A full-aperture characterization method based on dynamic time warping

Abstract

The present invention provides a full-pore characterization method based on dynamic time warping, belonging to the field of pore characterization technology. The method includes collecting nuclear magnetic resonance (NMR) T2 spectral data and pore size distribution data from different experiments, and preprocessing and standardizing the collected data; using a dynamic time warping algorithm, nonlinearly matching the processed NMR T2 spectral data with the pore size distribution experimental data; based on the matching results, integrating gas adsorption experiments, high-pressure mercury injection experiments, and NMR T2 spectral results to model the pore size distribution covering micropores, mesopores, and macropores, completing the full-pore characterization. The present invention solves the problems of existing shale oil reservoir pore structure characterization methods, which have low resolution within the micropore, mesopore, and macropore ranges and difficulty integrating experimental data.

Description

Technical Field

[0001] The present invention belongs to the technical field of aperture characterization, and in particular relates to a full-aperture characterization method based on dynamic time warping. Background Art

[0002] The pore size distribution characteristics of shale oil reservoirs have a significant impact on the reservoir's storage capacity and fluid migration characteristics. Traditional pore size characterization methods include nuclear magnetic resonance pore size spectroscopy, gas adsorption, and high-pressure mercury injection. These methods are based on different physical principles and have certain applicability and limitations:

[0003] Nuclear magnetic resonance pore size spectroscopy: The pore size distribution is derived using the NMRT2 spectrum, but the diffusion relaxation effect and the influence of the complex pore structure on the relaxation time are ignored, resulting in deviations in the characterization results.

[0004] Gas adsorption method: Nitrogen adsorption or carbon dioxide adsorption is used to characterize micropores and mesopores, but the ability to characterize macropore distribution is limited.

[0005] High-pressure mercury injection: Suitable for characterizing macropore distribution, but the destructive effect of high pressure on shale structure may lead to errors in the measurement results.

[0006] Although nuclear magnetic resonance, gas adsorption and high-pressure mercury injection methods are theoretically feasible for pore size characterization, when applied to shale oil reservoirs, these methods face significant challenges in parameter setting and data fusion due to the complex pore structure and the presence of multiphase fluids. Specifically, they are as follows: (1) Limitations of a single method: A single pore size characterization method is difficult to cover all scales of pores in shale reservoirs, and there is a problem of incomplete coverage of the pore size range. (2) Difficulty in data fusion: Data from different methods have overlapping and complementary pore size coverage, and the heterogeneity of data sources increases the difficulty of data fusion. Existing data fusion methods often rely on idealized assumptions and simplified models, making it difficult to fully utilize limited experimental data for accurate characterization of the full pore size range.

[0007] Traditional methods such as nuclear magnetic resonance (NMR) T2 spectroscopy, gas adsorption, and high-pressure mercury injection are theoretically feasible for pore size characterization. However, in practical application to shale oil reservoirs, these methods present significant challenges in parameter setting and data fusion due to the complex pore structure and presence of multiphase fluids. Existing technologies struggle to comprehensively and accurately characterize the full pore size range of shale oil reservoirs, from micropores to macropores. Summary of the Invention

[0008] In response to the above-mentioned deficiencies in the prior art, the present invention provides a full-pore characterization method based on dynamic time warping. The present invention solves the problems of low resolution of existing shale oil reservoir pore structure characterization methods in the micropore, mesopore and macropore ranges and difficulty in integrating experimental data.

[0009] In order to achieve the above objectives, the technical solution adopted by the present invention is: a full-aperture characterization method based on dynamic time warping, comprising the following steps:

[0010] S1. Data preparation: Collect NMR T2 spectrum data and pore size distribution data from different experiments, and preprocess and standardize the collected data;

[0011] S2, dynamic time warping matching: using the dynamic time warping algorithm, the NMR T2 spectrum data processed by S1 is nonlinearly matched with the pore size distribution experimental data;

[0012] S3. Characterization of full pore size distribution: Based on the matching results, the gas adsorption experiment, high-pressure mercury injection experiment and nuclear magnetic resonance T2 spectrum results are integrated to model the pore size distribution covering micropores, mesopores and macropores to complete the characterization of the full pore size.

