Pulse neutron logging gamma detector response matrix correction method

By collecting gamma energy spectrum in standard scale wells and correcting the response matrix using particle swarm optimization algorithm, the problem of mismatch between the numerical simulation model and the actual gamma detector response matrix is ​​solved, and the rapid and accurate correction and solution spectrum analysis of the response matrix is ​​realized. Improved accuracy.

CN118897939APending Publication Date: 2024-11-05SOUTHWEST PETROLEUM UNIV
View PDF 0 Cites 1 Cited by

Patent Information

Application Number
CN202410931194.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-12
Publication Date
2024-11-05

AI Technical Summary

Technical Problem

In the prior art, the numerical simulation model does not match the actual gamma detector response matrix, which causes the response matrix to be unable to accurately invert the gamma energy spectrum of the detector, especially when the environment changes, it is necessary to re-evaluate the response matrix, resulting in a large workload. And time-consuming.

Method used

Through physical experiments, gamma energy spectrum is collected in standard scale wells, element characteristic peaks are located, energy spectrum broadening coefficients are fitted, response matrix correction objective function is constructed, and correction coefficients are determined using particle swarm optimization algorithm to realize the response matrix for gamma detectors. Quick and accurate calibration.

Benefits of technology

The fast and accurate correction of the response matrix is ​​realized, so that the corrected matrix is ​​consistent with the actual detector response, shortens the numerical simulation time, and improves the accuracy and efficiency of understanding spectrum analysis.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118897939A_ABST
    Figure CN118897939A_ABST
Patent Text Reader

Abstract

The invention relates to a method for correcting a response matrix of a pulse neutron logging gamma detector, which is an effective means for correcting the response matrix to enable a simulation spectrum to be matched with an actual measurement spectrum, is mainly used for correcting the response matrix of the gamma detector for numerical simulation, and solves the problem that the response matrix is not matched with an actual instrument response matrix. The method comprises the following steps: performing peak searching and characteristic peak Gaussian fitting on a scale well energy spectrum, and determining an energy scale and a broadening coefficient; simulating a response matrix and broadening a single energy response by using a Monte Carlo method; on the basis of experimental spectrum and simulated spectrum errors, a response matrix correction objective function is constructed, a correction coefficient is determined through an optimization algorithm, and correction of a response matrix is achieved. According to the method, the response matrix can be rapidly and accurately corrected by using the optimization algorithm, the problem that the response matrix cannot accurately invert the detector response gamma-ray energy spectrum due to many factors is solved, and a foundation is laid for subsequent spectrum unfolding analysis processing and research of related interpretation methods of pulse neutron logging.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to a pulsed neutron logging gamma detector response matrix correction method, which is mainly used to correct the simulated response matrix when the logging instrument gamma detector response matrix obtained by numerical simulation does not match the actual instrument gamma detector response matrix, so as to ensure that the response gamma energy spectrum generated after the incident gamma photon flow enters the detector is consistent with the actual measured response gamma energy spectrum. The present invention first uses a pulsed neutron logging instrument to collect neutron-induced gamma energy spectrum in a standard calibration well with known mineral composition through physical experiments, then locates the characteristic peaks and corresponding channel addresses of the main elements from the measured energy spectrum, and calibrates the energy scale of the gamma detector according to the characteristic energy of the elements and the corresponding channel addresses, and obtains the corresponding relationship between the channel address and energy in the measured energy spectrum. Then, a Gaussian function is used to fit the characteristic peaks of the elements in the energy spectrum to determine the spectrum broadening coefficient, so as to achieve the broadening process of the simulated single energy response spectrum in the response matrix. Finally, the response matrix correction objective function is constructed by optimizing the error between the measured energy spectrum and the simulated energy spectrum obtained by convolution of the response matrix, and the correction coefficient of the gamma detector response matrix is ​​determined by the optimization algorithm, and finally the correction of the gamma detector response matrix is ​​realized. The present invention can quickly and accurately correct the gamma detector response matrix, so that the corrected response matrix can accurately match the actual detector response, and solves the problem that the response matrix cannot accurately invert the detector response gamma energy spectrum due to factors such as inaccurate numerical simulation models and the influence of the measurement environment, and lays the foundation for subsequent spectrum analysis and processing and related interpretation method research of pulsed neutron logging. Background Art

