Simulation matching method of multi-pulse measured gamma spectrum based on pulsed neutron instrument

Abstract

The present invention discloses a simulation matching method of a multi-pulse measured gamma energy spectrum based on a pulse neutron instrument, and belongs to the field of well logging technology. The present invention is based on the complete cycle timing of the pulse, and the acquisition time window within the cycle is set according to the pulse duty cycle, and the non-pop-up window and the capture window gamma counts are collected, and the collected energy spectrum contains the position and relative intensity information of the gamma characteristic peak under the current formation setting. Through the linear superposition of the non-pop-up window and the capture window gamma energy spectrum, the gamma energy spectrum under any source pulse cycle or shape setting is reproduced. On this basis, the superposition coefficient of the simulated gamma energy spectrum after broadening is solved to compensate for the accumulation of captured gamma in time caused by the lack of multiple pulse cycles in the simulated gamma energy spectrum; the difficulty of adding a pulse shape to the particle source setting module in the simulation is avoided, and the problems of high simulation difficulty and low precision caused by the diverse timing of the neutron source output pulse, unstable pulse shape and capture accumulation under multi-pulse measurement are solved.

Description

Technical Field

[0001] The invention belongs to the technical field of well logging, and in particular relates to a simulation matching method of a multi-pulse measured gamma energy spectrum based on a pulse neutron instrument. Background Art

[0002] Compared with traditional chemical source instruments, pulsed neutron instruments can avoid environmental protection and safety issues during the transportation and loading and unloading of radioactive sources. Due to the particularity of the neutron source emitting 14MeV high-energy neutrons in pulsed form, this type of instrument can use the time sequence of the interaction between high-energy neutrons and formation materials to collect gamma rays from inelastic scattering and radiation capture in different time windows, and can provide gamma energy spectrum information for rock physical analysis and reservoir quality evaluation, broadening the application scope of nuclear well logging.

[0003] The information obtained based on the above characteristics will be greatly affected by the pulse timing design, and there are many instabilities in the control of pulses. At the same time, due to the different decay characteristics of fast neutrons and thermal neutrons, in multi-pulse measurements, the captured gamma that is not completely decayed will accumulate in subsequent cycles, making it more difficult to match the simulated data with the measured data using conventional methods, which is specifically reflected in:

[0004] 1. It is necessary to simulate multiple cycles and collect the corresponding time window counts after the accumulation of captured counts in the non-pop-up window reaches stability in order to accurately reproduce the accumulation phenomenon of captured gamma rays in time.

[0005] 2. Actual pulses are often non-ideal and difficult to reproduce in simulations. At the same time, due to the influence of communication command delays, signal distortion, etc. during the actual measurement process, the neutron tube pulse may be unstable. The interference of such factors cannot be considered in the simulation.

[0006] 3. Once the pulse settings of the instrument change (such as duty cycle, period, delay time, etc.), it is necessary to modify the software simulation parameters to complete the new model calculation, wasting computing resources.

[0007] Therefore, it is necessary to develop a reliable and efficient method to match the measured energy spectrum using only single-shot, single-pulse simulation data when the actual instrument pulse is repeated many times and may change. Summary of the invention

[0008] The purpose of the present invention is to provide a simulation matching method for multi-pulse measured gamma energy spectrum data of a pulsed neutron instrument, so as to solve the problems of high simulation difficulty and low precision caused by the diversified timing of neutron source output pulses, unstable pulse shape and accumulation of capture counts under multi-pulse measurement.

[0009] To achieve the above object, the present invention adopts the following technical solution:

[0010] A simulation matching method for multi-pulse measured gamma energy spectrum based on a pulsed neutron instrument comprises the following steps:

[0011] Step S1, collecting parameters of the pulsed neutron logging instrument and environmental parameters of the measured formation, and using the collected parameters to construct a three-dimensional numerical model of the pulsed neutron logging instrument and the formation;

[0012] Step S2, obtaining a measured gamma energy spectrum and a simulated gamma energy spectrum; the simulated gamma energy spectrum data is the gamma energy spectrum obtained in the three-dimensional numerical model constructed in step S1, and the spectrum is obtained by:

[0013] S2.1. Assume that the emitted neutron source is 14 MeV fast neutrons and emits uniformly; set the data acquisition time window according to the pulse duty cycle;

