Method and system for predicting diesel pour point and cold filter plugging point

Abstract

The present application relates to the technical field of diesel oil freezing point and cold filter point prediction, and is a diesel oil freezing point and cold filter point prediction method and system, which comprises the following steps: analyzing and testing diesel oil by an analysis instrument to obtain an actual value of a first property; obtaining main homologues in diesel oil structure and a central molecule set according to the main homologues; expanding the central molecule set according to a preset rule to obtain virtual molecules; applying a preset structure and property relationship to the virtual molecules to obtain a virtual molecule content and a first property calculation function; optimizing the first property calculation function by an optimization algorithm to obtain a relative content of the virtual molecules; obtaining a diesel oil molecule matrix according to the relative content and a structure vector; and sending the diesel oil molecule matrix to a neural network model for prediction to obtain the freezing point and the cold filter point of the diesel oil. The present application can better describe diesel oil at a molecular level, and can more conveniently, economically and accurately predict the freezing point and the cold filter point.

Description

Technical Field

[0001] This invention relates to the field of diesel pour point and cold filter plugging point prediction technology, and provides a method and system for predicting diesel pour point and cold filter plugging point. Background Technology

[0002] Pour point and cold filter plugging point (CFPP) are important indicators characterizing the low-temperature performance of diesel fuel. Pour point (SP) indicates the highest temperature at which diesel fuel loses its fluidity in a cold environment. CFPP indicates the highest temperature at which diesel fuel can clog the fuel filter when passing through the diesel engine's fuel supply system. CFPP is directly related to the low-temperature performance of diesel fuel, while pour point is mainly related to the storage and transportation of diesel fuel. When the diesel fuel temperature drops to the CFPP temperature, the wax crystals formed will clog the filter, affecting the normal fuel supply and consequently impacting the normal operation of the engine.

[0003] With the development of computer and analytical technologies, the development of petroleum processing models has risen to the molecular level, enabling the construction of theoretical models with strong universality based on the chemical composition of raw materials. The rapid development of large-scale analytical techniques such as mass spectrometry, nuclear magnetic resonance, and infrared spectroscopy has provided the necessary conditions for establishing molecular-scale models, while computer technology has provided ample support for handling complex molecular dynamics calculations. Based on the molecular composition of diesel fuel, using a matrix to describe the molecular composition and component content, properties such as molecular weight, elemental content, density, and distillation range in diesel fuel can be obtained using the group contribution method with relatively small errors. However, the calculated pour point and cold filter plugging point deviate significantly from the actual values. Summary of the Invention

[0004] This invention provides a method and system for predicting the pour point and cold filter plugging point of diesel fuel, overcoming the shortcomings of the prior art and effectively solving the problem of large deviations between the calculated pour point and cold filter plugging point and the actual values.

[0005] One of the technical solutions of this invention is achieved through the following measures: a method for predicting the pour point and cold filter plugging point of diesel fuel, comprising the following steps:

[0006] Diesel fuel was analyzed and tested using analytical instruments to obtain the actual values ​​of its primary properties; the main homologues in the diesel fuel structure were identified, and the central molecule set was obtained based on the main homologues.

[0007] The central molecule set is expanded according to preset rules to obtain virtual molecules; the preset structure and property relationship is applied to the virtual molecules to obtain the virtual molecule content and the first property calculation function;

[0008] The calculation function of the first property is optimized by optimizing the algorithm, and the error between the actual value and the calculated value of the first property is limited to less than a preset value, so as to obtain the relative content of virtual molecules in virtual diesel.

[0009] The diesel fuel molecular matrix is ​​obtained based on the relative content and structure vector of the virtual molecules; the diesel fuel molecular matrix is ​​then fed into a neural network model for prediction to obtain the pour point and cold filter plugging point of diesel fuel.

[0010] The following are further optimizations and / or improvements to one of the above-mentioned technical solutions:

[0011] The above-mentioned method of feeding the diesel molecular matrix into a neural network model for prediction to obtain the pour point and cold filter plugging point of diesel may specifically include the following steps: dividing the diesel molecular matrix into a training set and a prediction set; inputting the training set into an initial neural network for training to obtain a neural network model; and inputting the prediction set into the neural network model for prediction to obtain the pour point and cold filter plugging point of diesel.

