Dose rate measurement method and device based on Poisson maximum likelihood estimation, equipment and storage medium

Abstract

The invention discloses a dose rate measurement method, device and equipment based on Poisson maximum likelihood estimation, and a storage medium, and the method comprises the steps: obtaining and preprocessing environment nuclear radiation monitoring data, and obtaining a fusion data sequence; dividing the fused data sequence into a plurality of time windows and determining an envelope spectrum sequence; performing radiation dose rate estimation on the envelope spectrum sequence through a Poisson maximum likelihood estimation method to obtain a preliminary dose rate estimation value; determining an environment correction factor according to the temperature data sequence, the humidity data sequence and the air pressure data sequence; correcting the initial dose rate estimation value to obtain a corrected dose rate estimation value; and inputting the corrected dose rate estimated value and the average dose rate and the maximum dose rate in each time window into a support vector machine model for prediction to obtain a future nuclear radiation dose rate prediction result, and updating the optimal estimated value of the dose rate in real time by introducing a Poisson maximum likelihood estimation algorithm. The accuracy and response speed of dose rate estimation are improved.

Description

Technical Field

[0001] This application relates to the field of dose rate measurement technology, and in particular to dose rate measurement methods, apparatus, devices and storage media based on Poisson maximum likelihood estimation. Background Technology

[0002] With the increasing demand for nuclear radiation safety management, accurate estimation and real-time prediction of nuclear radiation dose rates are of great significance in fields such as nuclear safety, environmental monitoring, and public health. The estimation of nuclear radiation dose rates is typically affected by multiple environmental factors; changes in natural environmental conditions such as air pressure, humidity, and temperature can interfere with radiation monitoring results to varying degrees. Therefore, how to perform environmental correction on the preliminary dose rate data to improve its accuracy and response speed has become a pressing technical challenge.

[0003] Traditional dose rate estimation methods rely heavily on theoretical models and simple statistical processing. However, these methods often have large errors when faced with complex environmental changes, resulting in inaccurate prediction results. Summary of the Invention

[0004] The main objective of this application is to provide a dose rate measurement method, apparatus, device, and storage medium based on Poisson maximum likelihood estimation, aiming to solve the technical problem that traditional dose rate estimation methods have large errors, resulting in inaccurate prediction results.

[0005] To achieve the above objectives, this application proposes a dose rate measurement method based on Poisson maximum likelihood estimation, wherein the dose rate measurement method based on Poisson maximum likelihood estimation includes: Environmental nuclear radiation monitoring data is acquired and preprocessed to obtain a fused data sequence, wherein the fused data sequence includes a nuclear radiation dose rate data sequence, a temperature data sequence, a humidity data sequence, and an air pressure data sequence. The fused data sequence is divided into multiple time windows, and the envelope spectrum sequence of each time window is determined. The radiation dose rate is estimated for the envelope spectrum sequence of each time window using the Poisson maximum likelihood estimation method, thus obtaining the preliminary dose rate estimate for each time window. The environmental correction factor is determined based on the temperature data sequence, humidity data sequence, and air pressure data sequence. The preliminary dose rate estimate for each time window is corrected based on the environmental correction factor to obtain the corrected dose rate estimate. The corrected dose rate estimate, the average dose rate within each time window, and the maximum dose rate are input into the support vector machine model to predict the nuclear radiation dose rate data, thereby obtaining the future nuclear radiation dose rate prediction results.

[0006] In one embodiment, the step of acquiring environmental nuclear radiation monitoring data and preprocessing the environmental nuclear radiation monitoring data to obtain a fused data sequence includes: Environmental nuclear radiation monitoring data is acquired, and outlier data points are removed from the environmental nuclear radiation monitoring data based on the isolated forest algorithm to obtain environmental nuclear radiation monitoring data after outlier removal. The environmental nuclear radiation monitoring data after outlier removal is interpolated using the LSTM algorithm to obtain an environmental nuclear radiation monitoring data sequence. Based on the environmental nuclear radiation monitoring data sequence, nuclear radiation dose rate data sequence, temperature data sequence, humidity data sequence, and air pressure data sequence are determined; The nuclear radiation dose rate data sequence, temperature data sequence, humidity data sequence, and air pressure data sequence are subjected to multi-source data normalization and fusion to obtain a fused data sequence.

[0007] In one embodiment, the outlier detection and removal of outlier data points in the environmental nuclear radiation monitoring data based on the isolated forest algorithm to obtain outlier-removed environmental nuclear radiation monitoring data includes: A training set containing multiple sample data is set up, wherein each sample data includes environmental nuclear radiation monitoring data and its corresponding label, and the label indicates whether the sample data is abnormal data; Feature extraction is performed on each sample data in the training set to obtain the feature vector of each sample data; The feature vectors are input into the isolated forest algorithm to construct multiple isolated trees. In each isolated tree, a portion of the feature vectors are randomly selected for node splitting until the stopping condition is met. The path length of each sample data in the multiple isolated trees is calculated, and the anomaly score of each sample data is calculated based on the path length. The sample data in the training set are sorted according to the abnormal scores, and the top N samples with the highest scores are selected as abnormal samples, where N is the preset number of abnormal samples. The abnormal samples are used to train the preset outlier detection model and adjust the model parameters until the convergence condition is met, thus obtaining the target outlier detection model. The environmental nuclear radiation monitoring data is input into the target anomaly detection model to obtain an anomaly score for each data point; An abnormal data threshold is determined based on the abnormal score. Data points with abnormal scores higher than the abnormal data threshold are removed as abnormal data points to obtain environmental nuclear radiation monitoring data after removing outliers.

[0008] In one embodiment, dividing the fused data sequence into multiple time windows and determining the envelope spectrum sequence of each time window includes: The length of the time window is determined based on the time series characteristics of the fused data sequence; The fused data sequence is divided according to a determined time window length to obtain multiple consecutive and non-overlapping time windows; Empirical mode decomposition is performed on the fused data sequence within each time window to obtain the set of intrinsic mode functions; Perform a Hilbert transform on each intrinsic mode function in the intrinsic mode function set to obtain the instantaneous amplitude sequence of each intrinsic mode function; The instantaneous amplitude sequences of all intrinsic mode functions within each time window are superimposed to obtain the envelope spectrum sequence of each time window.

[0009] In one embodiment, the step of estimating the radiation dose rate of the envelope spectrum sequence for each time window using the Poisson maximum likelihood estimation method to obtain a preliminary dose rate estimate for each time window includes: A Poisson distribution model is constructed, wherein the parameters of the Poisson distribution model are set according to the envelope spectrum sequence of each time window; Define an objective function, wherein the objective function is the likelihood function of the Poisson distribution model, and the objective function is calculated based on the envelope spectrum sequence of each time window and the parameters of the Poisson distribution model; The objective function is solved iteratively using a numerical optimization algorithm until a preset convergence condition or upper limit of the number of iterations is reached, so as to obtain the optimal estimate of the radiation dose rate for each time window. In the process of iterative solution, the gradient or second derivative of the objective function is calculated based on the current parameter estimate, and the current parameter estimate is updated based on the gradient or second derivative. The optimal estimate is used as the initial dose rate estimate for each time window.

