Method and system for detecting water content of petroleum oil product based on near infrared spectrum

Abstract

The present application relates to the technical field of oil water content detection, and particularly relates to a method and system for detecting water content of oil products based on near-infrared spectroscopy, wherein, when the near-infrared spectroscopy data is preprocessed, first, the sample spectroscopy data is divided into multiple point intervals based on the number of spectroscopy sampling points and the absorbance value of the maximum characteristic peak, then different total widths of sliding windows are set for each point interval, and when the total width of the sliding window is determined, the total width of the sliding window is determined according to the data quality of the sample spectroscopy data, so that in the total width determination process of the sliding window, not only the number of sampling points and the absorbance value of the characteristic peak are considered, but also the quality of the sampling data is considered, so that the total width of the sliding window not only has dynamic characteristics, but also has pertinence, and can adapt to spectroscopy data of different sampling types, so that the detection of the water content of oil products is applicable to spectroscopy data of different sampling types, and the universality of the method for detecting the water content of oil products is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of oil water cut detection, and particularly relates to a method and system for detecting oil water cut based on near-infrared spectroscopy. BACKGROUND

[0002] At present, onshore oil fields have entered the middle and late stages of development and production, and the natural energy contained in the reservoirs is decreasing, and some have even reached the stage of exhaustion. Continuous large-scale oil exploitation will lead to a decrease in well pressure, making it difficult to extract crude oil using formation pressure. In the face of this dilemma, most oil fields in the later stage will choose to use low-cost water injection for exploitation. By high-pressure water injection to the oil field to improve the formation pressure and formation energy, the pressure loss caused by oil production is maintained, therefore, the content of oil-water separation has become one of the important indicators of the oil industry, and water is also the main component of the present stage of oil well oil, and the detection of water content plays an extremely important role in the actual crude oil production and storage and transportation process of the oil field.

[0003] In the prior art, when detecting the water content of oil products by near-infrared spectroscopy, after obtaining the spectral data of the oil product, the spectral data needs to be preprocessed, for example, using a moving average smoothing method to preprocess the spectral data. However, in the prior art, a fixed sliding window value is generally determined by an empirical formula or experimental results. The above method has poor dynamic adaptability to different types of sampling data and cannot meet the smoothing processing needs of different types of spectral data. At the same time, the dynamic variable window moving average smoothing method in the prior art has poor adaptability when applied to oil product water content detection. SUMMARY

[0004] In order to solve the above technical problems, the present application provides a method and system for detecting the water content of oil products based on near-infrared spectroscopy, which solves the problems in the prior art.

[0005] According to one aspect of the present application, a method for detecting the water content of oil products based on near-infrared spectroscopy is provided, comprising the following steps:

[0006] S1: obtaining an oil product sample for water content detection;

[0007] S2: collecting the spectral data of the oil product sample to obtain sample spectral data;

[0008] S3: preprocessing the sample spectral data to obtain preprocessed sample spectral data; wherein the sample spectral data is preprocessed using an improved moving average smoothing method, specifically:

[0009] Sa: dividing the sample spectral data into a plurality of point intervals;

[0010] Sb: determining k of the i th point interval according to the data quality of the sample spectrum data of the i th point interval i value;

[0011] Sc: determining the total width of the sliding window of the i th point interval according to the k value of the i th point interval i

[0012] S4: performing a feature extraction operation on the preprocessed sample spectrum data to obtain sample spectrum feature data;

[0013] S5: establishing a quantitative analysis model of water content of petroleum oil products;

[0014] S6: inputting the sample spectrum feature data into the quantitative analysis model of water content of petroleum oil products to obtain a water content detection result.

[0015] Preferably, in the Sa, m sample spectrum data are included in each point interval;

[0016] The formula for determining m is:

[0017]

[0018] In the formula, round() is an integer function, a and b are coefficients, T is the number of sampling points of the sample spectrum data, and A is the absorbance value of the largest characteristic peak of the sample spectrum data.

