An oil product data analysis method, system and device based on spectral data

Abstract

The application discloses an oil product data analysis method, system and device based on spectral data, and the method comprises the following steps: determining simple evaluation data of a target oil product by using spectral data of the target oil product; searching for several oil product samples similar to the target oil product in an oil product database of an oil product category to which the target oil product belongs according to the simple evaluation data of the target oil product; determining a blending ratio of the several oil product samples by using a mixed integer nonlinear programming method, and mixing the several oil product samples into a blending oil product according to the blending ratio; determining molecular composition data and macroscopic physical property data of the blending oil product based on the blending ratio and the molecular composition data and the macroscopic physical property data of the several oil product samples; and taking the molecular composition data and the macroscopic physical property data of the blending oil product as analysis data of the target oil product. The method can quickly and accurately determine the molecular composition of the oil product and corresponding macroscopic physical property data, thereby providing technical support for subsequent production processes.

Description

Technical Field

[0001] This invention belongs to the field of oil processing technology, and in particular relates to an oil data analysis method, system and equipment based on spectral data. Background Technology

[0002] With increasingly stringent environmental protection requirements in my country, oil products are demanding stricter quality standards. Meanwhile, rising global oil prices necessitate stricter quality control measures for refineries to improve profitability.

[0003] Furthermore, in the precise control processes of petroleum processing, it is often necessary to conduct relevant simulations based on the macroscopic physical properties and even molecular composition of oil products to optimize the operating conditions for various oil products. However, if the macroscopic physical properties and molecular composition data of oil products are missing or inaccurate, it will lead to errors in the simulation process. If work is carried out according to the incorrect simulation results, oil products that cannot meet quality requirements may be produced. Reprocessing such oil products is time-consuming and labor-intensive, affecting the economic benefits of refineries. Therefore, how to accurately and quickly determine comprehensive macroscopic physical properties and even molecular composition data of oil products is a pressing problem that oil refining companies need to solve. Summary of the Invention

[0004] To address the problems existing in the prior art, the purpose of this invention is to provide a method, system, and device for oil data analysis based on spectral data.

[0005] To achieve the above objectives, the present invention provides the following four technical solutions:

[0006] In a first aspect, the present invention provides a method for oil product data analysis based on spectral data, comprising the following steps:

[0007] The short-term evaluation data of the target oil are determined using the spectral data of the target oil.

[0008] Based on the brief evaluation data of the target oil, search the oil database of the oil category to which the target oil belongs for several oil samples that are similar to the target oil.

[0009] The mixing ratio of the several oil samples is determined by the mixed integer nonlinear programming method, and the several oil samples are mixed into blended oil according to the mixing ratio.

[0010] Based on the mixing ratio and the molecular composition and macroscopic properties of the several oil samples, the molecular composition and macroscopic properties of the blended oil are determined.

[0011] The molecular composition data and macroscopic physical property data of the blended oils are used as detailed evaluation data for the target oils.

[0012] Preferably, both the brief evaluation data and the detailed evaluation data of the target oil product include the macroscopic physical property data of the target oil product, and the macroscopic physical property data in the detailed evaluation data of the target oil product is more than the macroscopic physical property data in the brief evaluation data of the target oil product.

[0013] Preferably, determining the short-term evaluation data of the target oil using its spectral data includes:

[0014] Obtain the spectral data of the target oil;

[0015] The spectral data of the target oil is input into the property prediction model to predict the macroscopic property data corresponding to the spectral data of the target oil.

[0016] The spectral data refers to the absorption intensity of light of a specified wavelength by the target oil after passing light of a specified wavelength through it. The property prediction model is one of the following: multiple linear regression model, principal component regression model, partial least squares model, artificial neural network model, deep learning model, and topological method model.

[0017] Preferably, when the property prediction model is an artificial neural network model whose input is spectral data and whose output is macroscopic property data, the property prediction model is trained through the following steps:

[0018] Establish an initial artificial neural network model;

[0019] Collect spectral data of known oil samples and measure macroscopic physical properties of known oil samples;

[0020] The initial artificial neural network model is trained using the spectral data and macroscopic physical property data of the known oil sample, and the trained artificial neural network model is used as the physical property prediction model.

[0021] Preferably, before the step of searching for several oil samples similar to the target oil in an oil database of the oil category to which the target oil belongs based on the brief evaluation data of the target oil, the method further includes:

[0022] Based on the brief evaluation data of the target oil, search for oil samples that match the target oil in the oil database of the oil category to which the target oil belongs;

[0023] If an oil sample matching the target oil is found in the oil database, the molecular composition data and macroscopic physical property data of the oil sample will be used as the detailed evaluation data of the target oil.

[0024] If no oil sample matching the target oil is found in the oil database, then the step of searching for several oil samples similar to the target oil in the oil database of the oil category to which the target oil belongs, based on the brief evaluation data of the target oil, is performed.

[0025] Preferably, the step of searching for several oil samples similar to the target oil in an oil database of the oil category to which the target oil belongs, based on the brief evaluation data of the target oil, includes:

[0026] Based on the brief evaluation data of the target oil, the physical property similarity between each oil sample in the oil database and the target oil is determined, and the oil samples in the oil database are sorted according to the physical property similarity.

