CN108956814A - A kind of method and property prediction technique of direct construction petrol molecule composition model - Google Patents
A kind of method and property prediction technique of direct construction petrol molecule composition model Download PDFInfo
- Publication number
- CN108956814A CN108956814A CN201810717484.XA CN201810717484A CN108956814A CN 108956814 A CN108956814 A CN 108956814A CN 201810717484 A CN201810717484 A CN 201810717484A CN 108956814 A CN108956814 A CN 108956814A
- Authority
- CN
- China
- Prior art keywords
- peak
- molecule
- gasoline
- property
- composition
- 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
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N30/00—Investigating or analysing materials by separation into components using adsorption, absorption or similar phenomena or using ion-exchange, e.g. chromatography or field flow fractionation
- G01N30/02—Column chromatography
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N30/00—Investigating or analysing materials by separation into components using adsorption, absorption or similar phenomena or using ion-exchange, e.g. chromatography or field flow fractionation
- G01N30/02—Column chromatography
- G01N30/62—Detectors specially adapted therefor
- G01N30/64—Electrical detectors
- G01N30/68—Flame ionisation detectors
Landscapes
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Investigating Or Analyzing Non-Biological Materials By The Use Of Chemical Means (AREA)
- Production Of Liquid Hydrocarbon Mixture For Refining Petroleum (AREA)
Abstract
The present invention provides a kind of method of direct construction petrol molecule composition model and property prediction techniques, construction method is the following steps are included: (1) carries out detailed hydrocarbon analysis to the gas chromatographic detection result of gasoline sample, to identify each peak may include in gas chromatogram molecule, and calculate the opposite point rate at each peak；(2) according to detailed hydrocarbon analysis as a result, and sorting out chromatographic peak by chromatographic peak type；(3) it to the chromatographic peak after classification, is analyzed according to chromatographic peak type and molecule type, obtains complete detailed hydrocarbon molecular composition result；(4) composition model of petrol molecule is directly generated by the complete detailed hydrocarbon molecular composition result.This method provided by the present invention is based on gas chromatographic detection result, algorithm, which is adjusted, with the peak based on statistical distribution rebuilds molecular composition, then it establishes petrol molecule composition model and predicts the macroscopic property of gasoline, this method, which can process and reconcile for gasoline, provides accurate data support.
Description
Technical field
The present invention relates to a kind of method of direct construction petrol molecule composition model and property prediction techniques, and in particular to
A kind of method and property prediction technique by gas chromatograph results direct construction petrol molecule composition model belongs to oil product composition
Analysis technical field.
Background technique
The processing of gasoline and the requirement reconciled to its quality and composition are very harsh, and traditional lumped model is increasingly
It is unable to satisfy the demand of a modernization refinery.In order to improve the benefit of refinery and improve target level of product quality, molecular level is developed
Model, processing and harmonic process to light-end products are implemented the other management of molecular level and are just become more and more important.And establish molecule
Grade model firstly the need of solves the problems, such as be acquisition oil product molecular composition.
The existing method for obtaining petrol molecule composition can be divided mainly into experimental method and with computer-assisted rebuilding molecular group
At method.Traditional analysis instrument such as gas-chromatography-hydrogen flame ionization detector (GC-FID), analyze result in comprising compared with
More total evolution peaks and fubaritic peak.The method of existing computer-assisted rebuilding petrol molecule composition is usually by gasoline
The problem of macroscopic property of sample is back-calculated to obtain its molecular composition, and calculated result does not have total evolution and result missing.But this
Class method can have uncertainty, impact to subsequent processing and reconciliation simulation process.Furthermore this method obtains
To gasoline property predicted value influenced by the experiment value inputted very big, and the also heavy dependence of the Part Methods in such methods point
The method of the quality and training data of son composition database.
Therefore it provides the method and property prediction technique of a kind of direct construction petrol molecule composition model have become this
The technical issues of field urgent need to resolve.
Summary of the invention
In order to solve the above shortcomings and deficiencies, the purpose of the present invention is to provide a kind of direct construction petrol molecule compositions
The method of model.
The object of the invention is also to provide a kind of systems of direct construction petrol molecule composition model.
The object of the invention is also to provide a kind of method and system for predicting gasoline macroscopic property.It is provided by the present invention
Technical solution adjusts algorithm using the peak based on statistical distribution and has rebuild the resulting molecular results of gas-chromatography, is completely divided
Son composition, for establishing petrol molecule composition model, and predicts its property.This method combines experimental method and computer reconstruction
Method, total evolution peak and the non-diagnostic peak in gas chromatograph results are split or speculated according to certain statistical distribution, is provided
Accurate, stable molecular composition overcomes existing method result it is not necessary to establish molecular composition database and be associated with training
Defect.Based on the macroscopic property of the prediction of molecular composition obtained by this method, is not influenced by experiment value, more there is reference value.