[0010] In one embodiment, after estimating the radiation dose rate of the envelope spectrum sequence for each time window using the Poisson maximum likelihood estimation method to obtain a preliminary dose rate estimate for each time window, the method further includes: A prior distribution is set for the parameters of the Poisson distribution model, wherein the prior distribution is determined based on domain knowledge or historical data; Based on the prior distribution and the preliminary dose rate estimate, a posterior distribution is constructed; Based on the posterior distribution, the parameters of the Poisson distribution model are sampled using the Markov chain Monte Carlo algorithm to obtain multiple parameter samples; The initial dose rate estimate for each time window is weighted and averaged according to the radiation dose rate estimate corresponding to each parameter sample to obtain the weighted average dose rate estimate. The weight of the weighted average is determined according to the posterior probability of each parameter sample. The final preliminary dose rate estimate for each time window is updated based on the weighted average dose rate estimate.

[0011] In one embodiment, determining the environmental correction factor based on the temperature data sequence, humidity data sequence, and air pressure data sequence includes: A multiple linear regression model was constructed, in which the independent variables are temperature data series, humidity data series, and air pressure data series, and the dependent variable is the environmental correction factor; Define a loss function, whereby the loss function represents the difference between the predicted value and the actual value of the multiple linear regression model; The gradient is calculated based on the loss function, and the parameters of the multiple linear regression model are iteratively updated using the gradient descent method until the preset convergence condition is met or the preset maximum number of iterations is reached. In each iteration, the gradient of the loss function is calculated based on the current parameter value, and the current parameter value is adjusted according to the gradient to gradually approach the optimal solution; The updated parameter values ​​are used as the regression coefficients of the multiple linear regression model. The environmental correction factor corresponding to each time window is calculated based on the regression coefficients and the temperature, humidity, and air pressure data sequences.

[0012] Furthermore, to achieve the above objectives, this application also proposes a dose rate measurement device based on Poisson maximum likelihood estimation, wherein the dose rate measurement device based on Poisson maximum likelihood estimation includes: A preprocessing module is used to acquire environmental nuclear radiation monitoring data and preprocess the environmental nuclear radiation monitoring data to obtain a fused data sequence, wherein the fused data sequence includes a nuclear radiation dose rate data sequence, a temperature data sequence, a humidity data sequence, and an air pressure data sequence. The determination module is used to divide the fused data sequence into multiple time windows and determine the envelope spectrum sequence of each time window; The estimation module is used to estimate the radiation dose rate of the envelope spectrum sequence for each time window using the Poisson maximum likelihood estimation method, and obtain the preliminary dose rate estimate for each time window. The determining module also determines an environmental correction factor based on the temperature data sequence, humidity data sequence, and air pressure data sequence; The correction module is used to correct the preliminary dose rate estimate for each time window based on the environmental correction factor, so as to obtain the corrected dose rate estimate. The prediction module is used to input the corrected dose rate estimate, the average dose rate and the maximum dose rate within each time window into the support vector machine model to predict the nuclear radiation dose rate data and obtain the future nuclear radiation dose rate prediction results.

[0013] Furthermore, to achieve the above objectives, this application also proposes a dose rate measurement device based on Poisson maximum likelihood estimation, the device comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the dose rate measurement method based on Poisson maximum likelihood estimation as described above.

[0014] In addition, to achieve the above objectives, this application also proposes a storage medium, which is a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it implements the steps of the dose rate measurement method based on Poisson maximum likelihood estimation as described above.

[0015] In addition, to achieve the above objectives, this application also provides a computer program product, which includes a computer program that, when executed by a processor, implements the steps of the dose rate measurement method based on Poisson maximum likelihood estimation as described above.

[0016] This application proposes one or more technical solutions to acquire environmental nuclear radiation monitoring data and preprocess the data to obtain a fused data sequence. The fused data sequence includes a nuclear radiation dose rate data sequence, a temperature data sequence, a humidity data sequence, and a pressure data sequence. The fused data sequence is divided into multiple time windows, and the envelope spectrum sequence of each time window is determined. The radiation dose rate of each time window's envelope spectrum sequence is estimated using the Poisson maximum likelihood estimation method to obtain a preliminary dose rate estimate for each time window. An environmental correction factor is determined based on the temperature, humidity, and pressure data sequences. The preliminary dose rate estimate for each time window is corrected using the environmental correction factor to obtain a corrected dose rate estimate. The corrected dose rate estimate, along with the average and maximum dose rates within each time window, is input into a support vector machine model to predict the nuclear radiation dose rate, resulting in a future nuclear radiation dose rate prediction. By introducing the Poisson maximum likelihood estimation algorithm to update the optimal dose rate estimate in real time, the accuracy and response speed of dose rate estimation are improved. Attached Figure Description

[0017] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0018] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a flowchart illustrating an embodiment of the dose rate measurement method based on Poisson maximum likelihood estimation provided in this application. Figure 2 This is a flowchart illustrating Embodiment 2 of the dose rate measurement method based on Poisson maximum likelihood estimation provided in this application; Figure 3 This is a schematic diagram of the module structure of the dose rate measurement device based on Poisson maximum likelihood estimation in an embodiment of this application; Figure 4 This is a schematic diagram of the hardware operating environment of the dose rate measurement device based on Poisson maximum likelihood estimation in the embodiments of this application.

[0020] The purpose, features, and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0021] It should be understood that the specific embodiments described herein are merely illustrative of the technical solutions of this application and are not intended to limit this application.

[0022] To better understand the technical solution of this application, a detailed description will be provided below in conjunction with the accompanying drawings and specific implementation methods.

[0023] The main solution of this application embodiment is as follows: Acquire environmental nuclear radiation monitoring data and preprocess the environmental nuclear radiation monitoring data to obtain a fused data sequence, wherein the fused data sequence includes a nuclear radiation dose rate data sequence, a temperature data sequence, a humidity data sequence, and a pressure data sequence; divide the fused data sequence into multiple time windows and determine the envelope spectrum sequence of each time window; estimate the radiation dose rate of the envelope spectrum sequence of each time window using the Poisson maximum likelihood estimation method to obtain a preliminary dose rate estimate for each time window; determine an environmental correction factor based on the temperature data sequence, humidity data sequence, and pressure data sequence; correct the preliminary dose rate estimate for each time window based on the environmental correction factor to obtain a corrected dose rate estimate; input the corrected dose rate estimate, the average dose rate within each time window, and the maximum dose rate into a support vector machine model to predict the nuclear radiation dose rate data, thereby obtaining a future nuclear radiation dose rate prediction result.

[0024] Traditional dose rate estimation methods rely heavily on theoretical models and simple statistical processing. However, these methods often have large errors when faced with complex environmental changes, resulting in inaccurate prediction results.

[0025] This application provides a solution that improves the accuracy and response speed of dose rate estimation by introducing a Poisson maximum likelihood estimation algorithm to update the optimal dose rate estimate in real time.

[0026] It should be noted that the executing entity in this embodiment can be a computing service device with data processing, network communication, and program execution functions, such as a tablet computer, personal computer, or mobile phone, or an electronic device capable of performing the above functions, such as a dose rate measurement device based on Poisson maximum likelihood estimation. The following description uses a dose rate measurement device based on Poisson maximum likelihood estimation as an example to illustrate this embodiment and the subsequent embodiments.

[0027] Based on this, embodiments of this application provide a dose rate measurement method based on Poisson maximum likelihood estimation, referring to... Figure 1 , Figure 1 This is a flowchart illustrating the first embodiment of the dose rate measurement method based on Poisson maximum likelihood estimation in this application.

[0028] In this embodiment, the dose rate measurement method based on Poisson maximum likelihood estimation includes steps S10~S60: Step S10: Acquire environmental nuclear radiation monitoring data and preprocess the environmental nuclear radiation monitoring data to obtain a fused data sequence, wherein the fused data sequence includes a nuclear radiation dose rate data sequence, a temperature data sequence, a humidity data sequence, and an air pressure data sequence.

