A method and system for analyzing petroleum fractionation tower products based on near infrared technology

Abstract

The present invention relates to the technical field of sample analysis, and discloses a method and system for analyzing products of a petroleum fractionator based on near-infrared technology, wherein different batches of samples are collected; near-infrared spectral scanning is performed on different batches of samples to obtain near-infrared spectral data of different batches of samples, and the near-infrared spectral data are preprocessed; based on the measured spectral intensity after preprocessing, the optimal correction value of different batches of samples is calculated, key spectral features of different batches of samples are obtained, and the measured spectral intensity and the optimal correction value corresponding to the key spectral features are combined into an input vector and input into a pre-trained calibration model to obtain chemical property information of different batches of samples. The chemical property information of the sample can be analyzed based on the change in the measured spectral intensity, and the optimal correction value is used as a correction factor of the analysis result to compensate for the deviation of the analysis process, and the optimal correction value is added to the analysis result to offset the analysis deviation.

Description

Technical Field

[0001] The present invention relates to the technical field of sample analysis, and in particular to a method and system for analyzing products of a petroleum fractionation tower based on near infrared technology. Background Art

[0002] In the oil refining process, the distillation tower is one of the key equipment to separate crude oil into different fractions. Traditional oil distillation tower product analysis methods mainly rely on chemical analysis methods, such as gas chromatography (GC), high performance liquid chromatography (HPLC), etc. Although these methods are accurate and reliable, they have problems such as long time consumption, high cost, and complex operation. With the development of the oil industry, higher requirements are placed on real-time monitoring and product quality control in the production process.

[0003] Near-infrared (NIR) spectroscopy has been widely used in industrial production and scientific research due to its rapid, non-destructive, solvent-free characteristics. Near-infrared spectroscopy can infer the composition or properties of a substance by measuring its absorption or reflection spectrum within a specific wavelength range, and is particularly suitable for online monitoring of liquid samples. In recent years, the application of this technology in the petrochemical field has received increasing attention and has become one of the effective means to improve production efficiency and reduce costs.

[0004] Despite this, there is still a lack of a method that can efficiently and accurately analyze the products of petroleum fractionation towers online. The existing near-infrared spectroscopy technology has the following major technical difficulties when applied to the analysis of petroleum fractionation tower products:

[0005] There may be large differences between different batches of petroleum samples. These changes may lead to changes in spectral characteristics, which in turn affect the accuracy of the analysis results. Specifically: There may be significant differences in the proportions of various compounds (such as hydrocarbons, sulfides, nitrides, etc.) in different batches of petroleum samples. These changes will cause changes in the absorption peak position and intensity of the near-infrared spectrum, thereby affecting the prediction accuracy of the model. Changes in environmental factors such as temperature and humidity may also affect the stability of the near-infrared spectrum, thereby introducing additional errors. Specifically: Changes in temperature will affect the physical state of the sample, such as viscosity, refractive index, etc., thereby changing the spectral signal. In addition, temperature will also affect the performance of the spectrometer itself, such as light source intensity, detector sensitivity, etc., further affecting the accuracy of the analysis results. Changes in humidity will cause changes in the content of water molecules in the air. Water molecules have a strong absorption effect on near-infrared light of certain wavelengths. This interference will lead to inconsistency in spectral data, thereby affecting the prediction effect of the model. Summary of the invention

[0006] In order to solve the above technical problems, the present invention provides a method and system for analyzing petroleum distillation tower products based on near-infrared technology, which analyzes the chemical property information of the sample based on the change of measured spectral intensity and the optimal correction value.