In order to achieve the above objectives, on the one hand, the present invention provides a kind of method of direct construction petrol molecule composition model,
In, it the described method comprises the following steps:
(1) detailed hydrocarbon analysis is carried out to the gas chromatographic detection result of gasoline sample, to identify each peak in gas chromatogram
The molecule that may include, and calculate the opposite point rate at each peak；
(2) according to detailed hydrocarbon analysis as a result, and sorting out chromatographic peak by chromatographic peak type；
(3) it to the chromatographic peak after classification, is analyzed according to chromatographic peak type and molecule type, obtains complete detailed hydrocarbon
Molecular composition result；
(4) composition model of petrol molecule is directly generated by the complete detailed hydrocarbon molecular composition result.
Method according to the present invention, it is preferable that the gasoline sample includes catalytically cracked gasoline, catalytic reforming vapour
Oil, direct steaming gasoline, catalytic cracking gasoline, hydrogasoline or coker gasoline.
Method according to the present invention, it is preferable that the chromatographic peak type includes known peak, escapes peak altogether and do not identify
Peak.
Method according to the present invention, it is preferable that analysis described in step (3) is using the peak tune based on statistical distribution
Algorithm is saved to the chromatographic peak after classification, is analyzed according to chromatographic peak type and molecule type.
Method according to the present invention, it is preferable that the peak adjusting algorithm based on statistical distribution specifically includes following
Step:
It is fitted the distribution at known peak: known peak being sorted out by molecule type and carbon number, each series data after classification is pressed
Statistical distribution fitting；
Altogether escape peak fractionation: total evolution peak is split, it is assumed that altogether evolution peak in each component relative amount, and according to
It is secondary to examine all it is assumed that choosing suitable hypothesis receiving；Above procedure is constantly repeated, until all total evolution peaks have all been split
At；
Infer non-diagnostic peak: assuming that the molecule type and carbon number of component included in non-diagnostic peak, and successively examine all
Receive it is assumed that choosing suitable hypothesis；Above procedure is constantly repeated, until completion is all inferred at all non-diagnostic peak peaks.
Method according to the present invention, wherein be fitted the distribution at known peak specifically includes the following steps: known peak is pressed
Molecule type and carbon number are sorted out, and sum to the content point rate of all molecules of identical molecule type and carbon number, obtain one
A data matrix；The data to molecular series each in data matrix relative to carbon number are fitted by statistical distribution again.
Method according to the present invention, it is preferable that the statistical distribution includes gamma distribution.
Method according to the present invention, it is preferable that molecule type described in step (3) includes n-alkane (NP), different
Structure alkane (IP), alkene (O), cycloalkane (NC) and aromatic hydrocarbons (A).Wherein, the isoparaffin further includes single branched chain isomer alkane
(MP), double branched chain isomer alkane (DP), three branched chain isomer alkane (TP)；The alkene further includes linear alkene (NO), isomery alkene
Hydrocarbon (BO).
Method according to the present invention, it is preferable that step directly generates petrol molecule composition model described in (4),
It include: to establish a molecular objects respectively after obtaining the complete detailed hydrocarbon molecular composition result for wherein each molecule, it should
Molecular objects are for executing inquiry molecular property and molecule content operation；
Gasoline object is resettled, which includes the molecular objects, and the gasoline object calculates vapour for executing
Oily macroscopic property operation.
On the other hand, the present invention also provides a kind of systems of direct construction petrol molecule composition model, wherein the system
System includes:
First unit, the first unit are used to carry out detailed hydrocarbon analysis to the gas chromatographic detection result of gasoline sample,
To identify each peak may include in gas chromatogram molecule, and calculate the opposite point rate at each peak；
Second unit, the second unit are used for according to detailed hydrocarbon analysis as a result, and returning chromatographic peak by chromatographic peak type
Class；
Third unit, the third unit are used for the chromatographic peak after classification, according to chromatographic peak type and molecule type into
Row analysis, obtains complete detailed hydrocarbon molecular composition result；
Unit the 4th, Unit the 4th are used to directly generate gasoline point by the complete detailed hydrocarbon molecular composition result
The composition model of son.
System according to the present invention, it is preferable that in third unit, the analysis is using the peak based on statistical distribution
Adjustment module analyzes the chromatographic peak after classification according to chromatographic peak type and molecule type.
