CN104484714A - Real-time prediction method for catalytic reforming device - Google Patents
Real-time prediction method for catalytic reforming device Download PDFInfo
- Publication number
- CN104484714A CN104484714A CN201410695886.6A CN201410695886A CN104484714A CN 104484714 A CN104484714 A CN 104484714A CN 201410695886 A CN201410695886 A CN 201410695886A CN 104484714 A CN104484714 A CN 104484714A
- Authority
- CN
- China
- Prior art keywords
- model
- real
- data
- yield
- catalytic reforming
- 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
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
Abstract
The invention relates to a real-time prediction method for a catalytic reforming device. On-site real-time data are processed through adopting a data reconciliation technology, and an improved differential evolution algorithm is combined for carrying out real-time correction on a reforming reaction kinetic parameter, so that a mechanism model can precisely describe the actual operation condition of the device. On the basis of the corrected model, the influence on the reforming product yield caused by key operation/process conditions such as the potential aromatic content, the feeding load, the inlet temperature, the reaction pressure and the hydrogen-oil ratio is analyzed. The segmented linearization is carried out according to the influence trends, a linear equation is resolved, corresponding Delta-Base yield data is obtained, a neural network modeling technology is combined for correlating the operation conditions with the Delta-Base data, a yield agency model is built, the yield data calculation speed is accelerated, the continuous yield real-time prediction of the catalytic reforming device is realized, and the theoretical support is provided for building a precise plan optimization PIMS model.
Description
Technical field
The present invention relates to the catalytic reforming unit yield real-time predicting method based on mechanism and operation characteristic, the method may be used for catalytic reforming process modeling and simulating, model real time correction and and PIMS production planning optimization model build in real time.
Background technology
Catalytic reforming (CR) is one of main machining method of modern oil refining and petrochemical process.Reforming technique is under uniform temperature, pressure, hydrogen-oil ratio and catalyzer existent condition, by catalytic dehydrogenation and isomerization, naphtha is transformed into the Reformed Gasoline the process of by-product high-purity hydrogen that are rich in aromatic hydrocarbons (benzene,toluene,xylene is called for short BTX).
Modern reforming catalyst by primary activity component (as platinum), promotor (as rhenium, tin etc.) and acid carrier (as halogen-containing γ-Al
2o
3) composition.The key reaction of catalytic reforming process is the reactions such as naphthenic hydrocarbon dehydrogenation, paraffin dehydrogenation and isomery.Naphtha component is converted into aromatic hydrocarbons and isomeric hydrocarbon by these reactions, is conducive to the octane value improving gasoline.
Catalytic reforming (CR) not only can provide premium, and can the considerable aromatic hydrocarbons of productive value.Benzene, toluene and dimethylbenzene are the base stocks of petroleum chemical industry, are mainly used in the products such as synthon, plastics and rubber.BTX needed for the whole world has nearly 70% from catalytic reforming (CR).Reformation by-product hydrogen is the source of cheap hydrogen, is the important source material of refinery's hydrogenation process.As can be seen here, catalytic reforming (CR) has vital role and position in petrochemical process.
The continuous reforming process flow process simplified as shown in Figure 1.As the naphtha of reformer feed primarily of the alkane of C4 to C12, naphthenic hydrocarbon and aromatic hydrocarbons composition, molecular composition reaches kind more than 300.Course of reaction need purity be 60% ~ 90% hydrogen to enter in reforming reactor together with naphtha feed as recyclegas and react.Reforming reactor is generally designed to the adiabatic reactor of four sections of series connection, equips the heat that a heating furnace consumes to compensate the preceding paragraph reaction between reactor.The discharging of last reactor sends into separation of products device through supercooling.An isolated gas part to mix with raw material as recycle hydrogen and continues to react, and another part hydrogen delivers to other hydrogenation plants, and the liquid phase part feeding piece-rate system of cooling is carried out purification and obtained aromatic hydrocarbon product.
In actual production process, the yield of aromatic hydrocarbons, liquefied gas, dry gas and hydrogen is subject to the impact of raw material and operating conditions, and actual output and the production schedule desired value of critical product often exist certain deviation.Trace it to its cause, key is that current production schedule PIMS model all adopts fixing yield, is difficult to accurate description device actual moving process.Therefore, how to set up an energy accurately, the yield model of quantitative description reformer is the key improving current oil refining apparatus Production planning model accuracy.