[0013] The beneficial effects of the present invention are as follows: based on the dynamic time warping (DTW) algorithm, the present invention obtains the optimal correspondence between the nuclear magnetic resonance T2 spectrum and the experimental pore size distribution data by standardizing and matching the two, and fits the mathematical conversion model to convert the nuclear magnetic resonance T2 spectrum into a full pore size distribution curve, thereby realizing accurate characterization and correction of the pore structure of shale oil reservoirs, realizing comprehensive characterization of the reservoir pore structure, and providing a high-precision data fusion solution.

[0014] Furthermore, the S1 includes the following steps:

[0015] S101, collecting nuclear magnetic resonance T2 spectrum data, gas adsorption experimental data and high-pressure mercury injection experimental data;

[0016] S102, performing Gaussian smoothing and normalization processing on the gas adsorption experimental data and the high-pressure mercury injection experimental data to obtain normalized pore size distribution experimental data;

[0017] S103, performing normalization processing on the collected nuclear magnetic resonance T2 spectrum data;

[0018] S104: The data processed by S102 and S103 are fused according to the same scale interval to complete the preprocessing and standardization.

[0019] The beneficial effect of this further approach is that, to eliminate scale differences in pore data from different experimental methods (such as carbon dioxide adsorption, nitrogen adsorption, and high-pressure mercury injection), the present invention employs standardization and normalization techniques to unify all pore distribution data to the same scale range (0–1). This processing step ensures data consistency while improving comparability, a prerequisite for the effective application of the dynamic time warping (DTW) algorithm. Furthermore, by integrating multi-source data, the present invention comprehensively considers data characteristics across different pore size ranges, addressing the inability of traditional methods to uniformly process pore data across a wide range. This technology can extract detailed pore characteristics of complex shale reservoirs, thereby supporting subsequent reservoir evaluation and fluid identification.

[0020] Furthermore, the S2 includes the following steps:

[0021] S201, calculating the Euclidean distance between the nuclear magnetic resonance T2 spectrum data sequence T processed by S1 and the pore size distribution experimental data P;

[0022] S202, based on the Euclidean distance, using the recursive formula of the dynamic time warping algorithm to calculate the cumulative distance matrix C;

[0023] S203 , based on the cumulative distance matrix C, backtrack the optimal path W, and construct a conversion relationship between the nuclear magnetic resonance T2 spectrum data sequence T and the pore size distribution experimental data P based on the optimal path W, completing the nonlinear matching process.

[0024] The beneficial effect of this further solution is that, by backtracking the optimal path of the dynamic time warping (DTW) algorithm, the present invention extracts the best matching point pairs within different pore size ranges and, based on these point pairs, establishes a conversion relationship between the T2 spectrum and pore size. This step not only optimizes the matching process but also establishes a precise mathematical model by minimizing fitting errors, further improving the accuracy of pore characterization.

[0025] Furthermore, the step S203 is specifically as follows:

[0026] Assume that the end point of the optimal path is , set the backtracking starting point to ,from Start by tracing back in the direction of minimum cumulative distance. , and get the optimal path W, where the next point of the backtracking path is obtained using the following formula:

[0027]

[0028] in, Indicates current location The previous point, Indicates current location Above, Indicates current location On the left side, Indicates the current The upper left corner;

[0029] Based on the optimal path W, a conversion relationship is established between the nuclear magnetic resonance T2 spectrum data sequence T and the pore size distribution experimental data P to complete the nonlinear matching process.

[0030] Furthermore, the optimal path W is expressed as follows:

[0031]

[0032]

[0033] in, represents the last coordinate in the optimal path W, represents the first The coordinates of the path points, Indicates the current optimal path point The next backtracking path point, Represents the index number of the point on the optimal path W, represents the total number of steps in the optimal path W. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 Flow chart of the method of the present invention.

[0035] Figure 2 Schematic diagram of normalized full pore size distribution characteristics for different experiments.

[0036] Figure 3 A schematic diagram of the matching results.

[0037] Figure 4 It is the integrated cross-plot of the sample pore size distribution and T2 spectrum matching results.

[0038] Figure 5 This is a fitting diagram of the functional relationship between the NMR T2 spectrum and the pore size distribution.

[0039] Figure 6 Schematic diagram of the calculated pore size spectrum for HQ1H.

[0040] Figure 7 Schematic diagram of the system structure of the present invention. DETAILED DESCRIPTION

[0041] The specific embodiments of the present invention are described below to facilitate understanding of the present invention by those skilled in the art. However, it should be clear that the present invention is not limited to the scope of the specific embodiments. For those skilled in the art, as long as various changes are within the spirit and scope of the present invention as defined and determined by the appended claims, these changes are obvious, and all inventions and creations utilizing the concepts of the present invention are protected.