[0027] Based on the ranking results, select several oil samples whose physical properties are closest to the target oil.

[0028] Preferably, the physical property similarity is equal to the weighted distance between the vectors composed of the macroscopic physical property data corresponding to the oil sample and the target oil.

[0029] Preferably, the step of determining the blending ratio of the plurality of oil samples using a mixed-integer nonlinear programming method includes:

[0030] The blending ratio of the several oil samples is determined based on mixed-integer nonlinear programming and a penalty function, wherein the penalty function includes:

[0031] The penalty function for the number of oil types is used to limit the number of oil types selected from the sorted oil samples for modeling.

[0032] The penalty function for the minimum mixing ratio is used to limit the minimum mixing ratio value in the final mixing ratio.

[0033] Preferably, the step of determining the molecular composition data and macroscopic property data of the blended oil based on the mixing ratio and the molecular composition data and macroscopic property data of the plurality of oil samples includes:

[0034] The molecular composition data of the blended oil is determined based on the mixing ratio and the molecular composition data of the several oil samples.

[0035] For a linear macroscopic property, the data of the linear macroscopic property of the several oil samples are weighted and summed according to the mixing ratio to obtain the data of the linear macroscopic property of the blended oil.

[0036] For a nonlinear macroscopic property, the data of the nonlinear macroscopic property of the blended oil are determined based on the molecular composition data of the blended oil and the corresponding property calculation model.

[0037] Preferably, after the step of determining the molecular composition data and macroscopic property data of the blended oil based on the mixing ratio and the macroscopic property data of the respective oil samples, and before the step of using the molecular composition data and macroscopic property data of the blended oil as detailed evaluation data of the target oil, the method further includes:

[0038] The mixing ratio is adjusted according to the difference in macroscopic properties between the blended oil and the target oil, and the several oil samples are remixed into a blended oil according to the adjusted mixing ratio until the difference in macroscopic properties between the blended oil and the target oil is minimized.

[0039] Preferably, when the weighted distance between the vectors composed of the corresponding macroscopic physical property data in the summary evaluation data of the blended oil and the target oil reaches the minimum value, it is determined that the difference in macroscopic physical properties between the blended oil and the target oil has reached the minimum.

[0040] Secondly, the present invention provides an oil product data analysis system based on spectral data, comprising:

[0041] An analysis module is used to determine the short-term evaluation data of the target oil using the spectral data of the target oil.

[0042] The search module is used to search for several oil samples that are similar to the target oil in the oil database of the oil category to which the target oil belongs, based on the brief evaluation data of the target oil.

[0043] A mixing module is used to determine the mixing ratio of the plurality of oil samples using a mixed integer nonlinear programming method, and to mix the plurality of oil samples into a blended oil according to the mixing ratio.

[0044] The determination module is used to determine the molecular composition data and macroscopic property data of the blended oil based on the mixing ratio and the molecular composition data and macroscopic property data of the respective oil samples.

[0045] The output module is used to use the molecular composition data and macroscopic physical property data of the blended oil as detailed evaluation data of the target oil.

[0046] Preferably, the analysis module includes:

[0047] The acquisition unit is used to acquire the spectral data of the target oil.

[0048] The prediction unit is used to input the spectral data of the target oil into the property prediction model and predict the macroscopic property data corresponding to the spectral data of the target oil.

[0049] The spectral data refers to the absorption intensity of light of a specified wavelength by the target oil after passing light of a specified wavelength through it. The property prediction model is one of the following: multiple linear regression model, principal component regression model, partial least squares model, artificial neural network model, deep learning model, and topological method model.

[0050] Preferably, the determining module includes:

[0051] A determining unit is used to determine the molecular composition data of the blended oil based on the mixing ratio and the molecular composition data of the plurality of oil samples.

[0052] The first calculation unit is used to perform a weighted summation of the data of the linear macroscopic property of the several oil samples according to the mixing ratio for a linear macroscopic property, so as to obtain the data of the linear macroscopic property of the blended oil.

[0053] The second calculation unit is used to determine the data of the nonlinear macroscopic property of the blended oil based on the molecular composition data and the corresponding property calculation model.

[0054] Preferably, the oil data analysis system based on spectral data further includes an adjustment module, which is disposed between the mixing module and the determination module. The adjustment module is used to adjust the mixing ratio according to the macroscopic property difference between the blended oil and the target oil, and to remix the several oil samples into a blended oil according to the adjusted mixing ratio until the macroscopic property difference between the blended oil and the target oil is minimized.

[0055] Thirdly, the present invention provides an oil analysis device based on spectral data, including a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus;

[0056] Memory, used to store computer programs;

[0057] The processor, when executing the program stored in the memory, implements the above-mentioned oil data analysis method based on spectral data.

[0058] Fourthly, the present invention provides a computer-readable storage medium storing one or more programs that can be executed by one or more processors to implement the above-described oil data analysis method based on spectral data.