System according to the present invention, it is preferable that the peak adjustment module based on statistical distribution specifically includes:
First module, first module are used to be fitted the distribution at known peak: known peak is returned by molecule type and carbon number
Class is fitted each series data after classification by statistical distribution；
Second module, second module for escaping the fractionation at peak altogether: total evolution peak is split, it is assumed that evolution altogether
The relative amount of each component in peak, and successively examine all it is assumed that choosing suitable hypothesis receiving；Above procedure is constantly repeated,
Until all total evolution peaks all split completion；
Third module, the third module is for inferring non-diagnostic peak: assuming that the molecule of component included in non-diagnostic peak
Type and carbon number, and successively examine all it is assumed that choosing suitable hypothesis receiving；Above procedure is constantly repeated, until all
All infer completion in non-diagnostic peak peak.
Another aspect, the present invention also provides it is a kind of predict gasoline macroscopic property method, wherein the method includes with
Lower step:
The composition model of petrol molecule is constructed, according to the method for the direct construction petrol molecule composition model to obtain
The property of the molecular composition of gasoline and each molecule；
According to the property of the molecular composition of gasoline and each molecule by corresponding macroscopic property mixing rule to gasoline
Macroscopic property is predicted.
The method of prediction gasoline macroscopic property according to the present invention, it is preferable that the macroscopic property includes density, evaporates
Journey, octane number, refractive index, molecular weight and Reid vapour pressure.
In another aspect, the present invention also provides a kind of systems for predicting gasoline macroscopic property, wherein the system comprises:
First unit, the first unit are used to be constructed according to the method for the direct construction petrol molecule composition model
The composition model of petrol molecule, to obtain the molecular composition of gasoline and the property of each molecule；
Second unit, the second unit are used to be passed through according to the molecular composition of gasoline and the property of each molecule corresponding
Macroscopic property mixing rule predicts the macroscopic property of gasoline.
This method provided by the present invention is adjusted with the peak based on statistical distribution and is calculated based on gas chromatographic detection result
Method (SPT algorithm) rebuilds molecular composition, then establishes petrol molecule composition model and predicts the macroscopic property of gasoline, this method can
Accurate data support is provided to process and reconcile for gasoline.
This method provided by the present invention compared with the conventional method, has the advantage that
1, this method provided by the present invention combines the advantages of experimental method and computer reconstruction method, has provided
Whole, stable molecular composition result；
2, this method does not need the wave spectrum of measurement great amount of samples and physical property is associated training, and workload is small, cost
It is cheap, it uses manpower and material resources sparingly；
3, this method can directly predict the property of gasoline from molecular composition, not influenced by macroscopic property experiment value, more
There is reference value；
4, the spectrogram fine tuning algorithm that this method provides is directly based upon statistical distribution, does not need macroscopic property and participates in amendment,
It only needs to can be realized using gas chromatograph results.
Detailed description of the invention
Fig. 1 is the method and property prediction of direct construction petrol molecule composition model provided by the embodiment of the present invention 1
The flow diagram of method；
Fig. 2 is that (known peak escapes peak and not altogether for the gas chromatogram of gasoline and three types peak in the embodiment of the present invention 1
Diagnostic peak) example；
Fig. 3 is the flow chart of SPT algorithm steps one；
Fig. 4 is the flow chart of SPT algorithm steps two；
Fig. 5 is the flow chart of SPT algorithm steps three；
Fig. 6 is normal alkane series (NP) molecule content SPT algorithm process result exemplary diagram in the embodiment of the present invention 1；
Fig. 7 is single serial (MP) molecule content SPT algorithm process result exemplary diagram of branched paraffin in the embodiment of the present invention 1；
Fig. 8 is double serial (DP) molecule content SPT algorithm process result exemplary diagrams of branched paraffin in the embodiment of the present invention 1；
Fig. 9 is serial (TP) molecule content SPT algorithm process result exemplary diagram of three branched paraffins in the embodiment of the present invention 1；
Figure 10 is serial (NO) molecule content SPT algorithm process result exemplary diagram of linear alkene in the embodiment of the present invention 1；
Figure 11 is serial (BO) molecule content SPT algorithm process result exemplary diagram of branched-chain alkene in the embodiment of the present invention 1；
Figure 12 is the embodiment of the present invention 1 middle ring alkane series (NC) molecule content SPT algorithm process result exemplary diagram；
Figure 13 is serial (A) molecule content SPT algorithm process result exemplary diagram of aromatic hydrocarbons in the embodiment of the present invention 1；
Figure 14 be the embodiment of the present invention 1 in gained gasoline sample property predicted value and gasoline sample property experiment value it
Between comparison diagram.
Specific embodiment
In order to which technical characteristic of the invention, purpose and beneficial effect are more clearly understood, now in conjunction in detail below
Embodiment carries out following detailed description to technical solution of the present invention, but should not be understood as the limit to enforceable range of the invention
It is fixed.
Embodiment 1
Present embodiments provide a kind of method and vapour by gas chromatograph results direct construction petrol molecule composition model
The prediction technique of oily macroscopic property, the flow diagram of this method is as shown in Figure 1, from figure 1 it appears that it includes following step
It is rapid:
To catalytically cracked gasoline sample carry out gas chromatographic detection, then to gained testing result carry out detailed hydrocarbon analysis (in
Magnificent people's republic's petrochemical industry standard SH/T 0714-2002), to identify, each peak may include in gas chromatogram
Molecule, and calculate the opposite point rate at each peak；
Wherein, gas chromatographic detection described in the present embodiment uses the Agilent 7890B gas phase of U.S. Agilent company
Chromatograph carries out, which is equipped with flame ionization ditector (FID).Wherein, chromatographic column is that PONA analysis is dedicated
Fused-silica capillary column, stationary phase be 100% methyl silicone, column length 50m, internal diameter 0.2mm, thickness of liquid film be 0.2 μ
m；Pressure is 86KPa before column；The initial temperature of chromatography temperature program is 35 DEG C, is kept for 5 minutes, and heating rate is 2 DEG C/min, finally
Temperature is 200 DEG C, and the final temperature residence time is 10 minutes；Sample injector temperature is 250 DEG C, split ratio 150:1, and sample volume is 0.5 μ
L；Detector temperature is 250 DEG C, and combustion gas is hydrogen, and flow velocity 35mL/min, combustion-supporting gas is air, and flow velocity 350mL/min is mended
Repaying gas is nitrogen, flow velocity 35mL/min；Carrier gas is nitrogen, and average linear speed is 12cm/s, the gas of the catalytically cracked gasoline sample
Phase chromatogram is as shown in Figure 2.
(2) according to detailed hydrocarbon analysis as a result, and chromatographic peak is sorted out by chromatographic peak type, be divided into known peak, escape peak altogether
And non-diagnostic peak；The example of all types of chromatographic peaks is as shown in Figure 2.
(3) algorithm (SPT) is adjusted to the chromatographic peak after classification using peak be distributed based on gamma, according to chromatographic peak type with
Molecule type is analyzed, and complete detailed hydrocarbon molecular composition result is obtained, wherein the peak based on gamma distribution adjusts algorithm
Specific flow chart is as in Figure 3-5, and the complete detailed hydrocarbon molecular composition result is as shown in figs. 6-13；
In step (3), the molecule type includes n-alkane (NP), isoparaffin, alkene, cycloalkane (NC) and aromatic hydrocarbons
(A)；Wherein, the isoparaffin includes single branched chain isomer alkane (MP), double branched chain isomer alkane (DP), three branched chain isomer alkane
(TP)；The alkene further includes linear alkene (NO), isomeric olefine (branched-chain alkene, BO).
Sorted detailed hydrocarbon (known peak) is handled by Fig. 3 step 1, obtains data matrix.Each series in data matrix
Content under carbon number and corresponding carbon number divides rate, is the data for fitting.It is distributed by gamma respectively to the number of each series in this example
According to being fitted, the parameter and root-mean-square error of each Series Molecules content distribution can be obtained.
It is split by Fig. 4 step 2 and escapes peak altogether: assuming initially that the relative amount of total evolution the included molecule in peak.Such as it escapes altogether
Include two kinds of components of molecule A and molecule B in appearance, might as well assume that in the relative amount for escaping molecule A in peak altogether be 1, molecule B
Relative amount be 0.This hypothesis can have many, we can establish a hypothesis list, and null hypothesis respectively.
After having inspected all hypothesis, chooses suitable it is assumed that receiving the relative amount for escaping component in this hypothesis altogether, sort out
And update data matrix.Gamma fitting is carried out with updated data, obtains new parameter and root-mean-square error.
Infer non-diagnostic peak by Fig. 5 step 3: assuming initially that non-diagnostic peak only includes a kind of component.It is then assumed that the component
Possible carbon number and possible molecule type.For example, its type of molecule for including in certain non-diagnostic peak might as well be assumed for virtue
Hydrocarbon, carbon number 10.Equally, we gather together various hypothesis, establish and assume list, and null hypothesis respectively.It has inspected
After all hypothesis, chooses suitable it is assumed that receiving the molecule type and carbon number in this hypothesis, sort out and update data matrix.With
Updated data carry out gamma fitting, obtain new parameter and root-mean-square error.
The content distribution data that Fig. 6-Figure 13 respectively shows each molecular series adjust algorithm steps 1 to step 3 place through peak
Variation schematic diagram after reason.In the present embodiment, molecule type will be divided into 8 classifications, respectively NP, MP, DP, TP, NO,
BO, NC, A.The processing result of this 8 types is corresponding with Fig. 6-Figure 13 respectively.Include three subgraph step 1- steps in each figure
Rapid 3, it is corresponding that three steps of algorithm are adjusted with peak respectively.The abscissa of these subgraphs is all carbon number, and ordinate is all quality
Divide rate.
Now by taking Fig. 8 of corresponding double branched paraffin data as an example, the result that peak adjusts three steps of algorithm is introduced.In order to state
It is convenient, three subgraphs of Fig. 8 are denoted as Fig. 8-1, Fig. 8-2 and Fig. 8-3 respectively herein.Scatterplot in Fig. 8-1 is that peak adjusts algorithm
Step 1 after the completion of in the data obtained matrix double branched paraffin series molecule content；Dotted line is that step 1 is fitted resulting parameter
Parameters1,DPThe gamma of representative is distributed.Similar with Fig. 8-1, scatterplot is respectively that peak is adjusted in algorithm in Fig. 8-2 and Fig. 8-3
After the completion of step 2 and step 3 in the data obtained matrix double branched paraffin series molecule content, blue dotted line is respectively
Parameters2,DPAnd Parameters3,DPRepresentative gamma distribution.There it can be seen that scatterplot and dotted line in Fig. 8-1
Departure degree is larger, is especially located at the scatterplot at C7 and is substantially less than dotted line.This is because some double branched paraffin molecules be with
The form for escaping peak or non-diagnostic peak altogether exists, thus the content of this moieties do not count enter step 1 data matrix
In.From as can be seen that working as the step 2 of peak adjusting algorithm, i.e., after the completion of evolution peak is split altogether, the distribution of scatterplot is in Fig. 8-2
Many has been approached with dotted line.
Fig. 8-3 does not have clear improvement relative to Fig. 8-2, and reason may have two aspects.First is that the content of non-diagnostic peak
Divide rate too low, keeps its variation unobvious；It on the other hand is that the content distribution that there are other molecule types deviates gamma distribution more
Far, it so that peak be allowed to adjust algorithm in the calculating process of step 3, is more prone to for non-diagnostic peak to be inferred as the molecule of the type.This
Kind situation can be found out in Fig. 9.Fig. 9-2 is illustrated, even if completing the fractionation at total evolution peak, the content point of three branched paraffins
Cloth is distributed still apart from each other with ideal gamma.Therefore, non-diagnostic peak, which is inferred as type, when algorithm can be more prone to step 3 is
The molecule of three branched paraffins.Then the distribution of scatterplot just has and is obviously improved in Fig. 9-3.
Figure 13 of corresponding aromatic hydrocarbons data is observed again.Figure 13-1 shows that the arene content at C7 is 0.This is because in GC-FID
In experimental situation, always with 2,3,3- trimethylpentanes flow out toluene altogether.The processing that algorithm is adjusted by peak, can obtain one
Relatively reasonable toluene level.In addition, it is noted that 11 n-alkanes all in library of molecules usually can whole quilts
GC-FID is identified, and does not flow out phenomenon altogether.Therefore, all images are the same in Fig. 6 of corresponding n-alkane data.
It can be seen that seeming more rationally by the distribution curve that peak adjusts the petrol molecule composition after algorithm is rebuild.
(4) software (property prediction module, more specifically molecule are used by the complete detailed hydrocarbon molecular composition result
Composition model) composition model that directly generates petrol molecule, it specifically includes: reading molecular composition information, instantiate gasoline pair
As each molecule is instantiated as molecular objects and is contained in gasoline object in molecular composition；
(5) according to the molecular composition of the composition model of petrol molecule acquisition gasoline and the property of each molecule, then by
The property of the molecular composition of the acquisition gasoline and each molecule is predicted by macroscopic property of the mixing rule to gasoline；It has
Body includes: that petrol molecule composition model passes through molecular composition and establishes gasoline object, contains in object and is established by each molecule
Molecular objects, molecular objects can execute inquiry molecular property, the sequence of operations such as the relative amount of molecule in the oil.Respectively
The property of molecule derives from NIST database.Gasoline object, which can execute, calculates the sequence of operations such as gasoline macroscopic property.Macroscopic view
Property is by the molecular property of each molecule, relative amount, and corresponding macroscopic property mixing rule is calculated.
Wherein, predicted value is shown in Table 1.The property predicted value of gasoline sample and the experiment value of gasoline sample property (use
Conventional method in that art measures) between comparison diagram it is as shown in figure 14.
Table 1
Note: the experiment value of every macroscopic property of catalytically cracked gasoline sample is using conventional method in that art in table 1
It measuring, specifically, wherein specific gravity is measured according to GBT 1884-2000, and boiling range is measured according to GBT 6536-2010,
Research octane number (RON) is measured according to GBT 5487-1995, and hydrocarbon content is measured according to GBT 11132-2008, Randt's steam
Pressing (KPa) is measured according to GBT 8017-2012.