[0029] It should be noted that environmental nuclear radiation monitoring data can be collected using various sensors, such as nuclear radiation sensors for collecting nuclear radiation dose rate data, temperature and humidity sensors for collecting temperature and humidity data, and barometric pressure sensors for collecting barometric pressure data. These data may be affected by noise interference or contain missing values ​​during the acquisition process, making preprocessing crucial. Preprocessing may include data cleaning, such as removing outliers and filling in missing values, as well as data normalization or standardization to ensure that data of different dimensions are treated equally in subsequent processing. The fused data sequence is the preprocessed and integrated dataset used for subsequent analysis.

[0030] It is understandable that the fused data sequence consists of nuclear radiation dose rate data sequence, temperature data sequence, humidity data sequence and air pressure data sequence. These data sequences are time-aligned, that is, various data at the same point in time are integrated together.

[0031] Step S20: Divide the fused data sequence into multiple time windows and determine the envelope spectrum sequence of each time window.

[0032] It should be noted that the time window division is to capture dynamic changes in the data, and the data within each time window is treated as an independent sample for analysis. The envelope spectrum sequence reflects the envelope characteristics of the nuclear radiation dose rate data and other environmental parameters (such as temperature, humidity, and air pressure) over time within each time window.

[0033] In one feasible implementation, step S20 may include: determining the length of a time window based on the time series characteristics of the fused data sequence; dividing the fused data sequence according to the determined time window length to obtain multiple continuous and non-overlapping time windows; performing empirical mode decomposition on the fused data sequence within each time window to obtain an intrinsic mode function set; performing Hilbert transform on each intrinsic mode function in the intrinsic mode function set to obtain the instantaneous amplitude sequence of each intrinsic mode function; and superimposing the instantaneous amplitude sequences of all intrinsic mode functions within each time window to obtain the envelope spectrum sequence of each time window.

[0034] It should be noted that the length of the time window needs to be determined based on the time series characteristics of the fused data sequence. The chosen time window length should ensure that it includes sufficient periodic fluctuation information and avoid being too large or too small. The window length can be selected to match the total length of the data sequence with the maximum period of the periodic fluctuation; this embodiment does not impose specific limitations on this.

[0035] It is understandable that if the data sequence is {x1,x2,…,xN}, where N is the total length of the data and the time window length is L, then the data sequence is divided into multiple time windows {W1,W2,…,WM}, each containing L data points that do not overlap.

[0036] It's worth noting that Empirical Mode Decomposition (EMD) is a method for decomposing nonlinear, non-stationary signals. It decomposes a signal into a series of Intrinsic Mode Functions (IMFs). Each IMF represents an intrinsic vibrational mode of the signal, arranged from high to low frequencies. The EMD process extracts IMFs from the data through continuous local extremum interpolation. Specifically, this involves: generating an envelope curve using extrema and interpolation to obtain the data residuals; detrending the residuals and generating new IMFs; repeating this process until all IMFs are obtained. For each window W... i N is obtained through EMD decomposition. IMF Each intrinsic mode function {IMF1, IMF2, ..., IMF} NIMF}, where each IMF represents a local oscillation mode in the signal.

[0037] The Hilbert transform is a linear transform that converts a real signal into an analytic signal, yielding information such as the instantaneous amplitude, instantaneous frequency, and instantaneous phase. In this embodiment, performing a Hilbert transform on each IMF yields its instantaneous amplitude sequence. These instantaneous amplitude sequences reflect the envelope characteristics of the data over time at different frequencies. Superimposing the instantaneous amplitude sequences of all IMFs within each time window yields the envelope spectrum sequence for that time window. The envelope spectrum sequence integrates the variation characteristics at different frequencies, providing a more comprehensive description of the dynamic changes of the data within that time window.

[0038] Step S30: Estimate the radiation dose rate of the envelope spectrum sequence for each time window using the Poisson maximum likelihood estimation method to obtain the preliminary dose rate estimate for each time window.

[0039] It should be noted that Poisson maximum likelihood estimation is a statistical method, particularly suitable for handling count data or low-frequency events. In the context of radiation dose rate estimation, since radiation events are often rare, the Poisson distribution provides a suitable mathematical model to describe these events. Poisson maximum likelihood estimation estimates the parameter, i.e., the radiation dose rate, by maximizing the likelihood function, minimizing the difference between the observed data (in this case, the envelope spectrum sequence) and the model's predictions.

[0040] In practice, the envelope spectrum sequence of each time window can be considered as observed data points, and then the Poisson maximum likelihood estimation method can be applied to estimate the radiation dose rate within that time window. This process may involve steps such as constructing the likelihood function, finding its derivative, and optimizing parameters. Ultimately, a preliminary dose rate estimate will be obtained for each time window, reflecting the changes in radiation dose rate over different time periods.

[0041] In one feasible implementation, step S30 may include: constructing a Poisson distribution model, wherein the parameters of the Poisson distribution model are set according to the envelope spectrum sequence of each time window; setting an objective function, wherein the objective function is the likelihood function of the Poisson distribution model, and the objective function is calculated based on the envelope spectrum sequence of each time window and the parameters of the Poisson distribution model; using a numerical optimization algorithm to iteratively solve the objective function until a preset convergence condition or an upper limit of the number of iterations is reached, to obtain the optimal estimate of the radiation dose rate for each time window, wherein during the iterative solution process, the gradient or second derivative of the objective function is calculated based on the current parameter estimate, and the current parameter estimate is updated based on the gradient or second derivative; and the optimal estimate is used as the preliminary dose rate estimate for each time window.

[0042] It should be noted that the Poisson distribution is typically used to describe the number of events occurring per unit of time. For the envelope spectrum sequence of each time window, the Poisson distribution can be used to estimate the occurrence rate of events (i.e., radiation dose rate) within that time window.

[0043] Understandably, the likelihood function of the Poisson distribution represents the probability of all observed data sequences given a parameter λ. In this implementation, it is assumed that there are multiple time windows W. i The envelope spectrum sequence within each window is S i (t). For a time window Wi, the likelihood function of the envelope spectrum sequence is:

[0044] Among them, T i It is the time window W i Length, S i (t) is the time window W i The envelope spectrum sequence value at time t. These are the Poisson distribution parameters that need to be estimated.

[0045] It is worth noting that, in order to estimate the radiation dose rate for each time window... We can take the logarithm of the likelihood function to obtain the log-likelihood function, and then maximize it. The maximum likelihood estimation problem is to estimate the likelihood by optimizing this log-likelihood function. The value of , i.e., radiation dose rate.

[0046] Since the likelihood function is about the parameters Nonlinear functions are solved using numerical optimization algorithms, such as gradient descent. The update formula for gradient descent is:

[0047] in, This is the updated estimate of the parameters for the i-th time window in the (k+1)-th iteration. This is an estimate of the parameter (radiation dose rate) for the i-th time window in the k-th iteration. The learning rate controls the magnitude of each update step. Log-likelihood function pairs The gradient.

[0048] The update process will continue iteratively until a convergence condition is met. The convergence condition can be set as follows: the norm of the gradient is less than a certain threshold (e.g., ...). (or reach the maximum number of iterations).

[0049] It's worth noting that, besides gradient descent, Newton's method can accelerate convergence by utilizing the second derivative of the log-likelihood function. By considering both the gradient and the second derivative simultaneously, Newton's method can update the parameters more significantly in each iteration.

