Heavy oil quality detection method and system based on multispectral fusion

Abstract

The invention relates to the field of heavy oil quality detection, in particular to a heavy oil quality detection method and system based on multispectral fusion. When a moving average detrending method is adopted for preprocessing multispectral data, firstly, the multispectral data are classified into extreme value interval multispectral data and common interval multispectral data; then data preprocessing is carried out by adopting different methods according to different data types, data preprocessing is carried out on common interval multi-spectral data by adopting a conventional moving average detrending method, preprocessing is carried out on extreme interval multi-spectral data by adopting a window attenuation mode, and when an extreme point is encountered, the multi-spectral data in the extreme interval is subjected to window attenuation; a mode of skipping an extreme point and then recovering an initial window is adopted, so that the influence of a detrending algorithm on the extreme point in the multispectral data can be effectively reduced, and the accuracy of data processing is improved; meanwhile, the mode of window attenuation and extreme point skipping also enables the data preprocessing time to be shortened to a certain extent, and the data preprocessing efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the field of heavy oil quality detection, and in particular to a heavy oil quality detection method and system based on multi-spectral fusion. Background Art

[0002] Heavy oil is an important energy resource, and its quality detection is of great significance for oil processing, transportation and use. Traditional heavy oil quality detection methods mainly rely on chemical analysis and physical testing, which are usually time-consuming, complicated to operate and destructive to samples. In recent years, spectral technology has been widely used in oil testing, but single spectral technology often has limitations and cannot fully reflect the complex characteristics of heavy oil.

[0003] Multi-spectral fusion technology can achieve complementary advantages by integrating different types of spectral data, thereby providing more comprehensive and reliable detection results. For example, in mineral oil detection, the multi-spectral fusion method combining infrared spectroscopy and Raman spectroscopy has been proven to significantly improve detection accuracy. However, the multi-spectral fusion technology for heavy oil quality detection is still in the development stage, and the data preprocessing process is accurate and efficient. Therefore, there is a lack of efficient and accurate detection systems and methods. Summary of the invention

[0004] In order to solve the above technical problems, the present invention provides a heavy oil quality detection method and system based on multi-spectral fusion, which are used to solve the problems existing in the prior art.

[0005] The present invention provides a heavy oil quality detection method based on multi-spectral fusion, comprising the following steps: S1: Perform multi-spectral acquisition on heavy oil samples to obtain multi-spectral data; S2: performing a data preprocessing operation on the multi-spectral data to obtain preprocessed multi-spectral data; The preprocessing operation includes standard normal variable transformation preprocessing and moving average detrending algorithm preprocessing; The data preprocessing operation of the multi-spectral data using the moving average detrending algorithm is specifically as follows: S2.a: Performing Fourier transform on the multi-spectral data to obtain frequency domain multi-spectral data; S2.b: identifying extreme points of the multi-spectral data according to the frequency domain multi-spectral data, and marking them; S2.c: The multi-spectral segment data within the preset range on both sides of the extreme value point are used as the extreme value interval multi-spectral segment data; the other multi-spectral segment data are used as the common interval multi-spectral segment data; S2.d: For multi-spectral data in common intervals, the first moving average method is used for data preprocessing; S2.e: For the multi-spectral data in the extreme value interval, the second moving average method is used for data preprocessing; S3: performing a data fusion operation on the preprocessed multi-spectral data; S4: Establish a heavy oil quality detection model; S5: Performing quality inspection on the heavy oil.

[0006] Preferably, in S2.b, the extreme points of the frequency domain multi-spectral data are identified by a threshold method; specifically: the highest value of the preset multiples of the frequency domain multi-spectral data is set as an upper threshold, and the lowest value of the preset multiples of the frequency domain multi-spectral data is set as a lower threshold, and the frequency domain multi-spectral data is traversed, and the local maximum points above the upper threshold are taken as extreme points, and the local minimum points below the lower threshold are taken as extreme points.

[0007] Preferably, the preset multiple is 0.5.

[0008] Preferably, in S2.c, the formula for determining the preset range is: ; In the formula, D is the number of data points in the preset range, a is the coefficient, is the sampling frequency of the i-th multi-spectral data, D<20.

[0009] Preferably, in S2.d, the formula of the first moving average method is:

[0010] In the formula, is the i-th multi-spectral data after preprocessing by the first moving average method, is the original i-th multi-spectral data, and n is the number of cycles of data movement.