[0007] The present invention provides a method for analyzing petroleum fractionation tower products based on near infrared technology, comprising:

[0008] Step 1: Collect different batches of samples from the petroleum fractionator;

[0009] Step 2: Use a near-infrared spectrometer to perform near-infrared spectrum scanning on different batches of samples to obtain near-infrared spectrum data of different batches of samples, where the near-infrared spectrum data at least includes wavelength and measured spectrum intensity;

[0010] Step 3: preprocessing the near infrared spectrum data;

[0011] Step 4: Calculate the optimal correction value of different batches of samples based on the measured spectral intensity after preprocessing;

[0012] Step 41: Obtain standardized values ​​of standard samples at different wavelengths;

[0013] Step 42: Calculate the spectral drift values ​​of different batches of samples at different wavelengths according to the standardized values ​​and the measured spectral intensity after preprocessing;

[0014] Step 43: Calculate the correction values ​​of different batches of samples according to the spectral drift value and the near infrared spectrum data;

[0015] Step 44: Optimizing the correction value to obtain the optimal correction value;

[0016] Step 5: Obtain the key spectral features of samples from different batches, and combine the measured spectral intensity and optimal correction value corresponding to the key spectral features into an input vector and input it into the pre-trained calibration model to obtain the chemical property information of samples from different batches.

[0017] Furthermore, the calculation formula of the spectral drift value is:

[0018] ;

[0019] In the formula, represents the spectral drift value of the kth batch of samples, represents the measured spectral intensity of the i-th wavelength of the k-th batch of samples, represents the spectral intensity of the i-th wavelength of the standard sample, Represents the weight coefficient of the i-th wavelength.

[0020] Furthermore, the calculation formula of the correction value is:

[0021] ;

[0022] In the formula, represents the calibration value of the kth batch of samples, Represents the original near-infrared spectral data of the kth batch of samples, Represents the spectral drift value of the kth batch of samples.

[0023] Furthermore, the correction value is optimized to obtain the optimal correction value, which specifically includes:

[0024] Step 441, constructing a primary optimization function according to the correction value, the spectral drift value, and the overall difference value, and calculating the optimized correction value;

[0025] Step 442, constructing an optimal optimization function according to the optimized correction value and the spectral drift value, and calculating the optimal correction value.

[0026] Further, obtaining key spectral features specifically includes the following steps:

[0027] Step 51: Calculate the correlation coefficient between the wavelength of each batch of samples and the standardized data in the standard sample;

[0028] Step 52: constructing a first threshold value for the correlation coefficient; determining whether the correlation coefficient is greater than or equal to the first threshold value, if so, defining the wavelength of the current batch of samples as a representative wavelength; if not, defining the wavelength of the current batch of samples as a non-representative wavelength; combining the representative wavelengths of all batches of samples into a representative data set;

[0029] Step 53: For the batch samples with representative wavelengths, calculating the coefficient of variation of the representative wavelengths;

[0030] Step 54: construct a second threshold value based on the representative data set; determine whether the coefficient of variation of the representative wavelength in the current batch of samples is less than or equal to the second threshold value, if so, use the representative wavelength of the current batch of samples as the key spectral feature, if not, use the representative wavelength of the current batch of samples as the non-key spectral feature.

[0031] Furthermore, the absolute value of the first threshold is 0.6.

[0032] Further, constructing a second threshold according to the representative data set specifically includes:

[0033] Step 541, obtaining the median of the representative wavelength coefficient of variation in each batch of samples according to the representative data set;

[0034] Step 542, obtaining the quartiles of the representative wavelength coefficient of variation in each batch of samples according to the representative data set;

[0035] Step 543, calculating a second threshold according to the median and the quartiles.

[0036] Furthermore, the calculation formula of the second threshold is:

[0037] ;

[0038] In the formula, represents the second threshold, represents the median value, represents the quartile, k represents the adjustment coefficient, Represents a dataset.

[0039] The present invention also provides a system for analyzing products of a petroleum fractionator based on near infrared technology, which is used to perform the above-mentioned method for analyzing products of a petroleum fractionator based on near infrared technology, comprising:

[0040] Sample collection module: used to collect different batches of samples from the petroleum distillation tower;

[0041] Scanning module: connected to the sample collection module, used to use a near-infrared spectrometer to perform near-infrared spectrum scanning on different batches of samples to obtain near-infrared spectrum data of different batches of samples, where the near-infrared spectrum data at least includes the measured spectrum intensity;

[0042] Preprocessing module: connected to the scanning module, used for preprocessing the near infrared spectrum data;

[0043] Calculation module: connected with the preprocessing module, used to calculate the optimal correction value of different batches of samples according to the measured spectral intensity after preprocessing;

[0044] Prediction module: connected to the calculation module, used to obtain the key spectral features of different batches of samples, and combine the measured spectral intensity and optimal correction value corresponding to the key spectral features into an input vector and input it into the pre-trained calibration model to obtain the chemical property information of different batches of samples.