[0014] S2.2, collect the gamma energy spectra in the non-pop-up window and the capture window within a pulse cycle, and obtain the energy spectrum counts in the non-pop-up window and the energy spectrum counts in the capture window respectively; wherein the non-pop-up window is used to collect the gamma rays generated by the inelastic scattering of high-energy neutrons and atoms during the emission period, and the capture window is used to collect the gamma rays generated by the radiation capture of thermal neutrons and atoms during the emission interval;

[0015] Step S3, respectively extracting characteristic peaks from the measured gamma energy spectrum and the simulated gamma energy spectrum, and identifying the energy-channel number corresponding to each characteristic peak; establishing a corresponding relationship between the two according to the identified energy-channel number, and obtaining a measured energy spectrum sequence that completely corresponds to the simulated gamma energy spectrum-channel number;

[0016] Step S4, normalizing the simulated gamma spectrum and the measured spectrum sequence obtained in step S3 respectively so that the sum of the counts in their respective valid channels is 1;

[0017] Step S5, matching the normalized simulated gamma spectrum with the normalized measured spectrum sequence to obtain a non-pop-up window spectrum suitable for multi-cycle complex pulse timing.

[0018] Furthermore, S3 also includes preprocessing the measured gamma spectrum before extracting the characteristic peaks, and the preprocessing steps are as follows:

[0019] S3.1. According to the 3σ principle, remove the data points with errors exceeding 3 times the standard deviation at each channel address to improve the accuracy of the measured data;

[0020] S3.2. Use Gaussian filtering to filter out the noise in the gamma spectrum data processed in step S3.1 to improve the matching degree between the measured gamma spectrum data and the ideal gamma spectrum data.

[0021] More preferably, the Gaussian filter is a 5-point Gaussian filter.

[0022] Furthermore, the corresponding relationship between the measured gamma energy spectrum and the simulated gamma energy spectrum in step S3 is established as follows:

[0023] S3.1. Extract characteristic peaks from the measured gamma energy spectrum data. The extracted characteristic peaks include hydrogen capture peak (2.23 MeV), carbon inelastic peak (4.43 MeV) and oxygen inelastic peak (6.13 MeV). According to the corresponding relationship between energy and channel address number, solve the channel address number corresponding to the energy of each characteristic peak of the measured gamma energy spectrum;

[0024] S3.2. Based on the channel address number solved in S3.1, interpolation processing is performed on the measured gamma energy spectrum data to obtain a measured energy spectrum sequence that completely corresponds to the simulated gamma energy spectrum-channel address.

[0025] Furthermore, the detailed matching process of step S5 is as follows:

[0026] S5.1. Establish the detector response function FWHM:

[0027]

[0028] Wherein, E is the simulated gamma ray energy, FWHM is the half-peak width of the Gaussian function performing the convolution operation when broadening the gamma spectral line with energy E; wherein, the coefficients a, b, and c are obtained by fitting the experimental data, and the initial values ​​of the above parameters are set to the commonly used empirical values;

[0029] S5.2, convolve the simulated gamma energy spectrum obtained in step S4 with the detector response function FWHM to complete the broadening of the simulated gamma energy spectrum so that it has a smooth characteristic that matches the measured energy spectrum;

[0030] S5.3, using the weighted least square method to solve the superposition coefficient of the broadened simulated gamma energy spectrum, to compensate for the temporal accumulation of captured gammas caused by the lack of multiple pulse periods in the simulated gamma energy spectrum;

[0031] S5.4, calculating the root mean square error between the simulated gamma spectrum obtained in step S5.3 and the measured spectrum sequence obtained in step S4;

[0032] S5.5. Repeat steps S5.2 to S5.4, output the FWHM parameter value and superposition coefficient with the minimum root mean square error, assign them to the simulated gamma energy spectrum, and you can get a non-pop-up window energy spectrum suitable for multi-cycle complex pulse timing.

[0033] Furthermore, the step S5 also includes optimizing the simulated gamma spectrum broadening, and the optimization process is:

[0034] Set the search range and step size around the initial values ​​of the three FWHM parameters a, b, and c; the search range and step size should cover the maximum fluctuation range caused by factors other than the detector itself; then use the exhaustive method to traverse all different values ​​of a, b, and c determined in S5.1; for each value, repeat step S5.2 to complete the simulation of gamma energy spectrum broadening and then enter step S5.3.

