Methods, apparatus, equipment and storage media for the detection of petroleum fraction composition

Abstract

This application provides a method, apparatus, equipment, and storage medium for detecting the composition of petroleum fractions, relating to the refining industry. Based on the physical property data and reference data of the petroleum to be tested, the petroleum fraction composition corresponding to the petroleum to be tested is initialized to obtain a pre-set molecular library, a combination of probability density functions, and a range of values ​​for the probability density function parameters. Within the range of values, a target value is selected and assigned to the corresponding probability density function parameter. Based on the pre-set molecular library, the combination of probability density functions, and the assigned probability density function parameters, the molecular content of the petroleum fraction composition is determined. Macroscopic physical property similarity is determined based on the molecular content and physical property data, and compositional similarity is determined based on the molecular content and reference data. The petroleum fraction composition of the petroleum to be tested is determined to be one that meets the similarity requirements of compositional similarity and macroscopic physical property similarity, thus better reflecting the distribution of petroleum molecular content in actual oil products.

Description

Technical Field

[0001] This application relates to the oil refining industry, and more particularly to a method, apparatus, equipment, and storage medium for detecting the composition of petroleum fractions. Background Technology

[0002] With the development of the oil refining industry, lumped-model-based refining process models are no longer sufficient to meet the needs of process unit control and optimization. The increasing quality specifications of petroleum products, growing environmental protection pressures, and the demand for multi-unit joint optimization have spurred the development of molecular-level process models. Compared to lumped-model models, molecular-level process models have stronger predictive capabilities, providing a research foundation for molecular refining and the development of new refining processes. The prerequisite for constructing a molecular-level process model is obtaining the composition of petroleum fractions; the accuracy of the petroleum fraction composition directly determines the accuracy of the molecular-level process model.

[0003] In related technologies, petroleum analysis techniques can usually only provide limited macroscopic properties of petroleum. Therefore, petroleum molecular reconstruction techniques can only use macroscopic properties of petroleum to invert the composition of petroleum fractions, resulting in the molecular content distribution of the obtained petroleum fraction composition not conforming to the molecular content distribution in actual oil products. Summary of the Invention

[0004] This application provides a method, apparatus, equipment, and storage medium for detecting the composition of petroleum fractions, in order to solve the problem that the molecular content distribution of petroleum fractions does not conform to the molecular content distribution in actual oil products.

[0005] In a first aspect, this application provides a method for detecting the composition of petroleum fractions, comprising: acquiring physical property data and reference data of the petroleum to be detected; initializing the composition of the petroleum fractions corresponding to the petroleum to be detected based on the physical property data and reference data to obtain initialization data, the initialization data including a pre-set molecular library, a combination form of probability density functions, and a range of values ​​for the probability density function parameters, the pre-set molecular library containing the molecular structures that the petroleum to be detected may contain; selecting a target value within the value range and assigning it to the corresponding probability density function parameter; determining the molecular content of the petroleum fraction composition based on the pre-set molecular library, the combination form of probability density functions, and the assigned probability density function parameter; determining the macroscopic physical property similarity based on the molecular content and physical property data, and determining the compositional similarity based on the molecular content and reference data; and determining the petroleum fraction composition of the petroleum to be detected that satisfies the similarity requirements of compositional similarity and macroscopic physical property similarity.

[0006] In one possible implementation, determining the petroleum fraction composition of the petroleum to be tested as having similarity requirements in both compositional similarity and macroscopic property similarity includes: determining whether the macroscopic property similarity is less than a reference macroscopic property similarity of the petroleum to be tested, where the reference macroscopic property similarity is the currently determined optimal macroscopic property similarity; if the macroscopic property similarity is not less than the reference macroscopic property similarity of the petroleum to be tested, then determining the macroscopic property similarity as the reference macroscopic property similarity of the petroleum to be tested; determining whether the compositional similarity is less than a reference compositional similarity of the petroleum to be tested, where the reference compositional similarity is the currently determined optimal compositional similarity; if the compositional similarity is not less than the reference compositional similarity of the petroleum to be tested, then determining the compositional similarity as the reference compositional similarity of the petroleum to be tested; and determining the petroleum fraction composition of the petroleum to be tested as having similarity requirements in both reference compositional similarity and reference macroscopic property similarity.

[0007] In one possible implementation, determining that the petroleum fractions of the petroleum to be tested meet the similarity requirements in terms of compositional similarity and macroscopic physical property similarity further includes:

[0008] If the macroscopic property similarity is less than the reference macroscopic property similarity of the petroleum to be tested, then the step of selecting a target value within the range and assigning it to the corresponding probability density function parameter is executed; if the composition similarity is less than the reference composition similarity of the petroleum to be tested, then the step of selecting a target value within the range and assigning it to the corresponding probability density function parameter is executed; if the reference composition similarity and the reference macroscopic property similarity do not meet the similarity requirements, then the step of selecting a target value within the range and assigning it to the corresponding probability density function parameter is executed.

[0009] In one possible implementation, determining that the petroleum fractions of the petroleum to be tested are petroleum fractions whose compositional similarity and macroscopic property similarity meet the similarity requirements further includes: determining the weighted sum of compositional similarity and macroscopic property similarity; determining that the petroleum fractions of the petroleum to be tested are petroleum fractions whose weighted sum meets the similarity requirements; if the weighted sum does not meet the similarity requirements, then performing the step of selecting a target value within the range and assigning it to the corresponding probability density function parameter.

[0010] One possible implementation involves selecting a target value within a range and assigning it to the corresponding probability density function parameter, including: selecting a target value within a range and assigning it to the corresponding probability density function parameter based on a bi-objective optimization algorithm, wherein the bi-objective optimization algorithm includes a genetic algorithm with non-dominated sorting using an elitist strategy; and / or, selecting a target value within a range and assigning it to the corresponding probability density function parameter based on a single-objective optimization algorithm, wherein the single-objective optimization algorithm includes a genetic algorithm.

[0011] In one possible implementation, macroscopic property similarity is determined based on molecular content and physical property data, including: the expression for macroscopic property similarity can be:

[0012] Prop Sim=-mean(δ|P ref -P pred |)

[0013] Among them, P ref It is the physical property data of petroleum, P pred δ represents the petroleum properties of the petroleum fraction composition, and δ is the weighting factor for the physical property data.

[0014] In one possible implementation, the compositional similarity is determined based on molecular content and reference data, including: the expression for compositional similarity can be:

[0015]

[0016] Wherein, COS represents the cosine similarity value between the molecular content of the petroleum fraction and the molecular content of the reference data, m ref，i It is the molecular content of the reference data, m pred，i It refers to the molecular content of petroleum fractions.