The core of accurate yield model is reaction mechanism dynamics accurately.As everyone knows, CONTINUOUS REFORMER reaction comprises the multiple reactions such as naphthenic hydrocarbon dehydrogenation, cycloalkane isomerization dehydrogenation, the dehydrocyclization of alkane, isomerization and hydrocracking, and mechanism is comparatively complicated.At present, for similar complex reaction system, the general method of lump that adopts carries out dynamic analysis.So, reaction core mechanism can be retained, can also reaction network be simplified simultaneously, reduce Chemical kinetic parameter estimation difficulty.In recent years, external many engineering corporatioies and scientific research institution, as Engerhard company, Esso company, Mobil oil company, British Petroleum Company p.l.c., Compaynie Francaise des Petroles, all reforming reaction is conducted in-depth research, propose respective catalytic reforming (CR) lumped watershed hydrologic model model.This patent considers model accuracy and counting yield, and the basis of 27 lumped models is set up reforming reactor mechanism model, for real-time estimate reformer critical product yield.
Summary of the invention
The invention provides a kind of reformer yield real-time predicting method based on mechanism and operation characteristic.
A real-time predicting method for catalytic reforming unit yield, comprises the following steps:
(1) with 27 lumped watershed hydrologic model for theoretical foundation sets up catalytic reforming reaction mechanism model, and set up radial adiabatic reforming reactor model by mathematical method;
(2) utilize .net interfacing to set up data communication between on-the-spot real-time data base and model, realize the plant running data of Real-time Collection catalytic reforming process; Set up data reconciliation standard in conjunction with field condition and knowhow, reject useless and wrong real time data; The manual analysis value (as different lumped component content and arene underwater content etc.) of current oil product is calculated according to raw material data;
(3) according to the real time data of the reformer gathered, mainly comprise feed loading, temperature of reaction, pressure, hydrogen-oil ratio and product information, minimum as optimization aim using the difference of two squares of the model predication value of reactor outlet and actual value, the differential evolution algorithm after improving is utilized to solve, fitted model parameters, realization mechanism model real time correction;
(4) based on the model after correction, for different production decision, analyze the critical product yield under different material quality, feed loading and operating conditions, set up product yield analytical database;
(5) product yield analytical database is utilized to train the neural network agent model that accurately can reflect actual condition.Use agent model operating conditions to be associated with Delta-Base data, realizing the real-time estimate of continuous reformer yield, providing theory support for setting up accurate planning optimization PIMS model.
Real time data described in step 2 is feed naphtha component information, feed loading, reactor inlet temperature, reaction pressure, hydrogen-oil ratio and product component information.
The differential evolution algorithm adopted in step 3 is the differential evolution algorithm with trigonometric mutation, and defines optimization aim and be:
In step 4, catalytic reforming (CR) production decision is divided into two classes: gasoline scheme and aromatic hydrocarbons scheme; Carry out operating characteristic analysis for raw material arene underwater content, device processing capacity, temperature of reaction, pressure and hydrogen-oil ratio, obtain and data are affected on the yield of critical product.
Employ neural network device agent model in step 5 to carry out alternative mechanism model and carry out prediction and calculation, thus meet the requirement calculated in real time for model counting yield.Wherein neural network model adopts backpropagation (Back propagation, BP) neural network, chooses neural network 5 input variables: arene underwater content, feed loading, entrance medial temperature, pressure, hydrogen-oil ratio; Output variable 5: hydrogen, reformation dry gas, reformation liquefied gas, reformation pentane material, reformed oil; Hidden layer is set to 7 layers; Utilize agent model to carry out real-time estimate to plant yield, obtain the Delta-Base value of product yield.
Method of the present invention is based on the catalytic reforming reaction kinetic model of 27 lumps and adiabatic radial reactor model, according to actual industrial data, application enhancements difference algorithm carries out real time correction to mechanism model, and based on this, operating characteristic analysis is carried out for key operation/process conditions, in conjunction with nerual network technique, set up agent model, operating conditions is associated with Delta-Base data, realizing the real-time estimate of continuous catalytic reforming plant yield, providing theory support for setting up accurate planning optimization PIMS model.
Accompanying drawing explanation
Fig. 1 is the continuous catalytic reforming process simplification process flow diagram of embodiment;
Fig. 2 is 27 lumped watershed hydrologic model model reaction networks;
Fig. 3 is the model real time correction simplified flow chart of embodiment;
Fig. 4 is the multilayer feedforward neural network structural representation of embodiment;
Fig. 5 is the neural network device agent model structural drawing of embodiment.
Embodiment
Below, the present invention is described in more detail in conjunction with the embodiments.
