A Gamma Energy Spectrum Calibration Method for Lanthanum Bromide Detectors Based on LSTM

Abstract

The present invention is applicable to the technical field of gamma energy spectrum calibration, and provides a gamma energy spectrum calibration method for a lanthanum bromide detector based on LSTM, including the following steps: performing data preparation and preprocessing on the gamma energy spectrometer of the lanthanum bromide detector; constructing an LSTM network model, training and testing the LSTM network model; importing the unknown measured original energy spectrum data of the gamma energy spectrometer of the lanthanum bromide detector into the trained LSTM network model to predict the corresponding nuclide energy spectrum data. The present invention effectively solves problems such as energy spectrum background, spectral line drift, and noise existing in the data collected by the gamma energy spectrometer of the lanthanum bromide detector. By utilizing the memory and processing capabilities of the LSTM network, rapid and stable calibration of the gamma energy spectrum is achieved, the nuclide spectrum line for determining the characteristic peak and peak area of the radioactive nuclide can be quickly obtained, the interference of human factors in the processing of energy spectrum data is significantly reduced, and the accuracy and efficiency of data processing are improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of gamma energy spectrum calibration, and particularly relates to a gamma energy spectrum calibration method for lanthanum bromide detectors based on LSTM. Background Technique

[0002] During the operation of a gamma spectrum instrument with a lanthanum bromide detector, due to factors such as the instrument operation mode and the environment, the acquired gamma energy spectrum data will show abnormalities, such as a series of data abnormalities including spectral line drift and data noise, which affect the inversion of the true element content. At the same time, affected by factors such as spectral line superposition caused by the Compton scattering effect, low-energy tails caused by small-angle scattering of gamma rays within the sensitive volume of the detector, as well as cosmic rays and electronics noise, the original data collected is mixed with background data that is not the measurement target. The presence of the background will make the overall gamma spectrum line higher and will mask some weak peaks, resulting in a large difference between the measured value and the true value.

[0003] When the prior art corrects the gamma energy spectrum data collected by a lanthanum bromide detector, it usually adopts a step-by-step method, that is, first corrects the spectral line drift, then performs noise correction, and finally performs background correction. The order and methods of correction are subjectively determined by the operator. Among them, the correction of spectral line drift adopts methods based on in-channel technology, invariant principle, correlation analysis, or fixed characteristic peak energy; noise correction adopts methods such as NASVD denoising method, LS-SVM segmented noise reduction method, or CUDA noise reduction method; background correction adopts methods such as sensitive non-linear background subtraction method or digital filter background subtraction method. However, this step-by-step correction method is not only time-consuming and laborious, requiring a large amount of computing resources, but also because these methods pursue universality for various gamma energy spectra, when applied to the gamma energy spectra of different detectors, they need to be adjusted and tested before use. Especially for gamma spectrometers with lanthanum bromide crystals as detectors, different from spectrometers with crystals such as sodium iodide (thallium) crystals and bismuth germanate as detectors, the spectra collected by them have the characteristics of high background and high characteristic peaks, which also requires the correction method to be adjusted specifically for them. In view of this, the present invention proposes a gamma energy spectrum calibration method for lanthanum bromide detectors based on long short-term memory network (LSTM). Summary of the Invention

[0004] The purpose of the present invention is to provide a gamma energy spectrum calibration method for lanthanum bromide detectors based on LSTM, aiming to solve the problems raised in the above background technique.

[0005] The purpose of the present invention is achieved through the following technical solutions:

[0006] A gamma energy spectrum calibration method for lanthanum bromide detectors based on LSTM includes the following steps:

[0007] Step S1: Perform data preparation and pre-processing for the lanthanum bromide detector gamma spectrometer, including the following sub-steps:

[0008] Step S11: preprocessing the measured original energy spectrum of the lanthanum bromide detector gamma spectrometer, including spectral line drift correction, noise correction and background correction, to obtain the corresponding measured nuclide energy spectrum;

[0009] Step S12: simulating the nuclide energy spectrum data of the random radioactive element content, and adding the background energy spectrum measured by the lanthanum bromide detector or simulated by the Monte Carlo method to the simulated nuclide energy spectrum, then adding random Gaussian noise to each spectral line, and performing random spectral line drift to obtain the corresponding simulated original energy spectrum;

[0010] Step S2: construct an LSTM network model, and train and test the LSTM network model;

[0011] Step S3: Import the unknown measured original energy spectrum data of the lanthanum bromide detector gamma spectrometer into the trained LSTM network model to predict the corresponding nuclide energy spectrum data.