[0059] Compared with the prior art, the above-mentioned technical solution of the present invention has the following advantages:

[0060] The method of this invention can quickly and accurately determine the detailed molecular composition and macroscopic properties of a target oil, providing technical support for subsequent production processes. This invention utilizes spectral data to analyze the macroscopic properties of the target oil, achieving rapid analysis of these properties. By comparing and verifying physical properties, this invention adjusts the molecular composition of the target oil, thereby determining its molecular composition and corresponding macroscopic properties, providing valuable guidance for optimizing subsequent process operations.

[0061] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures pointed out in the description, claims and drawings. Attached Figure Description

[0062] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0063] Figure 1 A flowchart illustrating the oil data analysis method based on spectral data provided in Embodiment 1 of the present invention;

[0064] Figure 2 This is a flowchart illustrating the oil data analysis method based on spectral data provided in Embodiment 2 of the present invention.

[0065] Figure 3 This is a flowchart illustrating the training process of the property prediction model provided in Embodiment 2 of the present invention.

[0066] Figure 4 This is a schematic diagram of the composition of the oil analysis equipment based on spectral data provided in Embodiment 3 of the present invention. Detailed Implementation

[0067] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0068] Example 1

[0069] like Figure 1As shown, this invention aims to provide a method for oil product data analysis based on spectral data, which mainly includes the following steps:

[0070] S1, using the spectral data of the target oil to determine the brief evaluation data of the target oil, wherein the brief evaluation data of the target oil includes some macroscopic physical property data of the target oil;

[0071] S10. Based on the brief evaluation data of the target oil, search the oil database of the oil category to which the target oil belongs for several oil samples that are similar to the target oil.

[0072] S20, the mixing ratio of the several oil samples is determined by the mixed integer nonlinear programming method, and the several oil samples are mixed into blended oil according to the mixing ratio.

[0073] S30, Based on the mixing ratio and the molecular composition data and macroscopic property data of the respective oil samples, determine the molecular composition data and macroscopic property data of the blended oil.

[0074] S40, the molecular composition data and macroscopic property data of the blended oil are used as detailed evaluation data (analytical data) of the target oil, wherein the macroscopic property data in the detailed evaluation data of the target oil is more than the macroscopic property data in the brief evaluation data of the target oil.

[0075] Example 2

[0076] The oil data analysis method based on spectral data of the present invention will be explained in detail below with reference to specific embodiments. Figure 2 A flowchart illustrating the oil data analysis method based on spectral data provided in Embodiment 2 of the invention. Figure 2 As shown, the oil data analysis method in this embodiment mainly includes the following steps:

[0077] S1, using the spectral data of the target oil to determine the brief evaluation data of the target oil, wherein the brief evaluation data of the target oil includes some macroscopic physical property data of the target oil.

[0078] In this embodiment, preliminary evaluation data of the target oil can be obtained first through instrument measurement or model calculation. This preliminary evaluation data may include macroscopic properties such as density, distillation range, sulfur content, octane number, and cetane number. In this invention, preliminary evaluation data is used in contrast to detailed evaluation data. Typically, the preliminary evaluation data contains fewer macroscopic properties than the detailed evaluation data.

[0079] In this embodiment, the spectral data can be near-infrared, mid-infrared, Raman, or nuclear magnetic resonance spectra.

[0080] S100.1, Based on the brief evaluation data of the target oil, determine the oil category to which the target oil belongs.

[0081] Using brief evaluation data of the target oil, such as density and / or distillation range, the target oil is collected and screened in a pre-established crude oil database, gasoline database, diesel database, and wax oil database to determine which oil category and corresponding oil database the target oil belongs to.

[0082] For example, density, initial boiling point and final boiling point, 5% distillation temperature and 95% distillation temperature can be used to determine which of the above oil databases the target oil belongs to.

[0083] First, determine the density range formed by the minimum and maximum density values ​​of each database, the first distillation range formed by the minimum and maximum values ​​of the 5% distillation temperature, and the second distillation range formed by the minimum and maximum values ​​of the 95% distillation temperature. Then, match the density of the target oil with the density ranges of each database. If there is a single range, select that database. If there is overlap, match the 5% distillation temperature of the target oil with the first distillation range of each database to determine if there is overlap; or match the 95% distillation temperature of the target oil with the second distillation range of each database to determine if there is overlap. If there is a single range, select that database. If there is overlap, continue to judge until the oil database to which it belongs is determined.

[0084] After determining the type of oil to which the target oil belongs, perform the following steps:

[0085] S100, based on the brief evaluation data of the target oil product, search the oil product database for oil samples that match the target oil product (target crude oil):

[0086] If an oil sample matching the target oil is found in the oil database, the molecular composition data and macroscopic physical property data of the oil sample will be used as the detailed evaluation data of the target oil.

[0087] If no oil sample matching the target oil is found in the oil database, proceed to step S200.

[0088] S200 is the step of searching for several oil samples similar to the target oil in the oil database based on the brief evaluation data of the target oil.

[0089] In this embodiment, step S200 mainly involves first analyzing the physical property similarity between each oil sample in the oil database and the target oil based on the brief evaluation data of the target oil, then sorting the oil samples in the oil database according to the physical property similarity, and finally selecting several oil samples with the closest physical properties to the target oil based on the sorting results.