[0035] Furthermore, the detailed solution process of the superposition coefficient in step S5.3 is:

[0036] S5.3.1. Use vectors to represent the energy spectrum counts of the non-pop-up window and the capture window of the simulated gamma energy spectrum;

[0037] Non-pop-up window: TP = (a t1 ,a t2 ,…,a tn ) T

[0038] The capture window is: CAP = (a c1 ,a c2 ,…,a cn ) T

[0039] Assuming that the linear superposition of the simulated gamma energy spectrum non-pop-up window and capture window counts is equal to the measured non-pop-up window count U, we get the equation:

[0040] U=(u1,u2,...,u n ) T =m1TP+m2CAP (1)

[0041] In equation (1), m1 and m2 represent the energy spectrum superposition coefficients of the simulated gamma energy spectrum non-pop-up window and the capture window, respectively.

[0042] S5.3.2. Order:

[0043]

[0044]

[0045] Rewriting equation (1) yields the matrix equation:

[0046] U=Am

[0047] S5.3.3. Rewrite U=Am using weighted least squares to obtain:

[0048] A T WU=A T WAm

[0049] m=(A T WA)-1 A T WU

[0050] S5.3.4. Solve the matrix obtained in step S53.3, and the solved m1 and m2 are the optimal superposition coefficients of the simulated gamma energy spectrum non-pop-up window and the capture window for matching the measured non-pop-up window counts; the superimposed energy spectrum: m1TP+m2CAP is used as a reproduction of the measured non-elastic energy spectrum to compensate for the lack of capture window time accumulation in the simulated gamma energy spectrum.

[0051] Furthermore, the root mean square error between the simulated gamma spectrum obtained in step S5.3 and the measured spectrum sequence obtained in step S4 is the root mean square error RMSE′ with a penalty factor:

[0052]

[0053] In the formula, m i 、s i denotes the counts of the measured energy spectrum sequence obtained in step S4 and the simulated gamma energy spectrum obtained in step S5.3 at the i-th channel address, and denotes the penalty factor p i , n represents the channel number of simulated and measured gamma-ray spectra.

[0054] The present invention provides a simulation matching method for multi-pulse measured gamma energy spectrum based on a pulse neutron instrument. Its simulated energy spectrum data is based on the timing of a complete pulse cycle (a pulse burst period followed by an interval period), and the acquisition time window within the cycle is set according to the pulse duty cycle. The non-pop-up window and the capture window gamma counts are collected, and the collected energy spectrum contains the position and relative intensity information of the gamma characteristic peak under the current formation setting. Since the collected simulated energy spectrum already contains the position and relative intensity information of the gamma characteristic peak set under the current formation environment, the gamma energy spectrum under any source pulse cycle or shape setting can be reproduced through the linear superposition of the non-pop-up window and the capture window gamma energy spectrum, including the source pulse frequency controlled by the neutron tube voltage signal, the duty cycle and the pulse delay and rise time determined by the neutron tube performance. On this basis, by solving the superposition coefficient of the simulated gamma energy spectrum after broadening, the temporal accumulation of captured gamma rays caused by the lack of multiple pulse periods in the simulated gamma energy spectrum is compensated; the difficulty of adding pulse shapes to the particle source setting module in the simulation is avoided; at the same time, the simulation of multiple periodic pulses is avoided, which significantly improves the efficiency of model calculations. It can also be used to evaluate the temporal accumulation effect of captured gamma rays in pulse periodicity measurements, which is of great reference significance for the optimization of neutron source pulse settings.

[0055] In summary, the simulation matching method of the present invention solves the problems of high simulation difficulty and low precision caused by the diversified timing of neutron source output pulses, unstable pulse shapes, and capture pileup under multi-pulse measurement. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 Simplified model for the application scenario of pulsed neutron instrument in gamma spectroscopy mode;

[0057] Figure 2 It is a schematic diagram of a pulse control signal and actual pulse emission;

[0058] Figure 3 are different pulse timing characteristics;

[0059] Figure 4 It is a schematic diagram of the gradual accumulation of capture counts in the non-pop-up window as the repeated measurement period increases when the neutron pulse is repeated periodically;

[0060] Figure 5 is a flow chart of an embodiment;

[0061] Figure 6 This is a comparison chart of the effect of matching the non-pop-up window energy spectrum measured over multiple pulse periods with single-period simulation data. DETAILED DESCRIPTION

[0062] The present invention is further described in detail below with reference to the accompanying drawings and embodiments.