[0012] Furthermore, in step S12, 20% of the data are randomly selected to perform random spectral line drift of about 5 lines before and after.

[0013] Furthermore, the specific process of step S2 is: constructing an LSTM network model, taking the measured original energy spectrum and the simulated original energy spectrum obtained in step S1 as the input of the LSTM network model, and taking the measured nuclide energy spectrum and the simulated nuclide energy spectrum as the output of the LSTM network model, the measured energy spectrum and the simulated energy spectrum both account for 50% of the training set data volume, and the model is trained and tested.

[0014] Furthermore, the LSTM network model includes an LSTM layer, a convolution layer and a regression layer;

[0015] The LSTM layer saves and transfers the information of the previous time step to the current time step through the recurrent connection of the hidden layer, processing data with time series dependency;

[0016] The convolution layer is used to integrate the features in the energy spectrum data, perform local connections and weight sharing;

[0017] The Regression layer is used to define the loss function layer of the regression problem, compare the predicted nuclide energy spectrum with the true nuclide energy spectrum, and calculate the loss value to update the network weights and biases to optimize the model performance.

[0018] Furthermore, the length of the hidden state of the LSTM layer is 100; the learning rate is 0.001, and the learning rate is adjusted according to the number of learning times. Every 10 training rounds, the learning rate is selectively multiplied by 0.5 according to the training effect; the regularization parameter is 0.001.

[0019] Furthermore, the Regression layer uses the root mean square error as the loss function.

[0020] Furthermore, the specific structure of the t-th neuron in the LSTM layer is A t , A t The specific working principle of the neuron is represented by the following mathematical equations:

[0021] ;

[0022] ;

[0023] ;

[0024] ;

[0025] ;

[0026] ;

[0027] In the formula, f t is the forget gate of neuron A t , W f and b f are both learnable parameters of the forget gate; σ is the Sigmoid function; a t-1 is the output value of the previous neuron; i t is the input gate of neuron A t , W i and b i are both learnable parameters of the input gate; is the new candidate value vector created by the tanh layer, W c and b c are both its learnable parameters; x t is the input value of the network at time t; o t is the output gate of neuron A t , W o and b o are both learnable parameters of the output gate; c t-1 is the long-term memory information output by the previous neuron, c t is the long-term memory information updated after passing through neuron A t ; a t is the output value of neuron A t .

[0028] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0029] The gamma energy spectrum correction method for lanthanum bromide detectors based on LSTM effectively solves problems such as energy spectrum background, spectral line drift, and noise in the data collected by lanthanum bromide detector gamma spectrometers. By utilizing the memory and processing capabilities of the LSTM network, rapid and stable correction of gamma energy spectra is achieved, including spectral line drift correction, noise correction, and background correction. Through this process, nuclide spectra for determining the characteristic peaks and peak areas of radionuclides can be quickly obtained, thereby significantly reducing the interference of human factors in energy spectrum data processing and improving the accuracy and efficiency of data processing. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 It is a flowchart of the method of the present invention.

[0031] Figure 2 It is a framework flowchart of the method of the present invention.

[0032] Figure 3 It is a comparison diagram of the predicted nuclide energy spectrum of the LSTM network model and the actual corrected true nuclide energy spectrum.

[0033] Figure 4 It is a preprocessing process diagram of the measured original energy spectrum.

[0034] Figure 5 It is a comparison diagram of the measured original energy spectrum and the measured nuclide energy spectrum before and after preprocessing.

[0035] Figure 6 It is a preprocessing process diagram of the simulated nuclide energy spectrum.