Embodiment 1
1. catalytic reforming (CR) modelling by mechanism
Fig. 1 is continuous catalytic reforming process simplification process flow diagram.Catalytic reforming process feed naphtha composition is complicated, and reaction and the subsidiary reaction of generation are numerous.Large quantifier elimination shows, the reaction that catalytic reforming process occurs can be divided into several large class, mainly contains: (1) hexatomic ring dehydrating alkanes generates aromatic hydrocarbons; (2) five-membered ring alkane isomerization generates hexatomic ring alkane; (3) alkane isomerization; (4) dehydrating alkanes cyclisation; (5) hydrocracking of alkane and naphthenic hydrocarbon; (6) demethylation (or claiming hydrogenolysis) of alkane and aromatic hydrocarbons; (7) aromatic hydrocarbons dealkylation; (8) carbon distribution reaction.
For such complex reaction system reaction kinetics modeling, need to use lumped reaction kinetics.So-called lump (Lumping) is exactly each quasi-molecule in complex reaction system is carried out merger by the principle that its dynamics is similar, be divided into several virtual component-lumped components, then sets up the reaction Kinetics Model of lumped component." human relations of lump reason " main results is as shown in table 1.
The present embodiment, in conjunction with actual industrial situation, have selected 27 lumped watershed hydrologic model theories and sets up reforming reaction model, and detailed lump classification is in table 2.
The reaction network mutually transformed between each lumped component is shown in Fig. 2, and for making reaction network figure concisely clear, only list the reaction network of 26 hydro carbons lumped components in fig. 2, hydrogen lump does not show at reaction network.For each reaction, its reaction rate equation is expressed as follows:
1) cyclization of paraffins reaction (reversible reaction)
2) cyclanes aromatization reaction (reversible reaction)
3) cycloalkane isomerization reaction (reversible reaction)
4) cracking reaction (non-reversible reaction)
For the modeling of reactor, suppose that the annulate shaft cross section catalyzer of each reactor, temperature and each concentration of component are evenly distributed, and without air-teturning mixed phenomenon, by desirable piston flow process, then obtain mass balance and heat balance equation is as follows:
2. data acquisition and mediation
Because industry spot situation is complicated, production run is subject to the impact of many factors, and the catalytic reforming (CR) model set up according to the mechanism of Experimental report completely cannot accurate simulation actual device, therefore needs to carry out calibration model in conjunction with field device actual motion characteristic.First be the collection and the reconciliation process process that realize field data.
1) on-site data gathering: in actual production process, most of factory all can use real-time data base to carry out the operation conditions of pen recorder, and provides the item at respective counts strong point so that image data.The present invention utilizes VB.net interfacing to develop catalytic reforming unit field data real-time acquisition system, can realize reading in on-the-spot real time data, and be stored in local data base.The data gathered are needed mainly to comprise base oil properties data, feed loading, reactor inlet temperatures, pressure, hydrogen-oil ratio, the hydrogen flowrate of separation vessel and the material composition data of generation oil.
2) data reconciliation process: by the limitation of Site Detection instrument reliability, often there is the problem such as material imbalance, heat imbalance in the data directly got from DCS, therefore can not be directly used in apparatus for establishing model.In order to ensure the accuracy of model sample data, be necessary to set up mediation standard to the data of Real-time Collection, the following several method of concrete use: (1) adopts day average to carry out calibration model; (2) according to the codomain of statistics and knowhow determination data, judge the accuracy of data according to this, misdata is deleted from local data base; (3) for the data that cannot gather in given period, setting up the computing formula of redundancy, deriving this point by gathering other data.
3) data calculate: the data after mediation, need just can obtain crucial service data through further process, and concrete processing mode is as follows:
Arene underwater content (massfraction)=benzene potential content+toluene potential content+C8 arene underwater content
Benzene potential content (massfraction)=C6 naphthenic hydrocarbon (massfraction) * 78/84+ benzene (massfraction)
Toluene potential content (massfraction)=C7 naphthenic hydrocarbon (massfraction) * 92/98+ toluene (massfraction)
C8 arene underwater content (massfraction)=C8 naphthenic hydrocarbon (massfraction) * 106/112+C8 aromatic hydrocarbons (massfraction)
78,84,92,98,106,112 relative molecular masses being respectively benzene, six naphthenic hydrocarbon, toluene, seven carbocyclic ring alkane, eight ring carbon alkane and eight carbon aromatic hydrocarbons in formula.
3. mechanism model real time correction
This step introduces the real time correction implementation procedure of mechanism model in conjunction with real case.
Model real time correction can be classified as Parameter Estimation Problem, and first Parameter Estimation Problem is converted into optimization problem by the present invention, that is:
Wherein, decision variable X comprises the sensing Summing Factor energy of activation of each reaction,
with
represent the mass percent of each component of product oil respectively.For such optimization aim, the present invention uses the difference algorithm of improvement to solve problem.