[0019] Preferably, in the Sb, the formula for calculating the k value of the i th point interval is: i

[0020]

[0021] In the formula, round() is an integer function, is the variance of the sample spectrum data of the i th point interval, is the maximum value in the variances of the sample spectrum data of all point intervals.

[0022] Preferably, in the Sc, the formula for calculating the total width K of the sliding window of the i th point interval is: i

[0023] .

[0024] Preferably, the S1 specifically comprises: stirring the petroleum oil to be tested and then placing the petroleum oil to be tested into a test tube, then placing the test tube into an ultrasonic cleaner, starting an ultrasonic vibration mode, and vibrating for 30 min, and then placing the test tube for 10 min after the vibration is completed, thereby obtaining a petroleum oil product sample.

[0025] ​​​Preferably, in the S2, the petroleum product sample is scanned by a near-infrared spectrometer.

[0026] Preferably, the near-infrared spectrometer is of the NIR512 type, with 512 pixels, an AMA905 fiber connector, a 1501 / mm grating, 900nm-1700nm, an optical resolution of 3.1nm w / 25mslit, a signal-to-noise ratio of >50000:1ms integration, and a dynamic range of 15x10 6 ; the acquisition waveband is set to 1000nm-2500nm, the resolution is ≤8nm, the exposure time is 1.27ms, the integration time is 20ms, and the average number of scans is set to 32.

[0027] Preferably, in the S5, the petroleum product water content quantitative analysis model is a BP neural network model.

[0028] Preferably, in the S4, the principal component analysis method is used to extract features from the preprocessed sample spectrum data to obtain sample spectrum feature data.

[0029] According to another aspect of the present application, a near-infrared spectrum-based petroleum product water content detection system is provided, which uses the above-mentioned near-infrared spectrum-based petroleum product water content detection method, and the system comprises:

[0030] A sample acquisition module is configured to acquire a petroleum product sample for water content detection.

[0031] A data acquisition module is configured to collect spectrum data of the petroleum product sample.

[0032] A data preprocessing module is configured to preprocess the sample spectrum data to obtain preprocessed sample spectrum data.

[0033] A feature data extraction module is configured to extract features from the preprocessed sample spectrum data to obtain sample spectrum feature data.

[0034] A model establishment module is configured to establish a petroleum product water content quantitative analysis model.

[0035] A water content detection module is configured to input the sample spectrum feature data into the petroleum product water content quantitative analysis model to obtain a water content detection result.

[0036] The present application has the following technical effects:

[0037] The present application divides the sample spectrum data into multiple point intervals based on the number of spectral sampling points and the absorbance value of the maximum characteristic peak when pre-processing the near-infrared spectrum data, then sets different total widths of the sliding window for each point interval, and when determining the total width of the sliding window, the total width of the sliding window is determined according to the data quality of the sample spectrum data, so that in the process of determining the total width of the sliding window, not only the number of sampling points and the absorbance value of the characteristic peak are considered, but also the quality of the sampling data is considered, so that the total width of the sliding window not only has dynamic characteristics, but also has pertinence and can adapt to spectrum data of different sampling types, so that the petroleum product water content detection is suitable for spectrum data of different sampling types, and the universality of the petroleum product water content detection method is improved. BRIEF DESCRIPTION OF DRAWINGS

[0038] In order to more clearly illustrate the specific embodiments of the present application or the technical solutions in the prior art, the drawings needed in the specific embodiments or the prior art description will be briefly introduced below. Obviously, the drawings in the following description are some embodiments of the present application, and those skilled in the art can also obtain other drawings according to these drawings without creative labor.

