illustrate that relationship between the two is more similar；Therefore, by the way that suitable threshold value λ is arranged, pass through formula
ηt,ni＞ λ (10)
For new input test sample point xt,iSelect suitable Similar moulding sample set；
(4.2) instant GPR is established using formula (1) and formula (2) based on the selected Similar moulding sample set of formula (10)
Model；Instant GPR models are calculated to sample point x from formula (3) and formula (4)t,iPredicted valueAnd its variance
(5) it is based on prediction probability information, it is automatic to select suitable prediction model for each new input test sample point；
(5.1) it is each new input test sample point xt,iThe suitable part GPR prediction models of selection；
Based on the probabilistic information that formula (6) provides, there is the model of maximal condition probability (MCP), i.e.,
MCPt,i=maxP (GPRm|xt,i), m=1 ..., M (11)
For new input test sample point xt,iMost suitable part GPR prediction models, corresponding predicted value and its variance are remembered respectively
ForWith
(5.2) from part, the instant GPR models of weighted sum, for each new input test sample point xt,iThe suitable prediction of selection
Model；
Prediction variance can be used for describing input test sample point xt,iWith the uncertainty of its prediction model；If one improper
Model to new input test sample xt,iIt is predicted, then corresponding variance yields is with regard to big；It, can be from based on this
WithThe middle model for selecting to have minimum variance (MV), i.e.,
As new input test sample point xt,iMost suitable prediction model；
(6) step (2) to (5) is repeated, from local GPR, weighting GPR and instant GPR models, each to be inputted under new operating mode
Test sample point finds most suitable prediction model, then obtains under new operating mode, the discharge flow rate curve of reciprocating mixing pump.
2. the reciprocating mixing pump discharge flow rate prediction technique that a kind of model according to claim 1 selects certainly, feature
It is in step (5), with the maximal condition probability and prediction variance of proposition, describes each input test sample point xt,iIt is pre- with its
The uncertainty for surveying model, to select suitable model to be predicted for it.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810288855.7A CN108549757B (en) | 2018-04-03 | 2018-04-03 | Model self-selection reciprocating type mixed transportation pump discharge flow rate prediction method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810288855.7A CN108549757B (en) | 2018-04-03 | 2018-04-03 | Model self-selection reciprocating type mixed transportation pump discharge flow rate prediction method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108549757A true CN108549757A (en) | 2018-09-18 |
| CN108549757B CN108549757B (en) | 2021-10-26 |
Family
ID=63513887
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810288855.7A Active CN108549757B (en) | 2018-04-03 | 2018-04-03 | Model self-selection reciprocating type mixed transportation pump discharge flow rate prediction method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108549757B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109446741A (en) * | 2018-12-21 | 2019-03-08 | 浙江工业大学 | A kind of modeling and prediction technique of mixing pump pump chamber transient temperature characteristic |
| CN113553673A (en) * | 2021-07-21 | 2021-10-26 | 浙江工业大学 | Centrifugal pump efficiency prediction method based on data-driven modeling |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20090150126A1 (en) * | 2007-12-10 | 2009-06-11 | Yahoo! Inc. | System and method for sparse gaussian process regression using predictive measures |
| CN103884075A (en) * | 2014-01-06 | 2014-06-25 | 浙江工业大学 | Computational fluid dynamics and energy prediction hybrid based greenhouse energy-saving control method |
| CN104699894A (en) * | 2015-01-26 | 2015-06-10 | 江南大学 | JITL (just-in-time learning) based multi-model fusion modeling method adopting GPR (Gaussian process regression) |
| CN104778298A (en) * | 2015-01-26 | 2015-07-15 | 江南大学 | Gaussian process regression soft measurement modeling method based on EGMM (Error Gaussian Mixture Model) |
| CN105913078A (en) * | 2016-04-07 | 2016-08-31 | 江南大学 | Multi-mode soft measurement method for improving adaptive affine propagation clustering |
| CN106056127A (en) * | 2016-04-07 | 2016-10-26 | 江南大学 | GPR (gaussian process regression) online soft measurement method with model updating |
| WO2017188501A1 (en) * | 2016-04-29 | 2017-11-02 | 경희대학교 산학협력단 | Method for recovering original signal in reduced complexity ds-cdma system |
| CN107451101A (en) * | 2017-07-21 | 2017-12-08 | 江南大学 | It is a kind of to be layered integrated Gaussian process recurrence soft-measuring modeling method |
-
2018
- 2018-04-03 CN CN201810288855.7A patent/CN108549757B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20090150126A1 (en) * | 2007-12-10 | 2009-06-11 | Yahoo! Inc. | System and method for sparse gaussian process regression using predictive measures |
| CN103884075A (en) * | 2014-01-06 | 2014-06-25 | 浙江工业大学 | Computational fluid dynamics and energy prediction hybrid based greenhouse energy-saving control method |