[0090] In specific applications, step S200 may include the following steps:

[0091] S210, For each oil sample in the oil database, calculate the physical property similarity between the oil sample and the target oil, wherein the physical property similarity is equal to the weighted distance between the vectors composed of the corresponding macroscopic physical property data of the oil sample and the target oil; wherein the weights can be determined in advance according to the importance of each macroscopic physical property.

[0092] S220, Sort the oil samples in the oil database according to the physical property similarity between each oil sample in the oil database and the target oil.

[0093] S230: Select several oil samples whose physical properties are closest to the target oil based on the sorting results.

[0094] For example, oil samples in the oil database are sorted in descending order of physical property similarity. Then, the top-ranked oil samples are selected as those with the closest physical properties to the target oil. The number of oil samples selected is not limited and is usually determined based on a trade-off between the accuracy of the calculation results and time constraints.

[0095] S300, the mixing ratio of the several oil samples is determined by using a mixed integer nonlinear programming method.

[0096] In this embodiment, the blending ratio of the plurality of oil samples is determined using a mixed-integer nonlinear programming method, mainly based on mixed-integer nonlinear programming and a penalty function. The penalty function includes:

[0097] The penalty function for the number of oil types is used to limit the number of oil types selected from the sorted oil samples for modeling.

[0098] The penalty function for the minimum mixing ratio is used to limit the minimum mixing ratio value in the final mixing ratio.

[0099] S400, the plurality of oil samples are mixed into blended oil according to the mixing ratio.

[0100] S500, based on the mixing ratio and the molecular composition data and macroscopic property data of the respective oil samples, determine the molecular composition data and macroscopic property data of the blended oil.

[0101] In this embodiment, step S500 can be further subdivided into the following steps:

[0102] S510, Determine the molecular composition data of the blended oil based on the mixing ratio and the molecular composition data of the plurality of oil samples;

[0103] S520, for a linear macroscopic property, the data of the linear macroscopic property of the several oil samples are weighted and summed according to the mixing ratio to obtain the data of the linear macroscopic property of the blended oil.

[0104] S530, for a nonlinear macroscopic property, the data of the nonlinear macroscopic property of the blended oil are determined based on the molecular composition data of the blended oil and the corresponding property calculation model.

[0105] For linear properties, the linear macroscopic property data of several oil samples involved in the blending and the blending ratio can be linearly added together to obtain the linear macroscopic property data of the blended oil.

[0106] For nonlinear physical properties, the nonlinear macroscopic physical property data of blended oils can be calculated based on the molecular composition data and physical property calculation formulas of the blended oils.

[0107] It should be noted that the execution order of steps S520 and S530 is not limited to this in actual application.

[0108] S600 analyzes the differences between the macroscopic physical property data of the blended oil and the macroscopic physical property data of the target oil (target crude oil).

[0109] S700, if the difference does not meet the preset threshold condition, the mixing ratio is adjusted, and the process returns to step S400. The several oil samples are remixed into blended oil according to the adjusted mixing ratio, so as to re-analyze the difference between the macroscopic physical property data of the blended oil and the macroscopic physical property data of the target oil.

[0110] In this embodiment, the difference meeting the preset threshold condition means that the difference between the macroscopic physical property data of the blended oil and the macroscopic physical property data of the target oil can be minimized.

[0111] Therefore, a quality assessment parameter can be preferably used to measure the difference between the macroscopic physical property data of the blended oil and the target oil. Specifically, this quality assessment parameter is equal to the weighted distance between the vectors composed of the corresponding macroscopic physical property data of the blended oil and the target oil.

[0112] In this embodiment, when the value of the quality assessment parameter reaches its minimum value, it is determined that the difference between the macroscopic physical property data of the blended oil and the macroscopic physical property data of the target oil has reached its minimum.

[0113] S800, if the difference meets a preset threshold condition, then the molecular composition data and macroscopic property data of the blended oil are used as the detailed evaluation data of the target oil, wherein the macroscopic property data in the detailed evaluation data of the target oil is more than the macroscopic property data in the brief evaluation data of the target oil.

[0114] The method in this embodiment can quickly and accurately determine the molecular composition and corresponding macroscopic physical property data of oils, providing useful guidance for subsequent process optimization.

[0115] The workflow of step S1 above is further illustrated below. Specifically, step S1 includes the following sub-steps:

[0116] S11, Obtain the spectral data of the target oil;

[0117] In this embodiment, the spectral data refers to the absorption intensity of light of a specified wavelength by the target oil after it passes through the target oil. The spectral data includes the relationship between the wavelength of the light and the absorption intensity of the oil. In practical applications, the spectral data can be a spectral curve with wavelength on the horizontal axis and absorption intensity on the vertical axis. Given a wavelength, the absorption intensity corresponding to that wavelength can be found on the spectral curve. The spectral data can be measured using a near-infrared absorption spectrometer or a mid-infrared absorption spectrometer.