[0063] Before describing in detail the simulation matching method of a multi-pulse measured gamma energy spectrum based on a pulsed neutron instrument provided by the present invention, the principle of the present invention is firstly described.

[0064] like Figure 1 As shown, in the conventional neutron instrument in the gamma spectrum measurement mode, the pulse neutron source 1 emits high-energy neutrons with an energy of 14 MeV. The high-energy neutrons mainly lose energy rapidly through inelastic scattering (process 3 in the figure) and elastic scattering, and slow down to thermal neutrons in a very short time (1 to 12 μs). The slowing down process is concentrated in the formation area 2 around the pulse neutron instrument. After the slowing down is completed, the neutrons are transported to the deeper formation area 5, and are finally absorbed by the formation in the form of radiation capture (process 4 in the figure). In this process, both inelastic scattering and radiation capture reactions are accompanied by the generation of neutron-induced gamma rays, which become inelastic gamma and captured gamma respectively. The two have obvious differences in time distribution due to the difference in the energy of the neutrons involved in the reaction. This embodiment utilizes this time characteristic. When the gamma detector (structure 6 in the figure) collects the gamma spectrum, the time window controlled by the external circuit is used to separately obtain these two types of gamma spectra, and more abundant information is extracted from the gamma spectrum.

[0065] Ideally, the division of the counting time window should strictly correspond to the pulse control signal. However, the actual controlled neutron pulse has great instability and non-ideality. Figure 2 Figure 1 is a schematic diagram of a pulse control signal and actual pulse emission. Figure 2As shown in the figure, in an ideal state, the pulse control signal period is 100μs and the duty cycle is 30%, while the duty cycle of the actual pulse output is only about half of that; at the same time, due to the delay characteristics of the ionization process in the neutron tube, the time when the neutron starts to emit is significantly later than the arrival time of the high-level control signal. Figure 1 In the application mode, the time difference related to the ionization process will change accordingly. Figure 3 As shown in the figure, pulses with different duty cycles have obvious differences in rise time and delay time. These time defects are often limited by the design of the neutron tube, which is difficult to completely eliminate and has great uncertainty. In the end, it is difficult for the pulse emitted by the neutron tube to rise synchronously with the control signal, and the control signal will also be distorted by the complex downhole operating environment. It can be seen that it is impossible to make an accurate estimate of the actual neutron output by relying solely on the control signal; even if a fast neutron detector is added near the neutron source to monitor the pulse timing characteristics in real time and accurately reproduce the non-ideal pulse output in the numerical model, there is still a problem of low efficiency. Not only does it need to set up a complex pulse time distribution function, but it is also necessary to modify the source particle module and recalculate after the pulse changes.

[0066] Secondly, this embodiment Figure 2 Only one pulse cycle of neutron source emission is fully displayed, while actual measurements often repeat dozens to hundreds of cycles. Figure 4 This is a schematic diagram of the gradual accumulation of capture counts in the non-pop-up window as the repeated measurement period increases when the neutron pulse is repeated periodically. Figure 4 As shown in the figure, since fast neutrons exist in the formation for a very short time (less than 12μs), theoretically, a clean capture energy spectrum can be obtained by collecting captured gamma rays generated by the interaction between thermal neutrons and formation materials 10μs after the pulse ends. In actual measurements, however, the existence time of thermal neutrons in the formation is relatively long, usually tens to hundreds of microseconds, which causes the thermal neutrons generated in the previous or even multiple cycles to remain within the time range of the current cycle and continue to interact with the formation to produce captured gamma rays, resulting in additional captured gamma ray counts in the inelastic time window. This phenomenon of captured gamma ray accumulation generated by multi-pulse periodic measurements can be reproduced in a three-dimensional numerical model by simulating multiple pulses, but this will lead to a significant increase in computing costs.