[0045] The embodiments of the present invention have the following technical effects:

[0046] The present invention performs spectral drift correction on different batches of samples to calculate the optimal correction value, and obtains the measured spectral intensity under key spectral features as analysis data. The change in the measured spectral intensity can reflect the chemical composition information. Based on the change in the measured spectral intensity, the analysis result of the chemical property information of the sample can be obtained. The optimal correction value is used as a correction factor for the analysis result to compensate for the deviation of the analysis process. The optimal correction value is added to the analysis result to offset the analysis deviation, so that the final analysis result is more accurate.

[0047] The present invention eliminates the influence of spectral drift caused by environmental factors on the analysis results by calculating the optimal correction value, thereby improving the accuracy and reliability of near-infrared spectral analysis; at the same time, key spectral features and optimal correction values ​​are used as input features of the correction model, which reduces the amount of data and retains key data for analysis, effectively reducing the influence of spectral feature changes caused by differences between sample batches on the analysis results, and also improving the accuracy of the analysis, more accurately reflecting the true chemical property information of samples from different batches.

[0048] The present invention calculates the correlation coefficient between each batch of samples and the standard sample, determines the batch of samples with representative wavelengths according to a first threshold, calculates the coefficient of variation for the representative wavelengths of the batch samples, and screens out key spectral features from the representative wavelengths according to a second threshold. The present invention uses the representative wavelengths that are representative and have high stability in each batch of samples as key spectral features, reduces the number of input features through two-step screening, and simplifies the complexity of the calibration model. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0050] Figure 1 This is a flow chart of a method for analyzing petroleum fractionation tower products based on near infrared technology provided by an embodiment of the present invention;

[0051] Figure 2 It is a structural schematic diagram of a method for analyzing petroleum fractionation tower products based on near infrared technology provided by an embodiment of the present invention;

[0052] Figure 3 It is a structural schematic diagram of an electronic device provided by the implementation of the present invention. DETAILED DESCRIPTION

[0053] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be described clearly and completely below. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work belong to the scope of protection of the present invention.

[0054] Figure 1 This is a flow chart of a method for analyzing petroleum fractionation tower products based on near infrared technology provided by an embodiment of the present invention. Figure 1 , including:

[0055] Step 1: Collect different batch samples from the petroleum fractionator.

[0056] The oil fractionation tower is an important refining equipment. It separates crude oil into fractions with different boiling point ranges by heating and distillation. In order to ensure product quality and optimize the production process, samples in the fractionation tower need to be collected and analyzed regularly. Open the valve of the sampling port to let the sample flow out, rinse the sampling bottle with a small amount of sample to remove possible contaminants, close the valve, wait for a while and then open it again to let fresh sample flow into the sampling bottle until it is full, quickly close the valve, and immediately seal the sampling bottle to prevent the sample from volatilizing or being contaminated. Mark each sampling bottle and indicate the sampling time, sampling location, batch number and other information for subsequent analysis and tracking. Set the collection interval. Samples collected in different collection intervals are samples from different batches.

[0057] Step 2: Use a near-infrared spectrometer to perform near-infrared spectrum scanning on different batches of samples to obtain near-infrared spectrum data of different batches of samples. The near-infrared spectrum data at least includes wavelength and measured spectrum intensity.

[0058] Near infrared spectroscopy data include but are not limited to:

[0059] Measuring spectral intensity: including reflectance, transmittance, and absorbance;

[0060] Wavelength: including wavelength range and wavelength frequency;

[0061] Scanning time, environmental conditions, sample information, background spectrum, etc.

[0062] Step 3: Preprocess the near-infrared spectral data.