[0118] S12, input the spectral data of the target oil into the property prediction model, and predict the macroscopic property data corresponding to the spectral data of the target oil, which is the brief evaluation data;

[0119] In this embodiment, the macroscopic physical property data is any one of boiling point, density, octane number, cloud point, pour point and aniline point. Since the quantity is small, it belongs to the brief evaluation data of the target oil.

[0120] like Figure 3 As shown: In this embodiment, the physical property prediction model is trained through the following steps:

[0121] S21, Establish an initial property prediction model, wherein the initial property prediction model may be a multiple linear regression, principal component regression, partial least squares, artificial neural network or deep learning and topological method model;

[0122] S22, Collect spectral data of known oil samples and measure macroscopic physical property data of known oil samples;

[0123] S23, using the spectral data and macroscopic property data of the known oil sample, the initial property prediction model is trained to obtain the trained property prediction model.

[0124] Specifically, when the property prediction model is an artificial neural network model whose input is spectral data and whose output is macroscopic property data, the above steps are also used to train the artificial neural network model to obtain the trained model as the property prediction model.

[0125] For example, when the spectral data of the known oil sample is spectral data acquired by a near-infrared spectrometer from 780nm to 2500nm with a step size of 4nm, and the macroscopic physical property data of the known oil sample's group composition includes the molecular physical property data of P-straight-chain alkanes, I-isoalkanes, N-cycloalkanes, O-olefins, and A-aromatics, specifically, the physical property data of the oil sample includes any one of boiling point, density, octane number, aromatics, olefins, benzene, flash point, refractive index, pour point, cloud point, condensation point, tumble point, aniline point, freezing point, viscosity index, viscosity, API gravity, and wax content. When the spectral data is a spectral curve with wavelength on the horizontal axis and absorption intensity on the vertical axis, the physical property prediction model is trained using the spectral data of the known oil sample and the physical property data of the known oil sample's group composition, including:

[0126] For each known oil sample containing PIONA group composition, boiling point, density, octane number, aromatics, olefins, benzene, flash point, refractive index, pour point, cloud point, condensation point, aniline point, freezing point, viscosity index, viscosity, API gravity, and wax content, absorption intensity data at different wavelengths of the spectrum are collected. Regression is used to establish a predictive model between the spectral data and the above-mentioned physical properties. The number of sample types is more than 80. The group composition data is the result of joint calculation and calibration by multiple sets of correlation models.

[0127] For each known oil sample containing p-straight-chain alkanes, the absorbance at the wavelength corresponding to the p-straight-chain alkanes is used as input, and the physical property data corresponding to the p-straight-chain alkanes is used as output to train the initial physical property prediction model, so that the trained physical property prediction model can predict the physical property data corresponding to the p-straight-chain alkanes based on the absorbance of the sample at the wavelength corresponding to the p-straight-chain alkanes.

[0128] For each known oil sample containing I-isoalkane among multiple known oil samples, collect its absorbance at the wavelength corresponding to I-isoalkane and the corresponding physical property data, where the number of sample types is more than 80.

[0129] For each known oil sample containing I-isoalkane, the absorbance at the wavelength corresponding to I-isoalkane is used as input, and the physical property data corresponding to I-isoalkane is used as output to train the initial physical property prediction model, so that the trained physical property prediction model can predict the physical property data corresponding to I-isoalkane based on the absorbance of the sample at the wavelength corresponding to I-isoalkane.

[0130] For each known oil sample containing N-cycloalkanes among a variety of known oil samples, collect its absorbance at the wavelength corresponding to N-cycloalkanes and the corresponding physical property data, wherein the number of sample types is more than 80.

[0131] For each known oil sample containing N-cycloalkane, the absorbance at the wavelength corresponding to N-cycloalkane is used as input, and the physical property data corresponding to N-cycloalkane is used as output to train the initial physical property prediction model, so that the trained physical property prediction model can predict the physical property data corresponding to N-cycloalkane based on the absorbance of the sample at the wavelength corresponding to N-cycloalkane.

[0132] For each known oil sample containing O-olefins among a variety of known oil samples, collect its absorbance at the wavelength corresponding to O-olefins and the corresponding physical property data, where the number of sample types is more than 80.

[0133] For each known oil sample containing O-olefins, the absorbance at the wavelength corresponding to O-olefins is used as input, and the physical property data corresponding to O-olefins is used as output to train the initial physical property prediction model, so that the trained physical property prediction model can predict the physical property data corresponding to O-olefins based on the absorbance of the sample at the wavelength corresponding to O-olefins.

[0134] For each known oil sample containing P-straight-chain alkanes among a variety of known oil samples, collect its absorbance at the wavelength corresponding to the P-straight-chain alkanes and the corresponding physical property data, where the number of sample types is more than 80.

[0135] For each known oil sample containing A-aromatics, the absorbance at the wavelength corresponding to A-aromatics is used as input, and the physical property data corresponding to A-aromatics is used as output to train the initial physical property prediction model. This allows the trained physical property prediction model to predict the physical property data corresponding to A-aromatics based on the absorbance of the sample at the wavelength corresponding to A-aromatics.