[0002] At present, most of my country's oil fields have entered the middle and late stages of development, and production wells are facing problems such as increased water content and lack of follow-up remaining resources. Monitoring and evaluating the remaining oil and gas content and distribution are of great significance to the design of subsequent development plans and the excavation of potential layers. Since oil and gas wells are mostly completed and perforated with steel casing, the resistivity logging method commonly used in exploration and logging to evaluate oil content is no longer applicable. The 14MeV fast neutrons emitted by the pulsed neutron source can penetrate the casing and react with the nuclei of formation elements. The secondary gamma rays generated contain a lot of formation information and can be used to evaluate the oil content of the reservoir. Therefore, pulsed neutron logging technology is often used in post-casing logging to monitor the remaining oil and gas, determine the distribution law of the remaining oil and gas, and thus adjust the development plan and improve the recovery rate.

[0003] Pulsed neutron logging technology has become one of the key technologies for post-casing formation parameter evaluation. It mainly generates secondary gamma rays through a series of nuclear reactions such as inelastic scattering, elastic scattering and radiation capture between neutrons and nuclei of formation elements. These secondary gamma rays are collected and recorded by the gamma detector of the instrument to form a corresponding gamma energy spectrum. By analyzing this secondary gamma energy spectrum, logging interpretation parameters reflecting formation information such as reservoir saturation and element content can be extracted. According to the different interpretation parameters extracted, pulsed neutron logging can realize carbon-oxygen ratio spectrum logging, neutron lifetime logging, and formation element logging.

[0004] Deconvolution analysis can be used to analyze the gamma spectrum collected by pulsed neutron logging. Deconvolution analysis converts the analysis of gamma spectrum into the inverse problem of solving the real gamma spectrum, and then uses the peak height or peak area of ​​the characteristic peak of the real gamma spectrum to characterize the element yield in the spectrum. Deconvolution processing requires the response matrix of the gamma detector to solve the real gamma spectrum. Only when the response matrix matches the real response of the detector can the spectrum be accurately analyzed to ensure the accuracy of interpretation.

[0005] In addition, in the study of pulsed neutron logging theory and interpretation methods, Monte Carlo numerical simulation methods are usually used to carry out research. A large number of formation simulation models need to be established to obtain the corresponding gamma detector response spectrum. If the numerical simulation of the response gamma spectrum is repeated every time the conditions are changed, it will not only be labor-intensive and time-consuming, but also labor-intensive. The response spectrum of the gamma detector can be quickly simulated by convolving the response matrix with the incident gamma energy. Therefore, under the premise of obtaining an accurate gamma detector response matrix, it is only necessary to simulate the energy deposition spectrum in the gamma detector under different conditions, and the gamma detector response gamma spectrum under the current formation conditions can be obtained by combining the gamma detector response matrix.

[0006] Both theoretical research and the interpretation of measured gamma energy spectra require the participation of the gamma detector response matrix. The response matrix is ​​generally obtained by Monte Carlo numerical simulation method, but due to certain differences between the simulation model and the actual model, the simulated response matrix cannot completely match the actual response of the gamma detector, and when the working environment of the instrument changes, its response matrix may also change, so it is necessary to calibrate the response matrix obtained by simulation. Summary of the invention

[0007] The purpose of the present invention is to provide a pulse neutron logging gamma detector response matrix correction method, so as to achieve effective correction of the simulated logging instrument gamma detector response matrix, ensure that the incident gamma ray energy distribution of the detector is consistent with the actual measured energy spectrum of the gamma detector after the convolution operation with the corrected response matrix, thereby laying a foundation for the rapid forward simulation and deconvolution operation of the pulse neutron logging energy spectrum.

[0008] In order to achieve the above technical purpose, the present invention provides the following technical solutions (technical process as shown in Figure 1 shown):

[0009] Step 1: Collect the energy spectrum of the standard calibration well experiment.

[0010] A standard calibration well is an artificially made simulation well with known formation composition. The pulsed neutron logging instrument is placed in the well to collect the corresponding neutron induced gamma energy spectrum according to the operating standards. The statistical accuracy of the measured energy spectrum should meet the requirements.

[0011] Step 2: Peak search for characteristic peaks of energy spectrum elements.

[0012] The gamma energy spectrum of the pulsed neutron gamma detector is peak-searched by the channel-by-channel comparison method. Based on the main elements in the formation of the standard calibration well, the total energy peak and the first and second escape peaks of the characteristic elements are determined, and the energy spectrum channel addresses corresponding to the characteristic peak positions are recorded.