[0042] Example 1

[0043] like Figure 1 As shown, the present invention provides a full-aperture characterization method based on dynamic time warping, and its implementation method is as follows:

[0044] S1. Data preparation: Collect NMR T2 spectrum data and pore size distribution data from different experiments, and preprocess and standardize the collected data as follows:

[0045] S101, collecting nuclear magnetic resonance T2 spectrum data, gas adsorption experimental data and high-pressure mercury injection experimental data;

[0046] S102, performing Gaussian smoothing and normalization processing on the gas adsorption experimental data and the high-pressure mercury injection experimental data to obtain normalized pore size distribution experimental data;

[0047] S103, performing normalization processing on the collected nuclear magnetic resonance T2 spectrum data;

[0048] S104: The data processed by S102 and S103 are fused according to the same scale interval to complete the preprocessing and standardization.

[0049] In this embodiment, the pore size distribution data obtained by carbon dioxide adsorption, nitrogen adsorption and high-pressure mercury injection need to be smoothed by Gaussian because of their large fluctuations. (corresponding to the first points), the smoothing process can be performed through the Gaussian function, the specific formula is as follows:

[0050]

[0051] in, represents the standard deviation of the Gaussian kernel, N represents the number of data points, represents the smoothed pore size distribution data, j Indicates the data point that currently needs to be smoothed.

[0052] In this example, in order to make the data of different experimental methods (such as carbon dioxide adsorption, nitrogen adsorption and high-pressure mercury injection) and nuclear magnetic resonance T2 spectrum comparable, all data need to be normalized. , whose minimum and maximum values ​​are and , the normalized data It can be calculated by the following formula:

[0053]

[0054] like Figure 2 As shown, all data are normalized to the interval [0,1] to eliminate the differences in data scales between different experimental methods. Figure 2 The pore size distributions obtained by different experimental methods are integrated into the full pore size distribution data. The X-axis represents the average pore size of the pores, in nanometers (nm), and the pore size range is from 1 nm to 1000 nm, covering the range of micropores (<2 nm), mesopores (2-50 nm) and macropores (>50 nm). The Y-axis represents the pore distribution characteristics of different samples in each pore size range, and the values ​​are normalized (standardized to the range of 0-1). The pore size is divided into three adsorption / measurement intervals by three red dotted lines, the leftmost is Adsorption pore size range (<10 nm), which mainly reflects the distribution of micropores (usually pores with a pore size less than 2 nm). The adsorption pore size range is (10-100 nm), which reflects the mesopore distribution. Adsorption measurements are used to analyze mesopores between 2 and 50 nm. The rightmost curve represents the high-pressure mercury porosimetry (HgP) pore size range (>100 nm). This range is primarily used to measure macropores, typically using high-pressure mercury porosimetry. The legend shows the sample numbers (e.g., 22CJ26-1, 22CJ26-2, etc.). Each curve represents the pore distribution characteristics of a specific sample within a different pore size range.

[0055] After completing the above data smoothing and normalization processing, the pore size distribution data obtained from different experimental methods (such as carbon dioxide adsorption, nitrogen adsorption, and high-pressure mercury intrusion) are integrated according to the same scale interval (0-1). The integrated data will be used for subsequent full pore size characterization and DTW algorithm-driven analysis.

[0056] S2, dynamic time warping matching: Using the dynamic time warping algorithm, nonlinear matching is performed on the NMR T2 spectrum data processed by S1 and the pore size distribution experimental data. The implementation method is as follows:

[0057] S201, calculating the Euclidean distance between the nuclear magnetic resonance T2 spectrum data sequence T processed by S1 and the pore size distribution experimental data P;

[0058] S202, based on the Euclidean distance, using the recursive formula of the dynamic time warping algorithm to calculate the cumulative distance matrix C;

[0059] S203, based on the cumulative distance matrix C, backtrack the optimal path W, and based on the optimal path W, construct a conversion relationship between the nuclear magnetic resonance T2 spectrum data sequence T and the pore size distribution experimental data P to complete the nonlinear matching, which is specifically:

[0060] Assume that the end point of the optimal path is , set the backtracking starting point to ,from Start by tracing back in the direction of minimum cumulative distance. , get the optimal path W;

[0061] Based on the optimal path W, a conversion relationship is established between the nuclear magnetic resonance T2 spectrum data sequence T and the pore size distribution experimental data P to complete the nonlinear matching process.