[0050] The W value for each time window is obtained by iteratively solving using numerical optimization methods (gradient descent or Newton's method). i The optimal estimate That is, the estimated radiation dose rate for each time window; ultimately, the optimal estimate for each time window. This corresponds to the preliminary dose rate estimate.

[0051] In one feasible implementation, after step S30, the method may further include: setting a prior distribution for the parameters of the Poisson distribution model, wherein the prior distribution is determined based on domain knowledge or historical data; constructing a posterior distribution based on the prior distribution and preliminary dose rate estimates; sampling the parameters of the Poisson distribution model using a Markov chain Monte Carlo algorithm based on the posterior distribution to obtain multiple parameter samples; performing a weighted average of the preliminary dose rate estimates for each time window based on the radiation dose rate estimates corresponding to each parameter sample to obtain a weighted average dose rate estimate, wherein the weights of the weighted average are determined based on the posterior probabilities of each parameter sample; and updating the final preliminary dose rate estimate for each time window based on the weighted average dose rate estimate.

[0052] It should be noted that the prior distribution is a hypothetical distribution of the parameters of the Poisson distribution model based on domain knowledge or historical data. By introducing the prior distribution, the accuracy of dose rate estimation can be improved by combining the experience of domain experts and historical data. In this embodiment, the selection of the prior distribution should reflect the possible range of parameter values ​​and their uncertainty.

[0053] Based on the prior distribution and preliminary dose rate estimates, a posterior distribution can be constructed. The posterior distribution integrates prior information and observational data, providing a more accurate estimate of the parameters. The process of constructing the posterior distribution typically involves applying Bayes' theorem, i.e., updating the prior distribution using observational data to obtain the posterior distribution.

[0054] Based on the posterior distribution, the Markov Chain Monte Carlo (MCMC) algorithm can be used to sample the parameters of the Poisson distribution model. The MCMC algorithm is a method that samples from complex probability distributions by constructing a Markov chain. In this implementation, the MCMC algorithm can be used to generate multiple parameter samples from the posterior distribution, reflecting the uncertainty of the parameters and their possible values. Based on the radiation dose rate estimates corresponding to each parameter sample, a weighted average can be performed on the preliminary dose rate estimates for each time window to obtain a weighted average dose rate estimate. The weights of the weighted average are determined based on the posterior probability of each parameter sample, i.e., the weights reflect the reliability of the parameter samples. By using a weighted average, the influence of different parameter samples on the dose rate estimation can be comprehensively considered, resulting in a more accurate dose rate estimate.

[0055] Finally, the final preliminary dose rate estimate for each time window can be updated based on the weighted average dose rate estimate. Applying the results of MCMC sampling to the dose rate estimation improves the accuracy and reliability of the estimation.

[0056] Step S40: Determine the environmental correction factor based on the temperature data sequence, humidity data sequence, and air pressure data sequence.

[0057] It should be noted that environmental parameters (such as temperature, humidity, and air pressure) can affect radiation dose rate measurements. For example, changes in temperature may alter sensor sensitivity, thus affecting the accuracy of the measurement results. Changes in humidity and air pressure may also affect radiation propagation and detection efficiency. Therefore, to obtain more accurate radiation dose rate estimates, corrections to these environmental factors are necessary.

[0058] In practice, empirical models or machine learning algorithms can be used to determine environmental correction factors. Empirical models may quantify the relationship between environmental factors such as temperature, humidity, and air pressure and radiation dose rate based on domain knowledge or experimental results. Machine learning algorithms can learn this relationship from historical data and use it to predict future data. In this embodiment, no restrictions are placed on the specific method for determining the environmental correction factor.

[0059] In one feasible implementation, step S40 may include: constructing a multiple linear regression model, wherein the independent variables of the model are temperature data series, humidity data series, and air pressure data series, and the dependent variable is an environmental correction factor; setting a loss function, wherein the loss function represents the difference between the predicted value and the actual value of the multiple linear regression model; calculating the gradient according to the loss function, and iteratively updating the parameters of the multiple linear regression model using the gradient descent method until a preset convergence condition is met or a preset upper limit for the number of iterations is reached; in each iteration, calculating the gradient of the loss function according to the current parameter value, and adjusting the current parameter value according to the gradient to gradually approach the optimal solution; using the iteratively updated parameter value as the regression coefficient of the multiple linear regression model. The environmental correction factor corresponding to each time window is calculated based on the regression coefficient and the temperature data series, humidity data series, and air pressure data series.

[0060] It should be noted that, Multiple linear regression is a commonly used statistical method to describe the linear relationship between a dependent variable and multiple independent variables. In this embodiment, temperature, humidity, and air pressure are used as independent variables, and an environmental correction factor is used as the dependent variable. By constructing a multiple linear regression model, the influence of these environmental factors on radiation dose rate measurement can be quantified. The loss function is used to evaluate the predictive performance of the model; it represents the difference between the model's predicted values ​​and the actual observed values. Commonly used loss functions include mean squared error (MSE) and root mean square error (RMSE). In this embodiment, an appropriate loss function can be selected to evaluate the model's predictive accuracy.

[0061] Gradient descent is a commonly used optimization algorithm for solving problems that minimize a loss function. It calculates the gradient of the loss function with respect to the model parameters and updates the parameter values ​​in the opposite direction of the gradient, thus gradually approaching the optimal solution. In each iteration, the gradient of the loss function is calculated based on the current parameter values, and the parameter values ​​are adjusted accordingly until a preset convergence condition is met (e.g., the norm of the gradient is less than a certain threshold) or a preset maximum number of iterations is reached. The updated parameter values ​​are the regression coefficients of the multiple linear regression model. The formula for calculating the gradient is:

[0062] in, This is a bias term. , , It is the regression coefficient. , , These are the temperature, humidity, and air pressure data for the i-th sample, respectively, where n is the number of samples. These are actual environmental corrective factors.

[0063] Based on regression coefficients and temperature, humidity, and air pressure data series, environmental correction factors corresponding to each time window can be calculated. These environmental correction factors reflect the degree of correction applied to radiation dose rate measurements under different environmental conditions. By applying these environmental correction factors to the initial dose rate estimate, a more accurate radiation dose rate estimate can be obtained, thereby improving the accuracy and reliability of the measurement.

[0064] Step S50: Correct the preliminary dose rate estimate for each time window according to the environmental correction factor to obtain the corrected dose rate estimate.

[0065] It should be noted that the environmental correction factor is introduced to correct for the potential impact of environmental factors on radiation dose rate measurements. Fluctuations in environmental factors such as temperature, humidity, and air pressure can cause slight changes in the performance of measuring equipment, thus affecting the accuracy of radiation dose rate measurements. To obtain a more accurate dose rate estimate, these environmental factors must be taken into account, and the initial dose rate estimate must be adjusted accordingly.

[0066] In practice, the preliminary dose rate estimate for each time window can be multiplied by an environmental correction factor to obtain a corrected dose rate estimate. This process essentially involves weighting the preliminary estimate, and the weighting coefficient is the environmental correction factor. The magnitude of this factor depends on environmental factors such as temperature, humidity, and air pressure within the current time window, reflecting the combined impact of these factors on radiation dose rate measurement.

[0067] Step S60: Input the corrected dose rate estimate, the average dose rate and the maximum dose rate within each time window into the support vector machine model to predict the nuclear radiation dose rate data and obtain the future nuclear radiation dose rate prediction results.