[0039] Figure 1 is a flowchart of a petroleum product water content detection method based on near-infrared spectrum provided by an embodiment of the present application;

[0040] Figure 2 is a flowchart of the sample spectrum data preprocessed by the improved moving average smoothing method provided by an embodiment of the present application;

[0041] Figure 3 is an effect diagram of partial spectrum data of different water contents when the sample spectrum data is preprocessed by the improved moving average smoothing method provided by an embodiment of the present application;

[0042] Figure 4 is a schematic diagram of a petroleum product water content detection system based on near-infrared spectrum provided by an embodiment of the present application. DETAILED DESCRIPTION

[0043] In order to make the purpose, technical solutions and advantages of the present application more clear, the technical solutions of the present application will be clearly and completely described below. Obviously, the described embodiments are only some of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0044] Embodiment 1, attached Figure 1 shows a flowchart of a petroleum product water content detection method based on near-infrared spectrum, as shown in the attachedFigure 1 As shown, a near-infrared spectroscopy-based water content detection method for petroleum oil products includes the following steps:

[0045] S1: Obtain a petroleum oil product sample for water content detection;

[0046] Wherein, the petroleum to be tested is uniformly stirred and then placed in a test tube, and then the test tube is placed in an ultrasonic cleaner, and the ultrasonic vibration mode is turned on to promote the homogeneous fusion of the petroleum oil product and remove the bubbles in the petroleum, thereby obtaining a petroleum oil product sample;

[0047] Wherein, the vibration process lasts for 30 min, and after the vibration is completed, it is placed for 10 min.

[0048] S2: Collect the spectrum of the petroleum oil product sample to obtain sample spectrum data;

[0049] In this step, the near-infrared spectrometer scans the petroleum oil product sample.

[0050] Specifically, the near-infrared spectrometer is a NIR512 type infrared spectrometer, the pixel is 512, the optical fiber connector is AMA905, the grating is 1501 / mm, 900nm-1700nm, the optical resolution is 3.1nm w / 25mslit, the signal-to-noise ratio is >50000:1ms integration, and the dynamic range is 15×10 6 ;

[0051] Set the collection wavelength range to 1000nm~2500nm, the resolution to ≤8nm, the exposure time to 1.27ms, the integration time to 20ms, and the average number of scans to 32 to improve the signal-to-noise ratio of the spectrum data;

[0052] At the same time, before the sample spectrum data collection starts, the near-infrared spectrometer is preheated, and at least 20 min later, the measurement starts to ensure that stable readings are obtained. Before the spectrum data of the petroleum oil product sample is collected, the spectrum data of the reference plate (polytetrafluoroethylene) and the dark current are first obtained to reduce the influence of dark current and uneven light on the spectrum of the petroleum oil product sample.

[0053] Further, different sampling parameters and sample processing methods are set for different petroleum oil products; for example, different preheating temperatures and different stirring methods are used for different petroleum oil products, and Table 1 shows the detection parameter configuration of typical petroleum oil product samples.

[0054] Table 1 Detection parameter configuration of typical petroleum oil product samples

[0055]

[0056] S3: performing a pretreatment operation on the sample spectrum data to obtain pretreated sample spectrum data;

[0057] Generally, no matter how advanced the near-infrared spectrometer used is, the measured sample spectrum data not only contains useful information reflecting the nature of the measured sample itself, but also contains a lot of noise signals affecting data analysis due to external environmental or internal structural factors. For example, due to the openness of the light path, the light path is easily disturbed by external conditions. Even in a darkroom environment, it is still affected by diffuse reflection and scattered light of the light source, and fluctuations in temperature in the experimental environment and mechanical vibration of the spectrometer will affect the spectrum data. These will produce noise signals that may sometimes mask the useful information of the spectrum, increasing the difficulty of spectral analysis. Therefore, it is necessary to process the original spectrum through corresponding pretreatment means to eliminate the interference from noise signals and improve the accuracy and reliability of the moisture content detection.

[0058] Specifically, the pretreatment operation on the sample spectrum data is specifically: using an improved moving average smoothing method to perform the pretreatment operation on the sample spectrum data.

[0059] The moving average smoothing method is a time domain signal processing technique that suppresses random noise by calculating the average value of adjacent points in the data window and retains trend information. Its core formula is

[0060] ;

[0061] In the formula, x i is the absorbance of the i-th wavelength point of the sample spectrum data, y i is the output value after averaging and smoothing, k is the half-width of the sliding window, and the total width of the sliding window is 2k+1. The basic principle is to select a window with a total width of 2k+1 wavelength points as a smoothing window, replace the measured value at the center wavelength point with the average value of all wavelength points in the window, and continuously move the window to complete the smoothing processing of all wavelength points. The moving average smoothing method has the characteristics of suppressing high-frequency noise and being suitable for real-time processing, and therefore has a relatively wide application in the field of petroleum product moisture content detection.