[0063] Smoothing: Use smoothing algorithms to smooth the spectral data to reduce the impact of random noise;

[0064] Background subtraction: Use the spectral data of the blank sample as the background for correction to eliminate background interference;

[0065] Baseline adjustment: Use a baseline correction algorithm to eliminate baseline drift;

[0066] Normalization: Normalize the spectral data to make the data of different batches of samples comparable;

[0067] Principal component analysis: Dimensionality reduction of spectral data, reducing data dimensions while retaining the main features;

[0068] Outlier Detection: Detect and remove outliers using statistical methods.

[0069] Step 4: Calculate the optimal correction values ​​for different batches of samples based on the measured spectral intensities after preprocessing.

[0070] In order to improve the accuracy and reliability of near-infrared spectral analysis, it is necessary to calibrate the spectral data of different batches of samples. By calculating the spectral drift value and applying the correction value, the influence of environmental factors on the spectral data can be eliminated and the accuracy of the analysis results can be improved. The specific performance is:

[0071] Step 41: Obtain standardized values ​​of standard samples at different wavelengths.

[0072] In near-infrared spectroscopy detection, samples are affected by their own chemical properties, environmental factors, and instrument parameters, which leads to changes in spectral data at different wavelengths, so that the spectral data of different batches of samples at different wavelengths may be different. Therefore, this embodiment obtains spectral data at different wavelengths to analyze the samples.

[0073] The wavelength of the sample is composed of several wavelength points, each of which has specific spectral data.

[0074] A standard sample is a sample with known chemical properties, which is usually used to calibrate instruments and verify the accuracy of analytical methods. In this embodiment, the standard sample refers to a sample with known composition prepared under the same conditions (such as temperature, humidity, etc.). The standardized values ​​include the spectral intensity data of the standard sample, which are usually obtained under ideal conditions (such as constant temperature and humidity) and used as a reference. The standardized values ​​also include the measured spectral intensity, reflectivity, transmittance, absorbance and other data of the standard sample at different wavelengths.

[0075] Step 42: Calculate the spectral drift values ​​of different batches of samples at different wavelengths according to the standardized values ​​and the measured spectral intensity after preprocessing;

[0076] The calculation formula of the spectral drift value is:

[0077] ;

[0078] In the formula, represents the spectral drift value of the kth batch of samples, represents the measured spectral intensity of the i-th wavelength of the k-th batch of samples, represents the spectral intensity of the i-th wavelength of the standard sample, Represents the weight coefficient of the i-th wavelength.

[0079] By calculating the spectral drift value, the data of different batches of samples can be adjusted to a relatively uniform standard, making the data of different batches more comparable. The spectral drift value can be used to compensate for the spectral drift caused by environmental changes, thereby improving the reliability of the analysis results, improving the prediction accuracy of the model, and ensuring that subsequent analysis results are more accurate and reliable.

[0080] Step 43: Calculate the correction values ​​of different batches of samples according to the spectral drift value and the near infrared spectrum data;

[0081] The calculation formula of the correction value is:

[0082] ;

[0083] In the formula, represents the calibration value of the kth batch of samples, represents the near infrared spectral data of the kth batch of samples, Represents the spectral drift value of the kth batch of samples.

[0084] The near-infrared spectral data are corrected by compensating for the spectral drift value, making the data of different batches of samples closer to the true value.

[0085] Step 44: Optimize the correction value to obtain the optimal correction value.

[0086] In near-infrared spectroscopy analysis, in order to further improve the accuracy and reliability of the correction value, it is necessary to optimize the correction value obtained by preliminary calculation to eliminate the residual errors and deviations. Through optimization, the optimal correction value closer to the true value can be obtained, thereby improving the accuracy of subsequent analysis. Specifically, it is manifested as:

[0087] Step 441, constructing a primary optimization function according to the correction value, the spectral drift value, and the overall difference value, and calculating the optimized correction value.

[0088] The primary optimization function is:

[0089] ;

[0090] In the formula, represents the optimized correction value of the kth batch of samples, represents the primary optimization function, and Represent the primary optimization coefficients, Represents the overall difference, specifically the difference between the calibration value of the kth batch of samples and the calibration values ​​of all batches of samples.