Difference algorithm (differential evolution, DE) be a kind of based on population random search algorithm, it has, and structure is simple, fast convergence rate, robustness high.The Variation mechanism of algorithm, the method namely generating filial generation is:
r′=r
1+F*(r
2-r
3) (8)
Wherein, r ' is newly-generated offspring individual, r
1, r
2, r
3be that the parent that three of random selecting in population are different is individual, F is differential evolution operator, is generally a constant.
Because this objective decision variables number is numerous, when causing Algorithm for Solving, calculated amount is very large, therefore needs to improve algorithm, accelerates its speed of convergence.The present invention have selected the improvement difference algorithm with trigonometric mutation, and the method is proved to be has remarkable effect in raising algorithm the convergence speed, and its Mutation Strategy improved can be expressed as:
r′=(r
1+r
2+r
3)/3+(p
2-p
1)(r
1-r
2)+(p
3-p
2)(r
2-r
3)+(p
1-p
3)(r
3-r
1) (9)
Wherein
p
1=|f(r
1)|/p′
p
2=|f(r
2)|/p′ (10)
p
3=|f(r
3)|/p′
P '=| f (r
1) |+| f (r
2) |+| f (r
3) | the simplified flow chart of (11) model real time correction is as shown in Figure 3.
This example chooses the real data of certain factory's continuous catalytic reforming device day as explanation, and the same day, the operation operating mode of device was: reactor pressure 0.35MPag, reactor weighted average temperature in (WAIT) 549 DEG C, trap pressure 0.24MPag, liquid hourly space velocity (LHSV) 1.25/h, hydrogen-oil ratio 0.25mol/mol.According to above-mentioned condition, obtain the model output data after correcting by Algorithm for Solving as shown in table 3, obtain the kinetic parameter of calibration model in table 4 simultaneously.
4. reformate yield analysis
By the realization of above step, mechanism model accurately can reflect the actual conditions of catalytic reforming unit by the correction of real time execution performance data.In this step, for actual production scheme need be divided into two alanysis, be respectively: produce gasoline and produce aromatic hydrocarbons.
The device critical product yield utilizing the mechanism model corrected to obtain for different material characteristic (being mainly arene underwater content), feed loading, reactor inlet temperature, pressure and hydrogen-oil ratio under specific production decision is analyzed, and wherein critical product mainly comprises: hydrogen, reformation dry gas, reformation liquefied gas, reformation pentane material and reformed oil.Set up local data library storage yield data, and the read-write of data can be realized.
Under table 5 and table 6 are respectively reactor inlet temperature and the change of raw material arene underwater content, crucial production yield rate analysis example (aromatics production scheme).Wherein, reactor inlet temperature weighted average inlet temperature (WAIT) represents, its computing formula is as follows:
WAIT=C
1T
1+C
2T
2+C
3T
3+C
4T
4(12)
C
1+C
2+C
3+C
4=1 (13)
Wherein, T
1~ T
4represent the temperature in of each reactor, C
1~ C
4represent the weights (between value 0 ~ 1) of respective reaction actuator temperature, choose relevant with the catalytic amount loaded in reactor.
As shown in table 5, in the process that WAIT raises gradually, the aromatics yield of reformed oil constantly rises, and this contributes to naphthenic hydrocarbon dehydrogenation and cyclization of paraffins dehydrogenation reaction owing to improving temperature, more can generate aromatic hydrocarbons.But also can find out from table, temperature is higher, the amplitude that under unit temperature, aromatics yield improves is less, and this is first that component owing to can obtain aromatic hydrocarbons in reactant reacts completely gradually; Secondly because too high temperature also makes hydrocracking reaction aggravate, produces more gas, liquid product yield decline, this is also the decline of reformation pentane material yield, and the main cause that reform dry gas and reformation yield of liquefied gas constantly rise.Meanwhile, the rising of temperature is also the yield decline of hydrogen, and this point has also showed out from analyzing data.
As shown in table 6, along with raw material arene underwater content increase, the corresponding raising of aromatics yield in reformed oil, device hydrogen yield is also improved.Meanwhile, reformation pentane material, reformation liquefied gas and reformation dry gas yield have reduction in various degree.Visible, in feed naphtha, arene underwater content has important impact for device product yield.
5. product yield real-time estimate
Also very high requirement is proposed to the counting yield of model while continuous catalytic reforming plant yield real-time estimate technical requirement mechanism model has high accuracy.The calibration model mechanism introduced above is very complicated, and solving speed is very slow, and does not easily restrain, and obviously cannot meet the needs of real-time estimate, therefore needs utilize nerual network technique to set up real-time estimate that agent model carrys out implement device yield badly.