[0017] Secondly, this application provides a device for detecting the composition of petroleum fractions, comprising: an acquisition module for acquiring physical property data and reference data of the petroleum to be detected; an initialization module for initializing the petroleum fraction composition corresponding to the petroleum to be detected based on the physical property data and reference data to obtain initialization data, the initialization data including a preset molecular library, a combination form of probability density functions, and a range of values ​​for the probability density function parameters, the preset molecular library containing the molecular structures that the petroleum to be detected may contain; a parameter assignment module for selecting target values ​​within the value range and assigning them to the corresponding probability density function parameters; a molecular content determination module for determining the molecular content of the petroleum fraction composition based on the preset molecular library, the combination form of probability density functions, and the assigned probability density function parameters; a similarity calculation module for determining the macroscopic physical property similarity based on the molecular content and physical property data, and determining the compositional similarity based on the molecular content and reference data; and a petroleum fraction composition determination module for determining the petroleum fraction composition of the petroleum to be detected that meets the similarity requirements of compositional similarity and macroscopic physical property similarity.

[0018] In one possible implementation, the petroleum fraction composition determination module can be specifically used to: determine the petroleum fraction composition of the petroleum to be tested that meets the similarity requirements of composition similarity and macroscopic property similarity, including: determining whether the macroscopic property similarity is less than the reference macroscopic property similarity of the petroleum to be tested, wherein the reference macroscopic property similarity is the currently determined optimal macroscopic property similarity; if the macroscopic property similarity is not less than the reference macroscopic property similarity of the petroleum to be tested, then the macroscopic property similarity is determined as the reference macroscopic property similarity of the petroleum to be tested; determining whether the composition similarity is less than the reference composition similarity of the petroleum to be tested, wherein the reference composition similarity is the currently determined optimal composition similarity; if the composition similarity is not less than the reference composition similarity of the petroleum to be tested, then the composition similarity is determined as the reference composition similarity of the petroleum to be tested; and determining the petroleum fraction composition of the petroleum to be tested that meets the similarity requirements of reference composition similarity and reference macroscopic property similarity.

[0019] In one possible implementation, the petroleum fraction composition determination module can also be specifically used to: determine the petroleum fraction composition of the petroleum to be tested that meets the similarity requirements of composition similarity and macroscopic property similarity, including: if the macroscopic property similarity is less than the reference macroscopic property similarity of the petroleum to be tested, then perform the step of selecting a target value within the value range and assigning it to the corresponding probability density function parameter; if the composition similarity is less than the reference composition similarity of the petroleum to be tested, then perform the step of selecting a target value within the value range and assigning it to the corresponding probability density function parameter; if the reference composition similarity and the reference macroscopic property similarity do not meet the similarity requirements, then perform the step of selecting a target value within the value range and assigning it to the corresponding probability density function parameter.

[0020] In one possible implementation, the petroleum fraction composition determination module can also be specifically used to: determine the petroleum fraction composition of the petroleum to be tested that meets the similarity requirements of composition similarity and macroscopic property similarity, including: determining the weighted sum of composition similarity and macroscopic property similarity; determining the petroleum fraction composition of the petroleum to be tested that meets the similarity requirements of the weighted sum; if the weighted sum does not meet the similarity requirements, then performing the step of selecting a target value within the value range and assigning it to the corresponding probability density function parameter.

[0021] In one possible implementation, the parameter assignment module is specifically used to: select a target value within a range and assign it to the corresponding probability density function parameter, including: selecting a target value within a range and assigning it to the corresponding probability density function parameter based on a bi-objective optimization algorithm, wherein the bi-objective optimization algorithm includes a genetic algorithm with elitist strategy and non-dominated sorting; and / or, selecting a target value within a range and assigning it to the corresponding probability density function parameter based on a single-objective optimization algorithm, wherein the single-objective optimization algorithm includes a genetic algorithm.

[0022] In one possible implementation, the similarity calculation module can be specifically used to: determine macroscopic property similarity based on molecular content and physical property data, including: determining macroscopic property similarity according to the following formula:

[0023] Prop Sim=-mean(δ|P ref -P pred |)

[0024] Where PropSim represents macroscopic property similarity, P ref It is a macroscopic property determined based on physical property data, P pred It is the macroscopic physical property of petroleum fraction composition determined based on molecular content, δ is the weighting factor of the physical property data, and mean represents the mean value.

[0025] In one possible implementation, the similarity calculation module can also be specifically used to: determine compositional similarity based on molecular content and reference data, including: determining compositional similarity according to the following formula:

[0026]

[0027] Where COS represents compositional similarity, m ref，i m represents the molecular content of molecule i in the reference data. pred，i ∑ represents the molecular content of molecule i in the composition of petroleum fraction, and ∑ represents the summation symbol.

[0028] Thirdly, this application provides a device for detecting the composition of petroleum fractions, comprising: a memory and a processor. The memory is used to store program instructions; the processor is used to invoke the program instructions in the memory to execute the petroleum fraction composition detection method of the first aspect.

[0029] Fourthly, this application provides a computer-readable storage medium storing computer-executable instructions, which, when executed, are used to implement the method for detecting the composition of petroleum fractions in the first aspect.

[0030] Fifthly, this application provides a computer program product comprising a computer program, which, when executed by a processor, is used to implement the method for detecting the composition of petroleum fractions in the first aspect.

[0031] The method, apparatus, equipment, and storage medium for detecting the composition of petroleum fractions provided in this application acquire physical property data and reference data of the petroleum to be detected; based on the physical property data and reference data, initialize the petroleum fraction composition corresponding to the petroleum to be detected to obtain a pre-set molecular library, a combination form of probability density functions, and a range of values ​​for probability density function parameters; select a target value within the range and assign it to the corresponding probability density function parameter; determine the molecular content of the petroleum fraction composition based on the pre-set molecular library, the combination form of probability density functions, and the assigned probability density function parameters; determine the macroscopic physical property similarity based on the molecular content and physical property data, and determine the compositional similarity based on the molecular content and reference data; and determine the petroleum fraction composition of the petroleum to be detected that meets the similarity requirements of compositional similarity and macroscopic physical property similarity. Since this application inverts the composition of petroleum fractions based on compositional similarity and macroscopic property similarity, the determined petroleum fractions of the petroleum to be tested become petroleum fractions whose compositional similarity and macroscopic property similarity meet the similarity requirements. Therefore, the molecular content distribution of the obtained petroleum fraction composition can conform to the actual petroleum molecular content distribution in the oil product. In addition, once the petroleum molecular content distribution is determined, the problem of multiple solutions for optimizing the petroleum fraction composition can also be solved. Attached Figure Description

[0032] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0033] Figure 1 This is a schematic diagram illustrating an application scenario of the method for detecting the composition of petroleum fractions provided in an embodiment of this application;

[0034] Figure 2 This is a schematic flowchart of a method for detecting the composition of petroleum fractions provided in an embodiment of this application;

[0035] Figure 3 This is a schematic diagram of the core molecular structure of diesel fuel provided in an embodiment of this application;

[0036] Figure 4 This is a schematic diagram of a combination of probability density functions provided in an embodiment of this application;

[0037] Figure 5 This is a schematic flowchart of a method for detecting the composition of petroleum fractions provided in another embodiment of this application;

[0038] Figure 6 This is a schematic flowchart of a method for detecting the composition of diesel fractions provided in an embodiment of this application;

[0039] Figure 7This is a detection result of diesel fraction composition using GC-FI TOF MS data as reference data, provided in one embodiment of this application;

[0040] Figure 8 This is a repeatability test result for detecting diesel fraction composition using GC-FI TOF MS data as reference data, provided in one embodiment of this application.