[0091] The calculation process is as follows: First, the average value of the calibration values ​​of all batches of samples is calculated, and the difference between the calibration value of each batch of samples and the average value is calculated to obtain the overall difference. .

[0092] Step 442, constructing an optimal optimization function according to the optimized correction value and the spectral drift value, and calculating the optimal correction value.

[0093] The optimal optimization function is:

[0094] ;

[0095] In the formula, represents the optimal correction value of the kth batch of samples, and Represent the optimal optimization coefficients, Represents the optimal optimization function.

[0096] Through two optimization processes, the residual error can be eliminated, making the optimal correction value closer to the true value, and the model more accurately reflects the actual chemical properties of the sample, and serves as the model input data to improve the prediction accuracy of the model.

[0097] Step 5: Obtain the key spectral features of samples from different batches, and combine the measured spectral intensity and optimal correction value corresponding to the key spectral features into an input vector and input it into the pre-trained calibration model to obtain the chemical property information of samples from different batches.

[0098] In order to improve the accuracy and reliability of near-infrared spectral analysis, it is necessary to extract key spectral features from a large amount of spectral data and input them into a pre-trained calibration model together with the optimal correction values ​​to obtain the chemical property information of different batches of samples.

[0099] Obtaining key spectral features includes the following steps:

[0100] Step 51: Calculate the correlation coefficient between the wavelength of each batch of samples and the standardized data in the standard sample.

[0101] Calculating the correlation coefficient can measure the linear relationship between the wavelengths of different batches of samples and the wavelength of the standard sample. A correlation coefficient close to 1 or -1 indicates a strong linear relationship between the two, while a correlation coefficient close to 0 indicates no obvious relationship. By calculating the correlation coefficient, wavelengths that show a strong correlation between different batches of samples and the standard sample can be screened out, that is, their changing trends in different batches of samples are similar, making them suitable for subsequent key spectral feature extraction.

[0102] The correlation coefficient of this embodiment is obtained by calculating the Pearson correlation coefficient.

[0103] Step 52: construct a first threshold; determine whether the correlation coefficient is greater than or equal to the first threshold, if so, define the wavelength of the current batch of samples as a representative wavelength; if not, define the wavelength of the current batch of samples as a non-representative wavelength; combine the representative wavelengths of all batches of samples into a representative data set.

[0104] In this embodiment, in order to screen out those wavelengths that show a strong correlation with the standard sample in different batches of samples, that is, to find wavelengths that are more consistent in different batches of samples. Preferably, the absolute value of the first threshold is assigned to 0.6. When the calculated absolute value of the correlation coefficient is greater than or equal to 0.6, it means that these wavelengths vary less in different batches of samples, that is, their performance in different batches is relatively stable, and the wavelength of the current batch of samples is defined as a representative wavelength; when the calculated absolute value of the correlation coefficient is less than 0.6, it means that these wavelengths vary more in different batches of samples, that is, their performance in different batches is unstable, and the wavelength of the current batch of samples is defined as a non-representative wavelength.

[0105] Step 53: For the batch samples with representative wavelengths, calculate the coefficient of variation of the representative wavelengths.

[0106] The coefficient of variation is used to measure the dispersion of representative wavelengths. It is obtained by the ratio of the standard deviation and the mean of the representative wavelengths. Calculating the coefficient of variation can further screen out wavelengths with higher stability in different batches of samples. The smaller the coefficient of variation, the smaller the change of the wavelength in different batches, that is, the higher its stability.

[0107] Step 54: construct a second threshold value based on the representative data set; determine whether the coefficient of variation of the representative wavelength of the current batch of samples is less than or equal to the second threshold value, if so, use the representative wavelength of the current batch of samples as the key spectral feature, if not, use the representative wavelength of the current batch of samples as the non-key spectral feature.

[0108] The median of the coefficient of variation is used as the second threshold. If the coefficient of variation is less than or equal to the second threshold, it means that the wavelength has little variation in different batches of samples, that is, it has little impact on different batches of samples and is suitable as a key spectral feature. In this embodiment, the key spectral features are retained and non-key spectral features are eliminated.