[0013] Step 3: Calibrate the gamma detector energy scale.

[0014] Linearly fit the corresponding channel address of the characteristic peak position of the searched element and its characteristic energy to determine the energy scale relationship of the gamma detector. The energy scale relationship can be expressed as:

[0015] E=kC+E z (1)

[0016] Where, E—energy, MeV;

[0017] k—gain, MeV;

[0018] C—channel address, dimensionless;

[0019] E z —Zero cut, energy corresponding to the 0th track, MeV.

[0020] Step 4: Solve the energy spectrum broadening coefficient.

[0021] Gaussian fitting is performed on the gamma counts in the peak region of the characteristic peak of the element determined in step 2 to determine the model parameters in the Gaussian model as shown in formula (2).

[0022]

[0023] Where, G—energy counting rate, cps;

[0024] E—energy, MeV;

[0025] E0—Energy corresponding to the peak position of characteristic peak, MeV;

[0026] C—the peak height of the characteristic peak corresponding to the counting rate, cps;

[0027] A—Gaussian peak width, MeV.

[0028] The Gaussian peak width A in the model parameters and the half-height width FWHM of the characteristic peak have a functional relationship as shown in formula (3). The half-height width FWHM corresponding to the characteristic peak of each element can be calculated by combining the peak width parameters obtained by Gaussian fitting with the characteristic energy of the element.

[0029]

[0030] Where, A—Gaussian width, MeV;

[0031] FWHM—Full width at half maximum of a characteristic peak, MeV.

[0032] Calculate the half-width (FWHM) of the characteristic peaks of at least three elements in the energy spectrum, establish a set of equations according to formula (4), and determine the broadening coefficients a, b, and c.

[0033]

[0034] Where, FWHM—full width at half maximum, MeV;

[0035] E—energy, MeV;

[0036] a, b, c—broadening coefficients, MeV, MeV 1 / 2 , MeV -1 .

[0037] Step 5: Gamma detector response matrix simulation.

[0038] The instrument gamma detector is modeled using Monte Carlo numerical simulation technology, and an isotropic gamma source is set. The energy of the gamma source is determined by the energy scale relationship of the gamma detector. The response energy spectrum of each gamma ray of corresponding energy in the detector is simulated, and the energy spectrum reflects the contribution of the gamma ray of this energy to each channel. The same simulation method is used to simulate n gamma energy spectra, where n is the number of channels in the energy spectrum. According to the gamma energy spectrum at each channel energy, an n×n response matrix R is constructed, as shown in formula (5).

[0039]

[0040] In the formula, R i —Contribution of the 1st to nth energies to the i-th channel;

[0041] E i —Energy corresponding to the ith track, MeV;

[0042] H ij —Detector count rate of the i-th energy in the j-th channel, cps.

[0043] Combined with the broadening coefficients a, b, and c determined in step 4, each response spectrum in the response matrix is ​​Gaussian broadened to obtain the broadened response matrix. The calculation method of spectrum broadening is shown in formula (6).

[0044]

[0045] Where, E(i)—energy of the i-th channel, MeV;

[0046] E(j)—energy of the jth channel, MeV;

[0047] G(i)—the count rate of the energy of the i-th channel without broadening, cps;

[0048] N(i)—normalized broadening count rate at the energy of the i-th channel, cps;

[0049] n—total energy channel number, dimensionless.

[0050] Step 6: Gamma detector response matrix correction.

[0051] Due to the difference between the simulation model and the actual instrument, the response matrix directly determined by numerical simulation cannot be completely consistent with the actual detector response, and the response matrix must be corrected. To this end, a Monte Carlo simulation model is established based on the structure and formation composition of the standard calibration well and the structure of the pulsed neutron logging instrument to simulate the energy distribution of gamma rays entering the gamma detector when the instrument is measuring in the simulated well. The convolution operation result of this energy distribution and the response matrix is ​​the response energy spectrum of the instrument. To correct the response matrix, a weight coefficient is assigned to each energy (each column) in the response matrix, and there are n weight coefficients in total, forming an n×1 weight coefficient matrix w. The simulated response matrix R is corrected by the weight coefficient matrix w to ensure that the convolution result of the simulated incident gamma energy distribution and the corrected response matrix can be consistent with the actual response energy spectrum. When calculating, the correction matrix w needs to be converted into an n×n diagonal matrix D for operation. At the same time, since the count rate in the low energy channel area of ​​the gamma detector is much higher than that in the high energy channel area, the statistical accuracy is very different. To solve this problem, the sigmoid function is introduced to form an n×n diagonal matrix Y as the weight matrix.