[0011] Preferably, in S2.e, the second moving average method is used to preprocess the data of the extreme value interval multi-spectral segment as follows: the initial window period number n is set 1 , and then use the window period number that decays in sequence to perform data preprocessing on the multi-spectral data; when moving to the marked extreme point, skip the extreme point, and then set the initial window period number n at the value after the extreme point 1 , and then the multi-spectral data is preprocessed using the window period number that decays in sequence. The formula for determining the period number n that decays in sequence is:

[0012]

[0013] Where s is the window cycle number attenuation coefficient s. Preferably, in S3, a wavelet feature fusion method is used to perform a data fusion operation on the preprocessed multi-spectral data.

[0014] Preferably, S3 specifically comprises: performing multi-layer wavelet transform on the multi-spectral data to obtain a plurality of high-frequency wavelet data and a plurality of low-frequency wavelet data, fusing the high-frequency wavelet data by setting a large absolute value, and fusing the low-frequency wavelet data by removing the average value.

[0015] Preferably, in S4, the heavy oil quality detection model is a support vector machine model.

[0016] According to another aspect of the present invention, a heavy oil quality detection system based on multi-spectral fusion is provided. The detection system adopts the above-mentioned heavy oil quality detection method based on multi-spectral fusion. The system comprises: A multi-spectral data acquisition module is used to perform multi-spectral acquisition on heavy oil samples to obtain multi-spectral data; A data preprocessing module, used for performing a data preprocessing operation on the multi-spectral data to obtain preprocessed multi-spectral data; A data fusion module, used for performing a data fusion operation on the preprocessed multi-spectral data; A heavy oil quality detection model building module is used to build a heavy oil quality detection model; The quality detection module is used to perform quality detection on the heavy oil.

[0017] The embodiments of the present invention have the following technical effects: When using the moving average detrending method to preprocess multi-spectral data, the multi-spectral data are first classified into extreme value interval multi-spectral data and ordinary interval multi-spectral data, and then different methods are used for data preprocessing according to different data types. For ordinary interval multi-spectral data, the conventional moving average detrending method is used for data preprocessing, and for extreme value interval multi-spectral data, the window attenuation method is used for preprocessing. When encountering extreme points, the extreme points are skipped and then the initial window is restored. This can effectively reduce the impact of the detrending algorithm on the extreme points in the multi-spectral data and improve the accuracy of data processing. At the same time, the window attenuation and extreme point skipping methods also reduce the time of data preprocessing to a certain extent and improve the efficiency of data preprocessing. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] 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.

[0019] Figure 1 It is a flow chart of a heavy oil quality detection method based on multi-spectral fusion provided by an embodiment of the present invention; Figure 2 It is a flowchart of a data preprocessing operation on the multi-spectral data using a moving average detrending algorithm provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0020] 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.

[0021] Embodiment 1, attached Figure 1 A flowchart of a heavy oil quality detection method based on multi-spectral fusion is shown in the attached figure. Figure 1 As shown, a heavy oil quality detection method based on multi-spectral fusion includes the following steps: S1: Perform multi-spectral acquisition on heavy oil samples to obtain multi-spectral data; In the process of multi-spectral acquisition of heavy oil samples, the multi-spectral bands include: near infrared spectrum band and Raman spectrum band; Among them, near-infrared light is an electromagnetic wave with a frequency between ultraviolet-visible light and mid-infrared light. The near-infrared spectrum mainly involves the vibration energy level transitions of chemical bonds such as CH, OH, NH in molecules. After absorbing near-infrared light of a specific wavelength, these bonds will transition from the ground state to the excited state to produce an absorption spectrum. The wavelength range of the near-infrared spectrum is usually 780-2500 nanometers, which is between visible light and infrared light. The light energy in this range is relatively low, which mainly triggers the frequency doubling and summation transitions of molecular vibrations. Heavy oil is mainly composed of hydrocarbons and a small amount of sulfur-containing, nitrogen-containing, and oxygen-containing compounds. The chemical bonds in these compounds (such as CH, OH, NH, etc.) will absorb light of specific wavelengths in the near-infrared region (780-2500 nm) to produce characteristic absorption peaks.

[0022] The principle of Raman spectroscopy for crude oil detection is based on the Raman scattering effect. When the laser irradiates the crude oil sample, the photons interact with the molecules in the sample, causing the photon energy to change, thereby generating scattered light. The difference between the frequency of the scattered light and the frequency of the incident light is called the Raman shift, which is related to the vibration mode and chemical bonds of the sample molecules. By analyzing the characteristic peaks and their intensities in the Raman spectrum, information on different chemical components in the crude oil can be obtained, such as hydrocarbon compounds, sulfur-containing compounds, nitrogen-containing compounds, etc. In practical applications, Raman spectroscopy technology can be used for composition analysis, quality control and quality detection of heavy oil.