[0012] When expanding the central molecule set according to the above-mentioned preset rules to obtain virtual molecules, the specific steps may include: adding methyl groups to the central molecules of the central molecule set according to the preset rules to obtain virtual molecules; wherein, the preset rules include one or a combination of the following: not considering isomerism, only changing the carbon number of alkanes, prioritizing the increase of the number of methyl groups in the ring structure, and cycloalkanes being preferred over aromatic rings when adding side chains to the ring.

[0013] The aforementioned set of central molecules may include 40 to 70 types of central molecules.

[0014] The aforementioned pre-defined structure-property relationship may include one or a combination of the following: structure and elemental content, structure and average molecular weight, structure and saturated fraction content, structure and aromatic fraction content, and structure and boiling point.

[0015] The optimization algorithms mentioned above may include one or a combination of the following: ant colony optimization, simulated annealing, genetic algorithm, and artificial neural network algorithm.

[0016] The aforementioned neural network model may include one or a combination of the following: single-layer neural network, feedforward neural network, radial basis function network, deep feedforward network, and recurrent neural network.

[0017] The second technical solution of the present invention is achieved through the following measures: a prediction system for diesel pour point and cold filter plugging point, comprising:

[0018] The analysis module is used to analyze and test diesel fuel using analytical instruments to obtain the actual values ​​of its primary properties.

[0019] The acquisition module is used to acquire the main homologues in the diesel fuel structure and obtain the central molecule set based on the main homologues.

[0020] The expansion module is used to expand the central molecule set according to preset rules to obtain virtual molecules;

[0021] The application module is used to apply preset structure and property relationships to virtual molecules to obtain virtual molecule content and first property calculation functions;

[0022] The optimization module is used to optimize the calculation function of the first property through an optimization algorithm, limiting the error between the actual value and the calculated value of the first property to be less than a preset value, and obtaining the relative content of virtual molecules in virtual diesel.

[0023] The matrix acquisition module is used to obtain the diesel fuel molecule matrix based on the relative content and structure vector of virtual molecules;

[0024] The prediction module is used to feed the diesel fuel molecular matrix into a neural network model for prediction, thereby obtaining the pour point and cold filter plugging point of diesel fuel.

[0025] The third technical solution of the present invention is achieved through the following measures: a computer device, including a memory and a processor, wherein the memory stores a program that can run on the processor, and the processor executes the program to implement the above-mentioned method for predicting the pour point and cold filter plugging point of diesel fuel.

[0026] The fourth technical solution of the present invention is achieved through the following measures: a storage medium storing one or more programs, which can be executed by one or more processors to realize the above-mentioned method for predicting the pour point and cold filter plugging point of diesel fuel.

[0027] This invention analyzes and tests diesel fuel using specialized instruments to obtain the actual values ​​of its primary properties. Based on the main homologues in the diesel fuel structure, molecules with similar reaction characteristics or structural features are grouped to obtain a set of central molecules. These central molecules are then used to generate virtual molecules, which are then used to calculate the virtual molecule content and primary properties using a preset structure-property relationship. An optimization algorithm is employed to ensure that the error between the calculated and actual property values ​​is less than a preset value, thus obtaining the relative content of each molecule in the virtual diesel fuel. A diesel fuel molecular matrix is ​​obtained by combining a structural vector with the molecular content relationship. Using a neural network, the diesel fuel molecular matrix is ​​input to obtain the pour point and cold filter plugging point (CPP) of the diesel fuel. This prediction method has high accuracy in predicting the pour point and CPP of diesel fuel. Utilizing the raw material molecular matrix allows for a better description of diesel fuel at the molecular level. This invention uses neural networks to analyze complex diesel fuel molecular structure information and predict molecular properties, enabling more convenient, economical, and accurate prediction of pour point and CPP. The matrix format facilitates editing and storage, and is easy for computers to recognize. This invention can better describe diesel fuel at the molecular level and can more conveniently, economically, and accurately predict pour point and CPP. Attached Figure Description

[0028] Appendix Figure 1This is a schematic flowchart of the method for predicting the pour point and cold filter plugging point of diesel fuel according to Embodiment 1 of the present invention.