[0041] Figure 9 This application provides a comparison of experimental and calculated values ​​of macroscopic physical properties of multiple diesel fractions, using GC-FI TOF MS data as reference data;

[0042] Figure 10 This is a comparison result of experimental and calculated values ​​of molecular content distribution of multiple diesel fractions detected using GC-FI TOF MS data as reference data, provided in one embodiment of this application.

[0043] Figure 11 This application provides another embodiment of the macroscopic physical property results of diesel fraction composition detection using historically detected diesel fractions as reference data;

[0044] Figure 12 This application provides another embodiment of the results for detecting the molecular content distribution of diesel fraction composition using historically detected diesel fraction composition as reference data;

[0045] Figure 13 This is a repeatability test result for detecting the composition of diesel fractions using historically detected diesel fractions as reference data, provided by another embodiment of this application;

[0046] Figure 14 This is a schematic diagram of the structure of a petroleum fraction composition detection device provided in an embodiment of this application;

[0047] Figure 15 This is a schematic diagram of the structure of a petroleum fraction composition detection device provided in one embodiment of this application.

[0048] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation

[0049] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0050] The terms “first,” “second,” etc., used in the specification and claims of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented, for example, in orders other than those illustrated or described herein. Furthermore, the terms “comprising” and “having,” and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, products, or apparatus.

[0051] In related technologies, petroleum molecule reconstruction technology based on analytical data is an important means of obtaining the composition of petroleum molecules. However, due to insufficient constraints on the molecular content in the composition of petroleum fractions from analytical data, multiple solutions to the optimization of petroleum fraction composition are easily caused. Effective utilization of petroleum molecule characterization technology holds promise for solving the problems of accuracy and multiple solutions in the molecular content distribution of petroleum fraction composition. However, due to the complexity of petroleum molecule composition, petroleum molecule characterization technology for diesel and heavier fractions still cannot provide the composition of all petroleum molecules (such as monomeric hydrocarbons), and advanced petroleum molecule characterization technology is very expensive. Therefore, the existing composition of diesel and heavier fractions is almost entirely derived from molecular reconstruction technology.

[0052] To address the aforementioned issues, this application proposes a method for detecting the composition of petroleum fractions, namely, constructing a petroleum fraction composition model. This method is based on a petroleum molecule reconstruction method using macroscopic property similarity and compositional similarity. Reference data is introduced during the petroleum fraction composition process. The compositional similarity between the molecular content of the petroleum fraction composition and the molecular content of the reference data is used as the optimization objective. Simultaneously, the macroscopic property similarity between the petroleum properties of the petroleum fraction composition and the petroleum properties of the physical property data is also used as the optimization objective. A dual-objective optimization algorithm and a single-objective optimization algorithm are employed to optimize the compositional similarity and macroscopic property similarity, ensuring that the molecular content distribution of the obtained petroleum fraction composition conforms to the actual petroleum molecule content distribution in petroleum products. Furthermore, once the petroleum molecule content distribution is determined, the problem of multiple solutions in petroleum fraction composition optimization can also be resolved.

[0053] Figure 1This is a schematic diagram illustrating an application scenario of the method for detecting the composition of petroleum fractions provided in an embodiment of this application. For example... Figure 1 As shown, this application scenario includes a first client 11, a server 12, and a second client 13, wherein the number of both the first client 11 and the second client 13 can be at least one. In practical applications, for example, after researchers and other relevant personnel obtain physical property data and reference data, they send the physical property data and reference data to the server 12 for storage through the first client 11; when the server 12 detects an instruction issued by a user through the second client 13 for detecting the petroleum fraction composition of a certain petroleum, it executes the petroleum fraction composition detection method provided in this application based on the stored physical property data and reference data, obtains the petroleum fraction composition of the petroleum, and sends the result of the petroleum fraction composition to the second client 13, thereby allowing the relevant personnel to know the petroleum fraction composition of the petroleum.

[0054] It should be noted that server 12 can also be replaced by a server cluster or other computing devices with a certain computing power. Both the first and second clients can be mobile phones, computers, laptops, or personal digital assistants (PDAs). Furthermore, the detection method provided in this application can also be used to detect the distillate composition of diesel fuel.

[0055] The following is combined with Figure 1 Application scenarios, refer to Figure 2 This application describes a method for detecting the composition of petroleum fractions according to exemplary embodiments thereof. It should be noted that the above application scenarios are shown only to facilitate understanding of the spirit and principles of this application, and the embodiments of this application are not limited to those described herein. Figure 1 The limitations of the application scenarios shown.

[0056] Figure 2 This is a schematic flowchart illustrating a method for detecting the composition of petroleum fractions according to an embodiment of this application. Figure 2 As shown, the method for detecting the composition of petroleum fractions in this application includes the following steps:

[0057] S201: Obtain the physical property data and reference data of the petroleum to be tested.

[0058] In this step, the physical property data and reference data of the petroleum to be tested can be obtained by analyzing existing petroleum samples. The physical property data represents the macroscopic properties of the petroleum, including its density, elemental composition, and distillation point. The reference data is molecular-level characterization data, including historically analyzed petroleum fraction compositions and / or gas chromatography-field ionization time-of-flight mass spectrometry (GC-FI TOF MS) data. Specifically, compared to the expensive quantitative molecular characterization technique for diesel fuel, full two-dimensional gas chromatography (GC×GC), semi-quantitative molecular characterization data (i.e., GC FI-TOF MS) is inexpensive and readily available.

[0059] S202: Based on the physical property data and reference data, the composition of the petroleum fraction corresponding to the petroleum to be tested is initialized to obtain initialization data. The initialization data includes a pre-set molecular library, a combination of probability density functions, and the range of values ​​for the probability density function parameters. The pre-set molecular library contains the molecular structures that the petroleum to be tested may contain.

[0060] In this step, the petroleum fraction composition corresponding to the petroleum to be tested is initialized to obtain an initial petroleum fraction composition. Subsequently, the initial petroleum fraction composition is optimized by optimization algorithms, such as direct optimization algorithms and / or indirect optimization algorithms, to finally determine an optimized petroleum fraction composition.

[0061] In this step, the pre-built molecular library refers to a pre-defined list of all molecular structures that the corresponding petroleum may contain. Specifically, the pre-built molecular library needs to conform to the petrochemical composition of the corresponding petroleum and include the main core molecular structures. Therefore, the construction of the pre-built molecular library can be completed by first defining the core molecular structures and then expanding the number of carbon atoms in the side chains based on the core molecular structures. For example, Figure 3 This is a schematic diagram of the core molecular structure of diesel fuel provided in an embodiment of this application.