[0067] Example 1

[0068] In order to solve the problem of numerical simulation caused by the complexity of pulse timing in gamma spectrum measurement of pulse neutron instrument. Figure 5 As shown, this embodiment adopts the following technical solution, including the following steps:

[0069] A method for simulating and matching multi-pulse measured gamma spectrum data of a pulsed neutron instrument, comprising the following steps:

[0070] Step S1, collecting environmental parameters of the pulsed neutron logging instrument and the formation it measures, and constructing a three-dimensional numerical model of the pulsed neutron logging instrument and the formation using the collected environmental parameters.

[0071] Step S2, obtaining a measured gamma energy spectrum and a simulated gamma energy spectrum; the simulated gamma energy spectrum data is the gamma energy spectrum obtained in the three-dimensional numerical model constructed in step S1, and the spectrum is obtained by:

[0072] In this embodiment, the neutron source is set to uniformly emit 14MeV fast neutrons during 0-30μs, and collects the non-windowed gamma spectrum during this time period. The neutron source stops emitting at 30-100μs, and collects the capture window gamma spectrum during the 40-100μs time period. The above setting is equivalent to simulating a pulse cycle of a neutron source with a frequency of 10kHz and a duty cycle of 30% and the counts of its corresponding two time windows, and the neutron source time distribution is a rationalized rectangular pulse.

[0073] Step S3, respectively extract the characteristic peaks in the measured gamma energy spectrum and the simulated gamma energy spectrum, and identify the energy-channel number corresponding to each characteristic peak; establish a corresponding relationship between the two according to the identified energy-channel number, and obtain a measured energy spectrum sequence that completely corresponds to the simulated gamma energy spectrum-channel address. Before extracting the characteristic peaks, the measured energy spectrum needs to be preprocessed to improve the matching degree between the measured gamma energy spectrum and the simulated gamma energy spectrum through preprocessing. The preprocessing process includes the following steps:

[0074] S3.1. According to the 3σ principle, remove the data points whose error exceeds 3 times the standard deviation at each channel address;

[0075] S3.2. Use Gaussian filtering to filter out the noise in the gamma spectrum data processed in step S3.1.

[0076] In the actual use of pulsed neutron instruments, due to the influence of the high temperature and high pressure environment underground, the corresponding relationship between the energy spectrum channel address and energy will drift, but the channel address i and energy E measured by the instrument are i There is still the following linear correspondence:

[0077] i=Z+E i / G

[0078] Where Z and G are coefficients to be determined, respectively. The two coefficients can be obtained based on any two characteristic peaks of determined energy on the measured energy spectrum and their actual corresponding channel address numbers. Based on the above correspondence, this embodiment achieves the acquisition of the measured energy spectrum sequence that completely corresponds to the simulated gamma energy spectrum-channel address, as follows:

[0079] Firstly, the characteristic peaks such as hydrogen capture peak (2.23MeV), C inelastic peak (4.43MeV) and oxygen inelastic peak (6.13MeV) were extracted from the measured and simulated gamma energy spectra respectively. According to the linear correspondence between energy and channel address, the channel address number corresponding to the energy of each characteristic peak was solved.

[0080] Secondly, based on the channel address number solved by S3.1, the measured gamma energy spectrum data is interpolated to obtain a measured energy spectrum sequence that completely corresponds to the simulated gamma energy spectrum-channel address.

[0081] Step S4: normalize the simulated gamma spectrum and the measured spectrum sequence obtained in step 3 respectively so that the sum of the counts in their respective valid channels is 1.

[0082] Step S5, matching the normalized simulated gamma spectrum with the normalized measured spectrum sequence to obtain a non-pop-up window spectrum suitable for multi-cycle complex pulse timing. The detailed process of S5 matching is:

[0083] S5.1. Establish the detector response function FWHM;

[0084]

[0085] Wherein, E is the simulated gamma ray energy, FWHM is the half-peak width of the Gaussian function that performs a convolution operation on the gamma spectral line with energy E when broadening it; wherein, the coefficients a, b, and c are obtained by fitting the experimental data, and the initial values ​​of the above parameters are set to commonly used empirical values.