| CN104699894A (en) * | 2015-01-26 | 2015-06-10 | 江南大学 | JITL (just-in-time learning) based multi-model fusion modeling method adopting GPR (Gaussian process regression) |
| CN104778298A (en) * | 2015-01-26 | 2015-07-15 | 江南大学 | Gaussian process regression soft measurement modeling method based on EGMM (Error Gaussian Mixture Model) |
| CN105913078A (en) * | 2016-04-07 | 2016-08-31 | 江南大学 | Multi-mode soft measurement method for improving adaptive affine propagation clustering |
| CN106056127A (en) * | 2016-04-07 | 2016-10-26 | 江南大学 | GPR (gaussian process regression) online soft measurement method with model updating |
| WO2017188501A1 (en) * | 2016-04-29 | 2017-11-02 | 경희대학교 산학협력단 | Method for recovering original signal in reduced complexity ds-cdma system |
| CN107451101A (en) * | 2017-07-21 | 2017-12-08 | 江南大学 | It is a kind of to be layered integrated Gaussian process recurrence soft-measuring modeling method |
Non-Patent Citations (2)
| Title |
|---|
| YI LIU 等: "Auto-Switch Gaussian Process Regression-Based Probabilistic Soft Sensors for Industrial Multigrade Processes with Transitions", 《INDUSTRIAL & ENGINEERING CHEMISTRY RESEARCH》 * |
| 袁小锋: "基于即时学习的复杂非线性过程软测量建模及应用", 《中国博士学位论文全文数据库 信息科技辑》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109446741A (en) * | 2018-12-21 | 2019-03-08 | 浙江工业大学 | A kind of modeling and prediction technique of mixing pump pump chamber transient temperature characteristic |
| CN109446741B (en) * | 2018-12-21 | 2023-04-18 | 浙江工业大学 | Modeling and predicting method for instantaneous temperature characteristic of pump cavity of mixed transportation pump |
| CN113553673A (en) * | 2021-07-21 | 2021-10-26 | 浙江工业大学 | Centrifugal pump efficiency prediction method based on data-driven modeling |
| CN113553673B (en) * | 2021-07-21 | 2023-03-21 | 浙江工业大学 | Centrifugal pump efficiency prediction method based on data-driven modeling |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108549757B (en) | 2021-10-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109740661B (en) | Oil well moving liquid level self-adaptive prediction method based on fuzzy evaluation | |
| CN107918709B (en) | Prediction method for transient opening height of one-way valve for multiphase mixed transportation pump | |
| CN107288617B (en) | Method and system for improving oil measuring precision of indicator diagram of pumping well | |
| KR101975436B1 (en) | Apparatus and method for forecasting production for shale gas well in transient flow using machine learning method | |
| Matthews et al. | Fitting and comparing competing models of the species abundance distribution: assessment and prospect | |
| CN105069093B (en) | A kind of Hydrological Time Series Similarity searching method based on embedded index | |
| CN108805215B (en) | Dynamic liquid level soft measurement method for sucker-rod pump pumping well based on improved drosophila algorithm | |
| CN109630095B (en) | A kind of rod-pumped well operating mode's switch method and system based on multi-angle of view study | |
| CN106951662B (en) | Method for diagnosing underground working condition of sucker-rod pump oil well based on valve working point | |
| CN104807589B (en) | A kind of ONLINE RECOGNITION method collecting flow pattern of gas-liquid two-phase flow in defeated-riser systems | |
| CN108549757A (en) | A kind of reciprocating mixing pump discharge flow rate prediction technique that model selects certainly | |
| CN108510120A (en) | A kind of hybrid modeling and prediction technique of reciprocating mixing pump discharge flow rate | |
| CN110458195A (en) | A kind of abnormal data discriminating method based on the fusion of more criterions | |
| CN106677763A (en) | Oil well dynamic liquid level prediction method based on dynamic integrated modeling | |
| CN109446741B (en) | Modeling and predicting method for instantaneous temperature characteristic of pump cavity of mixed transportation pump | |
| CN106022352A (en) | Submersible piston pump fault diagnosis method based on support vector machine | |
| Chen et al. | You only search once: A fast automation framework for single-stage dnn/accelerator co-design | |
| CN113627068A (en) | Method and system for predicting well testing productivity of fracture-cavity type oil and gas reservoir | |
| CN108399293A (en) | A kind of circulation system design with one-dimensional algorithm slitless connection analyzes multifunctional platform | |
| CN108694293A (en) | A kind of Active Modeling and prediction technique of reciprocating mixing pump flow rate | |
| CN116821828A (en) | Multi-dimensional time sequence prediction method based on industrial data | |
| CN103927438B (en) | Near-infrared wavelength Variable Selection based on successive projection algorithm | |
| Wu et al. | Study of Valve Motion in Reciprocating Refrigerator Compressors based on the 3-D Fluid–structure Interaction Model | |
| Xiaoxiao et al. | A self-evolution fault diagnosis method of sucker rod pump based on simulating dynamometer card | |
| CN114510880B (en) | Method for diagnosing working condition of sucker-rod pump based on Fourier transform and geometric characteristics |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |