Gasoline Engler distillation curve determination method and related device

Abstract

The invention discloses a gasoline Engler distillation curve determination method and a related device. The method comprises the following steps: determining a gasoline fingerprint of a real gasoline sample to be detected based on an established gasoline molecular component database; based on the gasoline fingerprint of the to-be-tested real gasoline sample and the trained Enn distillation machine learning model, the bubble point temperature and the distillate mass fraction of the to-be-tested real gasoline sample are obtained; and according to the bubble point temperature and the distillate mass content of each molecular component in the real gasoline sample to be detected, determining the Enn distillation curve of the real gasoline sample to be detected. The gasoline Engler distillation curve determination method disclosed by the invention has the advantages of rapider, more efficient and more accurate in the aspect of determining the bubble point temperature and the distillate mass content of any gasoline sample, and is convenient to widely apply in industrial real-time optimization.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of molecular refining, in particular to a method for determining Enny distillation curve of gasoline and a related device. BACKGROUND

[0002] The Enny distillation curve can reflect the distillation characteristics of the gasoline sample, including the composition, properties of the gasoline sample and various information of the gasoline sample in use and storage. The distillation characteristics of the gasoline sample will affect the starting, warming performance of the engine and the gas blocking trend under high temperature or high altitude conditions, so the Enny distillation curve has an important reference function for the selection of motor gasoline and aviation gasoline.

[0003] In the related art for determining the Enny distillation curve of the gasoline sample, the method for correlating the gas chromatography data and the Enny distillation curve of the gasoline to obtain the Enny distillation curve of the gasoline, compared with the method for determining the Enny distillation curve by using the experimental method, although the calculation time is saved, but the initial boiling point and the final boiling point still have errors. In addition, the correlation formula between different types of Enny distillation curves is established, but this correlation formula is an empirical model, the parameters in the empirical model are often difficult to determine, and the physical meaning is not clear. When the type of the gasoline sample changes, the related parameters in the empirical model need to be adjusted again, so this model does not have universality and has high maintenance cost. SUMMARY

[0004] In view of the above problems, the purpose of the present application is to provide a method for determining Enny distillation curve of gasoline and a related device.

[0005] In a first aspect, the present application provides a method for determining Enny distillation curve of gasoline, comprising:

[0006] Based on the established gasoline molecular component database, the gasoline fingerprint of the to-be-tested real gasoline sample is determined; the gasoline molecular component database is used to store the gasoline fingerprint representing the characteristics of the gasoline molecules;

[0007] Based on the gasoline fingerprint of the to-be-tested real gasoline sample and the trained Enny distillation machine learning model, the bubble point temperature and the distillation mass fraction of the to-be-tested real gasoline sample are obtained; the machine learning model is trained by using a deep neural network model according to a pre-constructed virtual gasoline sample and a molecular component sample set provided by a pre-established Enny distillation mechanism model; the Enny distillation mechanism model is an iterative model of multiple bubble point temperatures and multiple distillation mass fractions, which is established by referring to the process of measuring the Enny distillation curve of the real gasoline sample by using the experimental method and starting from the characteristics of the molecular components;

[0008] According to the bubble point temperature and the mass fraction of each molecular component of the real gasoline sample to be tested, an Engler distillation curve of the real gasoline sample to be tested is determined.

[0009] In one embodiment, the molecular component sample set is obtained by the following method:

[0010] The gasoline fingerprint of each molecular component of the virtual gasoline sample is substituted into the Engler distillation mechanism model to obtain the bubble point temperature and the mass fraction of each molecular component of the virtual gasoline sample.

[0011] The gasoline fingerprint of each molecular component of the virtual gasoline sample is taken as a molecular component sample feature, and the bubble point temperature and the mass fraction of each molecular component of the virtual gasoline sample are taken as a molecular component sample label.

[0012] Based on the molecular component sample feature and the molecular component sample label, the molecular component sample set is constructed.

[0013] The molecular component sample set is divided into a molecular component training sample set and a molecular component test sample set; the molecular component training sample set is used to train the machine learning model; and the molecular component test sample set is used to evaluate the generalization ability of the Engler distillation machine learning model.

[0014] In one embodiment, the virtual gasoline sample is constructed by the following method:

[0015] According to the group composition of different types of known gasoline, the molecular species contained in different types of virtual gasoline samples are determined.

[0016] According to the range of the mass fraction of each group, a constraint condition of the mass fraction of each molecular component in different types of virtual gasoline samples is determined.

[0017] According to the constraint condition, the molecular component mass fraction of the virtual gasoline sample is randomly generated.

[0018] The molecular component mass fraction is normalized to obtain the gasoline fingerprint of a plurality of virtual gasoline samples.

[0019] In one embodiment, the virtual gasoline sample is constructed by the following method:

[0020] According to the gas chromatography detection of different types of known gasoline, the mass fraction of each molecular component in the real gasoline sample is obtained.

[0021] A preset upper and lower limit is added to the mass fraction of each molecular component in the real gasoline sample to obtain a preset range of the mass fraction of each molecular component of the virtual gasoline sample.

[0022] The gasoline fingerprints of a plurality of virtual gasoline samples are randomly generated within the preset range of the mass fraction.

[0023] In one embodiment, the gasoline fingerprint of the real gasoline sample to be tested is determined based on the pre-established gasoline molecular component database, including:

[0024] For any molecular component in the real gasoline sample to be tested, the gasoline fingerprint of the same molecular component in the gasoline molecular component database is determined as the gasoline fingerprint of the any molecular component in the real gasoline sample to be tested.

[0025] In one embodiment, the Enny distillation mechanism model is pre-established by the following method:

[0026] In the virtual gasoline sample, the target bubble point temperature of the first distillation and the target distillation mass fraction corresponding to the target bubble point temperature of the first distillation are calculated according to the gasoline fingerprint of the molecular component of the virtual gasoline sample.

[0027] The virtual gasoline sample remaining after the first distillation is determined, and in the virtual gasoline sample remaining after the first distillation, the target bubble point temperature of the second distillation and the target distillation mass fraction corresponding to the target bubble point temperature of the second distillation are calculated according to the gasoline fingerprint of the molecular component of the virtual gasoline sample remaining after the first distillation.

[0028] Similarly, the target bubble point temperature and the target distillation mass fraction corresponding to the virtual gasoline sample remaining after each distillation are determined until the preset termination condition is stopped, and a plurality of target bubble point temperatures and a plurality of target distillation mass fractions are obtained.

[0029] In one embodiment, the preset termination condition is that the remaining volume corresponding to the virtual gasoline sample remaining after each distillation is less than a preset volume threshold.

[0030] In one embodiment, determining the target bubble point temperature and the target distillation mass fraction corresponding to each distillation of the pre-constructed virtual gasoline sample includes:

[0031] For any molecular component of the pre-constructed virtual gasoline sample, the first gas phase molar fraction of the molecular component is determined according to the gasoline fingerprint of the molecular component, the preset bubble point temperature, and the preset gas phase fugacity coefficient.

[0032] The gas phase fugacity coefficient adjustment value of the molecular component is determined by adjusting the preset gas phase fugacity coefficient according to the first gas phase molar fraction of the molecular component.

[0033] determining, according to the gas phase fugacity coefficient adjustment value, an adjusted mass yield of the molecular component at the preset bubble point temperature;

[0034] determining, according to a distillation balance condition, whether the preset bubble point temperature and the adjusted mass yield need to be updated;

[0035] if the distillation balance condition is met, taking the preset bubble point temperature and the adjusted mass yield as a target bubble point temperature and a target mass yield corresponding to the molecular component;

[0036] if the distillation balance condition is not met, iteratively updating the preset bubble point temperature and the adjusted mass yield according to a preset update strategy until an updated value of the adjusted mass yield meets the distillation balance condition; taking the updated value of the adjusted mass yield that meets the distillation balance condition and an updated value of the preset bubble point temperature as the target bubble point temperature and the target mass yield corresponding to the molecular component.

[0037] In one embodiment, the preset update strategy comprises:

[0038] iteratively updating the preset bubble point temperature of each molecular component for each distillation to obtain an updated value of the preset bubble point temperature after each iteration;

[0039] determining, according to the updated value of the preset bubble point temperature after each iteration and the gas phase fugacity coefficient adjustment value of each molecular component, an updated value of the adjusted mass yield corresponding to each molecular component after each iteration.

[0040] In one embodiment, the distillation balance condition comprises:

[0041] a difference between a sum of mass yields of each molecular component distilled in the virtual gasoline sample and 1 is less than or equal to a preset threshold value.

[0042] In one embodiment, for any molecular component of a pre-constructed virtual gasoline sample, determining a first gas phase mole fraction of the molecular component according to a gasoline fingerprint, a preset bubble point temperature, and a preset gas phase fugacity coefficient of the molecular component comprises:

[0043] for any molecular component of a virtual gasoline sample, determining a saturated vapor pressure of the molecular component at the preset bubble point temperature according to the preset bubble point temperature of the molecular component;

[0044] determining a liquid fugacity coefficient of the molecular component at the preset bubble point temperature according to the preset bubble point temperature of the molecular component and the gasoline fingerprint of the molecular component;

[0045] determining a liquid activity coefficient of the molecular component at the preset bubble point temperature according to the preset bubble point temperature of the molecular component and the gasoline fingerprint of the molecular component;

[0046] determining a first phase equilibrium constant of the molecular component at the preset bubble point temperature according to the preset gas fugacity coefficient, the saturated vapor pressure, the liquid fugacity coefficient, the liquid activity coefficient, and a phase equilibrium constant formula;

[0047] determining the first gas mole fraction of the molecular component at the preset bubble point temperature according to the first phase equilibrium constant and a gas mole fraction formula.

[0048] In one embodiment, the adjusting the preset gas fugacity coefficient according to the first gas mole fraction of the molecular component to determine a gas fugacity coefficient adjustment value of the molecular component comprises:

[0049] normalizing the first gas mole fraction of any molecular component of the virtual gasoline sample;

[0050] determining the gas fugacity coefficient adjustment value of the molecular component at the preset bubble point temperature according to the normalized first gas mole fraction and a Soave-Redlich-Kwong equation of state.

[0051] In one embodiment, the Soave-Redlich-Kwong equation of state includes a parameter q;

[0052] The value of the parameter q in the Soave-Redlich-Kwong equation of state is different when determining the liquid fugacity coefficient and determining the gas fugacity coefficient adjustment value.

[0053] In one embodiment, when determining the liquid fugacity coefficient, the parameter q in the Soave-Redlich-Kwong equation of state is determined by a table lookup method;

[0054] When determining the gas fugacity coefficient adjustment value, the parameter q in the Soave-Redlich-Kwong equation of state is determined by a preset equation.

[0055] In an embodiment, the determining the adjusted distillation mass fraction of the molecular component comprises:

[0056] For any molecular component of the virtual gasoline sample, according to the gas phase fugacity coefficient adjustment value, the saturated vapor pressure, the liquid phase fugacity coefficient, the liquid phase activity coefficient and the phase equilibrium constant formula of the molecular component, the second phase equilibrium constant of the molecular component at the preset bubble point temperature is determined;

[0057] According to the second phase equilibrium constant and the gas phase molar fraction formula, the second gas phase molar fraction and the adjusted distillation mass fraction of the molecular component are determined.

[0058] In a second aspect, an embodiment of the present application provides a gasoline Enny distillation curve determination device, comprising:

[0059] A gasoline fingerprint determination module is configured to determine a gasoline fingerprint of a to-be-tested real gasoline sample based on an established gasoline molecular component database; and the gasoline molecular component database is configured to store a gasoline fingerprint representing gasoline molecular characteristics.

[0060] A bubble point temperature and distillation mass fraction determination module is configured to obtain a bubble point temperature and a distillation mass fraction of the to-be-tested real gasoline sample based on the gasoline fingerprint of the to-be-tested real gasoline sample and a trained Enny distillation machine learning model; the machine learning model is trained by using a deep neural network model based on a pre-constructed virtual gasoline sample and a molecular component sample set provided by a pre-established Enny distillation mechanism model; and the Enny distillation mechanism model is an iterative model of multiple bubble point temperatures and multiple distillation mass fractions established by referring to a process of measuring an Enny distillation curve of a real gasoline sample by an experimental method and starting from characteristics of molecular components.

[0061] An Enny distillation curve determination module is configured to determine an Enny distillation curve of the to-be-tested real gasoline sample based on the bubble point temperature and the distillation mass fraction of each molecular component in the to-be-tested real gasoline sample.

[0062] In a third aspect, an embodiment of the present application provides a computing device, comprising a memory, a processor and a computer program stored in the memory and executable on the processor, and the processor implements the gasoline Enny distillation curve determination method described above when executing the program.

[0063] In a fourth aspect, an embodiment of the present application provides a computer readable storage medium, which stores a computer program, and the computer program is executable on a processor to implement the gasoline Enny distillation curve determination method described above.

[0064] In a fifth aspect, an embodiment of the present application provides a computer program product, comprising: a computer program, which, when executed by a processor, implements the above-mentioned gasoline Engler distillation curve determination method.

[0065] The above-mentioned technical solution provided by the embodiment of the present application has at least the following beneficial effects:

[0066] The embodiment of the present application provides a gasoline Engler distillation curve determination method and related device, and in the embodiment of the present application, the gasoline fingerprint of each molecular component in the real gasoline sample to be measured can be quickly determined by using the pre-established gasoline molecular component database; the gasoline fingerprint of each molecular component obtained is input into the Engler distillation machine learning model obtained by pre-training, and the bubble point temperature and the distillation mass fraction of each molecular component of the real gasoline sample to be measured are output; and the Engler distillation curve of the real gasoline sample to be measured can be drawn according to the bubble point temperature and the distillation mass fraction of each molecular component obtained. The gasoline Engler distillation curve determination method provided by the embodiment of the present application can directly and quickly output the bubble point temperature and the distillation mass fraction of any real gasoline sample to be measured.

[0067] The Engler distillation mechanism model provided by the embodiment of the present application is an iterative model of multiple bubble point temperatures and multiple distillation mass fractions established from the characteristics of molecular components with reference to the process of measuring the Engler distillation curve of the real gasoline sample by the experimental method, so the calculation process of the Engler distillation mechanism model is carried out from the microscopic level of molecular components, and compared with the existing macroscopic empirical model, the Engler distillation mechanism model has stronger adaptability, accuracy and reliability because it does not need to be adjusted again. Since the Engler distillation machine learning model is trained based on the molecular component sample set provided by the Engler distillation mechanism model, the bubble point temperature and the distillation mass fraction output by the Engler distillation machine learning model also have higher accuracy. At the same time, compared with the Engler distillation mechanism model, the Engler distillation machine learning model overcomes the problems of complex construction process and long running time of the Engler distillation mechanism model, and has the advantages of faster and higher efficiency, which is convenient for wide application in industrial real-time optimization.

[0068] The technical solutions of the present application will be further described in detail below with the aid of the drawings and embodiments. BRIEF DESCRIPTION OF DRAWINGS

[0069] The accompanying drawings are used to provide a further understanding of the present application, and constitute a part of the specification, together with the embodiments of the present application, to explain the present application, and do not constitute a limitation on the present application. In the drawings:

[0070] Figure 1 The flowchart of the gasoline Engler distillation curve determination method in the embodiment of the present application;

[0071] Figure 2A technical flowchart for determining a target bubble point temperature and a target mass distillation content ratio of a gasoline sample through a gasoline Engler distillation mechanism model in an embodiment of the present application is shown in FIG. 1.

[0072] Figure 3 A structural schematic diagram of a gasoline Engler distillation curve determination device in an embodiment of the present application is shown in FIG. 2. DETAILED DESCRIPTION

[0073] An embodiment of the present application provides a gasoline Engler distillation curve determination method and related device, although exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided so that the present disclosure can be more thoroughly understood and the scope of the present disclosure can be fully conveyed to those skilled in the art.

[0074] An embodiment of the present application provides a gasoline Engler distillation curve determination method, as shown in FIG. 1, which comprises the following steps: Figure 1

[0075] S11, based on the established gasoline molecular component database, determine the gasoline fingerprint of the real gasoline sample to be tested; the gasoline molecular component database is used to store the gasoline fingerprint representing the characteristics of the gasoline molecules.

[0076] S12, based on the gasoline fingerprint of the real gasoline sample to be tested, and the trained Engler distillation machine learning model, obtain the bubble point temperature and the distillation mass content ratio of the real gasoline sample to be tested; the machine learning model is trained by using a deep neural network model according to a pre-constructed virtual gasoline sample and a pre-established molecular component sample set provided by the Engler distillation mechanism model; the Engler distillation mechanism model is an iterative model of multiple bubble point temperatures and multiple distillation mass content ratios established from the characteristics of the molecular components in the process of measuring the Engler distillation curve of the real gasoline sample by the experimental method.

[0077] S13, determine the Engler distillation curve of the real gasoline sample to be tested according to the bubble point temperature and the distillation mass content ratio of each molecular component in the real gasoline sample to be tested.

[0078] The gasoline Engler distillation curve is determined by a plurality of bubble point temperatures of a gasoline sample and a distillation mass content ratio corresponding to each bubble point temperature. Through the pre-established gasoline molecular component database, any molecular component of the real gasoline sample to be tested can be quickly determined by the gasoline fingerprint. The determined gasoline fingerprint of any molecular component is input into the pre-trained Engler distillation machine learning model, and the corresponding bubble point temperature and distillation mass content ratio of the above-mentioned any molecular component can be directly output. After the bubble point temperature and the distillation mass content ratio of each molecular component in the real gasoline sample to be tested are determined, the Engler distillation curve of the real gasoline sample to be tested can be drawn.​

[0079] In step S11, the gasoline fingerprint refers to the unique characteristics of the molecular components reflected by specific chemical components and proportions in the gasoline resource. The gasoline molecular component database is pre-established to store the gasoline fingerprints of various gasoline molecules. Gasoline is a complex mixture of components, and different types of gasoline samples contain various molecular species. By using the gasoline fingerprints in the pre-established molecular component database, the characteristics of each molecular component in different types of gasoline samples can be quickly identified and determined.

[0080] In one embodiment, the gasoline fingerprint is the molecular component data reflecting the characteristics of pre-set different types of gasoline molecular components. The pre-set different types of gasoline molecular components can use, for example, molecular types that have a greater impact on the chemical properties of gasoline.

[0081] In one embodiment, chemical descriptors are used to describe and identify the gasoline fingerprints in the gasoline molecular component database. Any molecule in the real gasoline sample to be tested is mapped to the gasoline molecular component database to establish a mapping relationship with the same molecular component, and by reading the gasoline fingerprint of the molecular component in the gasoline molecular database, the gasoline fingerprint of the molecular component in the real gasoline sample to be tested can be quickly determined. For example, the gasoline molecular component database stores various types of gasoline fingerprints, including Simplified Molecular Input Line Entry System (SMILES) descriptors, molecular weight, critical temperature, critical pressure, eccentric factor, 25°C liquid molar volume, solubility, etc. When determining the gasoline fingerprint of any molecule in the real gasoline sample to be tested, it can be directly queried in the gasoline molecular component database.

[0082] In step S12, based on the gasoline fingerprints of each molecular component in the real gasoline sample to be tested determined in step S11 and the pre-trained Enny distillation machine learning model, the bubble point temperature and distillation mass fraction corresponding to each molecular component in the real gasoline sample to be tested are obtained. The Enny distillation machine learning model is trained using a molecular component sample set, which is obtained using a pre-constructed virtual gasoline sample and a pre-established Enny distillation mechanism model. The input of the Enny distillation machine learning model is the gasoline fingerprint of each molecular component in the real gasoline sample to be tested, and the output is the bubble point temperature and distillation mass fraction corresponding to each molecular component in the real gasoline sample to be tested. The output of each bubble point temperature and distillation mass fraction is used to draw the Enny distillation curve.

[0083] In one embodiment, the Enny distillation machine learning model is trained by building a deep neural network model, which includes a feature input layer, a fully connected layer, and a regression layer, etc.

[0084] In one embodiment, a molecular component training sample set is divided from the obtained molecular component sample set, and the obtained molecular component training sample set is used to train the deep neural network model to obtain an Engler distillation machine learning model meeting the requirements.

[0085] In one embodiment, the Engler distillation mechanism model is an iterative model of multiple bubble point temperatures and multiple distillation mass fractions established by referring to the process of measuring the Engler distillation curve of a real gasoline sample by an experimental method.

[0086] In one embodiment, the molecular component sample set used to train the Engler distillation machine learning model is obtained by: substituting the gasoline fingerprints of each molecular component in the virtual gasoline sample into the pre-established Engler distillation mechanism model to obtain the corresponding bubble point temperature and distillation mass fraction of each molecular component; using the gasoline fingerprints of each molecular component in the virtual gasoline sample as the molecular component sample features, and using the bubble point temperature and distillation mass fraction of each molecular component determined by the Engler distillation mechanism model as the sample labels of the molecular component; based on the obtained molecular component sample features and molecular component sample labels, the molecular component sample set can be obtained; the molecular component sample set is divided into a molecular component training sample set and a molecular component test sample set; wherein the molecular component training sample set is used to train the Engler distillation machine learning model, and the other molecular component test sample set is used to evaluate the generalization ability of the Engler distillation machine learning model.

[0087] In one embodiment, during the training of the deep neural network model, the sample features (i.e., gasoline fingerprints) of each molecular component in the molecular component training sample set are input into the established deep neural network model to obtain the predicted values of the corresponding bubble point temperature and distillation mass fraction of each molecular component; a loss function is established using the predicted values and sample labels of each molecular component to calculate the difference between the predicted values and the sample labels, for example, the loss function can use a root mean square error function. The parameters in the deep neural network model are continuously adjusted by using an optimization algorithm, and the value of the loss function is continuously reduced until a preset convergence condition is reached, so that the established deep neural network model converges, thereby obtaining the trained deep neural network model, which is the Engler distillation machine learning model.

[0088] Gasoline is a complex mixture mainly composed of various hydrocarbons, including straight-chain alkanes, branched alkanes, cycloalkanes, and aromatic hydrocarbons. The proportion and types of these different hydrocarbons in gasoline vary depending on their source, refining process, and use. The virtual gasoline sample is randomly generated under preset conditions based on the gasoline fingerprints (molecular component data) of various known types of gasoline.

[0089] In one embodiment, the virtual gasoline sample is constructed by the following method: since different types of gasoline are composed of various groups (structurally divided into four groups, namely paraffin group, olefin group, naphthene group and aromatic group), the molecular species contained in the virtual gasoline sample of different types can be determined according to the known components of each group of different types of gasoline; the constraint condition of the mass fraction of each molecule contained in the virtual gasoline sample of different types is determined according to the range of the mass fraction of each group of different types of gasoline; the molecular component mass fraction of the virtual gasoline sample is randomly generated under the constraint condition, and after normalization, the gasoline fingerprint of the virtual gasoline sample with various types is obtained.

[0090] In one embodiment, the virtual gasoline sample can also use experimental means to obtain a more realistic gasoline fingerprint or molecular component data of the virtual gasoline sample. For example, it can be constructed by the following method: the mass fraction of each molecular component in the real gasoline sample is obtained by detecting known different types of gasoline, including the intermediate distillate and finished gasoline commonly used in gasoline blending pool, by gas chromatography; the upper limit and lower limit range is added to the mass fraction of each molecular component in the real gasoline sample to form the preset range of the mass fraction of each molecular component of the virtual gasoline sample; the gasoline fingerprint of a plurality of virtual gasoline samples is randomly generated within the preset range of the mass fraction.

[0091] In one embodiment, the upper limit of the preset range of the mass fraction of each molecular component in the virtual gasoline sample formed is set to 130% of the mass fraction of each molecule, and the lower limit is set to 70% of the mass fraction of each molecule, but the upper and lower limit ranges are not limited in the embodiment.

[0092] In summary, by using the pre-established Enn distillation mechanism model, the gasoline fingerprint (molecular component data) of each molecular component in a large number of virtual gasoline samples is processed to obtain a molecular component training sample set containing a large number of training samples; the Enn distillation machine learning model is trained by using the molecular component training sample set; and by using the Enn distillation machine learning model and the gasoline fingerprint of each molecular component of the real gasoline sample to be tested, the corresponding bubble point temperature and distillation mass fraction of each molecular component in the gasoline sample can be obtained.

[0093] The pre-established Engler distillation mechanism model has high calculation accuracy and can be applied to different types of gasoline. The molecular component training sample set of the virtual gasoline sample can be accurately, conveniently and low-cost obtained, and then the Engler distillation machine learning model is trained by using the training sample set. The machine learning model can quickly and accurately predict the bubble point temperature and distillation mass fraction of each molecular component of any gasoline sample. Since the Engler distillation mechanism model has the problems of complex construction process and long running time, the Engler distillation machine learning model effectively overcomes the above shortcomings of the Engler distillation mechanism model, and can be widely applied in industrial real-time optimization.

[0094] In one embodiment, the Engler distillation mechanism model for constructing the molecular component sample set is established by the following method:

[0095] In the virtual gasoline sample, the target bubble point temperature of the first distillation and the target distillation mass fraction corresponding to the target bubble point temperature of the first distillation are calculated according to the gasoline fingerprint of the molecular components of the virtual gasoline sample.

[0096] The virtual gasoline sample remaining after the first distillation is determined. In the virtual gasoline sample remaining after the first distillation, the target bubble point temperature of the second distillation and the target distillation mass fraction corresponding to the target bubble point temperature of the second distillation are calculated according to the gasoline fingerprint of the molecular components of the virtual gasoline sample remaining after the first distillation.

[0097] Similarly, the target bubble point temperature and the target distillation mass fraction corresponding to the virtual gasoline sample remaining after each distillation are determined until the preset termination condition is stopped, and a plurality of target bubble point temperatures and a plurality of target distillation mass fractions are obtained.

[0098] The target bubble point temperature and the target distillation mass fraction are the bubble point temperature and the distillation mass fraction corresponding to each molecular component in the virtual gasoline sample required for determining the Engler distillation curve.

[0099] In one embodiment, the gasoline fingerprint of the molecular components of the virtual gasoline sample can be obtained by querying the pre-established gasoline molecular component database.

[0100] The Engler distillation mechanism model is an iterative model obtained by modeling the target bubble point temperature and the target distillation mass fraction corresponding to different molecular components at each distillation according to the process of measuring the Engler distillation curve of gasoline by the experimental method. After each distillation, the corresponding virtual gasoline sample remaining after distillation needs to be determined. The target bubble point temperature and the target distillation mass fraction corresponding to each distillation are calculated based on the virtual gasoline sample remaining after the last distillation.

[0101] In an embodiment, the preset termination condition is that the remaining volume of the virtual gasoline sample after each distillation is less than a preset volume threshold. The remaining volume after each distillation can be obtained by subtracting the distillation volume at the target bubble point temperature corresponding to the distillation from the volume of the gasoline sample before the distillation, or by subtracting the distillation volume at the target bubble point temperature corresponding to the distillation from the total volume of the virtual gasoline sample.

[0102] In an embodiment, the remaining volume after each distillation can be determined according to the distillation mass fraction. For example, after a distillation, the total distillation mass fraction of each molecular component in the virtual gasoline sample at the target bubble point temperature is 30%, and the total volume of the virtual gasoline sample is 1 liter, then the distillation volume at the target bubble point temperature of the distillation is 0.3 liters, and the remaining volume after the distillation is 0.7 liters.

[0103] In an embodiment, the preset volume threshold V residue is a small value, for example, 0.01. In the case that the remaining volume of the virtual gasoline sample after each distillation is greater than the preset volume threshold, the process of determining the target bubble point temperature and the target distillation mass fraction corresponding to each distillation is repeatedly executed based on the gasoline fingerprint of the molecular components of the remaining virtual gasoline sample until the preset termination condition is reached, that is, the remaining volume of the virtual gasoline sample after the last distillation is less than the preset volume threshold V residue , and finally a plurality of target bubble point temperatures and the target distillation mass fraction corresponding to each target bubble point temperature are obtained, and further the En distillation curve of the virtual gasoline sample can be plotted.

[0104] In an embodiment, determining the target bubble point temperature and the target distillation mass fraction corresponding to each distillation of the pre-constructed virtual gasoline sample comprises:

[0105] For any molecular component of the pre-constructed virtual gasoline sample, determining a first gas phase mole fraction of the molecular component according to the gasoline fingerprint of the molecular component, a preset bubble point temperature, and a preset gas phase fugacity coefficient;

[0106] Adjusting the preset gas phase fugacity coefficient according to the first gas phase mole fraction of the molecular component to determine a gas phase fugacity coefficient adjustment value of the molecular component;

[0107] Determining a to-be-adjusted distillation mass fraction of the molecular component at the preset bubble point temperature according to the gas phase fugacity coefficient adjustment value;

[0108] Determining whether the preset bubble point temperature and the to-be-adjusted distillation mass fraction need to be updated according to the distillation equilibrium condition;

[0109] If the distillation balance condition is met, the preset bubble point temperature and the to-be-adjusted distillation mass content are taken as the target bubble point temperature and the target distillation mass content corresponding to the molecular component;

[0110] If the distillation balance condition is not met, the preset bubble point temperature and the to-be-adjusted distillation mass content are iteratively updated according to a preset updating strategy until the updated value of the to-be-adjusted distillation mass content meets the distillation balance condition; and the updated value of the to-be-adjusted distillation mass content that meets the distillation balance condition and the updated value of the corresponding preset bubble point temperature are taken as the target bubble point temperature and the target distillation mass content corresponding to the molecular component.

[0111] In one embodiment, for any molecular component of the pre-constructed virtual gasoline sample, the first gas phase molar fraction of the molecular component is determined according to the gasoline fingerprint of the molecular component, the preset bubble point temperature, and the preset gas phase fugacity coefficient, including:

[0112] For any molecular component of the virtual gasoline sample, the saturated vapor pressure of the molecular component at the preset bubble point temperature is determined according to the preset bubble point temperature of the molecular component;

[0113] The liquid phase fugacity coefficient of the molecular component at the preset bubble point temperature is determined according to the preset bubble point temperature of the molecular component, the gasoline fingerprint of the molecular component, and the Soave-Redlich-Kwong (SRK) equation of state;

[0114] The liquid phase activity coefficient of the molecular component at the preset bubble point temperature is determined according to the preset bubble point temperature of the molecular component and the gasoline fingerprint of the molecular component;

[0115] The first phase equilibrium constant of the molecular component at the preset bubble point temperature is determined according to the preset gas phase fugacity coefficient, the saturated vapor pressure, the liquid phase fugacity coefficient, the liquid phase activity coefficient, and the phase equilibrium constant formula;

[0116] The first gas phase molar fraction of the molecular component at the preset bubble point temperature is determined according to the first phase equilibrium constant and the gas phase molar fraction formula.

[0117] The preset bubble point temperature of any molecular component in the virtual gasoline sample can be determined empirically. The first gas phase molar fraction at the preset bubble point temperature is determined by performing the following steps:

[0118] Specifically, the first gas phase molar fraction of any molecular component in the virtual gasoline sample can be determined by the following steps (1)-(5):

[0119] (1) According to the preset bubble point temperature of any molecular component, the saturation vapor pressure of the molecular component at the preset bubble point temperature is calculated and determined by using the Antoine equation, and the equation expression of the Antoine equation is as follows:

[0120]

[0121] Wherein: is the saturation vapor pressure of the i th molecular component in the virtual gasoline sample; T is the preset bubble point temperature (unit: Kelvin); A, B, C, D and E are all constants.

[0122] (2) According to the preset bubble point temperature of the molecular component, the gasoline fingerprint of the molecular component, the Soave-Redlich-Kwong (SRK) equation of state is solved, and the liquid fugacity coefficient of the molecular component at the preset bubble point temperature is calculated, wherein the calculation formula of the SRK equation of state is as follows:

[0123]

[0124]

[0125] Wherein: Z i is the compressibility factor of the i th molecular component in the virtual gasoline sample; T is the operating temperature; R is the gas constant; σ and ε are obtained by querying the SRK equation parameter table;

[0126] T r = T / T c , P r = P / P c , T c is the critical temperature, P c is the critical pressure, and P is the operating pressure;

[0127] a(T r ), Ψ and Ω are obtained by querying the SRK equation parameter table.

[0128] The equation parameter table is as follows: select the parameters corresponding to the Soave-Redlich-Kwong equation (SRK equation of state) column in the table.

[0129]

[0130] (3) According to the preset bubble point temperature of the molecular component, and the gasoline fingerprint of the molecular component, the liquid activity coefficient of the molecular component at the preset bubble point temperature is determined. The calculation formula of the liquid activity coefficient is as follows:

[0131]

[0132] wherein:

[0133]

[0134]

[0135] β ik =∑ m e mi τ mk ;

[0136]

[0137]

[0138]

[0139]

[0140] wherein: x i is the mole fraction of the i-th molecular component in the virtual gasoline sample; θ k is the surface area fraction of the i-th molecular component in the virtual gasoline sample; s k is the surface area fraction; e mi = nu i ·Q k / q i , q i is the surface area parameter of the i-th molecular component in the virtual gasoline sample; r i is the volume fraction of the i-th molecular component in the virtual gasoline sample; is the number of groups k in the molecular component i; Q k is the group surface area parameter; a ij is the group interaction parameter; T is the preset bubble point temperature; nu i is the number of groups contained in the i-th molecular component in the virtual gasoline sample.

[0141] (4) determining a first phase equilibrium constant of the molecular component at the preset bubble point temperature according to the preset gas phase fugacity coefficient, the saturated vapor pressure, the liquid phase fugacity coefficient, the liquid phase activity coefficient, and a phase equilibrium constant formula, wherein the phase equilibrium constant formula is as follows:

[0142]

[0143] wherein: K i is the first phase equilibrium constant; is the preset gas fugacity coefficient of the i-th molecular component in the virtual gasoline sample; P is the operating pressure.

[0144] (5) According to the first phase equilibrium constant and the gas phase molar fraction formula, the first gas phase molar fraction of the molecular component at the preset bubble point temperature is determined, and the formula of the gas phase molar fraction is as follows:

[0145]

[0146] Wherein: y i is the first gas phase molar fraction of the i-th molecular component; e is the gasification fraction of the virtual gasoline sample at the preset bubble point temperature; z i is the compressibility coefficient of the i-th molecular component.

[0147] In one embodiment, the preset gas phase fugacity coefficient of any molecular component in the virtual gasoline sample is adjusted to obtain a gas phase fugacity coefficient adjustment value. Determining the gas phase fugacity coefficient adjustment value includes: for any molecular component of the virtual gasoline sample, normalizing the first gas phase molar fraction of the molecular component; and determining the gas phase fugacity coefficient adjustment value of the molecular component at the preset bubble point temperature according to the normalized first gas phase molar fraction and the Soave-Redlich-Kwong (SRK) equation of state. At this time, the SRK equation of state also uses the above formulas (2)-(3), but the determination method of the parameter q therein is different from that of the liquid phase fugacity coefficient.

[0148] In one embodiment, the Soave-Redlich-Kwong (SRK) equation of state includes a parameter q; the value of the parameter q in the SRK equation of state is different when determining the liquid phase fugacity coefficient and determining the gas phase fugacity coefficient adjustment value.

[0149] In one embodiment, when determining the liquid phase fugacity coefficient, the parameter q in the Soave-Redlich-Kwong equation of state is determined by a table lookup method;

[0150] When determining the gas phase fugacity coefficient adjustment value, the parameter q in the Soave-Redlich-Kwong equation of state is determined by a preset equation.

[0151] For example: when determining the liquid phase fugacity coefficient, the parameter q in the SRK equation of state is calculated using the following formula:

[0152]

[0153] Wherein: Ψ, a(T r ), Ω parameter meaning introduction;

[0154] Ψ, a(T r ) and Ω are obtained from the SRK equation of state parameter table;

[0155] T r = T / T c , T c is the critical temperature, and T is the standard atmospheric pressure;

[0156] When determining the gas fugacity coefficient adjustment value, the parameter q in the SRK equation of state is calculated using the following formula:

[0157] q = a / (b R T) (8);

[0158] a = y i ' a m y i (9)

[0159] b = y i ' b i (10);

[0160]

[0161]

[0162] b i = Ω R T c / P c (13);

[0163] Wherein, T is a preset bubble point temperature;

[0164] R is a gas constant;

[0165] k is a symmetric matrix of binary interaction parameters;

[0166] yi is the first gas mole fraction of the i th molecular component in the virtual gasoline sample; pi is the saturated vapor pressure of the i th molecular component in the virtual gasoline sample;

[0167] P c is the critical pressure.

[0168] Specifically, when determining the gas fugacity coefficient adjustment, the first gas mole fraction after normalization is substituted into formulas (9)-(10), and parameters a and b are calculated according to formulas (9)-(13); the obtained parameters a and b are substituted into formula (8) to calculate the parameter q, and then the SRK equation of state, i.e. formulas (2)-(3), is used again to calculate the gas fugacity coefficient adjustment value of the molecular component at the preset bubble point temperature

[0169] In one embodiment, when normalizing the first gas phase mole fraction, the first gas phase mole fraction of any molecular component in the virtual gasoline sample is normalized according to the first gas phase mole fraction of the molecular component in the virtual gasoline sample, and the sum of the distillation mass content of all molecular components that are distilled at the preset bubble point temperature.

[0170] In one embodiment, determining the adjusted distillation mass content of any molecular component in the virtual gasoline sample includes: for any molecular component in the virtual gasoline sample, determining the second phase equilibrium constant of the molecular component at the preset bubble point temperature according to the gas phase fugacity coefficient adjustment value, the saturated vapor pressure, the liquid phase fugacity coefficient, the liquid phase activity coefficient, and the phase equilibrium constant formula of the molecular component; obtaining the second gas phase mole fraction according to the second phase equilibrium constant and the gas phase mole fraction formula, and then obtaining the adjusted distillation mass content of the molecular component through the relationship between the gas phase mole fraction and the distillation mass content. For example: the obtained gas phase fugacity coefficient adjustment value and liquid phase fugacity coefficient are substituted into the phase equilibrium constant formula (5) to calculate the second phase equilibrium constant at the preset bubble point temperature; the second phase equilibrium constant and the gasification fraction of the virtual gasoline sample at the preset bubble point temperature are substituted into the gas phase mole fraction formula (6) to obtain the corresponding second gas phase mole fraction, and then the adjusted distillation mass content is calculated according to the second gas phase mole fraction.

[0171] In one embodiment, the preset update strategy includes: iteratively updating the preset bubble point temperature of each molecular component for each distillation to obtain an updated value of the preset bubble point temperature after each iteration;

[0172] According to the updated value of the preset bubble point temperature after each iteration and the gas phase fugacity coefficient adjustment value of each molecular component, an updated value of the adjusted distillation mass content of each molecular component after each iteration is determined.

[0173] In one embodiment, the distillation equilibrium condition includes: the difference between the sum of the distillation mass content of each molecular component distilled in the virtual gasoline sample and 1 is less than or equal to a preset threshold value.

[0174] The distillation equilibrium condition is used to determine whether each molecular component in the virtual gasoline sample can reach a distillation equilibrium state. Whether each molecular component in the virtual gasoline sample can reach a distillation equilibrium state is determined by judging whether each distilled molecular component in the virtual gasoline sample is completely vaporized. For example: whether the distillation equilibrium condition is satisfied is determined by judging the relationship between f(x) and the preset threshold value, and the expression of f(x) is as follows:

[0175] f(x) = ∑y′ i -1 (14);

[0176] wherein: y′ iLet be the second gas phase mole fraction of the i-th component vaporized at the current value of the target bubble point temperature of the gasoline sample.

[0177] The distillation mass content calculation conditions are the parameters required to calculate the distillation mass content of each molecular component of the virtual gasoline sample at the target bubble point temperature.

[0178] The preset threshold can be a small value, such as 0.0001. If the sum of the adjusted distillate mass contents of each molecular component distilled from the virtual gasoline sample and the value 1 are all less than the preset threshold, the adjusted distillate mass contents of each molecular component distilled from the sample meet the distillation equilibrium condition. The preset bubble point temperature and the adjusted distillate mass contents do not need to be updated. At this time, the preset bubble point temperature and the second distillate mass are used as the target bubble point temperature and target distillate mass contents of the molecular component.

[0179] If the sum of the adjusted distillate mass contents of each molecular component distilled from the virtual gasoline sample differs from the value 1 by a factor not less than a preset threshold, then the adjusted distillate mass contents of the molecular components distilled in this instance do not meet the distillation equilibrium condition (refer to...). Figure 2 As shown, the adjusted distillate mass content corresponding to the molecular components distilled in each previous distillation already meets the distillation equilibrium condition. In this distillation, the preset bubble point temperature and the adjusted distillate mass content need to be updated. For example, Newton's iteration method can be used to update the preset bubble point temperature, obtaining the updated value of the preset bubble point temperature for each molecular component; simultaneously, the adjusted value of the vapor phase fugacity coefficient is used as the adjusted value of the vapor phase fugacity coefficient for each molecular component in the updated distillate mass content calculation conditions. Based on the preset update strategy, the updated value of the adjusted distillate mass content for each molecular component is iteratively determined until the distillation equilibrium condition is met, thereby obtaining the target bubble point temperature for each molecular component and the target distillate mass content corresponding to that target bubble point temperature.

[0180] The Engel distillation mechanism model pre-established in this embodiment of the invention refers to... Figure 2As shown, according to a preset iteration method, the target bubble point temperature of any molecular component and the corresponding target distillation mass fraction are calculated from the gasoline fingerprint, i.e., the microscopic molecular characteristics of the virtual gasoline sample. The Envergent distillation mechanism model is an iterative model obtained by referring to the process of measuring the Envergent distillation curve of the real gasoline sample by the experimental method. In the calculation process, the target bubble point temperature and the target distillation mass fraction of each molecular component in each distillation process are determined according to the distillation equilibrium condition; and whether all the molecular components in the virtual gasoline sample are distilled and vaporized is judged according to the preset termination condition. Since the material balance principle of the virtual gasoline sample is considered in the calculation process, the target bubble point temperature and the corresponding target distillation mass fraction data can be calculated more accurately. Based on the large number of molecular component training sample sets provided by the Envergent distillation mechanism model, the Envergent distillation machine learning model is trained. The gasoline fingerprint of the real gasoline sample to be measured is input into the Envergent distillation machine learning model, and the bubble point temperature and the distillation mass fraction corresponding to the bubble point temperature of each molecular component of the real gasoline sample to be measured are directly output.

[0181] In step S13, the Envergent distillation curve of the real gasoline sample to be measured is drawn according to the plurality of bubble point temperatures and the distillation mass fractions of the real gasoline sample to be measured obtained in step S12.

[0182] Based on the same inventive concept, the present application also provides a gasoline Envergent distillation curve determination device. Since the principle of the problem solved by the device is similar to the foregoing gasoline Envergent distillation curve determination method, the implementation of the device can be referred to the implementation of the foregoing method, and the repeated parts will not be described herein.

[0183] The present application provides a gasoline Envergent distillation curve determination device, which refers to Figure 3 As shown, the device comprises:

[0184] The gasoline fingerprint determination module 31 is configured to determine the gasoline fingerprint of the real gasoline sample to be measured based on the established gasoline molecular component database; and the gasoline molecular component database is configured to store the gasoline fingerprint representing the molecular characteristics of the gasoline.

[0185] The bubble point temperature and distillation mass fraction determination module 32 is configured to determine the bubble point temperature and the distillation mass fraction of the real gasoline sample to be measured based on the gasoline fingerprint of the real gasoline sample to be measured and the trained Envergent distillation machine learning model; the machine learning model is trained by using a deep neural network model according to the pre-constructed virtual gasoline sample and the molecular component sample set provided by the pre-established Envergent distillation mechanism model; and the Envergent distillation mechanism model is an iterative model of a plurality of bubble point temperatures and a plurality of distillation mass fractions established by referring to the process of measuring the Envergent distillation curve of the real gasoline sample by the experimental method from the characteristics of the molecular components.

[0186] The Engler distillation curve determination module 33 is configured to determine the Engler distillation curve of the real gasoline sample to be tested according to the bubble point temperature and the distillation mass fraction of each molecular component in the real gasoline sample to be tested.

[0187] The embodiment of the present application provides a computing device, which comprises a memory, a processor and a computer program stored in the memory and executable on the processor, and the processor implements the method for determining the Engler distillation curve of gasoline when executing the program.

[0188] The embodiment of the present application provides a computer readable storage medium, which stores a computer program, and the computer program is executable on a processor to implement the method for determining the Engler distillation curve of gasoline.

[0189] The embodiment of the present application provides a computer program product, which comprises a computer program, and the computer program is executable on a processor to implement the method for determining the Engler distillation curve of gasoline.

[0190] Obviously, those skilled in the art can make various modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications belong to the scope of the claims of the present application and their equivalent technologies, the present application also intends to include these modifications.

Claims

1. A method for determining a gasoline Engler distillation curve, characterized in that The method comprises the following steps: determining the gasoline fingerprint of the to-be-tested real gasoline sample based on the established gasoline molecular component database; the gasoline molecular component database is used to store the gasoline fingerprint representing the characteristics of the gasoline molecules; based on the gasoline fingerprint of the to-be-tested real gasoline sample and the trained Enny distillation machine learning model, the bubble point temperature and the distillation mass fraction of the to-be-tested real gasoline sample are obtained; the machine learning model is trained by using a deep neural network model according to a pre-constructed virtual gasoline sample and a molecular component sample set provided by a pre-established Enny distillation mechanism model; the Enny distillation mechanism model is an iterative model of multiple bubble point temperatures and multiple distillation mass fractions, which is established by referring to the process of measuring the Enny distillation curve of the real gasoline sample by the experimental method and starting from the characteristics of the molecular components; determining the Enny distillation curve of the to-be-tested real gasoline sample according to the bubble point temperature and the distillation mass fraction of each molecular component in the to-be-tested real gasoline sample.

2. The method of claim 1, wherein, The molecular component sample set is obtained by the following method: substituting the gasoline fingerprint of each molecular component of the virtual gasoline sample into the Enny distillation mechanism model to obtain the bubble point temperature and the distillation mass fraction of each molecular component of the virtual gasoline sample; taking the gasoline fingerprint of each molecular component of the virtual gasoline sample as the molecular component sample feature and taking the bubble point temperature and the distillation mass fraction of each molecular component of the virtual gasoline sample as the molecular component sample label; constructing the molecular component sample set based on the molecular component sample feature and the molecular component sample label; dividing the molecular component sample set into a molecular component training sample set and a molecular component test sample set; the molecular component training sample set is used to train the machine learning model; and the molecular component test sample set is used to evaluate the generalization ability of the Enny distillation machine learning model.

3. The method of claim 1, wherein, The virtual gasoline sample is constructed by the following method: determining the molecular species contained in different types of virtual gasoline samples according to the group composition of known different types of gasoline; determining the mass fraction constraint condition of each molecular component in different types of virtual gasoline samples according to the mass fraction range of each group; randomly generating the mass fraction of the molecular components of the virtual gasoline sample according to the constraint condition; normalizing the mass fraction of the molecular components to obtain the gasoline fingerprint of multiple virtual gasoline samples.

4. The method of claim 1, wherein, The virtual gasoline sample is constructed by the following method: obtaining the mass fraction of each molecular component in the real gasoline sample by detecting known different types of gasoline by gas chromatography; adding a preset upper and lower limit to the mass fraction of each molecular component in the real gasoline sample to obtain a preset range of the mass fraction of each molecular component of the virtual gasoline sample; randomly generating the gasoline fingerprint of multiple virtual gasoline samples within the preset mass fraction range.

5. The method of claim 1, wherein, The method for determining the gasoline fingerprint of the to-be-tested real gasoline sample based on the pre-established gasoline molecular component database comprises the following steps: For any molecular component in the real gasoline sample to be tested, the gasoline fingerprint of the same molecular component in the gasoline molecular component database is determined as the gasoline fingerprint of the any molecular component in the real gasoline sample to be tested.

6. The method of claim 1, wherein, The Envergent distillation mechanism model is pre-established in the following manner: In the virtual gasoline sample, the target bubble point temperature of the first distillation and the target distillation mass fraction corresponding to the target bubble point temperature of the first distillation are calculated according to the gasoline fingerprint of the molecular component of the virtual gasoline sample; The virtual gasoline sample remaining after the first distillation is determined; in the virtual gasoline sample remaining after the first distillation, the target bubble point temperature of the second distillation and the target distillation mass fraction corresponding to the target bubble point temperature of the second distillation are calculated according to the gasoline fingerprint of the molecular component of the virtual gasoline sample remaining after the first distillation; By analogy, the target bubble point temperature and the target distillation mass fraction corresponding to the virtual gasoline sample remaining after each distillation are determined until a preset termination condition is reached to stop, thereby obtaining a plurality of target bubble point temperatures and a plurality of target distillation mass fractions.

7. The method of claim 6, wherein, The preset termination condition is that the remaining volume corresponding to the virtual gasoline sample remaining after each distillation is less than a preset volume threshold.