[0062] For example, after the pre-built molecular library is constructed, the molecular content needs to be optimized to determine the composition of petroleum fractions. Specifically, optimizing the molecular content is the core of determining the composition of petroleum fractions. The goal of molecular content optimization is to find an accurate molecular content distribution so that the properties of the petroleum fraction composition closely resemble those of real petroleum.

[0063] For example, this study uses a combination of probability density functions to indirectly calculate the molecular content. The probability density function (PDF) includes the gamma probability density function (GammaPDF) and the histogram probability density function (HistogramPDF), as shown in formulas (1) and (2), respectively:

[0064]

[0065] f(x) = X i (2)

[0066] In GammaPDF, there are three adjustable parameters: α, β, and γ, used for continuous property constraints, such as carbon number and boiling point; x is the probability density function parameter. In HistogramPDF, X... i It refers to the content of a certain type of molecule, such as classifying petroleum molecules into hydrocarbons, sulfur-containing compounds, nitrogen-containing compounds, and oxygen-containing compounds, X i This corresponds to the content of hydrocarbons, sulfur-containing compounds, nitrogen-containing compounds, and oxygen-containing compounds.

[0067] For example, Figure 4 This is a schematic diagram illustrating a combination of probability density functions provided in an embodiment of this application. Figure 4 As shown, the molecular types of petroleum molecules include hydrocarbons, sulfur-containing compounds, nitrogen-containing compounds, and oxygen-containing compounds. Among them, the probability density function characterizing the boiling point of all petroleum molecules is GammaPDF; the probability density functions characterizing basic nitrogen and non-basic nitrogen, acids and phenols, alkanes, cycloalkanes, and aromatic hydrocarbons, and thioethers, thiophenes, benzothiophenes, and dibenzothiophenes are HistogramPDF; the probability density functions for monocyclic, bicyclic, and tricyclic cycloalkanes are HistogramPDF; and the probability density functions for alkylbenzenes, indene, indene, naphthalene, acenaphthene, acenaphthene, and tricyclic aromatic hydrocarbons are HistogramPDF.

[0068] For example, the range of values ​​for the probability density function parameters can be determined by combining physical property data, reference data, and probability density function parameters.

[0069] S203: Select the target value within the range and assign it to the corresponding probability density function parameter.

[0070] For example, in step S202, the range of values ​​for the probability density function parameters has been determined. In this step, a target value can be randomly selected from the determined range as the initial parameter of the probability density function.

[0071] S204: Determine the molecular content of petroleum fraction composition based on a pre-set molecular library, the combination of probability density functions, and the assigned probability density function parameters.

[0072] For example, in this step, the preset molecular library, the combination form of the probability density function, and the assigned probability density function parameters can be obtained from steps S202 and S203, respectively, thereby determining the molecular content of the petroleum fraction composition. Specifically, the molecular content of the petroleum fraction composition obtained here is an initial molecular content, which still needs to be optimized by an optimization algorithm.

[0073] S205: Determine the macroscopic property similarity based on molecular content and physical property data, and determine the compositional similarity based on molecular content and reference data.

[0074] For example, in this step, unlike the existing petroleum molecule reconstruction methods based on macroscopic property similarity which use macroscopic property similarity as the optimization objective, the method proposed in this application uses both macroscopic property similarity and compositional similarity as optimization objectives, forming a dual-objective optimization problem.

[0075] For example, based on molecular content and physical property data, the macroscopic physical property similarity is determined, including: determining the macroscopic physical property similarity according to the following formula (3):

[0076] Prop Sim=-mean(δ|P ref -P pred |) (3)

[0077] Where PropSim represents macroscopic property similarity, P ref It is a macroscopic property determined based on physical property data, P pred It is the macroscopic physical property of petroleum fraction composition determined based on molecular content, δ is the weighting factor of the physical property data, and mean represents the mean value.

[0078] For example, compositional similarity is determined based on molecular content and reference data, including: determining compositional similarity according to the following formula (4):

[0079]

[0080] Where COS represents compositional similarity, m ref，i m represents the molecular content of molecule i in the reference data. pred，i This represents the molecular content of molecule i in the composition of a petroleum fraction, and ∑ represents the summation symbol. Specifically, the COS value is between -1 and 1, and the larger the value, the more similar the two components are.

[0081] For example, compositional similarity refers to the degree of similarity between the molecular content of a petroleum fraction and the molecular content of a reference data. It can be quantitatively calculated using similarity algorithms. Calculation methods include distance-based methods and non-distance-based methods. Distance-based methods include Euclidean distance (ED), as shown in formula (5):

[0082] ED=(∑(m pred，i -m ref，i ) 2 ) 1 / 2 (5)

[0083] Wherein, ED represents the Euclidean distance between the molecular content of the petroleum fraction and the molecular content of the reference data, with a value between 0 and +∞. The smaller the value, the more similar the two are. pred，i This represents the molecular content of molecule i in the composition of a petroleum fraction, m. ref，i ∑ represents the molecular content of molecule i in the reference data, and ∑ represents the summation symbol.

[0084] The non-distance-based calculation methods include cosine similarity and Kullback-Leibler (KL) divergence. Cosine similarity is shown in formula (4) above, and KL divergence is shown in formula (6).

[0085]

[0086] Wherein, KL represents the KL divergence value between the molecular content of the petroleum fraction and the molecular content of the reference data, and its value ranges from 0 to 1. The smaller the value, the more similar the two are.

[0087] S206: Determine the composition of the petroleum fractions of the petroleum to be tested to meet the similarity requirements in terms of compositional similarity and macroscopic physical property similarity.

[0088] For example, in this step, determining the petroleum fraction composition of the petroleum to be tested to meet the similarity requirements of compositional similarity and macroscopic property similarity includes: determining whether the macroscopic property similarity is less than the reference macroscopic property similarity of the petroleum to be tested, where the reference macroscopic property similarity is the currently determined optimal macroscopic property similarity; if the macroscopic property similarity is not less than the reference macroscopic property similarity of the petroleum to be tested, then the macroscopic property similarity is determined as the reference macroscopic property similarity of the petroleum to be tested; determining whether the compositional similarity is less than the reference compositional similarity of the petroleum to be tested, where the reference compositional similarity is the currently determined optimal compositional similarity; if the compositional similarity is not less than the reference compositional similarity of the petroleum to be tested, then the compositional similarity is determined as the reference compositional similarity of the petroleum to be tested; determining the petroleum fraction composition of the petroleum to be tested to meet the similarity requirements of reference compositional similarity and reference macroscopic property similarity. For example, such as Figure 5 As shown, after determining the petroleum fractions of the petroleum to be tested as petroleum fractions that meet the similarity requirements of reference composition similarity and reference macroscopic physical property similarity, the petroleum fraction composition can be obtained. For example, there may be multiple petroleum fraction compositions that meet the requirements, so a set of petroleum fraction compositions will be obtained. Figure 5 A flowchart illustrating a method for detecting the composition of petroleum fractions provided in another embodiment of this application is shown below. Figure 2 Further refinement based on this. For example, such as... Figure 5 As shown, the input data includes physical property data and reference data. The reference data includes molecular characterization data and molecular composition models. The initialization of petroleum fraction composition includes a pre-set molecular library, a combination of probability density functions, and a range of probability density function parameters. The pre-set molecular library contains the molecular structures that the diesel fuel to be tested may contain. After the molecular structure is determined, the molecular properties can be determined. Based on the molecular properties and molecular content, the macroscopic physical properties (i.e., petroleum properties) of the petroleum fraction composition can be determined. The macroscopic physical properties include elemental content and group composition. Then, the macroscopic physical properties of the petroleum fraction composition are compared with the macroscopic physical properties of the physical property data to determine the similarity of macroscopic physical properties.

[0089] For example, in this step, such as Figure 5 As shown, determining the composition of a petroleum fraction of the petroleum to be tested as a petroleum fraction whose compositional similarity and macroscopic property similarity meet the similarity requirements further includes: if the macroscopic property similarity is less than the reference macroscopic property similarity of the petroleum to be tested, then performing the step of selecting a target value within the range and assigning it to the corresponding probability density function parameter; if the compositional similarity is less than the reference compositional similarity of the petroleum to be tested, then performing the step of selecting a target value within the range and assigning it to the corresponding probability density function parameter; if the reference compositional similarity and the reference macroscopic property similarity do not meet the similarity requirements, then performing the step of selecting a target value within the range and assigning it to the corresponding probability density function parameter.

[0090] For example, in this step, such as Figure 5 As shown, determining the petroleum distillate groups of the petroleum to be tested to meet the similarity requirements of compositional similarity and macroscopic property similarity further includes: determining the weighted sum of compositional similarity and macroscopic property similarity; determining the petroleum distillate groups of the petroleum to be tested to meet the similarity requirements of the weighted sum; if the weighted sum does not meet the similarity requirements, then the step of selecting a target value within the range and assigning it to the corresponding probability density function parameter is performed. For example, after determining that the petroleum distillate groups of the petroleum to be tested meet the similarity requirements of the weighted sum of reference compositional similarity and reference macroscopic property similarity, a petroleum distillate composition can be obtained. For example, multiple petroleum distillate compositions may meet the requirements, so a set of petroleum distillate compositions will be obtained.

[0091] Furthermore, such as Figure 5 As shown, selecting a target value within a range and assigning it to the corresponding probability density function parameter includes: selecting a target value within a range and assigning it to the corresponding probability density function parameter based on a bi-objective optimization algorithm, wherein the bi-objective optimization algorithm includes a genetic algorithm with elitist strategy and non-dominated sorting; and / or selecting a target value within a range and assigning it to the corresponding probability density function parameter based on a single-objective optimization algorithm, wherein the single-objective optimization algorithm includes a genetic algorithm.

[0092] For example, when using a direct optimization algorithm, the optimal compositional similarity and macroscopic property similarity can be iteratively calculated using a bi-objective optimization algorithm. During the iteration process, the optimization variable retains a series of optimal solutions that are dominated by either compositional similarity or macroscopic property similarity, and are not dominated by other solutions, forming the optimal solution set for petroleum fraction composition. When using an indirect optimization algorithm, the optimization variable strives to find the optimal weighted sum of compositional similarity and macroscopic property similarity during the iteration process, ultimately determining a single optimal solution. Specifically, the bi-objective optimization algorithm used in this application is a Genetic Algorithm for Non-dominated Sorting (NSGA-II) with an elitist strategy, and the single-objective optimization algorithm is a Genetic Algorithm (GA).

[0093] For example, taking cosine similarity as the optimization objective of composition similarity, the mathematical model of the optimization problem of the detection method of petroleum fraction composition using the direct optimization algorithm is shown in formula (7):

[0094]

[0095] For example, taking the weighted sum of compositional similarity and macroscopic property similarity as the optimization objective, the mathematical model of the optimization problem of the petroleum fraction composition detection method using the indirect optimization algorithm is shown in Equation (8):

[0096]

[0097] Where x is the optimization variable, x low and x up The minimum and maximum values ​​of the optimization variables are x and x', respectively, which can be determined using experimental data and the probability density function. low With x up ω is the weighting factor between the two objective functions.

[0098] In the petroleum fraction composition detection method provided in this application embodiment, composition similarity is added on the basis of macroscopic physical property similarity for petroleum fraction composition detection, which can make the detected petroleum molecule content distribution more consistent with the actual petroleum molecule content distribution in oil products; in addition, once the petroleum molecule content distribution is determined, it can also solve the problem of multiple solutions for petroleum fraction composition optimization.

[0099] In some embodiments, this application describes a method for detecting the composition of diesel fractions using GC-FI TOF MS data as reference data. Figure 6 A schematic flowchart of a method for detecting the composition of diesel fractions provided in an embodiment of this application includes the following steps:

[0100] S601: Obtain the physical property data and GC-FI TOF MS data of the diesel fuel to be tested.

[0101] The relevant instructions for this step are similar to those for step S201, and will not be repeated here.

[0102] S602: Based on the physical property data and GC-FI TOF MS data, the composition of the diesel fraction corresponding to the diesel to be tested is initialized to obtain initialization data. The initialization data includes a pre-set molecular library, the combination form of the probability density function, and the value range of the probability density function parameters. The pre-set molecular library contains the molecular structures that the diesel to be tested may contain.

[0103] The relevant instructions for this step are similar to those for step S202, and will not be repeated here.

[0104] Specifically, the probability density function combination is set as follows: ① The Histogram function is used to classify compounds of different types; ② The Histogram function is used to further classify compounds with different core structures under different types of compound classification; ③ The Gamma function is used to extract the boiling points of all molecules.

[0105] S603: Select the target value within the range of values ​​for the probability density function parameter and assign it to the corresponding probability density function parameter;

[0106] The relevant instructions for this step are similar to those for step S203, and will not be repeated here.

[0107] S604: Determine the molecular content of diesel fraction composition based on a pre-set molecular library, the combination of probability density functions, and the assigned probability density function parameters.

[0108] The relevant instructions for this step are similar to those for step S204, and will not be repeated here.

[0109] S605: Determine the macroscopic property similarity based on molecular content and physical property data, and determine the compositional similarity based on molecular content and GC-FITOF MS data.

[0110] The relevant instructions for this step are similar to those for step S205, and will not be repeated here.

[0111] Specifically, a pre-set molecular library contains the possible molecular structures of the diesel fuel to be tested. Once the molecular structure is determined, the molecular properties can be identified. Based on the molecular properties and molecular content, the macroscopic properties of the diesel fuel can be determined, including elemental content and group composition. The macroscopic properties of the diesel fuel are then compared with those in the physical property data to determine the similarity of the macroscopic properties.

[0112] S606: Determine the composition of the diesel fuel to be tested as a diesel fuel fraction whose compositional similarity and macroscopic physical property similarity meet the similarity requirements.

[0113] The relevant instructions for this step are similar to those for step S206, and will not be repeated here.

[0114] S607: Optimize the composition of diesel fractions using direct optimization algorithms and / or indirect optimization algorithms.

[0115] The relevant descriptions of the direct optimization algorithm and / or indirect optimization algorithm are as described in the above embodiments, and will not be repeated here.

[0116] Specifically, based on macroscopic property similarity and compositional similarity, direct optimization algorithms and indirect optimization algorithms were used to optimize the composition of diesel fractions, respectively. The results are as follows: Figure 7 As shown. Figure 7 The results show the composition of diesel fractions detected using GC-FI TOF MS data as reference data. The compositional similarity was determined using a cosine similarity algorithm, with a weight factor of 1 for the indirect optimization algorithm. A traditional method was used as the comparison group (i.e., petroleum molecule reconstruction based on macroscopic property similarity). Figure 7 In (1) and (2), the similarity of macroscopic physical properties is compared, including density, sulfur content, nitrogen content, and volume fraction-distillation temperature. Figure 7 Figures (3) to (6) compare compositional similarity, including the number of carbon atoms and the number of equivalent double bonds (DBE). Each circle in the figure represents a molecule, and the size of the circle indicates the molecular content. The results in the figure show that the macroscopic properties of the diesel fraction detected using the traditional method are consistent with the physical property data, but the molecular content distribution differs significantly from the GC-FI TOF MS data. The macroscopic properties of the diesel fraction detected using macroscopic property similarity and compositional similarity are consistent with the macroscopic properties of the physical property data, and the molecular content distribution is consistent with the GC-FI TOF MS data. This indicates that, using GC-FI TOF MS data as a reference, the method for detecting the composition of diesel fractions based on macroscopic property similarity and compositional similarity can accurately detect the composition of diesel fractions. Specifically, comparing the two optimization algorithms, the indirect optimization algorithm is more efficient than the direct optimization algorithm.

[0117] Specifically, to demonstrate the good repeatability of the method proposed in this application, the optimization process of the diesel fraction composition detection method was repeated three times, and the results are as follows: Figure 8 As shown. Figure 8 To determine the repeatability of diesel fraction composition test results using GC-FI TOF MS data as reference data, such as... Figure 8 As shown, the results of the three optimizations are consistent with the macroscopic physical properties and molecular content data of both the physical property data and the GC-FI TOF MS data. This indicates that the method for detecting the composition of diesel fractions based on macroscopic physical property similarity and compositional similarity, using GC-FI TOF MS data as reference data, has good repeatability.

[0118] Furthermore, this application uses GC-FI TOF MS data as reference data and employs a method for detecting the composition of petroleum fractions based on macroscopic property similarity and compositional similarity to detect the composition of different types of diesel fractions. The results are as follows: Figure 9 As shown. Figure 9 To compare the experimental and calculated macroscopic physical properties of multiple diesel fractions using GC-FI TOF MS data as reference data, such as... Figure 9 As shown, the sulfur content analysis data, density calculation values, distillation curve calculation values, and group composition data all show good agreement with the experimental values.

[0119] Specifically, the molecular content distribution of different types of diesel fractions was compared with GC-FI TOF MS data as follows: Figure 10 As shown, Figure 10 To compare the experimental and calculated values ​​of molecular content distribution of multiple diesel fractions using GC-FI TOF MS data as reference data, the following results were obtained: Figure 10(1), (3), and (5) are diesel samples obtained using different refining processes, resulting in different calculated carbon number-DBE distribution values. Figure 10 (2), (4), and (6) are experimental values ​​for detecting the molecular content distribution of diesel fractions in different diesel samples, such as... Figure 10 As shown, the molecular content distribution is consistent with the GC-FI TOF MS data, proving that the detection method for the composition of petroleum fractions based on macroscopic property similarity and compositional similarity has universality.

[0120] Furthermore, this application uses historically tested diesel fractions as reference data to detect diesel fraction composition. For example, when applying methods for detecting petroleum fraction composition based on macroscopic property similarity and compositional similarity to the actual refinery's oil fraction composition detection, a lack of detailed oil characterization data is encountered. Therefore, this application uses historically tested diesel fractions as reference data to optimize the current oil fraction composition. Taking diesel fraction composition detection as an example, in actual refinery production, macroscopic property data of diesel is periodically tested, but molecular characterization data is not frequently tested. Therefore, diesel samples with molecular characterization data can be used as reference oils to establish the diesel fraction composition of the reference oil, which is then used to optimize the diesel fraction composition for which molecular characterization data has not been tested at other times.

[0121] Specifically, based on actual refinery processing data, the diesel fraction of the reference oil was used as the reference data to construct the diesel fraction composition for the remaining days, and the diesel fraction composition detected by traditional methods was used as the control group. The results are as follows: Figure 11 , Figure 12 As shown. Figure 11 To use historically tested diesel fractions as reference data to determine the macroscopic physical properties of diesel fraction composition, such as... Figure 11 As shown, at the macroscopic property level, the molecular reconstruction method based on macroscopic property similarity and compositional similarity (i.e., the detection method for petroleum fraction composition based on macroscopic property similarity and compositional similarity) detects diesel fraction composition well, adapting to fluctuations in sulfur content and distillation curves. The diesel fraction composition detected by the molecular reconstruction method based on macroscopic property similarity also adapts to fluctuations in the macroscopic properties of diesel. The two methods have similar accuracy in macroscopic property detection. Figure 12 To determine the molecular content distribution of diesel fraction composition using historically tested diesel fractions as reference data, such as... Figure 12As shown, at the molecular composition level, the molecular content distribution in the diesel fraction detected by the molecular reconstruction method based on macroscopic property similarity and compositional similarity is consistent with the molecular content distribution in the diesel fraction detected by the reference oil. However, the molecular content distribution in the diesel fraction detected by the molecular reconstruction method based on macroscopic property similarity differs significantly from that in the diesel fraction detected by the reference oil. This indicates that the molecular reconstruction method based on macroscopic property similarity and compositional similarity can effectively ensure the accuracy of molecular content distribution when detailed oil characterization data is lacking, resulting in higher precision in the detected diesel fraction composition.

[0122] Furthermore, to demonstrate the good reproducibility of the method proposed in this study, the optimization process of the diesel fraction composition detection method was repeated three times, and the diesel fraction composition detected by the traditional method was used as a control group. The results are as follows: Figure 13 As shown. Figure 13 To verify the repeatability of diesel fraction composition test results using historically tested diesel fractions as reference data, such as... Figure 13 As shown, compared with traditional methods, the molecular reconstruction method based on macroscopic property similarity and compositional similarity showed only small fluctuations in the group composition content in the three optimization results, indicating that the detection of diesel fraction composition using historically detected diesel fraction composition as reference data has good repeatability.

[0123] The following are embodiments of the apparatus described in this application, which can be used to execute the embodiments of the method described in this application. For details not disclosed in the apparatus embodiments of this application, please refer to the embodiments of the method described in this application.