[0109] Step 541, obtaining the median of the representative wavelength coefficient of variation in each batch of samples according to the representative data set;

[0110] Step 542, obtaining the quartiles of the representative wavelength coefficient of variation in each batch of samples according to the representative data set;

[0111] Step 543, calculating a second threshold according to the median and the quartiles.

[0112] The calculation formula of the second threshold is:

[0113] ;

[0114] In the formula, represents the second threshold, represents the median value, represents the quartile, k represents the adjustment coefficient, Represents a dataset.

[0115] Using the median and quartiles to determine the second threshold can reduce the impact of outliers in the data on the second threshold. Because the median is not easily affected by extreme values, and the quartiles provide a summary of the data distribution, the calculation results are more robust, which helps to improve the stability of the data and ensure that the representative wavelengths screened out have high representativeness in different batches of samples. The representative wavelengths screened out by the second threshold have a smaller coefficient of variation and can better represent the common characteristics of different batches of samples. This can improve the accuracy and reliability of subsequent model predictions, because these wavelengths perform more consistently in different batches and are not easily affected by external factors.

[0116] Using the median and quartiles to determine the second threshold is not only simple and easy, but can also effectively eliminate wavelengths that vary greatly between different batches of samples, thereby simplifying the complexity of the model and improving the generalization ability of the model.

[0117] The calibration model of this embodiment is a pre-trained model, and no re-training is required before prediction. The chemical property information of different batches of samples can be obtained by inputting the input vector into the pre-trained calibration model.

[0118] Key spectral features usually contain information about the main chemical components of the sample, such as water, fat, protein, etc. Changes in these components can be reflected by changes in spectral intensity. For example, the absorption peak of water is usually very obvious near specific wavelengths, so the spectral intensity at these wavelengths can be used to infer the moisture content of the sample. Key spectral features are relatively consistent in different batches of samples, which means that their changes can be used to distinguish the differences between different samples, thereby helping to analyze chemical properties.

[0119] The optimal correction value is used as a correction factor in the calibration model analysis process to compensate for the deviation of the analysis process. The optimal correction value is added to the analysis result to offset the analysis deviation and make the final analysis result more accurate.

[0120] The chemical property information obtained by analysis includes but is not limited to: moisture content, fat content, protein content, cellulose content, ash content, sugar content, acidity, pH value, density, viscosity, color index, etc.

[0121] like Figure 3 As shown, this embodiment also discloses a petroleum fractionation tower product analysis system based on near infrared technology, comprising:

[0122] Sample collection module: used to collect different batches of samples from the petroleum distillation tower;

[0123] Scanning module: connected to the sample collection module, used to use a near-infrared spectrometer to perform near-infrared spectrum scanning on different batches of samples to obtain near-infrared spectrum data of different batches of samples, where the near-infrared spectrum data at least includes the measured spectrum intensity;

[0124] Preprocessing module: connected to the scanning module, used for preprocessing the near infrared spectrum data;

[0125] Calculation module: connected with the preprocessing module, used to calculate the optimal correction value of different batches of samples according to the measured spectral intensity after preprocessing;

[0126] Prediction module: connected to the calculation module, used to obtain the key spectral features of different batches of samples, and combine the measured spectral intensity and optimal correction value corresponding to the key spectral features into an input vector and input it into the pre-trained calibration model to obtain the chemical property information of different batches of samples.

[0127] Figure 3 Schematic diagram of the structure of an electronic device provided by the present invention. Figure 3 As shown, the electronic device 500 includes one or more processors 501 and a memory 502 .

[0128] The processor 501 may be a central processing unit (CPU) or other forms of processing units having data processing capabilities and / or instruction execution capabilities, and may control other components in the electronic device 500 to perform desired functions.