[0086] S5.2. Convolve the simulated gamma energy spectrum obtained in step S4 with the detector response function FWHM to complete the broadening of the simulated gamma energy spectrum so that it has a smooth characteristic that matches the measured energy spectrum.

[0087] The line broadening of the measured energy spectrum is not only related to the material selected for the detector itself, but also affected by the instrument, environment, and statistical errors. The preset FWHM broadening function coefficient selection is only determined by empirical values, and will fluctuate under actual circumstances. In addition, when performing gamma spectrum matching, because in the application of pulsed neutron gamma spectrum logging, the peak height of the characteristic peak determines the accuracy of the formation element information inversion, the counting difference of the channel where the characteristic peak is located is particularly critical, and sometimes even the matching degree of other energy segments can be sacrificed to reduce the counting difference between the measured and simulated data in the characteristic peak segment. Therefore, this embodiment optimizes the simulated spectrum broadening, specifically:

[0088] Set the search range and step size around the initial values ​​of the three FWHM parameters a, b, and c; the search range and step size should cover the maximum fluctuation range caused by factors other than the detector itself; then use the exhaustive method to traverse all different values ​​of a, b, and c determined in S5.1; for each value, repeat step S5.2 to complete the simulation of gamma energy spectrum broadening and then enter step S5.3.

[0089] S5.3. The weighted least square method is used to solve the superposition coefficient of the broadened simulated gamma energy spectrum, and the temporal accumulation of captured gamma caused by the lack of multiple pulse cycles in the simulated gamma energy spectrum is compensated. The detailed solution process of the superposition coefficient is:

[0090] S5.3.1. Use vectors to represent the energy spectrum counts of the non-pop-up window and the capture window of the simulated gamma energy spectrum;

[0091] Non-pop-up window: TP = (a t1 ,a t2 ,…,a tn ) T

[0092] The capture window is: CAP = (a c1 ,a c2 ,…,a cn ) T

[0093] Assuming that the linear superposition of the simulated gamma energy spectrum non-pop-up window and capture window counts is equal to the measured non-pop-up window count U, we get the equation:

[0094] U=(u1,u2,...,u n ) T =m1TP+m2CAP (1)

[0095] In equation (1), m1 and m2 represent the energy spectrum superposition coefficients of the simulated gamma energy spectrum non-pop-up window and the capture window, respectively.

[0096] S5.3.2. Order:

[0097]

[0098]

[0099] Rewriting equation (1) yields the matrix equation:

[0100] U=Am

[0101] S5.3.3. Since the counts of each channel in the measured energy spectrum are not equal, the statistical errors are also different. Therefore, the weighted least squares method is used to rewrite the matrix equation U = Am:

[0102] Weight matrix:

[0103]

[0104] Then the matrix equation required to be solved by the weighted least squares method should be rewritten as:

[0105] A T WU=A T WAm

[0106] m=(A T WA) -1 A T WU

[0107] S5.3.4. Solve the matrix obtained in step S53.3, and the solved m1 and m2 are the optimal superposition coefficients of the simulated gamma energy spectrum non-pop-up window and the capture window for matching the measured non-pop-up window counts; the superimposed energy spectrum: m1TP+m2CAP is used as a reproduction of the measured non-elastic energy spectrum to compensate for the lack of capture window time accumulation in the simulated gamma energy spectrum.

[0108] S5.4. Calculate the root mean square error RMSE′ with penalty factor between the model quasi-Mao spectrum obtained in step S5.3 and the measured spectrum sequence obtained in step S4:

[0109] The calculation formula of root mean square error RMSE is:

[0110]

[0111] In the formula, m i 、s i They respectively represent the counts of the measured energy spectrum sequence obtained in step S4 and the simulated gamma energy spectrum obtained in step S5.3 at the i-th channel address.

[0112] Introducing the penalty factor p i Rewrite the root mean square error RMSE to get the final error evaluation RMSE′ formula:

[0113]

[0114] Where n represents the channel number of simulated and measured gamma-ray spectra.

[0115] By introducing a penalty factor, the weight of the channel where the characteristic peak is located in the error calculation is increased, so that when the optimal match is achieved, the error between the simulated and measured data in the characteristic peak section is smaller.