[0052]

[0053]

[0054] The present invention adopts a particle swarm optimization algorithm to optimize the correction coefficient, but is not limited to using the optimization algorithm. The core formula of the particle swarm optimization algorithm consists of a particle speed update formula (9) and a position update formula (10).

[0055]

[0056]

[0057] In the formula, —The velocity value of the ith particle in the dth dimension after iteration k, dimensionless;

[0058] w—inertia weight;

[0059] —The optimal individual position of the ith particle in the dth dimension after k iterations, dimensionless;

[0060] —The global optimal position of the particle swarm in the dth dimension after k iterations, dimensionless;

[0061] —The position of the ith particle in the dth dimension after iteration k, dimensionless;

[0062] c1—weight coefficient of the particle tracking its own historical optimal value, dimensionless;

[0063] c2—weight coefficient of the particle to the optimal value of the group, dimensionless;

[0064] r1, r2—random numbers uniformly distributed in the interval [0,1], dimensionless;

[0065] k—Number of iterations, dimensionless.

[0066] The objective function is expressed as follows: (11) and (12).

[0067] K n×1 =Y n×n (N n×1 -R n×n ·D n×n ·X n×1 ) (11)

[0068]

[0069] Where, N is the normalized experimental gamma spectrum matrix;

[0070] R—simulated response matrix after broadening;

[0071] D—diagonal matrix of weight coefficients;

[0072] X—energy flux matrix;

[0073] Y—sigmoid function diagonal weight matrix;

[0074] K—difference matrix;

[0075] erro—Root mean square error between the experimental and simulated gamma spectrum.

[0076] When the root mean square error represented by the objective function is the smallest, the optimal weight coefficient matrix D can be obtained, and the response matrix can be corrected using this weight coefficient matrix. The corrected response matrix can be expressed as:

[0077]

[0078] In the formula, R c —corrected response matrix;

[0079] D—diagonal matrix of weight coefficients;

[0080] R—Simulation response matrix after broadening.

[0081] Beneficial Effects

[0082] Compared with the prior art, the present invention has the following significant advantages: (1) It can quickly and accurately correct the detector response matrix, solving the problem that the response matrix cannot accurately reflect the detector response due to the influence of the instrument itself, the working environment, etc. (2) It can be applied to numerical simulation research to solve the problem that the response matrix needs to be re-calculated when the simulation conditions change. By correcting the response matrix through the present invention, the time required for numerical simulation can be greatly shortened. (3) The present invention does not correct the response matrix as a whole, but corrects the distribution of each energy channel, which greatly improves the correction effect and makes the corrected response matrix more accurately match the detector response. BRIEF DESCRIPTION OF THE DRAWINGS

[0083] Figure 1 It is a technical flow chart of the present invention;

[0084] Figure 2 This is the simulation model of the limestone standard scale well in the example;

[0085] Figure 3 Normalized gamma spectrum measured for the limestone standard calibration well in the example;

[0086] Figure 4 is the energy calibration curve in the example;

[0087] Figure 5 is a schematic diagram of the response matrix after broadening in the example;

[0088] Figure 6 is the simulated gamma spectrum obtained by convolving the uncorrected response matrix with the energy flux in the example;

[0089] Figure 7 It is an iterative process using the method of the present invention in the examples;

[0090] Figure 8 is a schematic diagram of the response matrix after correction in the example;

[0091] Fig. 9 This is the simulated gamma spectrum obtained by convolving the corrected response matrix with the energy flux in the example. DETAILED DESCRIPTION

[0092] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and standard scale well experiments. It should be understood that the specific implementation examples described herein are only used to explain the present invention and are not used to limit the present invention.

[0093] Experimental plan: Place the pulsed neutron logging tool into Figure 2 Multiple measurements were made in the standard calibration well shown, and multiple gamma-ray energy spectra were obtained. The standard calibration well is in limestone formation, with a borehole radius of 10 cm and a height of 350 cm. The borehole fluid is water and there is a casing cement ring. The instrument is 236 cm long, the neutron source is located 57.5 cm below the instrument, and the neutron source pulse width is 40 μs.