Currently, the existing petrol molecule level composition model in this field is usually to be adjusted according to the experiment value of macroscopic property
Composition, then by the resulting macroscopic property for forming prediction gasoline.Therefore, which is influenced by the macroscopic property experiment value inputted
Very big, if the experimental error of the macroscopic property of gasoline is larger, the predicted value of this model also has large error.With
This model is different, and the model provided herein only adjusts algorithm according to GC-FID analysis and peak when predicting gasoline property
It is resulting as a result, therefore, not influenced by gasoline property experiment value, to more there is reference value after processing.
The prediction of the initial boiling point and the end point of distillation of ASTM D86 distillation curve is usually relatively difficult at present, can from table 1
Out, it can also obtain good result in model provided herein；Existing alkene, cycloalkane, the volume fraction of aromatic hydrocarbons
Experiment value is to be obtained by fluorescence measurement, but the error of this method is larger, and reproducibility is bad, and obtained by the application model prediction
Volume fraction is to analyze resulting detailed hydrocarbon mass fraction based on GC-FID to obtain, and result is more acurrate, and favorable reproducibility.Cause
This, the resulting volume fraction of the application model prediction should be more credible；In addition, it can also be seen that using the present invention from table 1
The critical natures such as octane number that provided method obtains and Reid vapour pressure, it may have have good prediction effect.
Claims (10)
1. a kind of method of direct construction petrol molecule composition model, which is characterized in that the described method comprises the following steps:
(1) detailed hydrocarbon analysis is carried out to the gas chromatographic detection result of gasoline sample, each peak may in gas chromatogram to identify
The molecule for including, and calculate the opposite point rate at each peak；
(2) according to detailed hydrocarbon analysis as a result, and sorting out chromatographic peak by chromatographic peak type；
(3) it to the chromatographic peak after classification, is analyzed according to chromatographic peak type and molecule type, obtains complete monomer hydrocarbon molecule
Form result；
(4) composition model of petrol molecule is directly generated by the complete detailed hydrocarbon molecular composition result.
2. the method according to claim 1, wherein the gasoline sample includes catalytically cracked gasoline, catalysis weight
Whole gasoline, direct steaming gasoline, catalytic cracking gasoline, hydrogasoline or coker gasoline.
3. the method according to claim 1, wherein the chromatographic peak type include known peak, altogether escape peak and
Non- diagnostic peak.
4. method according to claim 1 or 3, which is characterized in that analysis described in step (3) is using based on statistical
The peak of cloth adjusts algorithm to the chromatographic peak after classification, is analyzed according to chromatographic peak type and molecule type；
Preferably, the peak based on statistical distribution adjust algorithm specifically includes the following steps:
It is fitted the distribution at known peak: known peak being sorted out by molecule type and carbon number, to each series data after classification by statistics
Fitting of distribution；
The fractionation at peak is escaped altogether: total evolution peak is split, it is assumed that is escaped the relative amount of each component in peak altogether, and is successively examined
Test all it is assumed that choosing suitable hypothesis receiving；Above procedure is constantly repeated, until all total evolution peaks all split completion；
Infer non-diagnostic peak: assuming that the molecule type and carbon number of component included in non-diagnostic peak, and successively examine it is all it is assumed that
Suitable hypothesis is chosen to receive；Above procedure is constantly repeated, until completion is all inferred at all non-diagnostic peak peaks；
It is further preferred that the statistical distribution includes gamma distribution.
5. method according to claim 1 or 4, which is characterized in that molecule type described in step (3) includes N-alkanes
Hydrocarbon, isoparaffin, alkene, cycloalkane and aromatic hydrocarbons.
6. the method according to claim 1, wherein directly generating petrol molecule composition described in step (4)
Model, comprising: after obtaining the complete detailed hydrocarbon molecular composition result, establish a molecule pair respectively for wherein each molecule
As the molecular objects are for executing inquiry molecular property and molecule content operation；
Gasoline object is resettled, which includes the molecular objects, and the gasoline object is macro for executing calculating gasoline
See property operation.
7. a kind of system of direct construction petrol molecule composition model, which is characterized in that the system comprises:
First unit, the first unit is used to carry out detailed hydrocarbon analysis to the gas chromatographic detection result of gasoline sample, with mirror
Determine the molecule that each peak may include in gas chromatogram, and calculates the opposite point rate at each peak；
Second unit, the second unit are used for according to detailed hydrocarbon analysis as a result, and sorting out chromatographic peak by chromatographic peak type；
Third unit, the third unit are used to divide the chromatographic peak after classification according to chromatographic peak type and molecule type
Analysis, obtains complete detailed hydrocarbon molecular composition result；
Unit the 4th, Unit the 4th are used to directly generate petrol molecule by the complete detailed hydrocarbon molecular composition result
Composition model.