[0116] S5.5. After traversing all the values ​​of a, b, and c determined in S331, record the FWHM parameter value that minimizes RMSE′ and the superposition coefficient of the simulated inelastic and captured energy spectra, so as to achieve the matching of the single-cycle ideal pulse simulation data with the measured inelastic window energy spectrum under the multi-cycle complex pulse timing.

[0117] Figure 6 The figure is a schematic diagram of the matching results of the measured non-pop-up window counts (measured INN) using the simulated non-pop-up window energy spectrum (TP) and the simulated capture window energy spectrum (CAP). The simulated energy spectrum after linear superposition in the figure achieves better matching of the characteristic peak height and peak broadening in both the low-energy and high-energy ranges. However, the curve using only the non-pop-up window simulated count matching obviously has lower counts at the two characteristic peak positions around 2MeV. In this example, the timing characteristics of the neutron source pulse used in the measured data are unknown, and the simulated data uses the ideal single-cycle pulse design in S1, which shows that the present invention can use simple pulse simulation to achieve matching of neutron-induced gamma energy spectra under complex or even unknown timing designs.

Claims

1. A simulation matching method for multi-pulse measured gamma energy spectrum based on a pulsed neutron instrument, characterized in that: The following steps are involved: Step S1, collecting parameters of the pulsed neutron logging instrument and environmental parameters of the measured formation, and using the collected parameters to construct a three-dimensional numerical model of the pulsed neutron logging instrument and the formation; Step S2, obtaining a measured gamma energy spectrum and a simulated gamma energy spectrum; the simulated gamma energy spectrum data is the gamma energy spectrum obtained in the three-dimensional numerical model constructed in step S1, and the spectrum is obtained by: S2.

1. Assume that the emitted neutron source is 14 MeV fast neutrons and emits uniformly; set the data acquisition time window according to the pulse duty cycle; S2.2, collect the gamma energy spectra in the non-pop-up window and the capture window within a pulse cycle, and obtain the energy spectrum counts in the non-pop-up window and the energy spectrum counts in the capture window respectively; wherein the non-pop-up window is used to collect the gamma rays generated by the inelastic scattering of high-energy neutrons and atoms during the emission period, and the capture window is used to collect the gamma rays generated by the radiation capture of thermal neutrons and atoms during the emission interval; Step S3, respectively extracting characteristic peaks from the measured gamma energy spectrum and the simulated gamma energy spectrum, and identifying the energy-channel number corresponding to each characteristic peak; establishing a corresponding relationship between the two according to the identified energy-channel number, and obtaining a measured energy spectrum sequence that completely corresponds to the simulated gamma energy spectrum-channel number; Step S4, normalizing the simulated gamma spectrum and the measured spectrum sequence obtained in step S3 respectively so that the sum of the counts in their respective valid channels is 1; Step S5, matching the normalized simulated gamma spectrum with the normalized measured spectrum sequence to obtain a non-pop-up window spectrum suitable for multi-cycle complex pulse timing.

2. The method for simulating and matching a multi-pulse measured gamma spectrum based on a pulsed neutron instrument according to claim 1, characterized in that: The S3 also includes preprocessing of the measured gamma spectrum before extracting the characteristic peaks. The preprocessing steps are as follows: S3.

1. According to the 3σ principle, remove the data points whose error exceeds 3 times the standard deviation at each channel address; S3.

2. Use Gaussian filtering to filter out the noise in the gamma spectrum data processed in step S3.

1.

3. The method for simulating and matching a multi-pulse measured gamma spectrum based on a pulsed neutron instrument according to claim 2, characterized in that: The Gaussian filter is a 5-point Gaussian filter.

4. The method for simulating and matching a multi-pulse measured gamma spectrum based on a pulsed neutron instrument according to claim 1, characterized in that: The process of establishing the corresponding relationship between the measured gamma energy spectrum and the simulated gamma energy spectrum in step S3 is as follows: S3.

1. Extract characteristic peaks from the measured gamma energy spectrum data. The extracted characteristic peaks include hydrogen capture peak (2.23 MeV), carbon inelastic peak (4.43 MeV) and oxygen inelastic peak (6.13 MeV). According to the corresponding relationship between energy and channel address number, solve the channel address number corresponding to the energy of each characteristic peak of the measured gamma energy spectrum; S3.