[0094] Data collection: The multiple gamma-ray spectra obtained from the experiment were collected and sorted, and the 163 spectral lines obtained were averaged to eliminate statistical errors, thus obtaining the following Figure 3 Normalized experimental gamma spectrum shown.

[0095] Method verification: First, based on the limestone standard calibration well and the borehole environment, carbon, oxygen, iron and hydrogen were selected as characteristic elements, and the obtained experimental gamma ray spectrum was peak-searched for the full energy peaks and partial escape peaks of the above four elements. The results are shown in Table 1. The energy scale of the gamma detector was determined by fitting the relationship between the full energy peaks and partial escape peaks of the characteristic elements and the energy (e.g. Figure 4 As shown), the gamma counts in the peak areas of the element's full energy peak and some escape peaks were fitted with Gaussian fitting, and the three characteristic peak half-widths of carbon were selected to establish a set of equations for solution, and the broadening coefficients a, b, and c were obtained to be 0.063, 0.0075, and 36.8358.

[0096] Table 1 Peak search results

[0097]

[0098] Then, the Monte Carlo numerical simulation method is used to construct the gamma detector response matrix and the broadening coefficient is used to broaden it. The energy flux obtained by simulation is compared with the broadened response matrix (partial energy response spectrum such as Figure 5 As shown), the gamma energy spectrum simulating the gamma detector response is obtained ( Figure 6 ). However, Figure 3 By comparison, it is found that the simulated gamma detector response spectrum is partially inconsistent with the experimental measured spectrum, so the response matrix is ​​corrected. The response matrix after broadening is weighted and iterated through the optimization algorithm. The iterative process is as follows: Figure 7 As shown in , according to the optimization results, select the appropriate weights, and then determine the weight coefficient matrix D, and correct the response matrix. The corrected response matrix (partial energy response spectrum as shown in Figure 8 The energy flux is convolved with the simulated energy flux to obtain Fig. 9 The gamma energy spectrum of the simulated gamma detector response is shown and found to be consistent with Figure 3 The normalized gamma spectrum measured in the limestone standard calibration well shown in the figure is in good agreement with the measured gamma spectrum, and the root mean square error between the two is 5.419×10 -5 , indicating that this method of correcting the response matrix is ​​feasible and effective.