[0029] Appendix Figure 2 This is a schematic flowchart of the method for predicting the pour point and cold filter plugging point of diesel fuel according to Embodiment 2 of the present invention.

[0030] Appendix Figure 3 This is a schematic flowchart of the method for predicting the pour point and cold filter plugging point of diesel fuel in Embodiment 3 of the present invention.

[0031] Appendix Figure 4 This is a schematic block diagram of the diesel pour point and cold filter plugging point prediction system of Embodiment 4 of the present invention.

[0032] Appendix Figure 5 This is a schematic block diagram of the structure of the computer device according to Embodiment 6 of the present invention.

[0033] Appendix Figure 6 This is a schematic flowchart of the method for predicting the pour point and cold filter plugging point of diesel fuel in Embodiment 5 of the present invention.

[0034] Appendix Figure 7 This is a flowchart of the method for predicting the pour point and cold filter plugging point of diesel fuel according to Embodiment 5 of the present invention.

[0035] Appendix Figure 8 This is a conceptual diagram illustrating the method for predicting the pour point and cold filter plugging point of diesel fuel in Embodiment 5 of the present invention.

[0036] Appendix Figure 9 This is a schematic diagram of the central molecule of straight-run diesel fuel, which is the method for predicting the pour point and cold filter plugging point of diesel fuel in Example 5 of the present invention. Detailed Implementation

[0037] The present invention is not limited to the following embodiments, and the specific implementation can be determined according to the technical solution of the present invention and the actual situation.

[0038] The present invention will be further described below with reference to embodiments:

[0039] Example 1: As Figure 1 As shown, this embodiment of the invention provides a method for predicting the pour point and cold filter plugging point of diesel fuel, including the following steps:

[0040] Step S102: Analyze and test the diesel fuel using an analytical instrument to obtain the actual value of the first property;

[0041] Step S104: Expand the central molecule set according to preset rules to obtain virtual molecules;

[0042] Step S106: Apply the preset structure-property relationship to the virtual molecule to obtain the virtual molecule content and the first property calculation function;

[0043] Step S108: Optimize the calculation function of the first property through an optimization algorithm, and limit the error between the actual value and the calculated value of the first property to be less than a preset value, so as to obtain the relative content of virtual molecules in virtual diesel; wherein, the preset value is generally less than 0.01 to 0.05.

[0044] Step S110: Obtain the diesel fuel molecular matrix based on the relative content and structure vector of the virtual molecules;

[0045] Step S112: Input the diesel molecular matrix into the neural network model for prediction to obtain the pour point and cold filter plugging point of diesel.

[0046] According to the diesel pour point and cold filter plugging point prediction method provided in this embodiment, diesel is first analyzed and tested using a certain analytical instrument to obtain the actual value of the first property. Then, based on the main homologues in the diesel structure, molecules with similar reaction characteristics or structural features are grouped to obtain a certain set of central molecules. The set of central molecules is used to obtain virtual molecules according to certain rules. The virtual molecules are used to obtain the virtual molecule content and the first property calculation function by applying a preset structure-property relationship. A certain optimization algorithm is used to limit the error between the property calculation value and the actual value to be less than the preset value, and the relative content of each molecule in the virtual diesel is obtained. Combining a certain structure vector with the molecular content relationship, a diesel molecular matrix is ​​obtained. Using a neural network method, the diesel molecular matrix is ​​input to obtain the pour point and cold filter plugging point of diesel. This prediction method has high accuracy in predicting the pour point and cold filter plugging point of diesel. Using the raw material molecular matrix, diesel can be better described at the molecular level. Applying a neural network to analyze the complex diesel molecular structure information and predict molecular properties can more conveniently, economically and accurately predict the pour point and cold filter plugging point. The matrix form is easy to edit and store, and easy for computers to recognize.

[0047] Example 2: As Figure 2 As shown, the method for predicting the pour point and cold filter plugging point of diesel fuel involves feeding the diesel fuel molecular matrix into a neural network model for prediction, thereby obtaining the pour point and cold filter plugging point of the diesel fuel. Specifically, the method includes the following steps:

[0048] Step S202: Divide the diesel fuel molecular matrix into a training set and a prediction set;

[0049] Step S204: Input the training set into the initial neural network for training to obtain the neural network model;

[0050] Step S206: Input the prediction set into the neural network model for prediction to obtain the pour point and cold filter plugging point of diesel fuel.

[0051] In this embodiment, the diesel fuel molecular matrix is ​​fed into a neural network model for prediction to obtain the pour point and cold filter plugging point (CPP) of diesel fuel. Specifically, the diesel fuel molecular matrix is ​​first divided into a training set and a prediction set. Then, the training set is input into an initial neural network for training to obtain the neural network model. Finally, the prediction set is input into the neural network model for prediction to obtain the pour point and CPP of diesel fuel. By using a neural network to analyze complex diesel fuel molecular structure information and predict molecular properties, the pour point and CPP can be predicted more conveniently, economically, and accurately.

[0052] Example 3: As Figure 3 As shown, the method for predicting the pour point and cold filter plugging point of diesel fuel involves expanding the central molecule set according to preset rules to obtain virtual molecules. Specifically, it includes the following steps:

[0053] Step S302: Add methyl groups to the central molecules of the central molecule set according to preset rules to obtain virtual molecules;

[0054] The preset rules include one or a combination of the following: isomerism is not considered, alkanes only change their carbon number, the ring structure first satisfies the increase of the number of methyl groups, and cycloalkanes are preferred over aromatic rings when a side chain is added to the ring.

[0055] In this embodiment, the set of central molecules is expanded according to preset rules to obtain virtual molecules. Specifically, methyl groups are added to the central molecules according to preset rules, with a priority order for addition, including but not limited to ignoring isomerism, changing only the carbon number of alkanes, prioritizing the increase of the methyl number in ring structures, and prioritizing cycloalkanes over aromatic rings when adding side chains. After eliminating some unreasonable molecules, multiple virtual molecules are obtained, which serve as all molecules describing diesel fuel. Specifically, according to the preset rules, 40-70 central molecules can be expanded to 500-1000 molecules. Among them, there can be 803 virtual molecules.

[0056] In the above embodiments, diesel fuel includes, but is not limited to, straight-run diesel fuel, catalytic diesel fuel, coking diesel fuel, hydrotreated diesel fuel, hydrocracking diesel fuel, or one or more combinations thereof.

[0057] In some embodiments, the analytical instruments include, but are not limited to, one or more combinations of a two-dimensional gas chromatograph-time-of-flight mass spectrometer, a nuclear magnetic resonance spectrometer, a gas chromatograph-mass spectrometer, an inductively coupled plasma optical emission spectrometer, a Fourier transform infrared spectrometer, and an elemental analyzer. Specifically, by analyzing diesel feedstock molecules using a nuclear magnetic resonance spectrometer, the aromatic carbon content, aliphatic carbon content, degree of aromatic ring substitution, and degree of alkyl chain branching of the diesel fuel can be obtained. By using a two-dimensional gas chromatograph-time-of-flight mass spectrometer, the composition and content of the diesel fuel can be obtained.

[0058] In the above embodiments, the first property includes, but is not limited to, elemental content, average molecular weight, saturated fraction content, aromatic fraction content, and one or more combinations of distillate fractions.

[0059] In the above embodiments, the set of central molecules was determined to include 40 to 70 types of central molecules.

[0060] In some embodiments, the preset structure-property relationship includes, but is not limited to, one or more relationships between structure and elemental content, structure and average molecular weight, structure and saturated content, structure and aromatic content, and structure and boiling point.

[0061] In the above embodiments, the optimization algorithm includes, but is not limited to, one or more of the following: ant colony optimization, simulated annealing, genetic algorithm, and artificial neural network algorithm. Specifically, by employing the simulated annealing algorithm, the error between the calculated value and the actual value of the first property is limited to less than 1 × 10⁻⁶. -6 The relative content of each molecule in the virtual diesel was obtained.

[0062] In the above embodiments, the neural network model includes, but is not limited to, single-layer neural networks, feedforward neural networks, radial basis function networks, deep feedforward networks, recurrent neural networks, and other methods.

[0063] In some embodiments, the diesel fuel molecule matrix is ​​obtained based on the relative content and structure vector of virtual molecules, and structural fragments are selected. H2, There are 17 molecules in total, used to describe diesel molecules.

[0064] Example 4: Figure 4 As shown, this embodiment of the invention provides a prediction system for the pour point and cold filter plugging point of diesel fuel, comprising: an analysis module for analyzing and testing diesel fuel using an analytical instrument to obtain the actual value of a first property; an acquisition module for acquiring the main homologues in the diesel fuel structure and obtaining a set of central molecules based on the main homologues; an expansion module for expanding the set of central molecules according to preset rules to obtain virtual molecules; an application module for applying preset structure-property relationships to the virtual molecules to obtain a calculation function for the virtual molecule content and the first property; an optimization module for optimizing the calculation function for the first property using an optimization algorithm, limiting the error between the actual value and the calculated value of the first property to be less than a preset value, and obtaining the relative content of virtual molecules in the virtual diesel fuel; a matrix acquisition module for obtaining a diesel fuel molecule matrix based on the relative content and structure vector of the virtual molecules; and a prediction module for feeding the diesel fuel molecule matrix into a neural network model for prediction to obtain the pour point and cold filter plugging point of the diesel fuel.

[0065] The diesel pour point and cold filter plugging point prediction system provided in this embodiment includes an analysis module, an acquisition module, an expansion module, an application module, an optimization module, a matrix acquisition module, and a prediction module. The analysis module analyzes and tests the diesel fuel using analytical instruments to obtain the actual value of the first property. The acquisition module acquires the main homologues in the diesel fuel structure and obtains the central molecule set based on these homologues. The expansion module expands the central molecule set according to preset rules to obtain virtual molecules. The application module applies preset structure-property relationships to the virtual molecules to obtain the virtual molecule content and the calculation function for the first property. The optimization module optimizes the calculation function for the first property using an optimization algorithm, limiting the error between the actual value and the calculated value of the first property to less than a preset value, thus obtaining the relative content of virtual molecules in the virtual diesel fuel. The matrix acquisition module obtains the diesel fuel molecular matrix based on the relative content and structure vector of the virtual molecules. The prediction module feeds the diesel fuel molecular matrix into a neural network model for prediction to obtain the pour point and cold filter plugging point of the diesel fuel. Utilizing the raw material molecular matrix allows for a better description of diesel fuel at the molecular level. Applying neural networks to analyze complex diesel fuel molecular structures and predict molecular properties enables more convenient, economical, and accurate prediction of pour points and cold filter plugging points. The matrix format facilitates editing and storage, and is easily recognized by computers.

[0066] Example 5: Figure 6 , Figure 7 , Figure 8 and Figure 9 As shown, an embodiment of the present invention provides a specific method for predicting the pour point and cold filter plugging point of diesel fuel, including:

[0067] Step S402: The diesel fuel is analyzed and tested using a certain analytical instrument to obtain the actual value of property a;

[0068] Step S404: Based on the main homologues in the diesel fuel structure, obtain a certain set of central molecules;

[0069] Step S406: The central molecule set is used to obtain virtual molecules according to certain rules;

[0070] Step S408: Apply a certain structure-property relationship b to the virtual molecule to obtain the virtual molecule content and property a calculation function;

[0071] Step S410: Using a certain optimization algorithm, the error between the calculated value and the actual value of property a should be less than a certain value. Solve the function to obtain the relative content of each molecule in the virtual diesel.

[0072] Step S412: Obtain the diesel fuel molecular matrix by combining a certain structural vector with the molecular content relationship;

[0073] Step S414: Predict the freezing point and cold filter plugging point by applying a combination of molecular matrix and neural network.

[0074] Specifically, nuclear magnetic resonance (NMR) was used to analyze the molecules of straight-run diesel feedstock, obtaining the aromatic carbon content, lipid carbon content, degree of aromatic ring substitution, and degree of alkyl chain branching. Two-dimensional gas chromatography-time-of-flight mass spectrometry (GC-TOF-MS) was used to obtain the composition and content of the diesel. The average molecular weight, C, H, O, and elemental contents of heteroatoms N and S were also determined.

[0075] like Figure 9 As shown, based on the main homologues in the straight-run diesel structure, some molecules with similar reaction characteristics or structural features were merged to obtain 46 central molecules.