[0062] In this embodiment, the normalized data of the nuclear magnetic resonance T2 spectrum is set as the sequence , the pore size distribution experimental data is normalized to the sequence , Indicates the first m data, Indicates the first n The goal of the dynamic time warping algorithm is to minimize the matching error between sequence T and sequence P by finding the optimal path W. The specific steps are as follows:

[0063] 1) Distance matrix construction

[0064] Calculate the Euclidean distance between sequence T and sequence P and construct a distance matrix ,in, Indicates the i T point and j The Euclidean distance between points P:

[0065]

[0066] in, represents the value after normalization of the pore size distribution experimental data, Represents the normalized data of the NMR T2 spectrum.

[0067] 2) Cumulative distance matrix calculation

[0068] The cumulative distance matrix C is calculated according to the recursive formula of DTW, and its recursive formula is:

[0069]

[0070] in, Indicates that the cumulative distance matrix C is at position The value of the current point The cumulative distance when taking one step up (from the previous row), Indicates that the cumulative distance matrix C is at position The value of the current point The cumulative distance when taking one step to the left (from the previous column), Indicates that the cumulative distance matrix C is at position The value of the current point The cumulative distance transferred in the diagonal (upper left) direction.

[0071] The boundary conditions are:

[0072]

[0073] in, represents the starting point (boundary condition) of the cumulative distance matrix, Represents the original distance matrix 𝐷 at point The value at , that is, the Euclidean distance between the two sequences at the starting point.

[0074] for and :

[0075]

[0076] in, Represents the cumulative distance matrix in The value of the row and column 1, that is, the cumulative distance when recursively going down the first column, Represents the original distance matrix 𝐷 at position The value at , represents the first The local distance between the point and the first point in the sequence P, Represents the cumulative distance matrix in The value of the row and column 1, that is, the previous cumulative value when calculating step by step to the right along the first row, Indicates that the cumulative distance matrix is ​​in the first row and the The value of the column, that is, the cumulative distance when recursively going right along the first row, Represents the original distance matrix 𝐷 at position The value at , which indicates the difference between the first point in sequence T and the first point in sequence P The local distance between points, Indicates that the cumulative distance matrix is ​​in the first row and the The value of the column, that is, the previous cumulative value as you progress right along the first row.

[0077] 3) Optimal path backtracking

[0078] The purpose of optimal path backtracking is to find arrive The path that minimizes the cumulative distance on the entire path.

[0079] The end point of the optimal path is , the starting point is The backtracking process starts from Start by tracing back in the direction of minimum cumulative distance. .

[0080] The backtracking process follows the following rules: For each position , select the source point with the minimum cumulative distance to the current position. Specifically, the next point on the backtracking path is the one with the minimum distance from the three possible directions:

[0081]

[0082] in Indicates current location The previous point may be from above , left , or the upper left point.

[0083] from Start by going back to :

[0084]

[0085]

[0086] Until arrival So far, backtracking path This is the optimal matching path.

[0087] in, represents the last coordinate in the optimal path W, represents the first The coordinates of the path points, Indicates the current optimal path point The next backtracking path point, Represents the index number of the point on the optimal path W, Represents the total number of steps in the optimal path W, or the total number of points on the path.

[0088] like Figure 3As shown, the optimal path is found by performing dynamic time warping algorithm on the nuclear magnetic resonance T2 spectrum sequence and the pore size distribution sequence, thus obtaining the following Figure 4 The matching relationship between the values ​​​​between the two sequences shown. Figure 3 This figure shows the results of DTW matching of the NMR T2 spectrum and the full pore size distribution data. The X-axis, with Index as the coordinate, represents the index or classification index value of the pore size distribution and T2 distribution data. The Y-axis shows the normalized values ​​of the pore size distribution and T2 distribution. The data has been normalized (range 0-1). The blue pore size distribution curve represents the change in the normalized value in the pore size distribution of sample "22CJ26-22", and the yellow T2 distribution curve represents the change in the normalized value in the T2 distribution data of sample "22CJ26-22". The flexible matching line is a flexible connection line that is matched from the yellow T2 distribution curve to the blue pore size distribution curve along the optimal path in the figure. This figure intuitively shows the correlation between the pore size distribution and T2 distribution of sample "22CJ26-22". Through flexible matching technology, we can deeply analyze the correlation between NMR pore structure characteristics and actual pore size distribution data.