[0124] Figure 14 This is a schematic diagram of a petroleum fraction composition detection device provided in one embodiment of this application. For ease of explanation, only the parts relevant to the embodiment of this application are shown. Figure 14 As shown, the petroleum fraction composition detection device 1400 includes: an acquisition module 1401, an initialization module 1402, a parameter assignment module 1403, a molecular content determination module 1404, a similarity calculation module 1405, and a petroleum fraction composition determination module 1406. Among them,

[0125] The acquisition module 1401 is used to acquire the physical property data and reference data of the petroleum to be tested;

[0126] The initialization module 1402 is used to initialize the composition of the petroleum fraction corresponding to the petroleum to be tested based on the physical property data and reference data, and obtain the initialization data. The initialization data includes a preset molecular library, a combination form of probability density function and a range of values ​​for the probability density function parameters. The preset molecular library contains the molecular structures that the petroleum to be tested may contain.

[0127] The parameter assignment module 1403 is used to select a target value within the range of values ​​and assign it to the corresponding probability density function parameter;

[0128] The molecular content determination module 1404 is used to determine the molecular content of petroleum fraction composition based on a pre-set molecular library, the combination form of probability density functions, and the assigned probability density function parameters.

[0129] The similarity calculation module 1405 is used to determine the macroscopic physical property similarity based on molecular content and physical property data, and to determine the compositional similarity based on molecular content and reference data.

[0130] The petroleum fraction composition determination module 1406 is used to determine the petroleum fraction composition of the petroleum to be tested that meets the similarity requirements in terms of compositional similarity and macroscopic physical property similarity.

[0131] In one possible implementation, the petroleum fraction composition determination module 1406 can be specifically used to: determine the petroleum fraction composition of the petroleum to be tested that meets the similarity requirements of composition similarity and macroscopic property similarity, including: determining whether the macroscopic property similarity is less than the reference macroscopic property similarity of the petroleum to be tested, wherein the reference macroscopic property similarity is the currently determined optimal macroscopic property similarity; if the macroscopic property similarity is not less than the reference macroscopic property similarity of the petroleum to be tested, then the macroscopic property similarity is determined as the reference macroscopic property similarity of the petroleum to be tested; determining whether the composition similarity is less than the reference composition similarity of the petroleum to be tested, wherein the reference composition similarity is the currently determined optimal composition similarity; if the composition similarity is not less than the reference composition similarity of the petroleum to be tested, then the composition similarity is determined as the reference composition similarity of the petroleum to be tested; and determining that the petroleum fraction composition of the petroleum to be tested meets the similarity requirements of reference composition similarity and reference macroscopic property similarity.

[0132] In one possible implementation, the petroleum fraction composition determination module 1406 can also be specifically used to: determine the petroleum fraction composition of the petroleum to be tested that meets the similarity requirements of composition similarity and macroscopic property similarity, including: if the macroscopic property similarity is less than the reference macroscopic property similarity of the petroleum to be tested, then perform the step of selecting a target value within the value range and assigning it to the corresponding probability density function parameter; if the composition similarity is less than the reference composition similarity of the petroleum to be tested, then perform the step of selecting a target value within the value range and assigning it to the corresponding probability density function parameter; if the reference composition similarity and the reference macroscopic property similarity do not meet the similarity requirements, then perform the step of selecting a target value within the value range and assigning it to the corresponding probability density function parameter.

[0133] In one possible implementation, the petroleum fraction composition determination module 1406 can also be specifically used to: determine the petroleum fraction composition of the petroleum to be tested that meets the similarity requirements of composition similarity and macroscopic property similarity, including: determining the weighted sum of composition similarity and macroscopic property similarity; determining the petroleum fraction composition of the petroleum to be tested that meets the similarity requirements of the weighted sum; if the weighted sum does not meet the similarity requirements, then performing the step of selecting a target value within the value range and assigning it to the corresponding probability density function parameter.

[0134] In one possible implementation, the parameter assignment module 1403 is specifically used to: select a target value within a range and assign it to the corresponding probability density function parameter, including: selecting a target value within a range and assigning it to the corresponding probability density function parameter based on a bi-objective optimization algorithm, wherein the bi-objective optimization algorithm includes a genetic algorithm with elitist strategy and non-dominated sorting; and / or, selecting a target value within a range and assigning it to the corresponding probability density function parameter based on a single-objective optimization algorithm, wherein the single-objective optimization algorithm includes a genetic algorithm.

[0135] In one possible implementation, the similarity calculation module 1405 can be specifically used to: determine the macroscopic property similarity based on molecular content and physical property data, including: determining the macroscopic property similarity according to the following formula:

[0136] Prop Sim=-mean(δ|P ref -P pred |)

[0137] Where PropSim represents macroscopic property similarity, P ref It is a macroscopic property determined based on physical property data, P pred It is the macroscopic physical property of petroleum fraction composition determined based on molecular content, δ is the weighting factor of the physical property data, and mean represents the mean value.

[0138] In one possible implementation, the similarity calculation module 1405 can also be specifically used to: determine compositional similarity based on molecular content and reference data, including: determining compositional similarity according to the following formula:

[0139]

[0140] Where COS represents compositional similarity, m ref，i m represents the molecular content of molecule i in the reference data. pred，i ∑ represents the molecular content of molecule i in the composition of petroleum fraction, and ∑ represents the summation symbol.

[0141] The petroleum fraction composition detection device provided in this application embodiment has a similar implementation principle and technical effect to the above embodiments. For details, please refer to the above embodiments, which will not be repeated here.

[0142] Figure 15 This is a schematic diagram of the structure of a petroleum fraction composition detection device provided in one embodiment of this application. Figure 15 As shown, the petroleum fraction composition detection device 1500 includes: a processor 1510, which may include one or more; a memory 1520 for storing program instructions, such as data 1521 or application program 1522, which can be executed by the processor 1510, wherein the data includes physical property data and reference data, the memory 1520 may be temporary storage or persistent storage, and the processor 1510 is used to call the program instructions in the memory 1520 to execute the above-mentioned petroleum fraction composition detection method.

[0143] The petroleum fraction composition testing device 1500 may further include one or more power supply components 1530; one or more wired or wireless network interfaces 1540 for connecting the petroleum fraction composition testing device 1500 to a network; one or more input / output interfaces 1550; and / or one or more operating systems 1523, such as Windows Server™, Mac OS X™, Unix™, Linux™, FreeBSD™, etc. Those skilled in the art will understand that... Figure 15 The petroleum fraction composition detection device shown does not constitute a limitation on the petroleum fraction composition detection device, and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0144] This application also provides a computer-readable storage medium storing computer-executable instructions, which, when executed, are used to implement the above-mentioned method for detecting the composition of petroleum fractions.

[0145] This application also provides a computer program product, including a computer program that, when executed, implements the above-mentioned method for detecting the composition of petroleum fractions.