The present invention adopts backpropagation (Back propagation, BP) neural network, and this is one most widely used neural network structure in process control.BP neural network structure as shown in Figure 4.Total is made up of L layer neuron, and ground floor is input layer, and last one deck is output layer, and other layer is hidden layer, can obtain each neuronic model to be:
In formula:
Y
pjk: in j layer each neuron of kth under p group sample state and export;
X
pjk: in j layer, the sum functions of a kth neuron under p group sample state exports;
W
pjk: p neuronic link weights in i-th neuron to j layer in j-1 layer.
Wherein w
pjobe defined as a j layer kth neuronic threshold values.
F (y): neuronic nonlinear activation function.
The object of neural network learning finds out a series of weights, after making the often group input vector of sample act on network, the actual output vector of its network is consistent with the desired output vector of sample, whole learning process is the connection weights in adjustment network between each neuron, makes the error energy function of following network reach minimum:
Wherein:
O
pk: after p group sample input vector acts on network, a kth expectation value that neuronic function exports in network output layer.
The BP algorithm problem concerning study solving above-mentioned Multilayer Feedforward Neural Networks, this learning algorithm is made up of signal forward-propagating and error back propagation.Traditional BP algorithm can briefly be summarized as follows.
Wherein:
In formula:
δ
pjk: after p group sample input vector acts on network, the control information that in j layer, a kth neuron function exports;
T: learning time;
F ' (y): the single order derived function of neuron activation functions.
Formula (12) is the learning rules of weights in BP network to (16), formula learning speed and situation term coefficient are generally determined by experience, in traditional BP algorithm they can not with network structure, network state and external learning environment Auto-matching, need people for adjusting during network training.
According to above-mentioned neural network structure, choose neural network and input 5: arene underwater content, feed loading, entrance medial temperature, pressure, hydrogen-oil ratio; Export 5: hydrogen, reformation dry gas, reformation liquefied gas, reformation pentane material, reformed oil; Hidden layer is set to 7 layers.With this neural network training agent model, see Fig. 5.
According to analytical data is bright above, the reformer critical product yield under performance variable impact presents nonlinear trend.The present invention utilizes neural network agent predicts model to calculate, and non-linear object is carried out piece-wise linearization process, produces corresponding Delta-Base data, realizes yield real-time estimate.Result of calculation can be used for setting up PIMS model, improves the precision of model.
Utilize the Delta-Base data of neural network agent model calculation element critical product yield as shown in table 7, its computation process is divided into two steps:
The first step: calculate the prophetic yields of the every critical product of device as Base data according to given arene underwater content and operating conditions;
Second step: according to the free variable of setting, automatically calculate this variable change in set-point neighborhood under product yield predicted data, and calculate the variable quantity that unit variance changes lower yield, this variable quantity is namely as the Delta data of plant yield prediction, and computing formula is as follows:
Wherein, Val represents the free variable (as arene underwater content, reaction pressure etc.) of specifying, Y
keyproductrepresent critical product prophetic yields.
Table 7 take arene underwater content as free variable, respectively for the Delta_Base data that gasoline and aromatic hydrocarbons two schemes obtain, wherein BA represents the Base data of corresponding product yield, free variable 1 represents raw material arene underwater content change 1%, the situation of change of NA corresponding critical product yield under representing variable change, is the Delta data of product yield.The computing formula of critical product yield is as follows:
Y
product=C
product/C
Inlet(20)
Wherein, Y
productrepresent product yield, C
productrepresent the mass rate of product, C
inletrepresent the mass rate of feedstock.