[0089] Figure 4 The figure shows the relationship between the NMR T2 spectra and pore size distribution for multiple samples. The X-axis shows the NMR T2 spectrum time in milliseconds (ms), and the Y-axis shows the pore size in nanometers (nm), ranging from 0 nm to 2000 nm, including micropores, mesopores, and macropores. The legend shows multiple samples, including "22CJ26-1," "22CJ26-2," and "22CJ26-3." Each curve represents the relationship between T2 and pore size for a specific sample. The figure intuitively illustrates the relationship between the NMR T2 regression time and pore size of the sample.

[0090] The present invention generates a complete full pore size distribution curve by analyzing data derivatives and extreme points and fitting the conversion relationship of different experimental data using a mathematical model.

[0091] Based on the matching points in the optimal path W, a conversion relationship is established between the magnetic resonance T2 spectrum data sequence T and the pore size distribution experimental data P. Assume that the conversion relationship is a linear relationship:

[0092]

[0093] in, and All represent the parameters to be fitted. Solve by minimizing the following objective function and :

[0094]

[0095] in, represents the total number of matching point pairs, Indicates the first position of the pore size distribution experimental data in the optimal matching path The value of the matching point, Indicates the index number of the matching point pair, the value range is ,in is the total number of matching points, Indicates the first position of the NMR T2 spectrum data in the optimal matching path The value of the matching point. Solving this optimization problem can obtain the optimal conversion parameters and .

[0096] like Figure 5 As shown in the figure, the functional relationship between the NMR T2 spectrum and the pore size is obtained by solving the conversion parameters. The figure shows the functional relationship between the NMR T2 spectrum and the pore size distribution obtained by taking the average of the matching results of multiple samples. The X-axis shows the NMR T2 spectrum time in milliseconds (ms), and the Y-axis shows the pore size in nanometers (nm). The range is from 0 nm to 2500 nm, including micropores, mesopores and macropores. The yellow data points in the figure represent the corresponding relationship between the actual pore size and T2 regression time. The red dotted line is the data fitting curve, which shows the reasonable relationship between the NMR T2 time and pore size. The fitting equation given in the figure is , indicating that there is a functional relationship between pore size and T2 time. This model can be used to convert T2 measurement results into pore size distribution characteristics.

[0097] S3. Characterization of full pore size distribution: Based on the matching results, the gas adsorption experiment, high-pressure mercury injection experiment and nuclear magnetic resonance T2 spectrum results are integrated to model the pore size distribution covering micropores, mesopores and macropores to complete the characterization of the full pore size.

[0098] like Figure 6 As shown in the figure, the characterization of the T2 spectrum at a certain depth point in the HQ1H well to the full pore size distribution is completed. This figure shows the conversion of the nuclear magnetic T2 spectrum at multiple depths into a pore size distribution diagram. The X-axis shows the pore size in microns (μm). The Y-axis represents the pore component. The pore component value is used to represent the density and distribution characteristics of the pore structure within each pore size range. Figure 6 The data distribution curves of samples at different depths are displayed, and each curve is displayed in a different color to facilitate intuitive comparison of pore structure characteristics at different depths.

[0099] In this embodiment, the above method enables accurate characterization of different pore size ranges in shale oil reservoirs and extraction of fluid pore characteristics. The full-pore characterization technology proposed in this invention significantly improves the automation level of data processing, reduces the need for human intervention, and enhances the ability to identify and evaluate complex reservoir pore structures.

[0100] By combining data fusion with mathematical modeling, this method not only accurately characterizes the complex pore structure of shale oil reservoirs, but also provides reliable technical support for reservoir analysis and development. Practical examples have demonstrated that this method offers high accuracy and applicability in characterizing distributions across the entire pore size range, and is widely applicable for the characterization and evaluation of shale oil reservoirs and other unconventional oil and gas reservoirs.

[0101] Example 2

[0102] like Figure 7 As shown, the present invention provides a full-aperture characterization system based on dynamic time warping, and the full-aperture characterization system is used to perform the full-aperture characterization method described in Example 1, including:

[0103] The first processing module is used to collect NMR T2 spectrum data and pore size distribution data from different experiments, and preprocess and standardize the collected data;

[0104] The second processing module is used to perform nonlinear matching processing on the nuclear magnetic resonance T2 spectrum data processed by the first processing module and the pore size distribution experimental data using a dynamic time warping algorithm;

[0105] The third module is used to integrate the gas adsorption experiment, high-pressure mercury injection experiment and nuclear magnetic resonance T2 spectrum results based on the matching results, model the pore size distribution covering micropores, mesopores and macropores, and complete the characterization of the full pore size.