[0076] The central molecule set is formed by adding -CH3 to the central molecule according to rules. There is a priority order in the addition, including but not limited to not considering isomers, only changing the number of carbons in alkanes, increasing the number of methyl groups in ring structures first, and cycloalkane rings taking precedence over aromatic rings when adding side chains to the ring. After removing some unreasonable molecules, 803 virtual molecules are obtained and used as all molecules to describe diesel.

[0077] The virtual molecule is used to apply the group contribution method to obtain the relationship between structure and properties, including the relationship between structure and elemental content, structure and average molecular weight, structure and saturated fraction content, structure and aromatic fraction content, and structure and distillation fraction, thus obtaining the virtual molecule content and property calculation function.

[0078] The error between the calculated and actual values ​​of the properties, determined by the simulated annealing algorithm, should be less than 1 × 10⁻⁶. -6 The relative content of each molecule in the virtual diesel was obtained.

[0079] By combining the relationship between 16 structural vectors and molecular content, a diesel fuel molecular matrix was obtained, containing 803 molecules. As shown in Table 1, the error between experimental and calculated values ​​is within 5%, verifying the reliability of the matrix.

[0080] Table 1

[0081]

[0082] A neural network was used to analyze the complex molecular structure information of diesel fuel and predict its molecular properties. The molecular matrix was fed into the neural network for training to obtain the pour point and cold filter plugging point of diesel fuel, as shown in Table 2. The error between the calculated and experimental values ​​was within 0.6%.

[0083] Table 2

[0084]

[0085] The beneficial effects of this invention are as follows: (1) By utilizing the molecular matrix of raw materials, diesel fuel can be better described at the molecular level. (2) By applying neural networks, the complex molecular structure information of diesel fuel can be analyzed and molecular properties can be predicted, making it more convenient, economical, and accurate to predict pour point and cold filter plugging point. (3) The matrix format facilitates editing and storage, and is easy for computers to recognize.

[0086] Example 6: This embodiment of the invention provides a computer device, which includes a memory, a processor, a communication interface, and a communication bus. The memory stores a program that can run on the processor. When the processor executes the program, it implements the method for predicting the pour point and cold filter plugging point of diesel fuel as described in the above embodiment.

[0087] The processor can be a central processing unit, or it can be other general-purpose processors, digital signal processors, application-specific integrated circuits, field-programmable gate arrays or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or combinations of the above types of chips.

[0088] Memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs, non-transitory computer-executable programs, and units, such as the program units corresponding to the methods described above in this invention. The processor executes various functional applications and data processing by running the non-transitory software programs, instructions, and modules stored in the memory, thereby realizing the method for predicting the pour point and cold filter plugging point of diesel fuel as described in the above embodiments.

[0089] The memory may include a program storage area and a data storage area. The program storage area may store the operating system and applications required for at least one function; the data storage area may store data created by the processor, etc. Furthermore, the memory may include high-speed random access memory and non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. The memory may optionally include memory remotely located relative to the processor, which can be connected to the processor via a network. The one or more programs stored in the memory, when executed by the processor, perform the diesel pour point and cold filter plugging point prediction method described in the above embodiments.

[0090] Example 7: This embodiment of the invention provides a storage medium for computer-readable storage, which stores one or more programs that can be executed by one or more processors to implement the method for predicting the pour point and cold filter plugging point of diesel fuel as described in any of the above method embodiments.

[0091] The storage medium can be an internal storage unit of a computer device, such as a hard drive or memory. It can also be an external storage device, such as a plug-in hard drive, smart memory card, secure digital card, or flash memory card.

[0092] In this invention, the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance; the term "multiple" refers to two or more unless otherwise explicitly defined. The terms "install," "connect," "link," and "fix" should be interpreted broadly. For example, "connect" can be a fixed connection, a detachable connection, or an integral connection; "link" can be a direct connection or an indirect connection through an intermediate medium. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0093] In the description of this invention, it should be understood that the terms "upper," "lower," "front," "rear," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or module referred to must have a specific orientation or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0094] In the description of this specification, the terms "one embodiment," "some embodiments," "specific embodiment," etc., refer to a specific feature, structure, material, or characteristic described in connection with that embodiment or example, which is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0095] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

[0096] The above technical features constitute the embodiments of the present invention, which have strong adaptability and implementation effect. Unnecessary technical features can be added or removed according to actual needs to meet the needs of different situations.