[0068] It should be noted that Support Vector Machine (SVM) is a supervised learning model commonly used for classification and regression analysis. In this embodiment, the SVM model is used to predict future nuclear radiation dose rates based on historical dose rate data (corrected dose rate estimates, average dose rate, and maximum dose rate). SVM achieves classification or regression analysis of data by constructing a hyperplane in a high-dimensional space to maximize the margin between two classes of samples. In regression problems, the goal of SVM is to find a function that, given input features, can predict the value of the target variable as accurately as possible.

[0069] To build an SVM model, a training dataset is first required. In this embodiment, the training dataset includes corrected dose rate estimates, average dose rates, and maximum dose rates for multiple historical time windows, as well as the actual nuclear radiation dose rates over a period of time after these time windows. This data is used to train the SVM model, enabling it to learn the relationship between dose rate data and time window features.

[0070] During training, the SVM algorithm attempts to find an optimal hyperplane that best fits the dose rate variation trend in the training data. To find this optimal hyperplane, the SVM algorithm typically employs a method called the kernel trick, mapping the input features to a high-dimensional feature space that is linearly separable within this space. Commonly used kernel functions include linear kernels, polynomial kernels, and radial basis function (RBF) kernels. In this embodiment, a suitable kernel function can be selected to construct the SVM model.

[0071] After training, the SVM model can be used to predict future nuclear radiation dose rates. For each new time window, its corrected dose rate estimate, average dose rate, and maximum dose rate can be calculated, and these features are input into the trained SVM model. The model will output a predicted nuclear radiation dose rate value based on the input features, reflecting the expected radiation dose rate level over a future period.

[0072] This embodiment provides a dose rate measurement method based on Poisson maximum likelihood estimation. It acquires environmental nuclear radiation monitoring data and preprocesses the data to obtain a fused data sequence, which includes a nuclear radiation dose rate data sequence, a temperature data sequence, a humidity data sequence, and a pressure data sequence. The fused data sequence is divided into multiple time windows, and the envelope spectrum sequence of each time window is determined. The radiation dose rate of each time window's envelope spectrum sequence is estimated using Poisson maximum likelihood estimation to obtain a preliminary dose rate estimate for each time window. An environmental correction factor is determined based on the temperature, humidity, and pressure data sequences. The preliminary dose rate estimate for each time window is corrected based on the environmental correction factor to obtain a corrected dose rate estimate. The corrected dose rate estimate, along with the average and maximum dose rates within each time window, is input into a support vector machine model to predict the nuclear radiation dose rate data, resulting in a predicted future nuclear radiation dose rate. By introducing the Poisson maximum likelihood estimation algorithm to update the optimal dose rate estimate in real time, the accuracy and response speed of dose rate estimation are improved.

[0073] Based on the first embodiment of this application, in the second embodiment of this application, the content that is the same as or similar to that in the first embodiment described above can be referred to the above description, and will not be repeated hereafter. Based on this, please refer to... Figure 2 Step S10 includes steps S101 to S104: Step S101: Obtain environmental nuclear radiation monitoring data, and perform outlier detection and remove outlier data points based on the isolated forest algorithm to obtain environmental nuclear radiation monitoring data after removing outliers.

[0074] It should be noted that environmental nuclear radiation monitoring data may be affected by various factors, resulting in outliers. If these outliers are not removed, they may negatively impact subsequent dose rate estimation and prediction results. Therefore, after acquiring environmental nuclear radiation monitoring data, it is essential to first detect and process outliers.

[0075] Isolation Forest is a machine learning algorithm commonly used for outlier detection. Based on the "majority rules" principle, it recursively partitions the data space by randomly selecting a feature and a splitting value until every data point is isolated. Because outliers are sparsely distributed in the data space, they are usually isolated earlier. Based on this characteristic, the Isolation Forest algorithm can effectively identify outliers in a dataset.

[0076] In one feasible implementation, step S101, "detecting outliers and removing outlier data points from the environmental nuclear radiation monitoring data based on the isolated forest algorithm to obtain environmental nuclear radiation monitoring data after removing outliers," may include: setting a training set containing multiple sample data, wherein each sample data includes environmental nuclear radiation monitoring data and its corresponding label, the label indicating whether the sample data is outlier; extracting features from each sample data in the training set to obtain a feature vector for each sample data; inputting the feature vector into the isolated forest algorithm to construct multiple isolated trees, wherein each isolated tree randomly selects a portion of the feature vectors for node splitting until a stopping condition is met; and statistically analyzing each sample data in the isolated forest algorithm. The path lengths in multiple isolated trees are calculated, and anomaly scores are calculated for each sample data based on the path lengths. The sample data in the training set are sorted according to the anomaly scores, and the top N samples with the highest scores are selected as anomaly samples, where N is a preset number of anomaly samples. A preset anomaly detection model is trained using these anomaly samples, and the model parameters are adjusted until convergence conditions are met, resulting in a target anomaly detection model. The environmental nuclear radiation monitoring data is input into the target anomaly detection model to obtain anomaly scores for each data point. An anomaly data threshold is determined based on the anomaly scores, and data points with anomaly scores higher than the threshold are removed as anomaly data points, resulting in environmental nuclear radiation monitoring data after anomaly removal.

[0077] It should be noted that the training set containing multiple sample data can be derived from historical data, where each sample data includes environmental nuclear radiation monitoring data and its corresponding label. The labels are provided by domain experts based on experience or actual conditions to indicate whether the sample data is anomaly. By extracting features from the training set, feature vectors for each sample data can be obtained. These feature vectors reflect key information from the environmental nuclear radiation monitoring data. Subsequently, the feature vectors are input into the Isolation Forest algorithm to construct multiple isolated trees. Each isolated tree randomly selects a portion of the feature vectors for node splitting until a stopping condition is met, such as reaching a preset tree depth or the number of samples contained in a node falling below a certain threshold.

[0078] During the construction of the isolation trees, the algorithm calculates the path length of each sample data point across multiple isolation trees and then calculates an anomaly score for each sample data point based on the path length. The anomaly score reflects the degree to which a sample data point is isolated; a higher score indicates that the sample data is more likely to be an outlier. The formula for calculating the anomaly score is:

[0079] in, It is a sample Abnormal scores, It is a sample Average path length across all isolated trees It is a constant used for standardization, representing the expected path length for a sample dataset size n.

[0080] Next, the sample data in the training set are sorted according to the anomaly scores, and the top N samples are selected as anomaly samples. N is a preset number of anomaly samples, which can be adjusted according to actual needs. These anomaly samples are used to train a preset anomaly detection model. By continuously adjusting the model parameters until convergence conditions are met, the target anomaly detection model is obtained. This model can be used to identify anomalies in new environmental nuclear radiation monitoring data. Finally, the new environmental nuclear radiation monitoring data is input into the target anomaly detection model to obtain an anomaly score for each data point. An anomaly threshold is determined based on the anomaly scores, and data points with anomaly scores higher than the threshold are removed as anomalies, thus obtaining the environmental nuclear radiation monitoring data after anomaly removal. This process helps improve the accuracy and reliability of subsequent dose rate estimation and prediction.

[0081] Step S102: Perform time-series interpolation on the environmental nuclear radiation monitoring data after removing outliers based on the LSTM algorithm to obtain the environmental nuclear radiation monitoring data sequence.