[0062] As known from the above description, the selection of the sliding window width value is very important for smoothing. In the prior art, the fixed sliding window value is generally determined according to an empirical formula or experimental results. However, the above-mentioned method has poor dynamic adaptability for different types of sampling data and cannot meet the smoothing processing needs of different types of spectrum data. Therefore, the moving average smoothing method is improved in this embodiment, and a more adaptive scheme is proposed.

[0063] Specifically, as shown in FIG. 2, the improved moving average smoothing method includes the following steps:Figure 2 As shown, the improved moving average smoothing method is used to preprocess the sample spectrum data as follows:

[0064] Sa: Divide the sample spectrum data into several point intervals;

[0065] Wherein, in the Sa, each point interval includes m sample spectrum data;

[0066] The formula for determining m is:

[0067] ;

[0068] Wherein, round() is a rounding function, a and b are coefficients, T is the number of sampling points of the sample spectral data, and A is the absorbance value of the maximum characteristic peak of the sample spectral data;

[0069] Sb: Determine the k value of the i-th point interval according to the data quality of the sample spectrum data in the i-th point interval. i value;

[0070] Among them, k of the i-th point interval i The value is calculated as:

[0071] ;

[0072] In the formula, round() is the rounding function, is the variance of the sample spectrum data in the i-th point interval, is the maximum value of the variance of the sample spectrum data in all point intervals;

[0073] Sc: According to the k of the i-th point interval i The value determines the total width of the sliding window for the i-th point interval;

[0074] Among them, the total width of the sliding window of the i-th point interval is K i The calculation formula is:

[0075] .

[0076] In the existing research, there are many related researches on dynamic window denoising. On the basis of the existing research, the sample spectrum data is first divided into multiple point intervals based on the number of spectral sampling points and the absorbance value of the maximum characteristic peak in the embodiment. Then, different total widths of the sliding window are set for each point interval. When determining the total width of the sliding window, the total width of the sliding window is determined according to the data quality of the sample spectrum data, so that the total width of the sliding window is determined in the process of the total width of the sliding window. Not only the number of sampling points and the absorbance value of the characteristic peak are considered, but also the quality of the sampling data is considered, so that the total width of the sliding window not only has dynamic characteristics, but also has pertinence and can adapt to different types of spectral data.

[0077] The Figure 3 The effect diagram of the partial spectral data of different water contents of the sample spectrum data pretreated by the improved moving average smoothing method is shown, wherein the lines of different colors represent the spectral data of different water contents. By comparing the diagram with the original sample spectrum data, it is found that the peak value of the petroleum product sample spectrum data changes less than the peak value of the original spectrum data, which can meet the demand of subsequent water content detection.

[0078] As a preferred embodiment, the S3 further comprises: multivariate scatter correction and standard normal variable transformation; the above two methods are prior art, which are not discussed in detail in the embodiment.

[0079] S4: performing feature extraction on the pretreated sample spectrum data to obtain sample spectrum feature data;

[0080] In this step, the principal component analysis method is used to extract features from the pretreated sample spectrum data to obtain sample spectrum feature data.

[0081] S5: establishing a petroleum product water content quantitative analysis model;

[0082] In this step, the petroleum product water content quantitative analysis model is a BP neural network model. The BP neural network (BPNN) is a widely used fully connected feedforward neural network model, which is suitable for solving complex nonlinear problems. The basic architecture of the BP neural network model is divided into three layers: input layer, hidden layer and output layer. After the training process, the deep relationship between the input data and the output data can be gradually learned. The Sigmoid function is used as the neuron activation function during model training to enhance its ability to handle nonlinear problems. In order to optimize the performance of the model, the output result is calculated by forward propagation during training, and then the weights and biases in the network are adjusted by using the back propagation algorithm to reduce the prediction error value.