[0106] like Figure 7 The full-aperture characterization system provided in the illustrated embodiment can execute the technical solution shown in the full-aperture characterization method in the above-mentioned method embodiment 1. Its implementation principle and beneficial effects are similar and will not be repeated here.

[0107] In this embodiment, the present application can divide the functional units according to the full aperture characterization method. For example, each function can be divided into each functional unit, or two or more functions can be integrated into one processing unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of a software functional unit. It should be noted that the division of units in the present invention is schematic and is only a logical division. There may be other division methods in actual implementation.

[0108] In this embodiment, in order to realize the principles and beneficial effects of the full-aperture characterization method, the full-aperture characterization system includes hardware structures and / or software modules that perform corresponding functions. It should be easily appreciated by those skilled in the art that, in combination with the various schematic units and algorithm steps described in the embodiments disclosed in the present invention, the present invention can be implemented in the form of hardware and / or a combination of hardware and computer software. Whether a function is executed in a hardware- or computer-software-driven manner depends on the specific application and design constraints of the technical solution. Different methods can be used for each specific application to implement the described function, but such implementation should not be considered to be beyond the scope of this application.

Claims

1. A full-aperture characterization method based on dynamic time warping, characterized in that: The following steps are involved: S1. Data preparation: Collect NMR T2 spectrum data and pore size distribution data from different experiments, and preprocess and standardize the collected data; S2, dynamic time warping matching: using the dynamic time warping algorithm, nonlinear matching is performed on the NMR T2 spectrum data processed by S1 and the pore size distribution experimental data; S3. Characterization of full pore size distribution: Based on the matching results, the gas adsorption experiment, high-pressure mercury injection experiment and nuclear magnetic resonance T2 spectrum results are integrated to model the pore size distribution covering micropores, mesopores and macropores to complete the characterization of the full pore size.

2. The full-aperture characterization method based on dynamic time warping according to claim 1, characterized in that: Said S1 comprises the following steps: S101, collecting nuclear magnetic resonance T2 spectrum data, gas adsorption experimental data and high-pressure mercury injection experimental data; S102, performing Gaussian smoothing and normalization processing on the gas adsorption experimental data and the high-pressure mercury injection experimental data to obtain normalized pore size distribution experimental data; S103, performing normalization processing on the collected nuclear magnetic resonance T2 spectrum data; S104: The data processed by S102 and S103 are fused according to the same scale interval to complete the preprocessing and standardization.

3. The full-aperture characterization method based on dynamic time warping according to claim 1, characterized in that: The S2 comprises the following steps: S201, calculating the Euclidean distance between the nuclear magnetic resonance T2 spectrum data sequence T processed by S1 and the pore size distribution experimental data P; S202, based on the Euclidean distance, using the recursive formula of the dynamic time warping algorithm to calculate the cumulative distance matrix C; S203 , based on the cumulative distance matrix C, backtrack the optimal path W, and construct a conversion relationship between the nuclear magnetic resonance T2 spectrum data sequence T and the pore size distribution experimental data P based on the optimal path W, completing the nonlinear matching process.

4. The full-aperture characterization method based on dynamic time warping according to claim 3, characterized in that: The S203 is specifically as follows: Assume that the end point of the optimal path is , set the backtracking starting point to ,from Start by tracing back in the direction of minimum cumulative distance. , and get the optimal path W, where the next point of the backtracking path is obtained using the following formula: in, Indicates current location The previous point, Indicates current location Above, Indicates current location On the left side, Indicates the current The upper left corner; Based on the optimal path W, a conversion relationship is established between the nuclear magnetic resonance T2 spectrum data sequence T and the pore size distribution experimental data P to complete the nonlinear matching process.

5. The full-aperture characterization method based on dynamic time warping according to claim 4, characterized in that: The expression of the optimal path W is as follows: in, represents the last coordinate in the optimal path W, represents the first The coordinates of the path points, Indicates the current optimal path point The next backtracking path point, Represents the index number of the point on the optimal path W, represents the total number of steps in the optimal path W.

Images