[0082] It should be noted that after obtaining environmental radiation monitoring data after removing outliers, some missing values ​​or irregular sampling may be found. These incomplete data may affect the accuracy of subsequent dose rate estimation and prediction. To address this issue, a Long Short-Term Memory (LSTM) algorithm can be used for temporal imputation. LSTM is a special type of Recurrent Neural Network (RNN) that effectively solves the shortcomings of traditional RNNs in dealing with long-term dependencies by introducing mechanisms such as input gates, forget gates, and output gates. LSTM can capture long-term dependencies in the data sequence and accurately estimate missing values ​​accordingly. During temporal imputation, LSTM predicts missing data points based on known data points, thus obtaining a complete and regularly sampled environmental radiation monitoring data sequence.

[0083] Specifically, environmental radiation monitoring data after outlier removal can be input into an LSTM network. The network learns the dependencies between data points based on the temporal characteristics of the input data. Then, for each missing data point, the LSTM predicts its value based on the known data points before and after it. This process iterates through the entire data sequence until all missing values ​​are imputed. Ultimately, a complete and regularly sampled environmental radiation monitoring data sequence is obtained, which can be used for subsequent dose rate estimation and prediction. By introducing the LSTM algorithm for temporal imputation, the problem of incomplete data can be effectively solved, improving the accuracy and reliability of subsequent processing.

[0084] Step S103: Determine the nuclear radiation dose rate data sequence, temperature data sequence, humidity data sequence, and air pressure data sequence based on the environmental nuclear radiation monitoring data sequence.

[0085] It is important to note that after obtaining complete and regularly sampled environmental nuclear radiation monitoring data sequences, it is necessary to extract key information related to the nuclear radiation dose rate. This information includes nuclear radiation dose rate data sequences, temperature data sequences, humidity data sequences, and air pressure data sequences. These data sequences reflect the changes in the nuclear radiation dose rate in the environment and the changes in related environmental factors, respectively. By analyzing and processing these data sequences, we can further understand the relationship between nuclear radiation dose rate and environmental factors, providing a more accurate and reliable basis for subsequent dose rate estimation and prediction.

[0086] In practical implementation, appropriate data processing techniques can be employed to extract these data sequences. For example, a sliding window technique can be used to segment the data sequence, allowing for independent analysis of the data within each time window. Within each time window, specific algorithms or models can be used to extract nuclear radiation dose rate data, temperature data, humidity data, and air pressure data. This data will be organized into corresponding data sequences to facilitate subsequent dose rate estimation and prediction. By accurately extracting these data sequences, the accuracy and reliability of dose rate measurements can be further improved.

[0087] Step S104: Perform multi-source data normalization and fusion on the nuclear radiation dose rate data sequence, temperature data sequence, humidity data sequence and air pressure data sequence to obtain a fused data sequence.

[0088] It should be noted that the purpose of multi-source data normalization and fusion of nuclear radiation dose rate data series, temperature data series, humidity data series, and air pressure data series is to eliminate the dimensional differences between different data series, improve the consistency and comparability of the data, and thus provide a more accurate and reliable data foundation for subsequent dose rate estimation and prediction.

[0089] Multi-source data normalization refers to transforming data with different dimensions and value ranges to the same scale, enabling them to be compared and analyzed under the same standard. In this embodiment, a min-max normalization method can be used to scale each value in the data sequence to between 0 and 1. Specifically, for each data sequence, its maximum and minimum values ​​can be found, and then each value in the data sequence is normalized according to the following formula: Normalized value = (Original value - Minimum value) / (Maximum value - Minimum value). In this way, the values ​​in different data sequences are transformed to the same scale, facilitating subsequent data fusion and analysis.

[0090] Data fusion refers to integrating information from multiple data sequences to extract more comprehensive and accurate data. In this embodiment, a weighted average method can be used for data fusion. Specifically, each data sequence can be assigned a weight reflecting its importance in dose rate estimation and prediction. Then, based on these weights, the values ​​in each data sequence are weighted and averaged to obtain the fused data sequence. The fused data sequence will contain more comprehensive and accurate information, providing a more reliable foundation for subsequent dose rate estimation and prediction.

[0091] By normalizing and fusing multi-source data, a fused data sequence containing nuclear radiation dose rate data, temperature data, humidity data, and air pressure data can be obtained. This fused data sequence will be used in subsequent dose rate estimation and prediction steps to further improve the accuracy and reliability of dose rate measurement.

[0092] In this embodiment, environmental nuclear radiation monitoring data is acquired, and outlier data points are detected and removed using the isolated forest algorithm to obtain outlier-free environmental nuclear radiation monitoring data. The outlier-free environmental nuclear radiation monitoring data is then time-series interpolated using the LSTM algorithm to obtain an environmental nuclear radiation monitoring data sequence. Based on the environmental nuclear radiation monitoring data sequence, nuclear radiation dose rate data sequence, temperature data sequence, humidity data sequence, and air pressure data sequence are determined. Multi-source data normalization and fusion are then performed on the nuclear radiation dose rate data sequence, temperature data sequence, humidity data sequence, and air pressure data sequence to obtain a fused data sequence. By introducing the isolated forest algorithm and the LSTM algorithm, and combining them with multi-source data normalization and fusion processing, the accuracy and reliability of environmental nuclear radiation monitoring data can be effectively improved, thereby further enhancing the accuracy and reliability of dose rate measurement.

[0093] It should be noted that the above examples are only for understanding this application and do not constitute a limitation on the dose rate measurement method based on Poisson maximum likelihood estimation in this application. Any simple modifications based on this technical concept are within the scope of protection of this application.

[0094] This application also provides a dose rate measurement device based on Poisson maximum likelihood estimation, please refer to... Figure 3 The dose rate measurement device based on Poisson maximum likelihood estimation includes: The preprocessing module 10 is used to acquire environmental nuclear radiation monitoring data and preprocess the environmental nuclear radiation monitoring data to obtain a fused data sequence, wherein the fused data sequence includes a nuclear radiation dose rate data sequence, a temperature data sequence, a humidity data sequence, and an air pressure data sequence.

[0095] The determination module 20 is used to divide the fused data sequence into multiple time windows and determine the envelope spectrum sequence of each time window.

[0096] The estimation module 30 is used to estimate the radiation dose rate of the envelope spectrum sequence of each time window using the Poisson maximum likelihood estimation method, so as to obtain the preliminary dose rate estimate for each time window.

[0097] The determining module 20 also determines an environmental correction factor based on the temperature data sequence, humidity data sequence, and air pressure data sequence.

[0098] The correction module 40 is used to correct the preliminary dose rate estimate for each time window based on the environmental correction factor to obtain the corrected dose rate estimate.

[0099] The prediction module 50 is used to input the corrected dose rate estimate, the average dose rate and the maximum dose rate in each time window into the support vector machine model to predict the nuclear radiation dose rate data and obtain the future nuclear radiation dose rate prediction result.

[0100] The dose rate measurement device based on Poisson maximum likelihood estimation provided in this application, employing the dose rate measurement method based on Poisson maximum likelihood estimation in the above embodiments, can solve the technical problem that traditional dose rate estimation methods have large errors, resulting in inaccurate prediction results. Compared with the prior art, the beneficial effects of the dose rate measurement device based on Poisson maximum likelihood estimation provided in this application are the same as those of the dose rate measurement method based on Poisson maximum likelihood estimation provided in the above embodiments, and other technical features in the dose rate measurement device based on Poisson maximum likelihood estimation are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.

[0101] This application provides a dose rate measurement device based on Poisson maximum likelihood estimation. The dose rate measurement device based on Poisson maximum likelihood estimation includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the dose rate measurement method based on Poisson maximum likelihood estimation in the above embodiment 1.