[0136] When the known correspondence between physical property data and group composition includes the correspondence between p-straight-chain alkanes and their physical property data, I-isoalkanes and their physical property data, N-cycloalkanes and their physical property data, O-olefins and their physical property data, and A-isomers (or branched molecules) and their physical property data, the group composition corresponding to each physical property data is searched and predicted, including:

[0137] The predicted physical property data were compared with the physical property data corresponding to P-straight-chain alkanes, I-isoalkanes, N-cycloalkanes, O-olefins, and A-aromatics, respectively.

[0138] When any one of the predicted physical property data is consistent with any of the physical property data corresponding to P-straight-chain alkanes, I-isoalkanes, N-cycloalkanes, O-olefins, or A-aromatics, then the macroscopic physical property parameters at this time are the simplified evaluation data of the target oil that need to be determined using the spectral data of the target oil.

[0139] The method in this embodiment can quickly and accurately determine the detailed molecular composition and macroscopic property data of the target oil, providing technical support for subsequent production processes. This embodiment utilizes spectral data to analyze the macroscopic properties of the target oil, achieving rapid analysis of these properties. By comparing and verifying the properties, this embodiment adjusts the molecular composition of the target oil, thereby determining its molecular composition and corresponding macroscopic property data, providing valuable guidance for optimizing subsequent process operations.

[0140] Example 3

[0141] Based on the same inventive concept, Embodiment 3 of the present invention provides an oil data analysis system based on spectral data, comprising:

[0142] An analysis module is used to determine the short-term evaluation data of the target oil using the spectral data of the target oil.

[0143] The search module is used to search for several oil samples that are similar to the target oil in the oil database of the oil category to which the target oil belongs, based on the brief evaluation data of the target oil.

[0144] A mixing module is used to determine the mixing ratio of the plurality of oil samples using a mixed integer nonlinear programming method, and to mix the plurality of oil samples into a blended oil according to the mixing ratio.

[0145] The determination module is used to determine the molecular composition data and macroscopic property data of the blended oil based on the mixing ratio and the molecular composition data and macroscopic property data of the respective oil samples.

[0146] The output module is used to use the molecular composition data and macroscopic physical property data of the blended oil as detailed evaluation data of the target oil.

[0147] In this embodiment, the analysis module includes:

[0148] The acquisition unit is used to acquire the spectral data of the target oil.

[0149] The prediction unit is used to input the spectral data of the target oil into the property prediction model and predict the macroscopic property data corresponding to the spectral data of the target oil.

[0150] The spectral data refers to the absorption intensity of light of a specified wavelength by the target oil after passing light of a specified wavelength through it. The property prediction model is one of the following: multiple linear regression model, principal component regression model, partial least squares model, artificial neural network model, deep learning model, and topological method model.

[0151] In this embodiment, the determining module includes:

[0152] A determining unit is used to determine the molecular composition data of the blended oil based on the mixing ratio and the molecular composition data of the plurality of oil samples.

[0153] The first calculation unit is used to perform a weighted summation of the data of the linear macroscopic property of the several oil samples according to the mixing ratio for a linear macroscopic property, so as to obtain the data of the linear macroscopic property of the blended oil.

[0154] The second calculation unit is used to determine the data of the nonlinear macroscopic property of the blended oil based on the molecular composition data and the corresponding property calculation model.

[0155] In this embodiment, an adjustment module can also be provided, which is located between the mixing module and the determining module. The adjustment module is used to adjust the mixing ratio according to the difference in macroscopic properties between the blended oil and the target oil, and to remix the several oil samples into a blended oil according to the adjusted mixing ratio until the difference in macroscopic properties between the blended oil and the target oil is minimized.

[0156] The functions and implementation methods of each module in the oil data analysis system of this embodiment are consistent with the functions and implementation methods of each step in the oil data analysis method of Embodiment 1 or Embodiment 2 of the present invention. Therefore, they will not be described in detail here.

[0157] Example 4

[0158] Based on the same inventive concept, such as Figure 4 As shown, this embodiment of the invention provides an oil data analysis device based on spectral data, including a processor 1110, a communication interface 1120, a memory 1130, and a communication bus 1140, wherein the processor 1110, the communication interface 1120, and the memory 1130 communicate with each other through the communication bus 1140.

[0159] Memory 1130 is used to store computer programs;

[0160] When processor 1110 executes the program stored in memory 1130, it implements the following oil data analysis method based on spectral data:

[0161] The target oil is evaluated using spectral data, and the evaluation data includes some macroscopic physical properties of the target oil.

[0162] Based on the brief evaluation data of the target oil, search the oil database of the oil category to which the target oil belongs for several oil samples that are similar to the target oil.

[0163] The mixing ratio of the several oil samples is determined by the mixed integer nonlinear programming method, and the several oil samples are mixed into blended oil according to the mixing ratio.

[0164] Based on the mixing ratio and the molecular composition and macroscopic properties of the respective oil samples, the molecular composition and macroscopic properties of the blended oil are determined.

[0165] The molecular composition data and macroscopic property data of the blended oil are used as detailed evaluation data of the target oil, wherein the macroscopic property data in the detailed evaluation data of the target oil is more than the macroscopic property data in the brief evaluation data of the target oil.