[0036] Figure 7 It is a comparison diagram of the simulated nuclide energy spectrum and the preprocessed simulated original energy spectrum.

[0037] Figure 8 It is a schematic diagram of the basic framework of the LSTM network model.

[0038] Figure 9 It is a schematic diagram of the specific structure of the neural network of the LSTM layer.

[0039] Figure 10 It is a loss curve of the training process of the LSTM network model.

[0040] Figure 11 It is an RMSE curve of the training process of the LSTM network model. DETAILED DESCRIPTION OF THE INVENTION

[0041] In order to have a clearer understanding of the technical features, purposes and beneficial effects of the present invention, the technical solution of the present invention is now described in detail below, but it should not be construed as limiting the applicable scope of the present invention.

[0042] The specific implementation of the present invention is described in detail below in conjunction with specific embodiments.

[0043] An embodiment of the present invention provides a LSTM-based gamma spectrum correction method for a lanthanum bromide detector, and its flow chart and framework flow chart are respectively as follows: Figure 1 and Figure 2 As shown, the method comprises the following steps:

[0044] Step S1: Perform data preparation and pre-processing for the lanthanum bromide detector gamma spectrometer, including the following sub-steps:

[0045] Step S11: preprocessing the measured original energy spectrum of the lanthanum bromide detector gamma spectrometer, including spectral line drift correction, noise correction and background correction, to obtain the corresponding measured nuclide energy spectrum.

[0046] The original spectral line data collected by the gamma spectrometer of the lanthanum bromide detector consists of two parts: the nuclide energy spectrum and the background energy spectrum. The nuclide energy spectrum is the key data required to obtain the content of radioactive elements, while the background energy spectrum is a natural background caused by the superposition of spectral lines caused by the Compton scattering effect, the low-energy tailing caused by the small-angle scattering of gamma rays in the sensitive volume of the detector, and cosmic rays, electronic noise and other factors. The existence of the background will make the gamma spectrum line overall higher and will cover some weak peaks, resulting in a great difference between the measured value and the true value, so correction is required. At the same time, affected by factors such as the instrument status and the instrument operating environment, the collected original energy spectrum data may have spectral line drift, causing the characteristic peak to deviate from the energy channel to which it belongs, causing a great error in the final result, so drift correction is required. Noise is a measurement error caused by statistical fluctuations during energy spectrum measurement. This error is often unavoidable, so appropriate noise reduction processing is required to reduce its impact on the analysis results. After the above steps, the nuclide energy spectrum in the original energy spectrum can be extracted. The measured original energy spectrum and the measured nuclide energy spectrum are used as the input and output of the LSTM network model training respectively.

[0047] Figure 4Shows the preprocessing process of the measured original energy spectrum. In the figure, the spectral line drift correction module can use methods such as spectral line drift correction based on the technology within the channel boundary and the invariance principle, spectral line drift correction based on correlation, or spectral line drift correction based on the fixed energy of characteristic peaks, etc., to sequentially perform anomaly recognition on all the acquired energy spectra, and correct the gamma energy spectra with spectral line drift anomalies to be the same as the normal spectral lines. The spectral line noise reduction module can use methods such as NASVD denoising method, LS-SVM segmented denoising method, or CUDA denoising method, etc., to sequentially perform noise reduction processing on all the energy spectra after spectral drift correction. The background correction module can use methods such as sensitive non-linear background subtraction method, digital filter background subtraction method, or background subtraction method based on multivariate statistical analysis, etc., to sequentially perform background estimation and subtraction on all the energy spectra that have completed noise reduction processing. The finally obtained data is the nuclide energy spectrum corresponding to the original energy spectrum. In the present invention, the spectral line drift correction based on correlation, NASVD denoising method, and background subtraction method based on multivariate statistical analysis are used. The measured original energy spectrum and the measured nuclide energy spectrum before and after preprocessing are as Figure 5 shown.