[0102] The following is for reference. Figure 4 This document illustrates a schematic diagram of a dose rate measurement device suitable for implementing the Poisson maximum likelihood estimation-based method in the embodiments of this application. The dose rate measurement device based on Poisson maximum likelihood estimation in the embodiments of this application may include, but is not limited to, mobile terminals such as mobile phones, laptops, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Description), PMPs (Portable Media Players), and in-vehicle terminals (e.g., in-vehicle navigation terminals), as well as fixed terminals such as digital TVs and desktop computers. Figure 4 The dose rate measurement device based on Poisson maximum likelihood estimation shown is merely an example and should not impose any limitation on the functionality and scope of use of the embodiments of this application.

[0103] like Figure 4As shown, a dose rate measurement device based on Poisson maximum likelihood estimation may include a processing unit 1001 (e.g., a central processing unit, a graphics processor, etc.), which can perform various appropriate actions and processes according to a program stored in ROM (Read Only Memory) 1002 or a program loaded from storage device 1003 into RAM (Random Access Memory) 1004. RAM 1004 also stores various programs and data required for the operation of the dose rate measurement device based on Poisson maximum likelihood estimation. The processing unit 1001, ROM 1002, and RAM 1004 are interconnected via bus 1005. Input / output (I / O) interface 1006 is also connected to the bus. Typically, the following systems can be connected to I / O interface 1006: input devices 1007 including, for example, touchscreens, touchpads, keyboards, mice, image sensors, microphones, accelerometers, gyroscopes, etc.; output devices 1008 including, for example, LCDs (Liquid Crystal Displays), speakers, vibrators, etc.; storage devices 1003 including, for example, magnetic tapes, hard disks, etc.; and communication devices 1009. Communication device 1009 allows the dose rate measurement device based on Poisson maximum likelihood estimation to communicate wirelessly or wiredly with other devices to exchange data. Although the figure shows a dose rate measurement device based on Poisson maximum likelihood estimation with various systems, it should be understood that it is not required to implement or have all the systems shown. More or fewer systems can be implemented or have alternatively.

[0104] Specifically, according to the embodiments disclosed in this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments disclosed in this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from storage device 1003, or installed from ROM 1002. When the computer program is executed by processing device 1001, it performs the functions defined in the methods of the embodiments disclosed in this application.

[0105] The dose rate measurement device based on Poisson maximum likelihood estimation provided in this application, employing the dose rate measurement method based on Poisson maximum likelihood estimation in the above embodiments, can solve the technical problem that traditional dose rate estimation methods have large errors, resulting in inaccurate prediction results. Compared with the prior art, the beneficial effects of the dose rate measurement device based on Poisson maximum likelihood estimation provided in this application are the same as those of the dose rate measurement method based on Poisson maximum likelihood estimation provided in the above embodiments, and other technical features in this dose rate measurement device based on Poisson maximum likelihood estimation are the same as those disclosed in the previous embodiment method, and will not be repeated here.

[0106] It should be understood that the various parts disclosed in this application can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples.

[0107] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0108] This application provides a computer-readable storage medium having computer-readable program instructions (i.e., a computer program) stored thereon, the computer-readable program instructions being used to perform the dose rate measurement method based on Poisson maximum likelihood estimation in the above embodiments.

[0109] The computer-readable storage medium provided in this application may be, for example, a USB flash drive, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, RAM (Random Access Memory), ROM (Read Only Memory), EPROM (Erasable Programmable Read Only Memory or Flash Memory), optical fibers, CD-ROM (CD-Read Only Memory), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, system, or device. The program code contained on the computer-readable storage medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination thereof.

[0110] The aforementioned computer-readable storage medium may be included in a dose rate measurement device based on Poisson maximum likelihood estimation; or it may exist independently and not assembled into a dose rate measurement device based on Poisson maximum likelihood estimation.

[0111] The aforementioned computer-readable storage medium carries one or more programs that, when executed by a dose rate measurement device based on Poisson maximum likelihood estimation, cause the dose rate measurement device based on Poisson maximum likelihood estimation to: acquire environmental nuclear radiation monitoring data, and preprocess the environmental nuclear radiation monitoring data to obtain a fused data sequence, wherein the fused data sequence includes a nuclear radiation dose rate data sequence, a temperature data sequence, a humidity data sequence, and a pressure data sequence; divide the fused data sequence into multiple time windows, and determine the envelope spectrum sequence of each time window; and through Poisson maximum likelihood estimation... The slack maximum likelihood estimation method is used to estimate the radiation dose rate of the envelope spectrum sequence for each time window, obtaining a preliminary dose rate estimate for each time window. An environmental correction factor is determined based on the temperature, humidity, and air pressure data sequences. The preliminary dose rate estimate for each time window is corrected based on the environmental correction factor, resulting in a corrected dose rate estimate. The corrected dose rate estimate, along with the average and maximum dose rates within each time window, are input into a support vector machine model to predict the nuclear radiation dose rate data, yielding a future nuclear radiation dose rate prediction result.

[0112] Computer program code for performing the operations of this application can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including LAN (Local Area Network) or WAN (Wide Area Network)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0113] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0114] The modules described in the embodiments of this application can be implemented in software or hardware. The names of the modules do not necessarily limit the functionality of the unit itself.

[0115] The readable storage medium provided in this application is a computer-readable storage medium that stores computer-readable program instructions (i.e., a computer program) for executing the dose rate measurement method based on Poisson maximum likelihood estimation described above. This addresses the technical problem that traditional dose rate estimation methods suffer from large errors, leading to inaccurate prediction results. Compared to the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as those of the dose rate measurement method based on Poisson maximum likelihood estimation provided in the above embodiments, and will not be elaborated upon here.

[0116] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the dose rate measurement method based on Poisson maximum likelihood estimation as described above.

[0117] The computer program product provided in this application can solve the technical problem that traditional dose rate estimation methods have large errors, resulting in inaccurate prediction results. Compared with the prior art, the beneficial effects of the computer program product provided in this application are the same as those of the dose rate measurement method based on Poisson maximum likelihood estimation provided in the above embodiments, and will not be repeated here.

[0118] The above description is only a part of the embodiments of this application and does not limit the patent scope of this application. All equivalent structural transformations made under the technical concept of this application and using the contents of the specification and drawings of this application, or direct / indirect applications in other related technical fields, are included in the patent protection scope of this application.

Claims

1. A dose rate measurement method based on Poisson maximum likelihood estimation, characterized in that, The method includes: Environmental nuclear radiation monitoring data is acquired and preprocessed to obtain a fused data sequence, wherein the fused data sequence includes a nuclear radiation dose rate data sequence, a temperature data sequence, a humidity data sequence, and an air pressure data sequence. The fused data sequence is divided into multiple time windows, and the envelope spectrum sequence of each time window is determined. The radiation dose rate is estimated for the envelope spectrum sequence of each time window using the Poisson maximum likelihood estimation method, thus obtaining the preliminary dose rate estimate for each time window. The environmental correction factor is determined based on the temperature data sequence, humidity data sequence, and air pressure data sequence. The preliminary dose rate estimate for each time window is corrected based on the environmental correction factor to obtain the corrected dose rate estimate. The corrected dose rate estimate, the average dose rate within each time window, and the maximum dose rate are input into the support vector machine model to predict the nuclear radiation dose rate data, thereby obtaining the future nuclear radiation dose rate prediction results.