[0166] The aforementioned communication bus 1140 can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. This communication bus 1140 can be divided into an address bus, a data bus, a control bus, etc. For ease of illustration, it is represented by only one thick line in the figure, but this does not indicate that there is only one bus or one type of bus.

[0167] The communication interface 1120 is used for communication between the above-mentioned electronic device and other devices.

[0168] The memory 1130 may include random access memory (RAM) or non-volatile memory, such as at least one disk storage device. Optionally, the memory 1130 may also be at least one storage device located remotely from the aforementioned processor 1110.

[0169] The processor 1110 mentioned above can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0170] Example 5

[0171] Based on the same inventive concept, embodiments of the present invention provide a computer-readable storage medium storing one or more programs, which can be executed by one or more processors to implement the oil data analysis method based on spectral data in any of the above possible implementations.

[0172] Optionally, the storage medium may be a non-transitory computer-readable storage medium, such as a ROM, random access memory (RAM), CD-ROM, magnetic tape, floppy disk, and optical data storage device.

[0173] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented, in whole or in part, as a computer program product. A computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the flow or function according to the embodiments of the present invention is generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., a solid-state disk (SSD)).

[0174] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention. It should be noted that the terminology used herein is only for describing specific implementations and is not intended to limit the exemplary implementations according to this application. When the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0175] It should be noted that the terms "first," "second," etc., used in the specification, claims, and drawings of this application are used to distinguish similar objects and are not used to describe a specific order or sequence. It should be understood that such terms can be used interchangeably where appropriate so that the embodiments of this application described herein can be implemented, for example, in sequences other than those illustrated or described herein.

[0176] It should be understood that the exemplary embodiments described herein can be implemented in many different forms and should not be construed as being limited to the embodiments set forth herein. These embodiments are provided so that the disclosure of this application is thorough and complete, and that the concept of these exemplary embodiments is fully conveyed to those skilled in the art, and should not be construed as limiting the invention.

Claims

1. A method for oil product data analysis based on spectral data, characterized in that, Includes the following steps: The short-term evaluation data of the target oil are determined using the spectral data of the target oil. Based on the brief evaluation data of the target oil, several oil samples similar to the target oil are searched in the oil database of the oil category to which the target oil belongs; this includes: determining the physical property similarity between each oil sample in the oil database and the target oil based on the brief evaluation data of the target oil; sorting the oil samples in the oil database according to the physical property similarity; and selecting several oil samples with the closest physical properties to the target oil based on the sorting results; wherein the physical property similarity is equal to the weighted distance between the vectors composed of the corresponding macroscopic physical property data of the oil sample and the target oil. The blending ratio of the several oil samples is determined based on the mixed integer nonlinear programming method and the penalty function, and the several oil samples are blended into blended oil according to the blending ratio; wherein, the penalty function includes: a penalty function for the number of oil types, used to limit the number of oil types selected from the sorted oil samples for modeling; and a penalty function for the minimum blending ratio value, used to limit the minimum blending ratio value in the final obtained blending ratio. Based on the mixing ratio and the molecular composition and macroscopic properties of the several oil samples, the molecular composition and macroscopic properties of the blended oil are determined. The molecular composition data and macroscopic physical property data of the blended oils are used as detailed evaluation data for the target oils.

2. The oil product data analysis method based on spectral data according to claim 1, characterized in that, Both the brief evaluation data and the detailed evaluation data of the target oil product include the macroscopic physical property data of the target oil product, and the macroscopic physical property data in the detailed evaluation data of the target oil product is more than the macroscopic physical property data in the brief evaluation data of the target oil product.

3. The oil product data analysis method based on spectral data according to claim 1, characterized in that, The determination of the short-term evaluation data of the target oil using its spectral data includes: Obtain the spectral data of the target oil; The spectral data of the target oil is input into the property prediction model, and the macroscopic property data corresponding to the spectral data of the target oil is predicted as the brief evaluation data. The spectral data refers to the absorption intensity of light of a specified wavelength by the target oil after passing light of a specified wavelength through it. The property prediction model is one of the following: multiple linear regression model, principal component regression model, partial least squares model, artificial neural network model, deep learning model, and topological method model.

4. The oil product data analysis method based on spectral data according to claim 3, characterized in that, When the property prediction model is an artificial neural network model whose input is spectral data and whose output is macroscopic property data, the property prediction model is trained through the following steps: Establish an initial artificial neural network model; Collect spectral data of known oil samples and measure macroscopic physical properties of known oil samples; The initial artificial neural network model is trained using the spectral data and macroscopic physical property data of the known oil sample, and the trained artificial neural network model is used as the physical property prediction model.

5. The oil product data analysis method based on spectral data according to claim 1, characterized in that, Before the step of searching for several oil samples similar to the target oil in an oil database of the oil category to which the target oil belongs based on the brief evaluation data of the target oil, the method further includes: Based on the brief evaluation data of the target oil, search for oil samples that match the target oil in the oil database of the oil category to which the target oil belongs; If an oil sample matching the target oil is found in the oil database, the molecular composition data and macroscopic physical property data of the oil sample will be used as the detailed evaluation data of the target oil. If no oil sample matching the target oil is found in the oil database, then the step of searching for several oil samples similar to the target oil in the oil database of the oil category to which the target oil belongs, based on the brief evaluation data of the target oil, is performed.