[0083] The training process of the BP neural network model is as follows:

[0084] Set the number of neurons of input layer, hidden layer and output layer, set learning rate, momentum and other hyperparameters, initialize weights and biases of the network;

[0085] Prepare standard oil samples with different water content (0.1%-10%), collect spectra and label true values, and perform feature extraction to obtain a sample set of the BP neural network model;

[0086] Train the BP neural network model according to the sample set to obtain a quantitative analysis model of water content of petroleum products.

[0087] S6: input the sample spectral feature data into the quantitative analysis model of water content of petroleum products to obtain a water content detection result.

[0088] In embodiment 2, the present application further provides a near-infrared spectrum-based water content detection system for petroleum products, which adopts the near-infrared spectrum-based water content detection method for petroleum products in embodiment 1, and the system comprises:

[0089] A sample acquisition module is configured to acquire a petroleum product sample for water content detection.

[0090] A data acquisition module is configured to collect spectra of the petroleum product sample to obtain sample spectral data.

[0091] A data preprocessing module is configured to perform a preprocessing operation on the sample spectral data to obtain preprocessed sample spectral data.

[0092] A feature data extraction module is configured to perform a feature extraction operation on the preprocessed sample spectral data to obtain sample spectral feature data.

[0093] A model establishment module is configured to establish a quantitative analysis model of water content of petroleum products.

[0094] A water content detection module is configured to input the sample spectral feature data into the quantitative analysis model of water content of petroleum products to obtain a water content detection result.

[0095] In embodiment 3, the present application further provides an electronic device comprising one or more processors and memories.

[0096] The processor can be a central processing unit (CPU) or other forms of processing units with data processing capability and / or instruction execution capability, and can control other components in the electronic device to perform desired functions.

[0097] The memory can include one or more computer program products that can include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory, for example, can include random access memory (RAM), and / or a cache, and / or the like. The non-volatile memory, for example, can include read-only memory (ROM), hard disk, flash memory, and / or the like. The computer-readable storage media can store one or more computer program instructions executable by the processor to implement the method for detecting water content of petroleum oil product based on near-infrared spectrum of any embodiments of the present application and / or other desired functions. Various contents such as initial external parameters, threshold values, and the like can also be stored in the computer-readable storage media.

[0098] In one example, the electronic device can further include an input device and an output device, which are interconnected through a bus system and / or other forms of connection mechanism (not shown). The input device can include, for example, a keyboard, a mouse, and the like. The output device can output various information including pre-warning prompt information, braking force, and the like to the outside. The output device can include, for example, a display, a speaker, a printer, a communication network and a remote output device connected thereto, and the like.

[0099] Of course, components such as buses, input / output interfaces, and the like are omitted for simplicity. In addition, the electronic device can include any other appropriate components according to specific application cases.

[0100] In addition to the above method and device, the embodiments of the present application can also be a computer program product including computer program instructions that, when executed by a processor, cause the processor to implement the functions of the method for detecting water content of petroleum oil product based on near-infrared spectrum provided by any embodiments of the present application.

[0101] The computer program product can be written in any combination of one or more programming languages, including an object-oriented programming language such as Java, C++, and the like, and conventional procedural programming languages, such as the "C" programming language, or the like. The program code can execute entirely on the user's computing device, partly on the user's device, as a stand-alone software package, partly on the user's computing device and partly on a remote computing device, or entirely on the remote computing device or server.

[0102] In addition, the embodiments of the present application can also be a computer readable storage medium, which stores computer program instructions, and the computer program instructions make the processor realize the method for detecting water content of petroleum oil product based on near infrared spectrum provided by any embodiments of the present application when the processor runs.

[0103] 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 may, for example, include but is not limited to an electrical, a magnetic, an optical, an electromagnetic, an infrared, or a semiconductor system, device or apparatus, or any appropriate combination thereof. More specific examples (non-exhaustive list) of the readable storage medium include an electrical connection having one or more wires, a portable disc, 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 disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any appropriate combination thereof.