[0048] Step S12: Simulate nuclide energy spectrum data of the content of random radioactive elements, and add the background energy spectrum measured by the lanthanum bromide detector or simulated by the Monte Carlo method to the simulated nuclide energy spectrum. Then, add random Gaussian noise to each spectral line and perform random spectral line drift to obtain the corresponding simulated original energy spectrum.

[0049] Since only the objects collected by the gamma-ray spectrometer of the lanthanum bromide detector are used as training samples, the data volume is relatively small. At the same time, due to the uncertainty of the instrument operating state, it is difficult to determine the proportion of abnormal data in the original energy spectrum. If there is too little abnormal data in the acquired data, the corresponding anomaly correction information may not be learned during the training of the LSTM network model. Therefore, it is necessary to prepare simulated energy spectra with a higher proportion of abnormal data as training objects.

[0050] First, simulate including U according to the content of radioactive elements in various rocks 238 , Th 232 , K 40The nuclide energy spectra of the main measurement objects of the gamma-ray spectrometer are measured. Since there are inevitably anomalies such as scattered background and environmental background in each gamma-ray energy spectrum, it is necessary to add the characteristic background data of the lanthanum bromide detector to all the simulated nuclide energy spectra. This step can use the environmental background actually detected by the instrument or the instrument background simulated by the Monte Carlo method. On this basis, random Gaussian noise is added to simulate the noise generated by the instrument due to factors such as temperature changes, electronic device interference, and cosmic rays. Finally, 20% of the data is randomly selected for random spectral line drift of about 5 channels before and after, simulating the spectral line drift problem caused by temperature effects, component aging, and unstable measurement in the measurement of the lanthanum bromide detector gamma-ray spectrometer. The finally obtained simulated original energy spectrum is used as the input during the training of the LSTM network model, and the simulated nuclide energy spectrum is used as the output.

[0051] Figure 6 The preprocessing process of the simulated nuclide energy spectrum is shown. In the figure, the background simulation module adds the environmental background actually detected by the instrument or the background spectrum simulated by the Monte Carlo method to the energy spectrum line. The energy spectrum adding random noise module adds a random noise signal to the simulated nuclide energy spectrum after adding the background. Since the gamma-ray energy spectrum data consists of two parts: a pure signal and a Gaussian white noise signal, and the noise signal is not correlated with the intensity of the sampling signal, adding random Gaussian noise can simulate the gamma-ray energy spectrum. The random spectral line drift simulation module aims to artificially increase the proportion of spectral line drift anomalies in the test data set and enhance the LSTM network model's ability to identify and process spectral line drift. By randomly selecting 20% of the spectral lines and performing spectral line drift processing from +5 channels to -5 channels randomly, the spectral line drift simulation is realized. The data set obtained through the above steps is the original energy spectrum corresponding to the simulated nuclide energy spectrum. In the simulation process of the present invention, the measured environmental background and Gaussian noise are added, and spectral line drift processing of 1 channel is performed. The simulated nuclide energy spectrum and the simulated original energy spectrum obtained after its preprocessing are as Figure 7 shown.

[0052] Step S2: Construct an LSTM network model, train and test the LSTM network model, and adjust the model parameters to ensure that it can accurately correct the original energy spectrum.

[0053] The specific process of step S2 is as follows: Construct an LSTM network model, use the measured original energy spectrum and the simulated original energy spectrum obtained in step S1 as the input of the LSTM network model, and the measured nuclide energy spectrum and the simulated nuclide energy spectrum as the output of the LSTM network model. Both the measured energy spectrum and the simulated energy spectrum account for 50% of the training set data volume. Train and test the model; adjust the learnable parameters and hyperparameters of the model according to the training results to ensure that the model can accurately correct the original energy spectrum.