2. The method as described in claim 1, characterized in that, The acquisition of environmental nuclear radiation monitoring data, followed by preprocessing of the environmental nuclear radiation monitoring data to obtain a fused data sequence, includes: Environmental nuclear radiation monitoring data is acquired, and outlier data points are removed from the environmental nuclear radiation monitoring data based on the isolated forest algorithm to obtain environmental nuclear radiation monitoring data after outlier removal. The environmental nuclear radiation monitoring data after outlier removal is interpolated using the LSTM algorithm to obtain an environmental nuclear radiation monitoring data sequence. Based on the environmental nuclear radiation monitoring data sequence, nuclear radiation dose rate data sequence, temperature data sequence, humidity data sequence, and air pressure data sequence are determined; The nuclear radiation dose rate data sequence, temperature data sequence, humidity data sequence, and air pressure data sequence are subjected to multi-source data normalization and fusion to obtain a fused data sequence.

3. The method as described in claim 2, characterized in that, The method of detecting and removing outlier data points from the environmental nuclear radiation monitoring data based on the isolated forest algorithm, resulting in environmental nuclear radiation monitoring data after outlier removal, includes: A training set containing multiple sample data is set up, wherein each sample data includes environmental nuclear radiation monitoring data and its corresponding label, and the label indicates whether the sample data is abnormal data; Feature extraction is performed on each sample data in the training set to obtain the feature vector of each sample data; The feature vectors are input into the isolated forest algorithm to construct multiple isolated trees. In each isolated tree, a portion of the feature vectors are randomly selected for node splitting until the stopping condition is met. The path length of each sample data in the multiple isolated trees is calculated, and the anomaly score of each sample data is calculated based on the path length. The sample data in the training set are sorted according to the abnormal scores, and the top N samples with the highest scores are selected as abnormal samples, where N is the preset number of abnormal samples. The abnormal samples are used to train the preset outlier detection model and adjust the model parameters until the convergence condition is met, thus obtaining the target outlier detection model. The environmental nuclear radiation monitoring data is input into the target anomaly detection model to obtain an anomaly score for each data point; An abnormal data threshold is determined based on the abnormal score. Data points with abnormal scores higher than the abnormal data threshold are removed as abnormal data points to obtain environmental nuclear radiation monitoring data after removing outliers.

4. The method as described in claim 1, characterized in that, The step of dividing the fused data sequence into multiple time windows and determining the envelope spectrum sequence of each time window includes: The length of the time window is determined based on the time series characteristics of the fused data sequence; The fused data sequence is divided according to a determined time window length to obtain multiple consecutive and non-overlapping time windows; Empirical mode decomposition is performed on the fused data sequence within each time window to obtain the set of intrinsic mode functions; Perform a Hilbert transform on each intrinsic mode function in the intrinsic mode function set to obtain the instantaneous amplitude sequence of each intrinsic mode function; The instantaneous amplitude sequences of all intrinsic mode functions within each time window are superimposed to obtain the envelope spectrum sequence of each time window.

5. The method as described in claim 1, characterized in that, The radiation dose rate is estimated by using the Poisson maximum likelihood estimation method to calculate the envelope spectrum sequence for each time window, resulting in a preliminary dose rate estimate for each time window, including: A Poisson distribution model is constructed, wherein the parameters of the Poisson distribution model are set according to the envelope spectrum sequence of each time window; Define an objective function, wherein the objective function is the likelihood function of the Poisson distribution model, and the objective function is calculated based on the envelope spectrum sequence of each time window and the parameters of the Poisson distribution model; The objective function is solved iteratively using a numerical optimization algorithm until a preset convergence condition or upper limit of the number of iterations is reached, so as to obtain the optimal estimate of the radiation dose rate for each time window. In the process of iterative solution, the gradient or second derivative of the objective function is calculated based on the current parameter estimate, and the current parameter estimate is updated based on the gradient or second derivative. The optimal estimate is used as the initial dose rate estimate for each time window.

6. The method as described in claim 1, characterized in that, After estimating the radiation dose rate of the envelope spectrum sequence for each time window using the Poisson maximum likelihood estimation method to obtain the preliminary dose rate estimate for each time window, the method further includes: A prior distribution is set for the parameters of the Poisson distribution model, wherein the prior distribution is determined based on domain knowledge or historical data; Based on the prior distribution and the preliminary dose rate estimate, a posterior distribution is constructed; Based on the posterior distribution, the parameters of the Poisson distribution model are sampled using the Markov chain Monte Carlo algorithm to obtain multiple parameter samples; The initial dose rate estimate for each time window is weighted and averaged according to the radiation dose rate estimate corresponding to each parameter sample to obtain the weighted average dose rate estimate. The weight of the weighted average is determined according to the posterior probability of each parameter sample. The final preliminary dose rate estimate for each time window is updated based on the weighted average dose rate estimate.

7. The method as described in claim 1, characterized in that, The step of determining the environmental correction factor based on the temperature data sequence, humidity data sequence, and air pressure data sequence includes: A multiple linear regression model was constructed, in which the independent variables are temperature data series, humidity data series, and air pressure data series, and the dependent variable is the environmental correction factor; Define a loss function, whereby the loss function represents the difference between the predicted value and the actual value of the multiple linear regression model; The gradient is calculated based on the loss function, and the parameters of the multiple linear regression model are iteratively updated using the gradient descent method until the preset convergence condition is met or the preset maximum number of iterations is reached. In each iteration, the gradient of the loss function is calculated based on the current parameter value, and the current parameter value is adjusted according to the gradient to gradually approach the optimal solution; The updated parameter values ​​are used as the regression coefficients of the multiple linear regression model. The environmental correction factor corresponding to each time window is calculated based on the regression coefficients and the temperature, humidity, and air pressure data sequences.

8. A dose rate measurement device based on Poisson maximum likelihood estimation, characterized in that, The dose rate measurement device based on Poisson maximum likelihood estimation includes: A preprocessing module is used to acquire environmental nuclear radiation monitoring data and preprocess the environmental nuclear radiation monitoring data to obtain a fused data sequence, wherein the fused data sequence includes a nuclear radiation dose rate data sequence, a temperature data sequence, a humidity data sequence, and an air pressure data sequence. The determination module is used to divide the fused data sequence into multiple time windows and determine the envelope spectrum sequence of each time window; The estimation module is used to estimate the radiation dose rate of the envelope spectrum sequence for each time window using the Poisson maximum likelihood estimation method, and obtain the preliminary dose rate estimate for each time window. The determining module also determines an environmental correction factor based on the temperature data sequence, humidity data sequence, and air pressure data sequence; The correction module is used to correct the preliminary dose rate estimate for each time window based on the environmental correction factor, so as to obtain the corrected dose rate estimate. The prediction module is used to input the corrected dose rate estimate, the average dose rate and the maximum dose rate within each time window into the support vector machine model to predict the nuclear radiation dose rate data and obtain the future nuclear radiation dose rate prediction results.

9. A dose rate measurement device based on Poisson maximum likelihood estimation, characterized in that, The dose rate measurement device based on Poisson maximum likelihood estimation includes: a memory, a processor, and a dose rate measurement program based on Poisson maximum likelihood estimation stored in the memory and executable on the processor, the dose rate measurement program based on Poisson maximum likelihood estimation being configured to implement the dose rate measurement method based on Poisson maximum likelihood estimation as described in any one of claims 1 to 7.

10. A storage medium, characterized in that, The storage medium stores a dose rate measurement program based on Poisson maximum likelihood estimation, which, when executed by a processor, implements the dose rate measurement method based on Poisson maximum likelihood estimation as described in any one of claims 1 to 7.

Images