[0023] The near infrared spectrometer is used to collect data on the heavy oil. The acquisition parameters of the near infrared spectrometer are: the wavelength of the detection light source is 850nm near infrared light, the number of scans is 32 times, and the resolution is 3.5cm -1 ; The Raman spectrometer is used to collect data on the heavy oil. The collection parameters of the Raman spectrometer are as follows: the spectrum collection wave number range of the Raman spectrometer is 250-2340cm -1 , the central wavelength is 785nm, the laser power is 400mW, and the integration time is 500ms.

[0024] S2: performing a data preprocessing operation on the multi-spectral data to obtain preprocessed multi-spectral data; Multi-spectral data are more easily affected by external factors, such as temperature, noise, and light, which will affect the collected multi-spectral data, resulting in a certain difference between the multi-spectral spectrum collected by the acquisition instrument and the actual multi-spectral spectrum, thereby interfering with the extraction of effective information from the multi-spectral spectrum. Therefore, before performing multi-spectral data analysis, data preprocessing operations must be performed to reduce or eliminate the impact of interfering data.

[0025] In this step, the preprocessing operation includes standard normal variable transformation (SNV) preprocessing and moving average detrending algorithm preprocessing; The standard normal variable transformation preprocessing is to calculate the average reflectance value or absorbance value of the spectral data at each wavelength position in the multi-spectral data, then subtract the average value at the corresponding wavelength position from the spectral data of each sample, and calculate the standard deviation, and divide the spectral data of each sample by the standard deviation at the corresponding wavelength position to obtain the standard normal variable transformation preprocessing data; Detrending is a data processing method that is mainly used to eliminate long-term trends or periodic changes in time series data in order to more accurately analyze random fluctuations or other specific characteristics of the data. Its core principle is to remove the trend component by mathematically processing the original data, thereby obtaining a more stable data sequence. Moving average detrending algorithm is a type of detrending algorithm. For spectral data, the moving average detrending algorithm is generally used to preprocess multi-spectral data. However, the moving average detrending method is prone to change the amplitude of the extreme point too much when preprocessing multi-spectral data, resulting in distortion of the preprocessed multi-spectral data and errors in the quality detection of heavy oil. In this embodiment, if Figure 2 As shown, the data preprocessing operation of the multi-spectral data using the moving average detrending algorithm is specifically as follows: S2.a: Performing Fourier transform on the multi-spectral data to obtain frequency domain multi-spectral data; Fourier transform is a mathematical tool that can decompose complex signals into a series of sinusoidal components of different frequencies. In spectroscopy, Fourier transform converts data on light intensity varying with time or space into a spectrum where light intensity varies with frequency. In this way, features in the spectrum, such as extreme points such as absorption peaks and emission peaks, can be more intuitively observed and analyzed; S2.b: identifying extreme points of the multi-spectral data according to the frequency domain multi-spectral data, and marking them; In this step, the extreme points of the frequency domain multi-spectral data are identified by a threshold method; specifically, the highest value of the preset multiples of the frequency domain multi-spectral data is set as an upper threshold, and the lowest value of the preset multiples of the frequency domain multi-spectral data is set as a lower threshold, and the frequency domain multi-spectral data is traversed, and the local maximum points above the upper threshold are taken as extreme points, and the local minimum points below the lower threshold are taken as extreme points; As a preferred embodiment, the preset multiple is 0.5; S2.c: The multi-spectral segment data within the preset range on both sides of the extreme value point are used as the extreme value interval multi-spectral segment data; the other multi-spectral segment data are used as the common interval multi-spectral segment data; Among them, in this step, the formula for determining the preset range is: ; In the formula, D is the number of data points in the preset range, a is the coefficient, is the sampling frequency of the i-th multi-spectral data; in this embodiment, D<20; Therefore, the extreme value interval multi-spectral data consists of 2D+1 data points; S2.d: For multi-spectral data in common intervals, the first moving average method is used for data preprocessing; The formula of the first moving average method is:

[0026] In the formula, is the i-th multi-spectral data after preprocessing by the first moving average method, is the original i-th multi-spectral data, n is the number of cycles of data movement; S2.e: For the multi-spectral data in the extreme value interval, the second moving average method is used for data preprocessing; Among them, for the extreme value interval multi-spectral data, the second moving average method is used for data preprocessing: set the initial window period number n 1 , and then use the window period number that decays in sequence to perform data preprocessing on the multi-spectral data; when moving to the marked extreme point, skip the extreme point, and then set the initial window period number n at the value after the extreme point 1 , and then performing data preprocessing on the multi-spectral data using a number of window cycles that decays in sequence; The formula for determining the number of window cycles n that decays in sequence is:

[0027]

[0028] Among them, s is the window cycle number attenuation coefficient s, that is, each time the window moves, the value is attenuated by s based on the initial window cycle number; In this step, when the moving average detrending method is used to preprocess the multi-spectral data, the multi-spectral data is first classified into extreme value interval multi-spectral data and ordinary interval multi-spectral data, and then different methods are used for data preprocessing according to different data types. For ordinary interval multi-spectral data, the conventional moving average detrending method is used for data preprocessing, and for extreme value interval multi-spectral data, the window attenuation method is used for preprocessing, and when encountering extreme value points, the method of skipping the extreme value points and then restoring the initial window is used. This can effectively reduce the impact of the detrending algorithm on the extreme value points in the multi-spectral data and improve the accuracy of data processing. At the same time, the window attenuation and extreme value point skipping methods also reduce the time of data preprocessing to a certain extent, thereby improving the efficiency of data preprocessing.

[0029] S3: performing a data fusion operation on the preprocessed multi-spectral data; In this step, a wavelet feature fusion method is used to perform a data fusion operation on the preprocessed multi-spectral data; Specifically, S3 is as follows: performing multi-layer wavelet transform on the multi-spectral data to obtain a plurality of high-frequency wavelet data and a plurality of low-frequency wavelet data, fusing the high-frequency wavelet data by setting a large absolute value, and fusing the low-frequency wavelet data by removing the average value.

[0030] S4: Establish a heavy oil quality detection model; The heavy oil quality detection model is a support vector machine model. Support vector machine (SVM) is a powerful supervised learning algorithm that is widely used in classification and regression tasks. The core idea of ​​SVM is to separate data points of different categories by finding an optimal hyperplane and maximizing the classification interval.

[0031] The input of the support vector machine model is the fused data, and the output of the support vector machine model is the quality parameters of heavy oil, such as density, viscosity, sulfur content, etc.

[0032] S5: Performing quality testing on the heavy oil; The multi-spectral data of the heavy oil sample to be tested is input into the heavy oil quality detection model to obtain its quality detection result.

[0033] Embodiment 2, the present invention further provides a heavy oil quality detection system based on multi-spectral fusion, the detection system adopts the heavy oil quality detection method based on multi-spectral fusion of embodiment 1, and the system comprises: A multi-spectral data acquisition module is used to perform multi-spectral acquisition on heavy oil samples to obtain multi-spectral data; A data preprocessing module, used for performing a data preprocessing operation on the multi-spectral data to obtain preprocessed multi-spectral data; A data fusion module, used for performing a data fusion operation on the preprocessed multi-spectral data; A heavy oil quality detection model building module is used to build a heavy oil quality detection model; The quality detection module is used to perform quality detection on the heavy oil.

[0034] Embodiment 3, the present invention also provides an electronic device, including one or more processors and a memory.

[0035] The processor 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 to perform desired functions.

[0036] The memory may include one or more computer program products, and the computer program product 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 may run the program instructions to implement the heavy oil quality detection method based on multi-spectral fusion 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.

[0037] In one example, the electronic device may further include: an input device and an output device, and these components are interconnected through a bus system and / or other forms of connection mechanisms (not shown). The input device may include, for example, a keyboard, a mouse, etc. The output device may output various information to the outside, including early warning information, braking force, etc. The output device may include, for example, a display, a speaker, a printer, a communication network and a remote output device connected thereto, etc.

[0038] Of course, for the sake of simplicity, components such as buses, input / output interfaces, etc. are omitted. In addition, the electronic device may further include any other appropriate components according to specific application conditions.

[0039] In addition to the above-mentioned methods and devices, the embodiments 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 implement the functions of the heavy oil quality detection method based on multi-spectral fusion provided in any embodiment of the present application.

[0040] 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.

[0041] In addition, the embodiments 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 implements the heavy oil quality detection method based on multi-spectral fusion provided in any embodiment of the present application.

[0042] 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.