8. The method of claim 6, wherein, Determining the target bubble point temperature and the target distillation mass fraction corresponding to each distillation of the pre-constructed virtual gasoline sample includes: For any molecular component of the pre-constructed virtual gasoline sample, the first gas phase molar fraction of the molecular component is determined according to the gasoline fingerprint of the molecular component, a preset bubble point temperature, and a preset gas phase fugacity coefficient; The gas phase fugacity coefficient adjustment value of the molecular component is determined by adjusting the preset gas phase fugacity coefficient according to the first gas phase molar fraction of the molecular component; The adjusted distillation mass fraction of the molecular component at the preset bubble point temperature is determined according to the gas phase fugacity coefficient adjustment value; It is judged whether the preset bubble point temperature and the adjusted distillation mass fraction need to be updated according to the distillation equilibrium condition; If the distillation equilibrium condition is met, the preset bubble point temperature and the adjusted distillation mass fraction are taken as the target bubble point temperature and the target distillation mass fraction corresponding to the molecular component; If the distillation equilibrium condition is not met, the preset bubble point temperature and the adjusted distillation mass fraction are iteratively updated according to a preset update strategy until the updated value of the adjusted distillation mass fraction meets the distillation equilibrium condition; the updated value of the adjusted distillation mass fraction that meets the distillation equilibrium condition and the updated value of the corresponding preset bubble point temperature are taken as the target bubble point temperature and the target distillation mass fraction corresponding to the molecular component.

9. The method of claim 6, wherein, The preset update strategy includes: The preset bubble point temperature of each molecular component of each distillation is iteratively updated to obtain the updated value of the preset bubble point temperature after each iteration; The target bubble point temperature and the target distillation mass fraction corresponding to each distillation of the pre-constructed virtual gasoline sample are determined in the following manner: For any molecular component of the pre-constructed virtual gasoline sample, the first gas phase molar fraction of the molecular component is determined according to the gasoline fingerprint of the molecular component, a preset bubble point temperature, and a preset gas phase fugacity coefficient; The gas phase fugacity coefficient adjustment value of the molecular component is determined by adjusting the preset gas phase fugacity coefficient according to the first gas phase molar fraction of the molecular component; The adjusted distillation mass fraction of the molecular component at the preset bubble point temperature is determined according to the gas phase fugacity coefficient adjustment value; It is judged whether the preset bubble point temperature and the adjusted distillation mass fraction need to be updated according to the distillation equilibrium condition; If the distillation equilibrium condition is met, the preset bubble point temperature and the adjusted distillation mass fraction are taken as the target bubble point temperature and the target distillation mass fraction corresponding to the molecular component; If the distillation equilibrium condition is not met, the preset bubble point temperature and the adjusted distillation mass fraction are iteratively updated according to a preset update strategy until the updated value of the adjusted distillation mass fraction meets the distillation equilibrium condition; the updated value of the adjusted distillation mass fraction that meets the distillation equilibrium condition and the updated value of the corresponding preset bubble point temperature are taken as the target bubble point temperature and the target distillation mass fraction corresponding to the molecular component. The preset update strategy includes: The preset bubble point temperature of each molecular component of each distillation is iteratively updated to obtain the updated value of the preset bubble point temperature after each iteration; According to the updated value of the preset bubble point temperature after each iteration, the updated value of the vapor fugacity coefficient of each molecular component, and the updated value of the to-be-adjusted mass fraction of each molecular component after each iteration is determined.

10. The method of claim 8 or 9, wherein, The distillation equilibrium condition comprises: The difference between the sum of the mass fractions of the molecular components distilled in the virtual gasoline sample and 1 is less than or equal to a preset threshold.

11. The method of claim 8, wherein, The method comprises the following steps: According to the preset bubble point temperature of the molecular component, the saturation vapor pressure of the molecular component at the preset bubble point temperature is determined. According to the preset bubble point temperature of the molecular component, the liquid fugacity coefficient of the molecular component at the preset bubble point temperature is determined. According to the preset bubble point temperature of the molecular component, the liquid activity coefficient of the molecular component at the preset bubble point temperature is determined. According to the preset bubble point temperature of the molecular component, the liquid activity coefficient of the molecular component at the preset bubble point temperature is determined. According to the first phase equilibrium constant and the gas molar fraction formula, the first gas molar fraction of the molecular component at the preset bubble point temperature is determined.

12. The method of claim 8, wherein, The method comprises the following steps: The first gas molar fraction of the molecular component is normalized. According to the normalized first gas molar fraction and the Soave-Redlich-Kwong equation of state, the adjusted value of the gas fugacity coefficient of the molecular component at the preset bubble point temperature is determined.

13. The method of claim 11 or 12, wherein, The Soave-Redlich-Kwong equation of state comprises a parameter q. The value of the parameter q in the Soave-Redlich-Kwong equation of state is different when determining the liquid fugacity coefficient and determining the adjusted value of the gas fugacity coefficient.

14. The method of claim 13, wherein, When determining the liquid fugacity coefficient, the parameter q in the Soave-Redlich-Kwong equation of state is determined by a table lookup method. When determining the adjusted value of the gas fugacity coefficient, the parameter q in the Soave-Redlich-Kwong equation of state is determined by a preset equation.

15. The method of claim 8, wherein, The method comprises the following steps: For any molecular component of the virtual gasoline sample, according to the gas phase fugacity coefficient adjustment value, the saturated vapor pressure, the liquid phase fugacity coefficient, the liquid phase activity coefficient and the phase equilibrium constant formula of the molecular component, the second phase equilibrium constant of the molecular component at the preset bubble point temperature is determined; According to the second phase equilibrium constant and the gas phase molar fraction formula, the second gas phase molar fraction of the molecular component and the to-be-adjusted distillation mass fraction are determined.

16. A gasoline engine distillation curve determining apparatus characterized by comprising: It comprises: A gasoline fingerprint determination module is configured to determine the gasoline fingerprint of a to-be-tested real gasoline sample based on an established gasoline molecular component database; The gasoline molecular component database is configured to store the gasoline fingerprint representing the characteristics of the gasoline molecules; A bubble point temperature and distillation mass fraction determination module is configured to obtain the bubble point temperature and the distillation mass fraction of the to-be-tested real gasoline sample based on the gasoline fingerprint of the to-be-tested real gasoline sample and a trained Enny distillation machine learning model; the machine learning model is trained by using a deep neural network model based on a pre-constructed virtual gasoline sample and a molecular component sample set provided by a pre-established Enny distillation mechanism model; the Enny distillation mechanism model is an iterative model of multiple bubble point temperatures and multiple distillation mass fractions, which is established by referring to the process of measuring the Enny distillation curve of the real gasoline sample by the experimental method and starting from the characteristics of the molecular components; An Enny distillation curve determination module is configured to determine the Enny distillation curve of the to-be-tested real gasoline sample based on the bubble point temperature and the distillation mass fraction of each molecular component in the to-be-tested real gasoline sample.

17. A computing device, comprising: It comprises: A memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor executes the program to implement the gasoline Enny distillation curve determination method according to any one of claims 1-15.

18. A computer-readable storage medium, characterized in that, The computer readable storage medium stores a computer program, and the computer program is executed by the processor to implement the gasoline Enny distillation curve determination method according to any one of claims 1-15.

19. A computer program product, characterised in that, The computer program product comprises a computer program, and the computer program is executed by the processor to implement the gasoline Enny distillation curve determination method according to any one of claims 1-15.

Images