[0129] The memory 502 may include one or more computer program products, which may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may include, for example, random access memory (RAM) and / or cache memory (cache), etc. The non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc. One or more computer program instructions may be stored on the computer-readable storage medium, and the processor 501 may run the program instructions to implement a method for analyzing petroleum fractionation tower products based on near-infrared technology and / or other desired functions of any embodiment of the present application described above. Various contents such as initial external parameters, thresholds, etc. may also be stored in the computer-readable storage medium.

[0130] In one example, the electronic device 500 may further include: an input device 503 and an output device 504, which are interconnected via a bus system and / or other forms of connection mechanisms (not shown). The input device 503 may include, for example, a keyboard, a mouse, etc. The output device 504 may output various information to the outside, including early warning prompt information, braking force, etc. The output device 504 may include, for example, a display, a speaker, a printer, a communication network and a remote output device connected thereto, etc.

[0131] Of course, to simplify, Figure 3 Only some of the components related to the present application in the electronic device 500 are shown, and components such as a bus, an input / output interface, etc. are omitted. In addition, according to specific application situations, the electronic device 500 may also include any other appropriate components.

[0132] In addition to the above-mentioned methods and devices, an embodiment of the present application may also be a computer program product, which includes computer program instructions, which, when executed by a processor, enable the processor to execute the steps of a method for analyzing petroleum distillation tower products based on near-infrared technology provided in any embodiment of the present application.

[0133] The computer program product may be written in any combination of one or more programming languages ​​to write program codes for performing the operations of the embodiments of the present application, including object-oriented programming languages, such as Java, C++, etc., and conventional procedural programming languages, such as "C" language or similar programming languages. The program code may be executed entirely on the user computing device, partially on the user device, as an independent software package, partially on the user computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0134] In addition, an embodiment of the present application may also be a computer-readable storage medium on which computer program instructions are stored. When the computer program instructions are executed by a processor, the processor executes the steps of a method for analyzing petroleum distillation tower products based on near-infrared technology provided in any embodiment of the present application.

[0135] The computer readable storage medium can adopt any combination of one or more readable media. The readable medium can be a readable signal medium or a readable storage medium. The readable storage medium can include, for example, but is not limited to, a system, device or device of electricity, magnetism, light, electromagnetic, infrared, or semiconductor, or any combination of the above. More specific examples (non-exhaustive list) of readable storage media include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0136] It should be noted that the terms used in the present invention are only for describing specific embodiments, rather than limiting the scope of the present application. As shown in the present specification, unless the context clearly indicates an exception, the words "one", "a", "a kind of" and / or "the" do not specifically refer to the singular, but may also include the plural. The terms "include", "comprise" or any other variant thereof are intended to cover non-exclusive inclusion, so that the process, method or device including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method or device. In the absence of more restrictions, the elements defined by the sentence "include one..." do not exclude the presence of other identical elements in the process, method or device including the elements.

[0137] It should also be noted that the terms "center", "up", "down", "left", "right", "vertical", "horizontal", "inside", "outside", etc., indicating the orientation or positional relationship, are based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention. Unless otherwise clearly specified and limited, the terms "installed", "connected", "connected", etc. should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be an indirect connection through an intermediate medium, or it can be a connection between the two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0138] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the technical solutions of the embodiments of the present invention.