[0104] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement to part or all of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the technical solutions of the embodiments of the present application.

Claims

1. A method for detecting water content in petroleum oil products based on near infrared spectroscopy, characterized by, The method comprises the following steps: S1: obtaining a petroleum product sample for water content detection; S2: performing spectrum acquisition on the petroleum product sample to obtain sample spectrum data; In the step S2, the petroleum product sample is scanned by a near-infrared spectrometer; S3: performing pretreatment operation on the sample spectrum data to obtain pretreated sample spectrum data; wherein the sample spectrum data is pretreated by using an improved moving average smoothing method, and the pretreatment operation is specifically as follows: Sa: dividing the sample spectrum data into a plurality of point intervals; in the step Sa, each point interval includes m sample spectrum data; The formula for determining m is as follows: ； In the formula, round() is an integer function, a and b are coefficients, T is the number of sampling points of the sample spectrum data, and A is the absorbance value of the maximum characteristic peak of the sample spectrum data; Sb: Determine the k value of the i-th point interval according to the data quality of the sample spectrum data in the i-th point interval. i Value; In the Sb, the k value of the i-th point interval i The value is calculated as: ； where round() is a rounding function, is a variance of the sample spectral data for the i-th point interval, is a maximum value among the variances of the sample spectral data for all point intervals; Sc: k of the ith point interval i determines the total width of the sliding window of the ith point interval; S4: performing feature extraction operation on the pretreated sample spectrum data to obtain sample spectrum feature data; S5: establishing a petroleum product water content quantitative analysis model; in the step S5, the petroleum product water content quantitative analysis model is a BP neural network model; S6: inputting the sample spectrum feature data into the petroleum product water content quantitative analysis model to obtain a water content detection result.

2. The near infrared spectroscopy-based water content detection method for petroleum oil products according to claim 1, characterized by, In the Sc, the total width K of the sliding window of the i-th point interval i The calculation formula is: 。 3. The near infrared spectroscopy-based water content detection method for petroleum oil products according to claim 1, characterized by, In the step S1, the to-be-detected petroleum is stirred and then placed in a test tube, the test tube is placed in an ultrasonic cleaner, an ultrasonic vibration mode is started, the vibration process lasts for 30 minutes, and after the vibration is completed, the test tube is placed for 10 minutes, thereby obtaining the petroleum product sample.

4. The near infrared spectroscopy-based water content detection method for petroleum oil products according to claim 1, characterized by, The near infrared spectrometer is NIR512 type near infrared spectrometer, the pixel is 512, the optical fiber connector is AMA905, the grating is 1501 / mm, 900nm-1700nm, the optical resolution is 3.1nm w / 25mslit, the signal-to-noise ratio is >50000:1msintegration, the dynamic range is 15x10 6 ; the acquisition wave band is set to 1000nm-2500nm, the resolution is ≤8nm, the exposure time is 1.27ms, the integration time is 20ms, and the average number of scanning is set to 32 times.

5. The near infrared spectroscopy-based water content detection method for petroleum oil products according to claim 1, characterized by, In the step S4, the pretreated sample spectrum data is extracted by using a principal component analysis method to obtain sample spectrum feature data.

6. A near infrared spectroscopy-based water content detection system for petroleum products, characterized by comprising: The system adopts the petroleum product water content detection method based on near-infrared spectrum according to any one of claims 1-5, and the system comprises: a sample acquisition module configured to acquire a petroleum product sample for water content detection; a data acquisition module configured to perform spectrum acquisition on the petroleum product sample to obtain sample spectrum data; a data preprocessing module configured to perform pretreatment operation on the sample spectrum data to obtain pretreated sample spectrum data; a feature data extraction module configured to perform feature extraction operation on the pretreated sample spectrum data to obtain sample spectrum feature data; a model establishment module configured to establish a petroleum product water content quantitative analysis model; a water content detection module configured to input the sample spectrum feature data into the petroleum product water content quantitative analysis model to obtain a water content detection result.

Images