[0146] This application also provides a chip for executing instructions, which is used to perform the method for detecting the composition of petroleum fractions as described in any of the above method embodiments.

[0147] Other embodiments of this application will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of this application that follow the general principles of this application and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of this application are indicated by the following claims.

[0148] It should be understood that this application is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this application is limited only by the appended claims.

Claims

1. A method for detecting the composition of petroleum fractions, characterized in that, include: Obtain physical property data and reference data of the petroleum to be tested; Based on the physical property data and the reference data, the composition of the petroleum fraction corresponding to the petroleum to be tested is initialized to obtain initialization data. The initialization data includes a preset molecular library, a combination of probability density functions, and a range of values ​​for the probability density function parameters. The preset molecular library contains the molecular structures that the petroleum to be tested may contain. Select a target value within the range of values ​​and assign it to the corresponding probability density function parameter; Based on the pre-set molecular library, the combination form of the probability density function, and the assigned probability density function parameters, the molecular content of the petroleum fraction composition is determined; Based on the molecular content and the physical property data, the macroscopic physical property similarity is determined, and based on the molecular content and the reference data, the compositional similarity is determined. The petroleum fractions of the petroleum to be tested are determined to be petroleum fractions whose compositional similarity and macroscopic physical property similarity meet the similarity requirements; The process of determining the petroleum fraction composition of the petroleum to be tested to meet the similarity requirements of compositional similarity and macroscopic physical property similarity includes: Determine whether the macroscopic property similarity is less than the reference macroscopic property similarity of the oil to be tested, wherein the reference macroscopic property similarity is the currently determined optimal macroscopic property similarity; If the macroscopic property similarity is not less than the reference macroscopic property similarity of the oil to be tested, then the macroscopic property similarity is determined as the reference macroscopic property similarity of the oil to be tested. Determine whether the compositional similarity is less than the reference compositional similarity of the petroleum to be detected, wherein the reference compositional similarity is the currently determined optimal compositional similarity; If the compositional similarity is not less than the reference compositional similarity of the petroleum to be tested, then the compositional similarity is determined to be the reference compositional similarity of the petroleum to be tested. The petroleum fractions of the petroleum to be tested are determined to be petroleum fractions whose similarity to reference composition and reference macroscopic physical properties meets the similarity requirements.

2. The detection method according to claim 1, characterized in that, Also includes: If the macroscopic property similarity is less than the reference macroscopic property similarity of the petroleum to be detected, then the step of selecting the target value within the range and assigning it to the corresponding probability density function parameter is performed. If the compositional similarity is less than the reference compositional similarity of the petroleum to be detected, then the step of selecting a target value within the range and assigning it to the corresponding probability density function parameter is performed. If the reference composition similarity and reference macroscopic property similarity do not meet the similarity requirements, then the step of selecting the target value within the range and assigning it to the corresponding probability density function parameter is performed.

3. The detection method according to claim 1, characterized in that, The process of determining the petroleum fraction composition of the petroleum to be tested to meet the similarity requirements of compositional similarity and macroscopic physical property similarity includes: Determine the weighted sum of compositional similarity and macroscopic property similarity; The petroleum fractions of the petroleum to be tested are determined to be the weighted sum of petroleum fractions that meet the similarity requirements; If the weighted sum does not meet the similarity requirement, then the step of selecting the target value within the range and assigning it to the corresponding probability density function parameter is performed.

4. The detection method according to any one of claims 1 to 3, characterized in that, The step of selecting a target value within the range and assigning it to the corresponding probability density function parameter includes: Based on the bi-objective optimization algorithm, an objective value is selected within the range of values ​​and assigned to the corresponding probability density function parameter. The bi-objective optimization algorithm includes a genetic algorithm with elitist strategy and non-dominated sorting. And / or, based on a single-objective optimization algorithm, a target value is selected within the range of values ​​and assigned to the corresponding probability density function parameter, wherein the single-objective optimization algorithm includes a genetic algorithm.

5. The detection method according to any one of claims 1 to 3, characterized in that, The step of determining the macroscopic property similarity based on the molecular content and the physical property data includes: The macroscopic property similarity is determined according to the following formula: Where PropSim represents the macroscopic property similarity, P ref It is a macroscopic property determined based on the aforementioned physical property data, P pred These are the macroscopic physical properties of the petroleum fraction composition determined based on the aforementioned molecular content. It is the weighting factor of the physical property data, and mean represents the sign of the mean.

6. The detection method according to any one of claims 1 to 3, characterized in that, The determination of compositional similarity based on the molecular content and the reference data includes: The compositional similarity is determined according to the following formula: Where COS represents the compositional similarity, m ref，i The m represents the molecular content of molecule i in the reference data. pred，i The value represents the molecular content of molecule i in the composition of the petroleum fraction, and ∑ represents the summation symbol.

7. A device for detecting the composition of petroleum fractions, characterized in that, include: The acquisition module is used to acquire the physical property data and reference data of the petroleum to be tested; An initialization module is used to initialize the petroleum fraction composition corresponding to the petroleum to be tested based on the physical property data and the reference data to obtain initialization data. The initialization data includes a preset molecular library, a combination of probability density functions, and a range of values ​​for the probability density function parameters. The preset molecular library contains the molecular structures that the petroleum to be tested may contain. The parameter assignment module is used to select a target value within the range of values ​​and assign it to the corresponding probability density function parameter; The molecular content determination module is used to determine the molecular content of the petroleum fraction composition based on the preset molecular library, the combination form of the probability density function, and the assigned probability density function parameters. The similarity calculation module is used to determine the macroscopic property similarity based on the molecular content and the physical property data, and to determine the compositional similarity based on the molecular content and the reference data. A petroleum fraction composition determination module is used to determine the petroleum fraction composition of the petroleum to be tested that meets the similarity requirements of composition similarity and macroscopic property similarity. The petroleum fraction composition determination module is specifically used for: determining whether the macroscopic property similarity is less than the reference macroscopic property similarity of the petroleum to be tested, wherein the reference macroscopic property similarity is the currently determined optimal macroscopic property similarity; if the macroscopic property similarity is not less than the reference macroscopic property similarity of the petroleum to be tested, then determining the macroscopic property similarity as the reference macroscopic property similarity of the petroleum to be tested; determining whether the composition similarity is less than the reference composition similarity of the petroleum to be tested, wherein the reference composition similarity is the currently determined optimal composition similarity; if the composition similarity is not less than the reference composition similarity of the petroleum to be tested, then determining the composition similarity as the reference composition similarity of the petroleum to be tested; and determining the petroleum fraction composition of the petroleum to be tested to satisfy the similarity requirements of the reference composition similarity and the reference macroscopic property similarity.

8. A device for detecting the composition of petroleum fractions, characterized in that, include: Memory and processor; The memory is used to store program instructions; The processor is used to call program instructions in the memory to execute the method for detecting the composition of petroleum fractions as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed, are used to implement the method for detecting the composition of petroleum fractions as described in any one of claims 1 to 6.

Images