[0054] The LSTM network model includes LSTM layer, convolution layer and regression layer. Figure 8 The figure is a basic framework diagram of the LSTM network model. The LSTM layer saves and transfers the information of the previous time step to the current time step through the cyclic connection of the hidden layer, thereby processing data with time series dependency. The hidden state length of the LSTM layer is set to 100; the learning rate is set to 0.001, and the learning rate is adjusted according to the number of learning times. After every 10 rounds of training, the learning rate is selectively multiplied by 0.5 according to the training effect (you can also choose not to change the learning rate or terminate early to shorten the training time of the model); to avoid overfitting, the regularization parameter is set to 0.001. The convolution layer is used to integrate the features in the energy spectrum data, perform local connections and weight sharing, which can reduce the number of parameters and improve training efficiency. The regression layer is used to define the loss function layer of the regression problem. It compares the predicted nuclide energy spectrum with the output nuclide energy spectrum and calculates the loss value. It is used to update the weights and biases of the network to optimize the model performance. The regression layer uses the root mean square error (RMSE) as the loss function.

[0055] Figure 9 The figure is a schematic diagram of the specific structure of the neural network of the LSTM layer in the LSTM network model. The LSTM network is a special recurrent neural network (RNN). The goal of RNN design is to process and analyze sequence data. Unlike traditional feedforward neural networks, RNN has a recurrent connection, allowing information to be transmitted cyclically in the network. Its characteristic is that it can use the previous input information to affect the current output information, thereby capturing the temporal relationship in the sequence data. It can use the output of the previous time step as the input of the current time step, thereby introducing a memory mechanism for the neural network, so that RNN can process sequence data of any length and can predict or generate each element in the sequence. However, RNN is prone to gradient disappearance or gradient explosion problems when processing long sequences, resulting in the model being unable to effectively learn long-term information. LSTM is designed to avoid long-term dependency problems. It is different from the simple structure of RNN where each neuron is composed of only one layer of tanh activation function, but has a four-layer neural structure (forget gate, input gate, candidate value vector and output gate).

[0056] Figure 9 Middle A t is the specific structure of the tth neuron in the LSTM layer, A t-1 and A t+1 are the previous and next neurons respectively, and A tThey have exactly the same structure, so they will not be shown repeatedly. Among the neurons in the LSTM layer, σ1 represents the Sigmoid neural network layer of the forget gate, which is used to determine what information to discard from the previous input and the current input; σ2 represents the Sigmoid neural network layer of the input gate, which determines what information to update for the current module, then creates a new candidate value vector through the tanh neural network layer, and finally multiplies the output of Sigmoid by the output of tanh to determine which information is important and needs to be retained; σ3 represents the Sigmoid neural network layer of the output gate, which determines what information the current neuron outputs. In addition, the memory unit running through the top of the LSTM is used to store and update long-term memory information, and it controls the flow of information through the forget gate, input gate, and output gate mechanisms, so as to capture long-term dependencies. Through the above structure, the LSTM can well overcome the problem that traditional RNNs cannot have long-term memory, making it more suitable for the correction of gamma energy spectrum data where abnormal information appears irregularly. A t The specific working principle of the neuron can be expressed by the following mathematical equations:

[0057] ;

[0058] ;

[0059] ;

[0060] ;

[0061] ;

[0062] ;

[0063] In the formula, f t is the forget gate of the A t neuron, and both W f and b f are learnable parameters of the forget gate; σ is the Sigmoid function; a t-1 is the output value of the previous neuron; i t is the input gate of the A t neuron, and both W i and b i are learnable parameters of the input gate; is the new candidate value vector created by the tanh layer, and both W c and b c are its learnable parameters; x t is the input value of the network at time t; o t is the output gate of the A t neuron, and both W o and b oare all learnable parameters of the output gate; c t-1 is the long-term memory information output by the previous neuron, c t is after A t The long-term memory information updated by the neuron; a t is the output value of the A t neuron.

[0064] Figure 10 and Figure 11 are respectively the loss curve and the RMSE curve in the training process of the LSTM network model. It can be seen from the figure that after 15,000 iterations, the prediction result of the model is relatively stable, its loss number approaches 0, and the RMSE value is stable at about 1, indicating that the prediction result of the model is relatively accurate.

[0065] Step S3: Import the unknown measured original energy spectrum data of the lanthanum bromide detector gamma-ray spectrometer into the trained LSTM network model to predict the corresponding nuclide energy spectrum data.