Claims

1. A method for analyzing petroleum fractionator products based on near infrared technology, characterized in that: include: Step 1: Collect different batches of samples from the petroleum fractionator; Step 2: Use a near-infrared spectrometer to perform near-infrared spectrum scanning on different batches of samples to obtain near-infrared spectrum data of different batches of samples, where the near-infrared spectrum data at least includes wavelength and measured spectrum intensity; Step 3: preprocessing the near infrared spectrum data; Step 4: Calculate the optimal correction value for different batches of samples based on the measured spectral intensity after preprocessing, including: Step 41: Obtain standardized values ​​of standard samples at different wavelengths; Step 42: Calculate the spectral drift values ​​of different batches of samples at different wavelengths according to the standardized values ​​and the measured spectral intensity after preprocessing; Step 43: Calculate the correction values ​​of different batches of samples according to the spectral drift value and the near infrared spectrum data; Step 44: Optimizing the correction value to obtain the optimal correction value; Step 5: Obtain key spectral features of samples from different batches, and combine the measured spectral intensity and optimal correction value corresponding to the key spectral features into an input vector and input it into the pre-trained calibration model to obtain the chemical property information of samples from different batches; obtaining key spectral features specifically includes the following steps: Step 51: Calculate the correlation coefficient between the wavelength of each batch of samples and the standardized data in the standard sample; Step 52: constructing a first threshold value for the correlation coefficient; determining whether the correlation coefficient is greater than or equal to the first threshold value, if so, defining the wavelength of the current batch of samples as a representative wavelength; if not, defining the wavelength of the current batch of samples as a non-representative wavelength; combining the representative wavelengths of all batches of samples into a representative data set; Step 53: For the batch samples with representative wavelengths, calculating the coefficient of variation of the representative wavelengths; Step 54: construct a second threshold value based on the representative data set; determine whether the coefficient of variation of the representative wavelength in the current batch of samples is less than or equal to the second threshold value, if so, use the representative wavelength of the current batch of samples as the key spectral feature, if not, use the representative wavelength of the current batch of samples as the non-key spectral feature.

2. A method for analyzing petroleum fractionator products based on near infrared technology according to claim 1, characterized in that: The calculation formula of the spectral drift value is: ； In the formula, represents the spectral drift value of the kth batch of samples, represents the measured spectral intensity of the i-th wavelength of the k-th batch of samples, represents the spectral intensity of the i-th wavelength of the standard sample, Represents the weight coefficient of the i-th wavelength.

3. A method for analyzing petroleum fractionator products based on near infrared technology according to claim 1, characterized in that: The calculation formula of the correction value is: ； In the formula, represents the calibration value of the kth batch of samples, Represents the original near-infrared spectral data of the kth batch of samples, Represents the spectral drift value of the kth batch of samples.

4. A method for analyzing petroleum fractionator products based on near infrared technology according to claim 1, characterized in that: The correction value is optimized to obtain the optimal correction value, which specifically includes: Step 441, constructing a primary optimization function according to the correction value, the spectral drift value, and the overall difference value, and calculating the optimized correction value; Step 442, constructing an optimal optimization function according to the optimized correction value and the spectral drift value, and calculating the optimal correction value.

5. A method for analyzing petroleum fractionator products based on near infrared technology according to claim 1, characterized in that: The absolute value of the first threshold is 0.

6.

6. The method for analyzing petroleum fractionator products based on near infrared technology according to claim 1, characterized in that: Constructing a second threshold based on the representative data set specifically includes: Step 541, obtaining the median of the representative wavelength coefficient of variation in each batch of samples according to the representative data set; Step 542, obtaining the quartiles of the representative wavelength coefficient of variation in each batch of samples according to the representative data set; Step 543, calculating a second threshold according to the median and the quartiles.

7. A method for analyzing petroleum fractionator products based on near infrared technology according to claim 6, characterized in that: The calculation formula of the second threshold is: ； In the formula, represents the second threshold, represents the median value, represents the quartile, k represents the adjustment coefficient, Represents a dataset.

8. A system for analyzing products of a petroleum fractionator based on near infrared technology, used for executing a method for analyzing products of a petroleum fractionator based on near infrared technology as claimed in any one of claims 1 to 7, characterized in that: include: Sample collection module: used to collect different batches of samples from the petroleum distillation tower; Scanning module: connected to the sample collection module, used to use a near-infrared spectrometer to perform near-infrared spectrum scanning on different batches of samples to obtain near-infrared spectrum data of different batches of samples, where the near-infrared spectrum data at least includes the measured spectrum intensity; Preprocessing module: connected to the scanning module, used for preprocessing the near infrared spectrum data; Calculation module: connected with the preprocessing module, used to calculate the optimal correction value of different batches of samples according to the measured spectral intensity after preprocessing; Prediction module: connected to the calculation module, used to obtain the key spectral features of different batches of samples, and combine the measured spectral intensity and optimal correction value corresponding to the key spectral features into an input vector and input it into the pre-trained calibration model to obtain the chemical property information of different batches of samples.

Images