8. system according to claim 7, which is characterized in that in third unit, the analysis is using based on statistical
The peak adjustment module of cloth analyzes the chromatographic peak after classification according to chromatographic peak type and molecule type；
Preferably, the peak adjustment module based on statistical distribution specifically includes:
First module, first module are used to be fitted the distribution at known peak: known peak sorted out by molecule type and carbon number, it is right
Each series data after classification is fitted by statistical distribution；
Second module, second module for escaping the fractionation at peak altogether: total evolution peak is split, it is assumed that escapes in peak altogether
The relative amount of each component, and successively examine all it is assumed that choosing suitable hypothesis receiving；Above procedure is constantly repeated, until
All total evolution peaks all split completion；
Third module, the third module is for inferring non-diagnostic peak: assuming that the molecule type of component included in non-diagnostic peak
And carbon number, and successively examine all it is assumed that choosing suitable hypothesis receiving；Above procedure is constantly repeated, until all do not reflect
All infer completion in the peak Ding Feng.
9. a kind of method for predicting gasoline macroscopic property, which is characterized in that the described method comprises the following steps:
The group of the method building petrol molecule of direct construction petrol molecule composition model according to claim 1-6
At model, to obtain the molecular composition of gasoline and the property of each molecule；
According to the property of the molecular composition of gasoline and each molecule by corresponding macroscopic property mixing rule to the macroscopic view of gasoline
Property is predicted；
Preferably, the macroscopic property includes density, boiling range, octane number, refractive index, molecular weight and Reid vapour pressure.
10. a kind of system for predicting gasoline macroscopic property, which is characterized in that the system comprises:
First unit, the first unit are formed for direct construction petrol molecule according to claim 1-6
The composition model of the method building petrol molecule of model, to obtain the molecular composition of gasoline and the property of each molecule；
Second unit, the second unit are used to pass through corresponding macroscopic view according to the molecular composition of gasoline and the property of each molecule
Property mixing rule predicts the macroscopic property of gasoline.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810717484.XA CN108956814B (en) | 2018-07-03 | 2018-07-03 | Method for directly constructing gasoline molecular composition model and property prediction method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810717484.XA CN108956814B (en) | 2018-07-03 | 2018-07-03 | Method for directly constructing gasoline molecular composition model and property prediction method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108956814A true CN108956814A (en) | 2018-12-07 |
| CN108956814B CN108956814B (en) | 2020-07-03 |
Family
ID=64485035
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810717484.XA Active CN108956814B (en) | 2018-07-03 | 2018-07-03 | Method for directly constructing gasoline molecular composition model and property prediction method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108956814B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109507352A (en) * | 2019-01-02 | 2019-03-22 | 杭州辛孚能源科技有限公司 | The molecular method of stock is arbitrarily flowed in a kind of prediction petrochemical iy produced |
| CN109859805A (en) * | 2019-03-22 | 2019-06-07 | 杭州辛孚能源科技有限公司 | A kind of gasoline reconciliation optimization method based on molecular composition |
| CN114724644A (en) * | 2022-03-25 | 2022-07-08 | 中国石油大学(北京) | Method and equipment for predicting gasoline octane number based on intermolecular interaction |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH07253425A (en) * | 1994-03-15 | 1995-10-03 | Cosmo Sogo Kenkyusho:Kk | Property analyzing method for gasoline |
| CN101685087A (en) * | 2008-09-27 | 2010-03-31 | 中国石油化工股份有限公司 | Method for automatically analyzing gasoline composition by utilizing gas chromatography |
| CN101876648A (en) * | 2009-04-30 | 2010-11-03 | 中国石油化工股份有限公司 | Method for automatically analyzing gasoline detailed hydrocarbon compositions by utilizing gas chromatography |
| CN107704720A (en) * | 2017-09-18 | 2018-02-16 | 中国石油大学(北京) | The construction method and property Forecasting Methodology of a kind of oil average molecular structure model |
| CN107817302A (en) * | 2017-09-27 | 2018-03-20 | 中国石油天然气集团公司 | A kind of gas chromatography analysis method of petroleum hydrocarbon component |