Table 1 " human relations of lump reason " main results
Lump number | Stoichiometric number | Time | Researcher |
3 | 4 | 1959 | Smith |
31 | 78 | 1980 | Jenkins |
28 | 81 | 1987 | Froment |
35 | 36 | 1997 | Taskar |
26 | 48 | 1997 | Padmavathi |
24 | 71 | 2000 | Ancheyta-Juarez |
17 | 17 | 2004 | Hu |
21 | 51 | 2004 | Hu |
20 | 31 | 2006 | Weifeng |
18 | 17 | 2006 | Weifeng |
27 | 52 | 2010 | Hongjun |
38 | 86 | 2012 | Wang |
The detailed component of table 2 27 lumped watershed hydrologic model model
The prediction of table 3 calibration model exports the contrast with real data
composition | actual | predict | difference |
P1 | 0.02% | 0.03% | -0.01% |
P2 | 0.09% | 1.08% | -1.00% |
P3 | 5.40% | 1.28% | 4.12% |
NP4 | 5.09% | 5.36% | -0.27% |
IP5 | 2.35% | 2.35% | 0.00% |
NP5 | 1.42% | 0.00% | 1.42% |
SBP6 | 3.12% | 1.88% | 1.23% |
NP6 | 1.66% | 0.66% | 0.99% |
A6 | 7.07% | 7.11% | -0.04% |
NP7 | 0.19% | 0.11% | 0.07% |
A7 | 19.25% | 19.31% | -0.06% |
EB | 4.37% | 4.39% | -0.02% |
OX | 6.08% | 5.85% | 0.23% |
MX | 9.78% | 9.70% | 0.08% |
PX | 4.61% | 4.47% | 0.14% |
A9 | 29.50% | 29.72% | -0.22% |
Lumped model reactive kinetics parameters after table 4 corrects
The yield of the lower critical product of table 5 temperature in (WAIT) change
The yield of the lower critical product of table 6 raw material arene underwater content change
Table 7 device real-time yield prediction Delta-base value
By the carrying out of above step, the present invention can realize the real-time estimate of the catalytic reforming unit yield based on mechanism and operation characteristic.The method with 27 lumped watershed hydrologic model models for theoretical foundation, actual industrial data are utilized to carry out real time correction to mechanism model, and the analysis data obtained by calculating calibration model carry out neural network training agent model, overcome the limitation that mechanism model computing velocity is slow.Neural network agent model is utilized to calculate the yield of catalytic reforming unit critical product, realize associating of operating conditions and Delta-Base data, reaching the effect of the real-time estimate of continuous catalytic reforming plant yield, providing theory support for setting up accurately planning optimization PIMS model.
Claims (5)
1. a real-time predicting method for catalytic reforming unit yield, is characterized in that, comprises the following steps:
(1) with 27 lumped watershed hydrologic model for theoretical foundation sets up catalytic reforming reaction mechanism model, and set up radial adiabatic reforming reactor model by mathematical method;
(2) utilize .net interfacing to set up data communication between on-the-spot real-time data base and model, realize the plant running data of Real-time Collection catalytic reforming process; Set up valid data mediation standard in conjunction with field condition and knowhow, reject useless and wrong real time data; The characteristic of current oil product is calculated according to raw material data;
(3) according to the real time data of the reformer gathered, minimum as optimization aim using the difference of two squares of the model predication value of reactor outlet and actual value, the differential evolution algorithm after improving is utilized to solve, fitted model parameters, realization mechanism model real time correction;
(4) based on the model after correction, for different production decision, analyze the critical product yield under different material quality, feed loading and operating conditions, set up product yield analytical database;
(5) product yield analytical database is utilized to train the neural network agent model that accurately can reflect actual condition, and operating conditions is associated with Del ta-Base data, realizing the real-time estimate of continuous catalytic reforming plant yield, providing theory support for setting up accurate planning optimization PIMS model.
2. real-time predicting method according to claim 1, is characterized in that, real time data described in step 2 is feed naphtha component information, feed loading, reactor inlet temperature, reaction pressure, hydrogen-oil ratio and product component information.
3. real-time predicting method according to claim 1, is characterized in that, the differential evolution algorithm adopted in step 3 is the differential evolution algorithm with trigonometric mutation, and defines optimization aim and be:
4. real-time predicting method according to claim 1, is characterized in that, in step 4, catalytic reforming (CR) production decision is divided into two classes: gasoline scheme and aromatic hydrocarbons scheme; Carry out operating characteristic analysis for raw material arene underwater content, device processing capacity, temperature of reaction, pressure and hydrogen-oil ratio, obtain and data are affected on the yield of critical product.