[0066] Deploy the trained LSTM network model in practical applications. Use the trained LSTM network model to correct any unprocessed measured original energy spectrum with problems of both noise and spectral line drift. Compare the obtained predicted nuclide energy spectrum with the nuclide energy spectrum (i.e., the true nuclide energy spectrum) after artificial spectral line drift correction, background correction, and noise correction. The result is as Figure 3 shown. It can be seen that the predicted nuclide energy spectrum of the LSTM network model coincides highly with the true nuclide energy spectrum after actual correction, indicating that it can complete the tasks of fast spectral line drift correction, background correction, and noise correction for gamma-ray spectra.

[0067] The above is only the preferred implementation mode of the present invention. It should be noted that for those skilled in the art, without departing from the concept of the present invention, several deformations and improvements can still be made, and these should also be regarded as the protection scope of the present invention, and these will not affect the implementation effect of the present invention and the practicality of the patent.

Claims

1. A gamma spectrum correction method for lanthanum bromide detector based on LSTM, characterized in that: The following steps are involved: Step S1: Perform data preparation and pre-processing for the lanthanum bromide detector gamma spectrometer, including the following sub-steps: Step S11: preprocessing the measured original energy spectrum of the lanthanum bromide detector gamma spectrometer, including spectral line drift correction, noise correction and background correction, to obtain the corresponding measured nuclide energy spectrum; Step S12: simulating the nuclide energy spectrum data of the random radioactive element content, and adding the background energy spectrum measured by the lanthanum bromide detector or simulated by the Monte Carlo method to the simulated nuclide energy spectrum, then adding random Gaussian noise to each spectral line, and performing random spectral line drift to obtain the corresponding simulated original energy spectrum; Step S2: construct an LSTM network model, and train and test the LSTM network model; The measured original energy spectrum and the simulated original energy spectrum obtained in step S1 are used as the input of the LSTM network model, and the measured nuclide energy spectrum and the simulated nuclide energy spectrum are used as the output of the LSTM network model. The measured energy spectrum and the simulated energy spectrum both account for 50% of the training set data volume, and the model is trained and tested; The LSTM network model includes an LSTM layer, a convolution layer and a regression layer; The LSTM layer saves and transmits the information of the previous time step to the current time step through the loop connection of the hidden layer, and processes the data with time series dependency; the hidden state length of the LSTM layer is 100; the learning rate is 0.001, and the learning rate is adjusted according to the number of learning times. After every 10 rounds of training, the learning rate is selectively multiplied by 0.5 according to the training effect; the regularization parameter is 0.001; the specific structure of the t-th neuron in the LSTM layer is A t , A t The specific working principle of neurons is expressed by the following mathematical equation: ； ； ； ； ； ； In the formula, f t A t The forget gate of the neuron, W f and b f are all learnable parameters of the forget gate; σ is the Sigmoid function; a t-1 is the output value of the previous neuron; i t A t The input gate of the neuron, W i and b i These are all learnable parameters of the input gate; The new candidate value vector created for the tanh layer, W c and b c are all learnable parameters; x t is the input value of the network at time t; o t A t The output gate of the neuron, W o and b o are all learnable parameters of the output gate; c t-1 is the long-term memory information output by the previous neuron, c t For passing A t Long-term memory information after neurons are updated; a t A t The output value of the neuron; The convolution layer is used to integrate the features in the energy spectrum data, perform local connections and weight sharing; The Regression layer is used to define the loss function layer of the regression problem, compare the predicted nuclide energy spectrum with the true nuclide energy spectrum, and calculate the loss value to update the weights and biases of the network and optimize the model performance; the Regression layer uses the root mean square error as the loss function; Step S3: Import the unknown measured original energy spectrum data of the lanthanum bromide detector gamma spectrometer into the trained LSTM network model to predict the corresponding nuclide energy spectrum data.

2. The LSTM-based gamma spectrum correction method for lanthanum bromide detectors according to claim 1, characterized in that: In step S12, 20% of the data are randomly selected to perform random spectral line drift of the five preceding and following tracks.

Images