-
2018
- 2018-07-03 CN CN201810717484.XA patent/CN108956814B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH07253425A (en) * | 1994-03-15 | 1995-10-03 | Cosmo Sogo Kenkyusho:Kk | Property analyzing method for gasoline |
| CN101685087A (en) * | 2008-09-27 | 2010-03-31 | 中国石油化工股份有限公司 | Method for automatically analyzing gasoline composition by utilizing gas chromatography |
| CN101876648A (en) * | 2009-04-30 | 2010-11-03 | 中国石油化工股份有限公司 | Method for automatically analyzing gasoline detailed hydrocarbon compositions by utilizing gas chromatography |
| CN107704720A (en) * | 2017-09-18 | 2018-02-16 | 中国石油大学(北京) | The construction method and property Forecasting Methodology of a kind of oil average molecular structure model |
| CN107817302A (en) * | 2017-09-27 | 2018-03-20 | 中国石油天然气集团公司 | A kind of gas chromatography analysis method of petroleum hydrocarbon component |
Non-Patent Citations (1)
| Title |
|---|
| 崔晨 等: "汽油分子组成模型构建", 《石油化工高等学校学报》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109507352A (en) * | 2019-01-02 | 2019-03-22 | 杭州辛孚能源科技有限公司 | The molecular method of stock is arbitrarily flowed in a kind of prediction petrochemical iy produced |
| CN109859805A (en) * | 2019-03-22 | 2019-06-07 | 杭州辛孚能源科技有限公司 | A kind of gasoline reconciliation optimization method based on molecular composition |
| CN109859805B (en) * | 2019-03-22 | 2021-01-15 | 广东辛孚科技有限公司 | Gasoline blending optimization method based on molecular composition |
| CN114724644A (en) * | 2022-03-25 | 2022-07-08 | 中国石油大学(北京) | Method and equipment for predicting gasoline octane number based on intermolecular interaction |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108956814B (en) | 2020-07-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CA2630146C (en) | Micro-hydrocarbon analysis | |
| US9201053B2 (en) | Method for measuring the properties of petroleum fuels by distillation | |
| CN107636124B (en) | Process for the thermal cracking production of bio-hydrocarbons from biorenewable feedstocks | |
| Kapur et al. | Determination of aromatics and naphthenes in straight run gasoline by 1H NMR spectroscopy. Part I | |
| CN108956814A (en) | A kind of method and property prediction technique of direct construction petrol molecule composition model | |
| Berrier et al. | Predictive modeling of aerospace fuel properties using comprehensive two-dimensional gas chromatography with time-of-flight mass spectrometry and partial least squares analysis | |
| Heyne et al. | Towards fuel composition and properties from Two-dimensional gas chromatography with flame ionization and vacuum ultraviolet spectroscopy | |
| US7501285B1 (en) | Detection and classification of heavy hydrocarbon contamination in refinery process streams via spectrofluorometry | |
| CN109507352B (en) | Method for predicting molecular composition of any stream in petrochemical production | |
| Al-Dahhan et al. | Classification of crude oils and its fractions on the basis of paraffinic, naphthenic and aromatics | |
| Cui et al. | Computer-aided gasoline compositional model development based on GC-FID analysis | |
| CA2281269A1 (en) | A method and apparatus for predicting a distillation temperature range of a hydrocarbon-containing compound | |
| JP2018509594A (en) | Relative evaluation method of naphtha flow | |
| Hall et al. | Quantifying isomeric effects: A key factor in aviation fuel assessment and design | |
| JP6030231B2 (en) | Method for monitoring refiner reforming unit utilizing detailed analysis of hydrocarbon composition | |
| JPH0750101B2 (en) | Method and apparatus for measuring cetane number or cetane index | |
| US4904604A (en) | Hydrocarbon mixture analysis | |
| Cheng et al. | Qualitative and quantitative analysis of refined F–T wax on GC–MS and GC | |
| Lee et al. | Investigation of methods for determining aromatics in middle-distillate fuels | |
| Mondal et al. | A Rapid 1H NMR-Based Estimation of PONA for Light and Narrow Cut Naphtha Samples of Refinery Streams toward BS-VI Gasoline | |
| Herink et al. | Application of hydrocarbon cracking experiments to ethylene unit control and optimization | |
| CN109406642B (en) | Method for analyzing strength of hydrogen supply capability of narrow fraction of petroleum cracking product | |
| Ikonen et al. | Different Hydrocarbon Types in Jet Fuel as Determined by GCxGC/FID-An Alternative Method to D2425 | |
| Iob et al. | Hydrocarbon group (PONA) analysis of reformate by FT-ir spectroscopy | |
| Chapman et al. | Comparison of the Particulate Matter Index and Particulate Evaluation Index Numbers Calculated by Detailed Hydrocarbon Analysis by Gas Chromatography (Enhanced ASTM D6730) and Vacuum Ultraviolet Paraffin, Isoparaffin, Olefin, Naphthene, and Aromatic Analysis (ASTM D8071) |
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 |