6. The oil product data analysis method based on spectral data according to claim 1, characterized in that, The step of determining the molecular composition and macroscopic properties of the blended oil based on the mixing ratio and the molecular composition and macroscopic properties of the plurality of oil samples includes: The molecular composition data of the blended oil is determined based on the mixing ratio and the molecular composition data of the several oil samples. For a linear macroscopic property, the data of the linear macroscopic property of the several oil samples are weighted and summed according to the mixing ratio to obtain the data of the linear macroscopic property of the blended oil. For a nonlinear macroscopic property, the data of the nonlinear macroscopic property of the blended oil are determined based on the molecular composition data of the blended oil and the corresponding property calculation model.

7. The oil product data analysis method based on spectral data according to claim 1, characterized in that, After the step of determining the molecular composition and macroscopic properties of the blended oil based on the mixing ratio and the molecular composition and macroscopic properties of the plurality of oil samples, and before the step of using the molecular composition and macroscopic properties of the blended oil as detailed evaluation data of the target oil, the method further includes: The mixing ratio is adjusted according to the difference in macroscopic properties between the blended oil and the target oil, and the several oil samples are remixed into a blended oil according to the adjusted mixing ratio until the difference in macroscopic properties between the blended oil and the target oil is minimized.

8. The oil product data analysis method based on spectral data according to claim 7, characterized in that, When the weighted distance between the vectors composed of the corresponding macroscopic physical property data in the summary evaluation data of the blended oil and the target oil reaches the minimum value, it is determined that the difference in macroscopic physical properties between the blended oil and the target oil has reached the minimum.

9. An oil product data analysis system based on spectral data, characterized in that, include: An analysis module is used to determine the short-term evaluation data of the target oil using the spectral data of the target oil. The search module is used to search for several oil samples that are similar to the target oil in an oil database of the oil category to which the target oil belongs, based on the brief evaluation data of the target oil. This includes: determining the physical property similarity between each oil sample in the oil database and the target oil based on the brief evaluation data of the target oil; sorting the oil samples in the oil database according to the physical property similarity; and selecting several oil samples with the closest physical properties to the target oil based on the sorting results. The physical property similarity is equal to the weighted distance between the vectors composed of the macroscopic physical property data corresponding to the oil samples and the target oil. A mixing module is used to determine the blending ratio of the plurality of oil samples based on a mixed integer nonlinear programming method and a penalty function, and to mix the plurality of oil samples into a blended oil according to the blending ratio; wherein, the penalty function includes: a penalty function for the number of oil types, used to limit the number of oil types selected from the sorted oil samples for modeling; and a penalty function for the minimum blending ratio value, used to limit the minimum blending ratio value in the final obtained blending ratio. The determination module is used to determine the molecular composition data and macroscopic property data of the blended oil based on the mixing ratio and the molecular composition data and macroscopic property data of the plurality of oil samples. The output module is used to use the molecular composition data and macroscopic physical property data of the blended oil as detailed evaluation data of the target oil.

10. The oil product data analysis system based on spectral data according to claim 9, characterized in that, The analysis module includes: The acquisition unit is used to acquire the spectral data of the target oil. The prediction unit is used to input the spectral data of the target oil into the property prediction model and predict the macroscopic property data corresponding to the spectral data of the target oil. The spectral data refers to the absorption intensity of light of a specified wavelength by the target oil after passing light of a specified wavelength through it. The property prediction model is one of the following: multiple linear regression model, principal component regression model, partial least squares model, artificial neural network model, deep learning model, and topological method model.

11. The oil product data analysis system based on spectral data according to claim 9, characterized in that, The determining module includes: A determining unit is used to determine the molecular composition data of the blended oil based on the mixing ratio and the molecular composition data of the plurality of oil samples. The first calculation unit is used to perform a weighted summation of the data of the linear macroscopic property of the several oil samples according to the mixing ratio for a linear macroscopic property, so as to obtain the data of the linear macroscopic property of the blended oil. The second calculation unit is used to determine the data of the nonlinear macroscopic property of the blended oil based on the molecular composition data and the corresponding property calculation model.

12. The oil product data analysis system based on spectral data according to claim 9, characterized in that, It also includes an adjustment module, which is located between the mixing module and the determining module. The adjustment module is used to adjust the mixing ratio according to the difference in macroscopic properties between the blended oil and the target oil, and to remix the several oil samples into a blended oil according to the adjusted mixing ratio until the difference in macroscopic properties between the blended oil and the target oil is minimized.

13. An oil product data analysis device based on spectral data, characterized in that, It includes a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus; Memory, used to store computer programs; A processor, when executing a program stored in a memory, implements the oil data analysis method based on spectral data as described in any one of claims 1 to 8.

14. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores one or more programs, which can be executed by one or more processors to implement the oil data analysis method based on spectral data as described in any one of claims 1 to 8.

Images