5. real-time predicting method according to claim 1, is characterized in that, employs neural network device agent model and carry out alternative mechanism model and carry out prediction and calculation in step 5; Wherein neural network model adopts backpropagation (Back propagation, BP) neural network, chooses neural network 5 input variables: arene underwater content, feed loading, entrance medial temperature, pressure, hydrogen-oil ratio; Output variable 5: hydrogen, reformation dry gas, reformation liquefied gas, reformation pentane material, reformed oil; Hidden layer is set to 7 layers; Utilize agent model to carry out real-time estimate to plant yield, obtain the Delta-Base value of product yield.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410695886.6A CN104484714B (en) | 2014-11-26 | 2014-11-26 | A kind of real-time predicting method of catalytic reforming unit yield |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410695886.6A CN104484714B (en) | 2014-11-26 | 2014-11-26 | A kind of real-time predicting method of catalytic reforming unit yield |
Publications (2)
Publication Number | Publication Date |
---|---|
CN104484714A true CN104484714A (en) | 2015-04-01 |
CN104484714B CN104484714B (en) | 2018-05-11 |
Family
ID=52759255
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201410695886.6A Active CN104484714B (en) | 2014-11-26 | 2014-11-26 | A kind of real-time predicting method of catalytic reforming unit yield |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN104484714B (en) |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106202910A (en) * | 2016-07-07 | 2016-12-07 | 华东理工大学 | The yield real-time predicting method of a kind of residual hydrocracking device and application thereof |
CN108287474A (en) * | 2017-12-27 | 2018-07-17 | 上海交通大学 | Based on the probabilistic catalytic reforming reactor robust operation optimization method of raw material |
CN108846161A (en) * | 2018-05-03 | 2018-11-20 | 华东理工大学 | A kind of modeling method of catalytic gasoline adsorpting desulfurization device |
CN109698014A (en) * | 2018-09-25 | 2019-04-30 | 中国石油天然气股份有限公司 | The prediction technique of catalytic reformate |
CN110556167A (en) * | 2019-09-06 | 2019-12-10 | 北京赛普泰克技术有限公司 | MTO reaction kinetic model, MTO reaction regeneration integration model and application thereof |
CN111475957A (en) * | 2020-04-13 | 2020-07-31 | 华东理工大学 | Oil refining process production plan optimization method based on device mechanism |
CN111899795A (en) * | 2020-06-12 | 2020-11-06 | 中国石油天然气股份有限公司 | Molecular-level oil refining processing full-flow optimization method, device and system and storage medium |
CN113628692A (en) * | 2021-05-31 | 2021-11-09 | 杭州电子科技大学 | Catalytic reforming process modeling method based on mixing of data and mechanism |
EP4089680A1 (en) * | 2020-06-12 | 2022-11-16 | PetroChina Company Limited | Method, apparatus and system for real-time optimization of molecular-level device, and storage medium |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1570035A (en) * | 2004-05-12 | 2005-01-26 | 浙江大学 | Catalytic reforming 17 lump reaction model modelling approach |
CN1655165A (en) * | 2005-01-12 | 2005-08-17 | 浙江中控软件技术有限公司 | Modeling method for residual oil catalytic cracking reaction mechanism model |
US20100160148A1 (en) * | 2008-12-23 | 2010-06-24 | Brian Nabozny | Catalytic Reformer Catalyst Collector Including Purge Gas Stream |
CN102289579A (en) * | 2011-08-03 | 2011-12-21 | 浙江大学 | Modeling method for 38-lumping continuous reforming device reactor |
CN103087753A (en) * | 2011-10-28 | 2013-05-08 | 中国石油化工股份有限公司 | Heavy raw material industrial cracking furnace value maximization model construction method |
-
2014
- 2014-11-26 CN CN201410695886.6A patent/CN104484714B/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1570035A (en) * | 2004-05-12 | 2005-01-26 | 浙江大学 | Catalytic reforming 17 lump reaction model modelling approach |
CN1655165A (en) * | 2005-01-12 | 2005-08-17 | 浙江中控软件技术有限公司 | Modeling method for residual oil catalytic cracking reaction mechanism model |
US20100160148A1 (en) * | 2008-12-23 | 2010-06-24 | Brian Nabozny | Catalytic Reformer Catalyst Collector Including Purge Gas Stream |
CN102289579A (en) * | 2011-08-03 | 2011-12-21 | 浙江大学 | Modeling method for 38-lumping continuous reforming device reactor |
CN103087753A (en) * | 2011-10-28 | 2013-05-08 | 中国石油化工股份有限公司 | Heavy raw material industrial cracking furnace value maximization model construction method |
Non-Patent Citations (3)
Title |
---|
ABOALFAZL ASKARI 等: "Simulation and modeling of catalytic reforming process", 《PETROLEUM & COAL》 * |
侯卫锋 等: "催化重整集总动力学模型的建立及其在线应用", 《化工学报》 * |
周红军 等: "芳烃型催化重整集总反应动力学模型", 《石油学报(石油加工)》 * |
Cited By (18)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106202910A (en) * | 2016-07-07 | 2016-12-07 | 华东理工大学 | The yield real-time predicting method of a kind of residual hydrocracking device and application thereof |
CN108287474A (en) * | 2017-12-27 | 2018-07-17 | 上海交通大学 | Based on the probabilistic catalytic reforming reactor robust operation optimization method of raw material |
CN108287474B (en) * | 2017-12-27 | 2021-11-09 | 上海交通大学 | Robust operation optimization method of catalytic reforming reactor based on uncertainty of raw materials |