Claims

1. A method for predicting the freezing point and cold filter plugging point of a diesel fuel, characterized in that The method comprises the following steps: The diesel oil is analyzed by an analysis instrument to obtain actual values of first properties; main homologues in a diesel oil structure are obtained, and a central molecule set is obtained according to the main homologues; The central molecule set is expanded according to a preset rule to obtain virtual molecules; a preset structure-property relationship is applied to the virtual molecules to obtain a virtual molecule content and a first property calculation function; An optimization algorithm is used to optimize the first property calculation function, so that an error between the actual values of the first properties and calculated values of the first properties is less than a preset value, and relative contents of the virtual molecules in virtual diesel oil are obtained; A diesel oil molecule matrix is obtained according to the relative contents of the virtual molecules and a structure vector; the diesel oil molecule matrix is input into a neural network model for prediction to obtain a freezing point and a cold filter plugging point of the diesel oil; The diesel oil molecule matrix is divided into a training set and a prediction set; the training set is input into an initial neural network for training to obtain the neural network model; the prediction set is input into the neural network model for prediction to obtain the freezing point and the cold filter plugging point of the diesel oil; The preset structure-property relationship includes one or a combination of the following: a structure-element content, a structure-average molecular weight, a structure-saturated fraction content, a structure-aromatic fraction content, and a structure-boiling point; The central molecule set is expanded according to the preset rule to obtain the virtual molecules, specifically including the following steps: a methyl group is added to a central molecule in the central molecule set according to the preset rule to obtain the virtual molecules; the preset rule includes one or a combination of the following: no isomerism, only changing a carbon number of a paraffin, first satisfying an increase in a methyl number of a ring structure, and a naphthenic ring being better than an aromatic ring when a side chain is added to the ring.

2. The method of predicting the freezing point and cold filter plugging point of diesel according to claim 1, characterized by The central molecule set includes 40 to 70 central molecules.

3. The method of predicting the freezing point and cold filter plugging point of diesel according to claim 1 or 2, characterized in that The optimization algorithm includes one or a combination of the following: an ant colony algorithm, a simulated annealing method, a genetic algorithm, and an artificial neural network algorithm.

4. The method of predicting the freezing point and cold filter plugging point of diesel according to claim 1 or 2, characterized by The neural network model includes one or a combination of the following: a single-layer neural network, a feedforward neural network, a radial basis function network, a deep feedforward network, and a recurrent neural network.

5. A system for predicting the pour point and cold filter plugging point of a diesel fuel, characterized in that The method for predicting the freezing point and the cold filter plugging point of the diesel oil comprises the following steps: An analysis module is configured to analyze the diesel oil by an analysis instrument to obtain actual values of first properties; An acquisition module is configured to obtain main homologues in a diesel oil structure, and a central molecule set is obtained according to the main homologues; An expansion module is configured to expand the central molecule set according to a preset rule to obtain virtual molecules; An application module is configured to apply a preset structure-property relationship to the virtual molecules to obtain a virtual molecule content and a first property calculation function; An optimization module is configured to optimize the first property calculation function by an optimization algorithm, so that an error between the actual values of the first properties and calculated values of the first properties is less than a preset value, and relative contents of the virtual molecules in virtual diesel oil are obtained; A matrix acquisition module is configured to obtain a diesel oil molecule matrix according to the relative contents of the virtual molecules and a structure vector; A prediction module is configured to input the diesel molecule matrix into a neural network model to obtain the freezing point and the cold filter plugging point of the diesel. 6.A computer device, comprising a memory and a processor, wherein the memory stores a program capable of running on the processor, and the program comprises the following steps of: The processor implements the method for predicting the freezing point and the cold filter plugging point of diesel according to any one of claims 1 to 4 when executing the program.

7. A storage medium, characterized by The storage medium stores one or more programs, and the one or more programs are executable by one or more processors to implement the method for predicting the freezing point and the cold filter plugging point of diesel according to any one of claims 1 to 4.

Images