2. Based on the channel address number solved in S3.1, interpolation processing is performed on the measured gamma energy spectrum data to obtain a measured energy spectrum sequence that completely corresponds to the simulated gamma energy spectrum-channel address.

5. The method for simulating and matching a multi-pulse measured gamma spectrum based on a pulsed neutron instrument according to claim 1, characterized in that: The detailed matching process of step S5 is as follows: S5.

1. Establish the detector response function FWHM: Wherein, E is the simulated gamma ray energy, FWHM is the half-peak width of the Gaussian function performing the convolution operation when broadening the gamma spectral line with energy E; wherein, the coefficients a, b, and c are obtained by fitting the experimental data, and the initial values ​​of the above parameters are set to the commonly used empirical values; S5.2, convolve the simulated gamma energy spectrum obtained in step S4 with the detector response function FWHM to complete the broadening of the simulated gamma energy spectrum so that it has a smooth characteristic that matches the measured energy spectrum; S5.3, using the weighted least square method to solve the superposition coefficient of the broadened simulated gamma energy spectrum to compensate for the temporal accumulation of captured gammas caused by the lack of multiple pulse cycles in the simulated gamma energy spectrum; S5.4, calculating the root mean square error between the simulated gamma spectrum obtained in step S5.3 and the measured spectrum sequence obtained in step S4; S5.

5. Repeat steps S5.2 to S5.4, output the FWHM parameter value and superposition coefficient with the minimum root mean square error, assign them to the simulated gamma energy spectrum, and you can get a non-pop-up window energy spectrum suitable for multi-cycle complex pulse timing.

6. The method for simulating and matching a multi-pulse measured gamma spectrum based on a pulsed neutron instrument according to claim 5, characterized in that: The step S5 also includes optimizing the simulated gamma spectrum broadening, and the optimization process is: Set the search range and step size around the initial values ​​of the three FWHM parameters a, b, and c; the search range and step size should cover the maximum fluctuation range caused by factors other than the detector itself; then use the exhaustive method to traverse all different values ​​of a, b, and c determined in S5.1; for each value, repeat step S5.2 to complete the simulation of gamma energy spectrum broadening and then enter step S5.

3.

7. The method for simulating and matching a multi-pulse measured gamma spectrum based on a pulsed neutron instrument according to claim 6, characterized in that: The detailed solution process of the superposition coefficient in step S5.3 is: S5.3.

1. Use vectors to represent the energy spectrum counts of the non-pop-up window and the capture window of the simulated gamma energy spectrum; Non-pop-up window: TP = (a t1 ,a t2 ,…,a tn ) T The capture window is: CAP = (a c1 ,a c2 ,…,a cn ) T Assuming that the linear superposition of the simulated gamma energy spectrum non-pop-up window and capture window counts is equal to the measured non-pop-up window count U, we get the equation: U=(u1,u2,...,u n ) T =m1TP+m2CAP (1) In equation (1), m1 and m2 represent the energy spectrum superposition coefficients of the simulated gamma energy spectrum non-pop-up window and the capture window, respectively. S5.3.

2. Order: Rewriting equation (1) yields the matrix equation: U=Am S5.3.

3. Rewrite U=Am using weighted least squares to obtain: A T WU=A T WAm m=(A T WA) -1 A T WOW S5.3.

4. Solve the matrix obtained in step S5.3.3, and the solved m1 and m2 are the optimal superposition coefficients of the simulated gamma energy spectrum non-pop-up window and the capture window for matching the measured non-pop-up window counts; the superimposed energy spectrum: m1TP+m2CAP is used as a reproduction of the measured non-elastic energy spectrum to compensate for the lack of capture window time accumulation in the simulated gamma energy spectrum.

8. The method for simulating and matching a multi-pulse measured gamma spectrum based on a pulsed neutron instrument according to claim 7, characterized in that: The root mean square error between the simulated gamma spectrum obtained in step S5.3 and the measured spectrum sequence obtained in step S4 is the root mean square error RMSE′ with penalty factor: In the formula, m i 、s i represents the counts of the measured energy spectrum sequence obtained in step S4 and the simulated gamma spectrum obtained in step S5.3 at the i-th channel address, p i represents the penalty factor, and n represents the number of channels of the simulated and measured gamma energy spectra.

Images