[0043] 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 heavy oil quality detection method based on multi-spectral fusion, characterized in that: The following steps are involved: S1: Perform multi-spectral acquisition on heavy oil samples to obtain multi-spectral data; S2: performing a data preprocessing operation on the multi-spectral data to obtain preprocessed multi-spectral data; The preprocessing operation includes standard normal variable transformation preprocessing and moving average detrending algorithm preprocessing; The data preprocessing operation of the multi-spectral data using the moving average detrending algorithm is specifically as follows: S2.a: Performing Fourier transform on the multi-spectral data to obtain frequency domain multi-spectral data; S2.b: identifying extreme points of the multi-spectral data according to the frequency domain multi-spectral data, and marking them; S2.c: The multi-spectral segment data within the preset range on both sides of the extreme value point are used as the extreme value interval multi-spectral segment data; the other multi-spectral segment data are used as the common interval multi-spectral segment data; S2.d: For multi-spectral data in common intervals, the first moving average method is used for data preprocessing; S2.e: For the multi-spectral data in the extreme value interval, the second moving average method is used for data preprocessing; S3: performing a data fusion operation on the preprocessed multi-spectral data; S4: Establish a heavy oil quality detection model; S5: Performing quality testing on the heavy oil sample.

2. The heavy oil quality detection method based on multi-spectral fusion according to claim 1 is characterized in that: In S2.b, the extreme points of the frequency domain multi-spectral data are identified by a threshold method; specifically: the highest value of the preset multiples of the frequency domain multi-spectral data is set as an upper threshold, and the lowest value of the preset multiples of the frequency domain multi-spectral data is set as a lower threshold, and the frequency domain multi-spectral data is traversed, and the local maximum points above the upper threshold are taken as extreme points, and the local minimum points below the lower threshold are taken as extreme points.

3. The heavy oil quality detection method based on multi-spectral fusion according to claim 2 is characterized in that: The preset multiple is 0.

5.

4. The heavy oil quality detection method based on multi-spectral fusion according to claim 2 is characterized in that: In S2.c, the formula for determining the preset range is: ； In the formula, D is the number of data points in the preset range, a is the coefficient, is the sampling frequency of the i-th multi-spectral data, D<20.

5. The heavy oil quality detection method based on multi-spectral fusion according to claim 4 is characterized in that: In S2.d, the formula of the first moving average method is: ； In the formula, is the i-th multi-spectral data after preprocessing by the first moving average method, is the original i-th multi-spectral data, and n is the number of cycles of data movement.

6. The heavy oil quality detection method based on multi-spectral fusion according to claim 4 is characterized in that: In S2.e, the second moving average method is used to preprocess the data of the multi-spectral segment data in the extreme value interval as follows: the initial window period number n1 is set, and then the multi-spectral segment data is preprocessed by using the window period number that decays in sequence; the extreme point is skipped when moving to the marked extreme point, and then the initial window period number n1 is set at the value after the extreme point, and then the multi-spectral segment data is preprocessed by using the window period number that decays in sequence; wherein the formula for determining the period number n that decays in sequence is: ； ； Where s is the window cycle number attenuation coefficient s.

7. The heavy oil quality detection method based on multi-spectral fusion according to claim 1 is characterized in that: In S3, a wavelet feature fusion method is used to perform a data fusion operation on the preprocessed multi-spectral data.

8. The heavy oil quality detection method based on multi-spectral fusion according to claim 7 is characterized in that: Specifically, S3 is as follows: performing multi-layer wavelet transform on the multi-spectral data to obtain a plurality of high-frequency wavelet data and a plurality of low-frequency wavelet data, fusing the high-frequency wavelet data by setting a large absolute value, and fusing the low-frequency wavelet data by removing the average value.

9. The heavy oil quality detection method based on multi-spectral fusion according to claim 1 is characterized in that: In S4, the heavy oil quality detection model is a support vector machine model.

10. A heavy oil quality detection system based on multi-spectral fusion, characterized in that: The detection system adopts the heavy oil quality detection method based on multi-spectral fusion according to any one of claims 1 to 9, and the system comprises: A multi-spectral data acquisition module is used to perform multi-spectral acquisition on heavy oil samples to obtain multi-spectral data; A data preprocessing module, used for performing a data preprocessing operation on the multi-spectral data to obtain preprocessed multi-spectral data; A data fusion module, used for performing a data fusion operation on the preprocessed multi-spectral data; A heavy oil quality detection model building module is used to build a heavy oil quality detection model; The quality detection module is used to perform quality detection on the heavy oil.

Images