Claims

1. A pulsed neutron logging gamma detector response matrix correction method comprises the following steps: Step 1: Acquisition of energy spectrum of standard calibration well experiment; The standard calibration well is an artificial simulation well with known formation composition. The pulsed neutron logging instrument is placed in the well to collect the corresponding neutron-induced gamma energy spectrum according to the operating standard. The statistical accuracy of the measured energy spectrum should meet the requirements. Step 2: Peak search for characteristic peaks of energy spectrum elements; The gamma energy spectrum of the pulsed neutron gamma detector is peak-searched by the channel-by-channel comparison method. Based on the main elements in the formation of the standard calibration well, the full energy peak and the first and second escape peaks of the characteristic elements are determined, and the energy spectrum channel address corresponding to the characteristic peak position is recorded; Step 3: Calibration of gamma detector energy scale; Linearly fit the corresponding channel address of the characteristic peak position of the searched element and its characteristic energy to determine the energy scale relationship of the gamma detector. The energy scale relationship can be expressed as: E=kC+E z (1) Where, E—energy, MeV; k—gain, MeV; C—track address, dimensionless; E z —Zero cut, the energy corresponding to the 0th track, MeV; Step 4: Solve the spectrum broadening coefficient; Performing Gaussian fitting on the gamma counts in the peak region of the characteristic peak of the element determined in step 2 to determine the model parameters in the Gaussian model as shown in formula (2); Where, G—energy counting rate, cps; E—energy, MeV; E 0 —Energy corresponding to the characteristic peak position, MeV; C—the peak height of the characteristic peak corresponding to the counting rate, cps; A—Gaussian peak width, MeV; The Gaussian peak width A in the model parameters and the half-height width FWHM of the characteristic peak have a functional relationship as shown in formula (3). The half-height width FWHM corresponding to the characteristic peak of each element can be calculated by combining the peak width parameter obtained by Gaussian fitting with the characteristic energy of the element; Where, A—Gaussian width, MeV; FWHM—Full width at half maximum of characteristic peak, MeV; Calculate the half-height width (FWHM) of at least three characteristic peaks of elements in the energy spectrum, establish a system of equations according to formula (4), and determine the broadening coefficients a, b, and c; Where, FWHM—full width at half maximum, MeV; E—energy, MeV; a, b, c—broadening coefficients, MeV, MeV 1 / 2 , MeV -1 ; Step 5: Gamma detector response matrix simulation; The instrument's gamma detector is modeled using Monte Carlo numerical simulation technology, and an isotropic gamma source is set up. The energy of the gamma source is determined by the energy scale relationship of the gamma detector. ; Simulate the response energy spectrum of each gamma ray of corresponding energy in the detector, which reflects the contribution of the gamma ray of this energy to each channel; simulate n gamma energy spectra in the same simulation method, where n is the number of channels of the energy spectrum; construct an n×n response matrix R according to the gamma energy spectrum at each channel energy, as shown in formula (5); In the formula, R i —Contribution of the 1st to nth energies to the i-th channel; E i —Energy corresponding to the ith track, MeV; H ij —The count rate of the detector at the i-th energy on the j-th channel, cps; Combined with the broadening coefficients a, b, and c determined in step 4, each response spectrum in the response matrix is ​​Gaussian broadened to obtain the broadened response matrix; the calculation method of spectrum broadening is shown in formula (6); Where, E(i)—energy of the i-th channel, MeV; E(j)—energy of the jth channel, MeV; G(i)—the count rate of the energy of the i-th channel without broadening, cps; N(i)—normalized broadening count rate at the energy of the i-th channel, cps; n—total energy channel number, dimensionless; Step 6: Gamma detector response matrix correction; Due to the difference between the simulation model and the actual instrument, the response matrix directly determined by numerical simulation cannot be completely consistent with the actual detector response, and the response matrix must be corrected. To this end, a Monte Carlo simulation model is established based on the structure and formation composition of the standard calibration well and the structure of the pulsed neutron logging instrument to simulate the energy distribution of gamma rays entering the gamma detector when the instrument is measuring in the simulated well. The convolution operation result of this energy distribution and the response matrix is ​​the response energy spectrum of the instrument. To correct the response matrix, a weight coefficient is assigned to each energy (each column) in the response matrix, with a total of n weight coefficients, forming an n×1 weight coefficient matrix w; The simulated response matrix R is corrected by the weight coefficient matrix w to ensure that the convolution result of the simulated incident gamma energy distribution and the corrected response matrix is ​​consistent with the actual response energy spectrum; During calculation, the correction matrix w needs to be converted into an n×n diagonal matrix D for operation. At the same time, since the counting rate in the low-energy channel area of ​​the gamma detector is much higher than that in the high-energy channel area, the statistical accuracy is very different. To solve this problem, the sigmoid function is introduced to form an n×n diagonal matrix Y as the weight matrix. The present invention adopts a particle swarm optimization algorithm to optimize the correction coefficient, but is not limited to using this optimization algorithm. The core formula of the particle swarm optimization algorithm is composed of a particle speed update formula (9) and a position update formula (10); In the formula, —The velocity value of the ith particle in the dth dimension after iteration k, dimensionless; w—inertia weight; —The optimal individual position of the ith particle in the dth dimension after k iterations, dimensionless; —The global optimal position of the particle swarm in the dth dimension after iteration k, dimensionless; —The position of the ith particle in the dth dimension after iteration k, dimensionless; c 1 —The weight coefficient of the particle tracking its own historical optimal value, dimensionless; c 2 —The weight coefficient of the particle to the optimal value of the group, dimensionless; r 1 、r 2 —A random number uniformly distributed in the interval [0,1], dimensionless; k—number of iterations, dimensionless; The objective function is shown in the following equations (11) and (12): K n×1 =Y n×n (N n×1 -R n×n ·D n×n ·X n×1 ) (11) Where, N is the normalized experimental gamma spectrum matrix; R—simulated response matrix after broadening; D—diagonal matrix of weight coefficients; X—energy flux matrix; Y—sigmoid function diagonal weight matrix; K—difference matrix; erro—Root mean square error between the experimental gamma spectrum and the simulated gamma spectrum; When the root mean square error represented by the objective function is the smallest, the optimal weight coefficient matrix D can be obtained. The weight coefficient matrix can be used to correct the response matrix. The corrected response matrix can be expressed as: In the formula, R c —corrected response matrix; D—diagonal matrix of weight coefficients; R—Simulation response matrix after broadening.

Citation Information

Cited By

  • Calculation method of pulse neutron logging gamma-ray spectrum broadening coefficient

    CN119761188A