CN108846161B (en) * | 2018-05-03 | 2022-10-21 | 华东理工大学 | Modeling method of catalytic gasoline adsorption desulfurization device |
CN108846161A (en) * | 2018-05-03 | 2018-11-20 | 华东理工大学 | A kind of modeling method of catalytic gasoline adsorpting desulfurization device |
CN109698014A (en) * | 2018-09-25 | 2019-04-30 | 中国石油天然气股份有限公司 | The prediction technique of catalytic reformate |
CN109698014B (en) * | 2018-09-25 | 2021-01-01 | 中国石油天然气股份有限公司 | Method for predicting catalytic reformate |
CN110556167A (en) * | 2019-09-06 | 2019-12-10 | 北京赛普泰克技术有限公司 | MTO reaction kinetic model, MTO reaction regeneration integration model and application thereof |
CN110556167B (en) * | 2019-09-06 | 2022-02-25 | 北京赛普泰克技术有限公司 | Construction method of MTO reaction kinetic model and MTO reaction regeneration integration model |
CN111475957A (en) * | 2020-04-13 | 2020-07-31 | 华东理工大学 | Oil refining process production plan optimization method based on device mechanism |
CN111475957B (en) * | 2020-04-13 | 2024-02-02 | 华东理工大学 | Oil refining process production plan optimization method based on device mechanism |
EP4089680A1 (en) * | 2020-06-12 | 2022-11-16 | PetroChina Company Limited | Method, apparatus and system for real-time optimization of molecular-level device, and storage medium |
EP4089680A4 (en) * | 2020-06-12 | 2023-08-09 | PetroChina Company Limited | Method, apparatus and system for real-time optimization of molecular-level device, and storage medium |
EP4089681A4 (en) * | 2020-06-12 | 2023-08-09 | PetroChina Company Limited | Method, apparatus and system for whole-process optimization of molecular-level oil refinery processing, and storage medium |
CN111899795B (en) * | 2020-06-12 | 2023-12-22 | 中国石油天然气股份有限公司 | Molecular-level oil refining processing full-flow optimization method, device and system and storage medium |
CN111899795A (en) * | 2020-06-12 | 2020-11-06 | 中国石油天然气股份有限公司 | Molecular-level oil refining processing full-flow optimization method, device and system and storage medium |
CN113628692A (en) * | 2021-05-31 | 2021-11-09 | 杭州电子科技大学 | Catalytic reforming process modeling method based on mixing of data and mechanism |
CN113628692B (en) * | 2021-05-31 | 2024-01-30 | 杭州电子科技大学 | Modeling method for catalytic reforming process based on data and mechanism mixture |
Also Published As
Publication number | Publication date |
---|---|
CN104484714B (en) | 2018-05-11 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN104484714A (en) | Real-time prediction method for catalytic reforming device | |
CN104789256B (en) | A kind of yield real-time predicting method of catalytic cracking unit | |
CN104765346B (en) | A kind of oil refining process whole process modeling method | |
CN100409232C (en) | Method and system for operating a hydrocarbon production facility | |
CN104804761B (en) | A kind of yield real-time predicting method of hydrocracking unit | |
CN109815616B (en) | Molecular-level hydrogenation device simulation and optimization method | |
CN104392098A (en) | Method for predicting yield of catalytically cracked gasoline | |
CN103914595A (en) | Modeling method of medium-temperature coal tar total-fraction hydrogen cracking lumping kinetic model | |
CN104965967A (en) | Yield real-time prediction method for atmospheric and vacuum distillation unit | |
CN103524284B (en) | Forecasting and optimizing method for ethylene cracking material configuration | |
CN104463343A (en) | Method for predicting catalytic cracking light oil yield | |
CN104765347B (en) | Yield real-time predicting method in a kind of residual oil delayed coking | |
CN111598306B (en) | Method and device for optimizing production plan of oil refinery | |
CN111475957B (en) | Oil refining process production plan optimization method based on device mechanism | |
CN104361153A (en) | Method for predicting coking amount of heavy oil catalytic cracking settler | |
Sadighi et al. | Predictive modeling for an industrial naphtha reforming plant using artificial neural network with recurrent layers | |
Sadighi et al. | Optimizing an industrial scale naphtha catalytic reforming plant using a hybrid artificial neural network and genetic algorithm technique | |
CN106202910A (en) | The yield real-time predicting method of a kind of residual hydrocracking device and application thereof | |
Wei et al. | Dynamic modeling and economic model predictive control with production mode switching for an industrial catalytic naphtha reforming process | |
CN108287474B (en) | Robust operation optimization method of catalytic reforming reactor based on uncertainty of raw materials | |
Polovina et al. | Modeling a reaction section of a commercial continuous catalytic reformer | |
CN115312130A (en) | Mixed modeling method for simulation of yield-increasing catalytic cracking high-added-value products | |
CN115938502A (en) | Chemical product characteristic prediction method and system based on molecular-level reaction mechanism | |
JP7380021B2 (en) | Systems, methods and programs | |
Sadighi et al. | Comparison of kinetic-based and artificial neural network modeling methods for a pilot scale vacuum gas oil hydrocracking reactor |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |