A smart system for personal dose acquisition and monitoring of nuclear radiation

Abstract

This invention discloses an intelligent personal radiation dose acquisition and monitoring system, specifically relating to the field of nuclear radiation dose monitoring. It addresses the problems of existing dose monitoring systems, such as coarse radiation signal analysis, delayed exposure risk identification, and difficulty in prospective assessment under complex radiation fields and dynamic body positioning conditions. The system performs refined waveform analysis on the raw pulse signals acquired by the personal detection unit, dynamically corrects scattering characteristics and energy deposition based on real-time personal positioning data, constructs an intermediate energy deposition spectrum reflecting the interaction between radiation and human tissue, and predicts the exposure situation within future monitoring time windows based on the evolution of spectral characteristics. This enables intelligent assessment and early warning of personal radiation exposure risks, improves the real-time performance and reliability of nuclear radiation protection, is suitable for nuclear emergency response and high-risk operational scenarios, and supports continuous monitoring and rapid deployment.

Description

Technical Field

[0001] This invention relates to the field of nuclear radiation dose monitoring technology, and more specifically, to an intelligent system for acquiring and monitoring individual nuclear radiation doses. Background Technology

[0002] Current personal radiation dose monitoring is mainly used in scenarios such as nuclear emergency response, radioactive accident rescue, nuclear facility inspection, and high-risk industrial site operations. In these environments, individuals are typically exposed to radiation fields with complex spatial distribution, rapid intensity changes, uncertain incident directions, and significant influence from body posture and movement. Traditional personal dosimeters often employ post-hoc statistical methods using cumulative dose or dose rate, typically based on the average signal output from the detector or simple counting results. This approach struggles to reflect the differences in energy deposition during the actual interaction between radiation particles and human tissue, and also fails to promptly identify the dynamic trend of the radiation field evolving from low-risk to high-risk.

[0003] In addition, complex radiation fields are often accompanied by phenomena such as pulse accumulation and changes in the dominant scattering mechanism. Simply relying on dose rate thresholds to trigger early warnings can easily lead to false alarms or delayed responses, which cannot meet the needs of prospective assessment and dynamic management of individual exposure risks.

[0004] Therefore, there is an urgent need for an intelligent personal dose acquisition and monitoring technology that can perform detailed analysis of radiation signals under personal wear and movement conditions, characterize energy deposition features in combination with body position factors, and predict changes in exposure status on a short time scale, so as to improve the real-time, reliability and proactiveness of nuclear radiation protection. Summary of the Invention

[0005] To overcome the aforementioned deficiencies of the prior art, embodiments of the present invention provide an intelligent personal dose acquisition and monitoring system for nuclear radiation to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] A smart system for acquiring and monitoring personal dose of nuclear radiation, comprising:

[0008] The acquisition module is used to acquire raw pulse signals through the personal detection unit and convert them into digital waveform data, simultaneously acquire real-time personal body position data, and jointly establish a raw data stream;

[0009] The waveform analysis module is used to dynamically configure pulse waveform separation parameters based on the noise statistics characteristics and body position data of the original data stream, and to parse discrete single-pulse event sequences from the digitized waveform data.

[0010] The correction module is used to extract the waveform feature parameters of each single pulse event and perform position-related Compton scattering correction on the feature parameters based on the body position data and the ratio of the pulse leading edge to the trailing edge time.

[0011] The spectrum construction module is used to input the corrected pulse characteristic parameters into the preset energy deposition response model to generate an intermediate energy deposition spectrum that reflects the interaction between radiation and human tissue.

[0012] The spectral prediction module is used to predict the spectral characteristics of intermediate energy deposition spectra within a future monitoring time window based on the characteristic evolution sequence of intermediate energy deposition spectra.

[0013] The dose monitoring module is used to assess an individual's exposure risk based on spectral feature predictions of intermediate energy deposition spectra.

[0014] As a further aspect of the present invention, the establishment of the original data stream in the acquisition module specifically includes:

[0015] The personal detection unit includes a plastic scintillator detector and a photomultiplier tube;

[0016] The scintillation light signal induced by radiation particles is collected and converted into an analog electrical pulse signal. The analog electrical pulse signal is sampled at equal intervals to convert the continuous voltage change into a discrete digital sequence, forming the original digital waveform data.

[0017] Simultaneously, real-time body position data of an individual's body axis relative to a set reference coordinate system is acquired, and a unified timestamp is assigned to the original digital waveform data and the real-time body position data. Based on this timestamp, the two types of data are aligned and spliced ​​to generate the original data stream.

[0018] As a further aspect of the present invention, the waveform analysis module dynamically configures pulse waveform separation parameters to parse discrete single-pulse event sequences from digitized waveform data, specifically including:

[0019] From the digitized waveform data of the original data stream, select a local waveform of the non-pulse signal segment, and calculate the amplitude standard deviation of the waveform sample as the real-time baseline noise level;

[0020] Extract the real-time angle between the axis of the personal detection unit and the direction of gravity from the raw data stream. Calculate the equivalent solid angle correction coefficient of the radiation incident direction relative to the sensitive volume of the detector based on the angle. Set the reference time window for pulse width discrimination based on the reciprocal of the equivalent solid angle correction coefficient.

[0021] The real-time baseline noise level is used as the reference threshold for pulse detection, and the equivalent solid angle correction coefficient is used as the position-related response scaling factor to proportionally adjust the reference threshold.

[0022] Using a reference threshold and a reference time window as pulse waveform separation parameters, the digitized waveform data is scanned point by point. When the amplitude of consecutive data points exceeds the reference threshold and falls within the reference time window, a complete single-pulse event waveform is captured and arranged according to the timestamp to form the single-pulse event sequence.

[0023] As a further aspect of the present invention, the correction module extracts waveform feature parameters for each single-pulse event, and performs position-related Compton scattering correction on the feature parameters based on body position data and the ratio of pulse leading edge to trailing edge time. Specifically, this includes:

[0024] The specific feature parameters are the peak amplitude, waveform integral area, and the ratio of the leading edge time to the trailing edge time of the pulse, which are calculated from the digital waveform of each single pulse event read from the single pulse event sequence.

[0025] Based on the real-time sine value of the angle between the axis of the personal detection unit and the direction of gravity, as well as the ratio of the leading and trailing edges, the classification of the current pulse event and the corresponding Compton scattering correction weight coefficient are determined according to the set event classification mapping relationship. Then, scalar multiplication is performed on the waveform integral area to obtain the corrected pulse integral area.

[0026] As a further aspect of the present invention, in the spectrum construction module, the energy deposition response model establishes a quantitative correspondence between the pulse signal amplitude and the actual deposited energy in human soft tissue by testing the response pulse amplitude of the personal detection unit to different known energy monoenergetic radiation sources, and combining the energy deposition share in the tissue equivalent material calculated by Monte Carlo simulation, thereby mapping the integral area of ​​each pulse to a discrete equivalent deposition energy range.

[0027] As a further aspect of the present invention, the generation of an intermediate energy deposition spectrum reflecting the interaction between radiation and human tissue in the spectrum construction module specifically includes:

[0028] The corrected pulse characteristic parameters are input into the energy deposition response model. The distribution of all input single-pulse events in the discrete equivalent deposition energy range is counted. The count value of each energy range is normalized to generate an intermediate energy deposition spectrum with equivalent deposition energy as the horizontal axis and normalized count rate as the vertical axis.

[0029] As a further aspect of the present invention, the spectral prediction module, based on the characteristic evolution sequence of the intermediate energy deposition spectrum, specifically includes predicting the spectral characteristics of the intermediate energy deposition spectrum within a future monitoring time window, including:

[0030] Calculate the spectral centroid energy of the intermediate energy deposition spectrum within the current monitoring time window to obtain the first spectral characteristic parameter;

[0031] Calculate the total count rate of single-pulse events within the same time window to obtain the second spectral characteristic parameters;

[0032] The proportion of event counts in the interval above the average energy in the intermediate energy deposition spectrum is calculated to obtain the third spectral characteristic parameters;

[0033] The first, second and third spectral feature parameters are combined into the current spectral feature state vector. The current spectral feature state vector and the historical state vector sequence stored in chronological order are input into the recursive state predictor.

[0034] The recursive state predictor fits a first-order dynamic model of the spectral feature state changing over time using a recursive least squares algorithm with an embedded forgetting factor.

[0035] The first-order dynamic model is used to perform forward prediction, calculate the predicted value of the spectral feature state of the future monitoring time window, apply boundary constraints and physical non-negativity constraints based on historical extreme values ​​to the predicted value, and output the final prediction result.

[0036] As a further aspect of the present invention, the dose monitoring module, based on the spectral feature prediction results of the intermediate energy deposition spectrum, specifically includes the following:

[0037] Obtain the characteristic predicted values ​​of the intermediate energy deposition spectrum. If both the predicted value of the spectral centroid energy and the predicted count rate show a monotonically increasing trend, it is determined that the individual is in a state of increased exposure risk within the future monitoring time window.

[0038] When the risk of exposure is rising, the incremental estimate of the individual's expected cumulative dose within the future monitoring time window is calculated based on the predicted spectral centroid energy and total count rate. The incremental estimate is compared with the incremental estimate based on the preset baseline, and the individual's exposure risk assessment result is output.

[0039] The technical effects and advantages of the intelligent personal dose acquisition and monitoring system for nuclear radiation of this invention are as follows:

[0040] This invention analyzes the raw pulse signal acquired by the personal detection unit into discrete single-pulse events and introduces body position data to dynamically correct waveform separation and scattering characteristics, enabling radiation signals to be stably and accurately identified and characterized even under complex posture and motion conditions. By constructing an intermediate energy deposition spectrum oriented towards equivalent energy deposition in human tissue, it avoids the systematic bias caused by dose conversion based solely on detector material response, improving the physical consistency of personal exposure assessment. This invention is not limited to passive statistics of the current dose, but achieves forward-looking judgment of the changing trend of personal exposure risk by predicting the evolution of spectral characteristics on a short timescale, thus providing early warning before significant dose accumulation. In application scenarios with complex radiation fields, rapidly changing incident directions, and strong signal randomness, this technical solution can effectively reduce false alarm and hysteresis risks, improve the real-time performance, reliability, and proactive prevention and control capabilities of personal radiation protection, and has good engineering practical value. Attached Figure Description

[0041] Figure 1 This is a schematic diagram of the structure of a personal dose intelligent acquisition and monitoring system for nuclear radiation according to the present invention. Detailed Implementation

[0042] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0043] Example 1

[0044] Figure 1 This invention discloses an intelligent personal dose acquisition and monitoring system for nuclear radiation, comprising:

[0045] The acquisition module is used to acquire raw pulse signals through the personal detection unit and convert them into digital waveform data, simultaneously acquire real-time personal body position data, and jointly establish a raw data stream;

[0046] The waveform analysis module is used to dynamically configure pulse waveform separation parameters based on the noise statistics characteristics and body position data of the original data stream, and to parse discrete single-pulse event sequences from the digitized waveform data.

[0047] The correction module is used to extract the waveform feature parameters of each single pulse event and perform position-related Compton scattering correction on the feature parameters based on the body position data and the ratio of the pulse leading edge to the trailing edge time.

[0048] The spectrum construction module is used to input the corrected pulse characteristic parameters into the preset energy deposition response model to generate an intermediate energy deposition spectrum that reflects the interaction between radiation and human tissue.

[0049] The spectral prediction module is used to predict the spectral characteristics of intermediate energy deposition spectra within a future monitoring time window based on the characteristic evolution sequence of intermediate energy deposition spectra.

[0050] The dose monitoring module is used to assess an individual's exposure risk based on spectral feature predictions of intermediate energy deposition spectra.

[0051] In the acquisition module, a raw data stream is established.

[0052] The personal detection unit explicitly employs a tightly coupled detection structure consisting of a plastic scintillator detector and a photomultiplier tube. The plastic scintillator is fixedly installed within the detection cavity of the wearable device. Its material is an organic scintillator material with short decay time and high light yield, enabling it to generate a scintillator light signal with intensity related to the deposited energy in a very short time when exposed to radiation particles such as gamma rays or beta particles. The photocathode of the photomultiplier tube maintains stable optical coupling with the light-emitting end of the plastic scintillator. Optical loss is reduced through optical adhesive or direct bonding, ensuring efficient conversion of the scintillator light into an electrical signal. The scintillator light signal triggered by the radiation particles entering the scintillator is amplified step-by-step within the photomultiplier tube, ultimately forming an analog electrical pulse signal at the output with an amplitude corresponding to the scintillator light intensity. This analog electrical pulse signal undergoes impedance matching and amplitude shaping by a front-end signal conditioning circuit before being sent to an analog-to-digital converter for equal-time sampling. The sampling interval is set to a fixed value, such as 10 nanoseconds or 20 nanoseconds, based on the pulse rise time and decay characteristics, thus fully covering the leading edge, peak, and trailing edge changes of a single pulse. By continuously sampling, the voltage change process of the analog electrical pulse over time is converted into a set of discrete digital sampling points arranged in time sequence, forming raw digital waveform data containing pulse accumulation, baseline drift and noise components.

[0053] To achieve synchronous correlation between radiation signals and individual body posture, a unified body posture reference coordinate system is clearly constructed to acquire real-time body posture data. This reference coordinate system uses the Earth's gravity direction as the global reference axis, with its vertical direction defined as the Z-axis, the horizontal direction pointing directly forward defined as the X-axis, and the direction perpendicular to the X-axis and located in the horizontal plane defined as the Y-axis, thus forming a fixed three-dimensional rectangular coordinate system. The posture sensing unit within the personal wearable device maintains a fixed installation orientation with the detection unit's axis. It continuously collects the components of gravitational acceleration along each axis through a built-in acceleration sensing element and calculates the angle between the individual's body axis and the gravity direction in real time based on these components. This body posture data is continuously output at a fixed sampling period and is time-stamped using the same clock source as the digitized waveform data, ensuring the consistency of the time reference for both types of data at the hardware level. During data acquisition, each segment of raw digitized waveform data and the corresponding body posture data is assigned a timestamp in a unified format, with the timestamp accuracy consistent with the analog-to-digital conversion sampling period. Subsequently, the original digitized waveform data and real-time body position data are aligned and spliced ​​point by point according to the timestamp. The waveform sampling point sequence within the same time slice is combined and stored with the corresponding body position parameters to form an original data stream containing waveform sampling data, body position information, and time stamp.

[0054] In the waveform analysis module, pulse waveform separation parameters are dynamically configured to parse discrete single-pulse event sequences from digitized waveform data.

[0055] The digitized waveform data acquired from the raw data stream is a continuously sampled time series. This series includes both effective pulse signals induced by radiated particles and baseline noise fluctuations formed under particle-free conditions. To accurately characterize the noise level under the current acquisition conditions, the continuous sampling points in the digitized waveform data are first identified, and segments without obvious pulse rising edge characteristics are selected as non-pulse signal segments. The selection of this segment is based on the characteristic that the waveform remains stable over a long period of time without continuous significant amplitude abrupt changes, avoiding the erroneous inclusion of pulse leading or trailing edges in noise calculations. For the set of sampling points within the non-pulse signal segments, their amplitude variations are directly statistically analyzed. By calculating the dispersion of the amplitude of each sampling point in this local waveform sample relative to the average amplitude, the amplitude standard deviation, which reflects the combined effects of electronic noise, environmental interference, and baseline drift under the current acquisition environment, is obtained. This amplitude standard deviation is determined and output as the real-time baseline noise level. Since the detector is a plastic scintillator with finite geometric dimensions and a fixed installation orientation, its response to radiation is not isotropic. When an individual's body position changes, the angle between the radiation incident direction and the detector's sensitive volume changes, which leads to changes in the effective action path length of the radiation particles in the scintillator, the energy deposition probability, and the response intensity of the effective trigger pulse. Therefore, it is necessary to introduce an equivalent solid angle correction coefficient through the body position angle.

[0056] The real-time angle information between the personal detection unit's axis and the direction of gravity is synchronously read from the raw data stream. This angle is continuously output by the body position sensing unit and kept in time synchronized with the waveform sampling data. This angle information is used to determine the detector's current spatial attitude relative to the incident radiation direction. Based on the pre-defined detector sensitive volume geometry and scintillator light-receiving characteristics, an equivalent solid angle correction coefficient relative to the detector sensitive volume is calculated based on the angle. This correction coefficient characterizes the change in the detector's effective response to incident radiation under the current body position. By default, the cosine of the angle is used as the correction coefficient. When the detector axis is nearly perpendicular to the incident radiation direction, the equivalent solid angle correction coefficient takes a smaller value; when the axis and incident direction tend to be aligned, the correction coefficient takes a larger value. Furthermore, based on the reciprocal relationship of the equivalent solid angle correction coefficient, a reference time window length for pulse width discrimination is set. This means the time window is widened when the detector's effective response weakens and narrowed when the effective response strengthens, ensuring the pulse width discrimination rule aligns with changes in body position and avoiding misjudgments caused by pulse broadening or compression due to changes in the incident direction. The real-time baseline noise level is directly used as the reference threshold for pulse detection. This threshold reflects the statistical scale of noise amplitude under the current sampling conditions, ensuring that pulse judgment is not triggered within the noise fluctuation range. To further incorporate the influence of body position on the detector's response amplitude, the obtained equivalent solid angle correction coefficient is used as a body position-related response scaling factor to proportionally adjust the reference threshold. When the equivalent solid angle correction coefficient is small, it indicates a decrease in the detector's effective response to incident radiation under the current body position, thus increasing the pulse detection threshold and suppressing noise-induced false triggering. When the correction coefficient is large, it indicates an enhanced effective response, thus decreasing the threshold to ensure that low-amplitude effective pulses can still be captured.

[0057] The baseline threshold, corrected for body position, and the corresponding pulse width discrimination reference time window are used together as pulse waveform separation parameters to perform point-by-point scanning processing on the digitized waveform data. During the scanning process, the amplitude of each sampling point is sequentially checked to see if it exceeds the current dynamic threshold. When the amplitude of multiple consecutive sampling points is consistently higher than the threshold and its duration falls within the set reference time window, the sampling point sequence is determined to correspond to a complete single-pulse event waveform. For each determined single-pulse event, the set of sampling points corresponding to its start point, peak segment, and decay end point is extracted to completely preserve the waveform morphology information of the event. All identified single-pulse events are arranged and numbered according to their timestamp order to form a structured single-pulse event sequence.

[0058] In the correction module, waveform feature parameters of each single pulse event are extracted, and position-related Compton scattering correction is performed on the feature parameters based on body position data and the ratio of pulse leading edge to trailing edge time.

[0059] From the established sequence of single-pulse events, the digitized waveform data corresponding to each single-pulse event is read sequentially in chronological order. The digitized waveform is a voltage-time variation sequence recorded at a fixed sampling interval, completely covering the pulse's initiation, rise, peak, and decay processes. For each single-pulse event, firstly, the sampling point with the largest amplitude is identified in its waveform data, and the amplitude of this sampling point is determined as the peak amplitude. This peak amplitude reflects the instantaneous light output intensity generated by the radiation interaction in the scintillator, characterizing the instantaneous intensity features of energy deposition in a single event. Secondly, the amplitudes of all sampling points within the effective time range of the single-pulse event waveform are summed to obtain the waveform integral area. This integral area reflects the overall output of the scintillating light signal in the time dimension and has a stable correspondence with the total energy deposition of radiating particles in the probe medium, serving as an important parameter for energy equivalence mapping. The time segments corresponding to the pulse leading edge and pulse trailing edge are determined in the single-pulse event waveform. The leading edge time from pulse initiation to reaching the peak and the trailing edge time from peak decay to baseline are calculated, and the ratio of the leading edge time to the trailing edge time is obtained accordingly. This ratio is used to characterize the symmetry and broadening properties of pulse waveforms, and can reflect the influence of radiation interaction type and scattering mechanism on waveform morphology.

[0060] To eliminate the systematic bias in energy estimation results caused by individual body position changes and differences in scattering mechanisms, a Compton scattering correction process based on joint judgment of body position and waveform morphology is introduced. First, the real-time angle between the axis of the individual detection unit and the direction of gravity is obtained from synchronously acquired body position data, and a sine value is calculated for this angle to characterize the degree of deviation of the radiation incident direction relative to the sensitive volume under the current detector attitude. This sine value numerically reflects the magnitude of the lateral component of the incident direction and the detector axis, used to characterize the difference in directional response caused by body position changes. Subsequently, the sine value of the angle is jointly analyzed with the obtained ratio of the leading edge time to the trailing edge time of the pulse. Based on a pre-defined event classification mapping relationship, the current single-pulse event is classified. This event classification mapping relationship uses the degree of body position deviation and the pulse waveform broadening characteristics as the criteria to divide single-pulse events into different scattering-dominant categories, while extracting the event category dominated by Compton scattering. For different categories, corresponding Compton scattering correction weight coefficients are pre-defined. These weight coefficients are used to quantify the influence of the scattering process and body position factors on the energy deposition estimation results. Finally, a scalar multiplication operation is performed on the waveform integral area of ​​the single-pulse event, and the integral area is multiplied by the corresponding correction weight coefficient to obtain the corrected pulse integral area.

[0061] The spectrum construction module generates an intermediate energy deposition spectrum that reflects the interaction between radiation and human tissue.

[0062] The energy deposition response model is established based on the actual response characteristics of the personal detection unit, and is completed through a combination of experimental calibration and simulation calculation. Specifically, multiple monoenergetic radiation sources with known energies are selected as calibration sources. These sources are applied sequentially to the personal detection unit under standard experimental conditions. While maintaining consistency in detector installation, body position reference conditions, and signal acquisition parameters, waveform data of single-pulse events induced by the monoenergetic radiation sources are collected. For the collected single-pulse events, characteristic parameters such as peak amplitude and waveform integral area are extracted. The distribution characteristics of the pulse integral area under different known energy conditions are statistically analyzed to form the energy response curve of the detection unit in the detection medium. Simultaneously, for the aforementioned monoenergetic radiation sources, the energy deposition process of radiated particles in equivalent materials of human soft tissue is simulated using the Monte Carlo simulation method. The simulation uses equivalent material parameters consistent with the composition and density of human soft tissue to calculate the average energy deposition share of particles in the tissue under corresponding radiation energy conditions. The statistical results of the measured pulse integral area are correlated with the tissue energy deposition share calculated by the Monte Carlo simulation to establish a quantitative correspondence between pulse signal characteristic parameters and the actual deposited energy in human soft tissue. Based on this correspondence, the continuous energy range is discretized. According to the pre-set energy resolution requirements, the tissue equivalent deposition energy is divided into multiple adjacent discrete intervals, and the corresponding pulse integral area value range is determined for each interval, thereby forming a complete energy deposition response model and realizing the stable mapping of the integral area of ​​each single pulse event to a clear equivalent deposition energy interval.

[0063] The characteristic parameters of single-pulse events, corrected for body position and scattering, are used as input data and processed one by one into the established energy deposition response model. For each single-pulse event, based on its corrected waveform integral area, and according to the correspondence between the integral area and the equivalent deposition energy interval defined in the response model, the discrete equivalent deposition energy interval to which the event belongs is determined, and the event counts in that interval are accumulated. Within a complete monitoring time window, the above interval mapping and counting operations are performed on all single-pulse events to obtain the event distribution number in each equivalent deposition energy interval. To eliminate the influence of changes in the total number of events within different time windows on the spectral analysis, the event count values ​​of each energy interval are normalized. The normalization method is to divide the number of events in each interval by the total number of single-pulse events in that time window to obtain the normalized count rate corresponding to each energy interval. Finally, the representative value of the equivalent deposition energy interval is used as the horizontal axis, and the corresponding normalized count rate is used as the vertical axis to construct an intermediate energy deposition spectrum. This intermediate energy deposition spectrum statistically reflects the energy deposition distribution characteristics of radiation particles in the equivalent material of human tissue under the current body position and radiation environment conditions, while preserving the relative contribution relationship of different energy ranges.

[0064] In the spectral prediction module, the spectral characteristics of the intermediate energy deposition spectrum are predicted within the future monitoring time window based on the characteristic evolution sequence of the intermediate energy deposition spectrum.

[0065] For the intermediate energy deposition spectrum already constructed within the current monitoring time window, the statistical characteristics of the spectrum are quantitatively calculated to form spectral feature parameters for prediction. First, the spectral centroid energy of the intermediate energy deposition spectrum is calculated. Specifically, within the discretized equivalent deposition energy intervals, a representative energy value for each energy interval is selected, and this representative energy value is weighted and accumulated with the corresponding normalized count rate for each interval. The result is used as the spectral centroid energy within the monitoring time window. This spectral centroid energy characterizes the average energy deposition level per unit event for an individual in the tissue equivalent sense, reflecting whether the radiation energy distribution has shifted overall towards higher or lower energy directions. Second, the total number of single-pulse events separated within the same monitoring time window is counted, and combined with the fixed statistical duration of the time window, the total count rate of single-pulse events per unit time is calculated. This count rate serves as the second spectral feature parameter, reflecting the density level of interaction events experienced by an individual under the current radiation environment. Next, based on the distribution of energy ranges in the intermediate energy deposition spectrum, the average energy value of the spectrum is calculated. The event counts corresponding to all energy ranges above this average energy value are then accumulated. This accumulated value is then compared to the total number of events within the time window to obtain the proportion of high-energy events, which serves as the third spectral characteristic parameter. This parameter characterizes the relative proportion of high-energy deposition events among all events under the current radiation conditions. Through the above calculation process, a complete set of spectral characteristic parameters is formed, consisting of the spectral centroid energy, event count rate, and the proportion of high-energy events.

[0066] The calculated first, second, and third spectral feature parameters are combined in a fixed order to form the spectral feature state vector corresponding to the current monitoring time window. This spectral feature state vector is stored chronologically in a historical state vector sequence to characterize the dynamic process of spectral feature evolution over time. Before prediction, the current spectral feature state vector and the historical state vector sequence are input into a recursive state predictor. The recursive state predictor employs a recursive least squares algorithm with an embedded forgetting factor to fit a first-order dynamic model of spectral feature state changes over time online. Specifically, this first-order dynamic model uses the linear mapping relationship between spectral feature states within adjacent monitoring time windows as its modeling object. By continuously introducing the latest observed spectral feature state vector and gradually reducing the influence weight of earlier historical data on the model parameters, the model parameters can quickly adapt to the changing trends of spectral features under current radiation environment and individual body position conditions. The introduction of the forgetting factor clearly limits the model's sensitivity range to recent state changes, ensuring that the dynamic model remains effective only on short timescales and does not rely on assumptions of long-term stability. By continuously iteratively updating, a set of parameters for a first-order dynamic model describing the relationship between changes in spectral feature states between adjacent time windows is obtained. This model is used to characterize the evolution direction and magnitude of changes in spectral feature states under the current environment.

[0067] Based on the fitted first-order dynamic model, a forward prediction operation is performed on the current spectral feature state vector to calculate the predicted spectral feature state value for the next monitoring time window. This predicted value also includes the predicted spectral centroid energy, count rate, and high-energy event proportion, reflecting the potential changes in energy deposition distribution characteristics for an individual within the next monitoring time window. To ensure the rationality of the prediction results at both the physical and engineering levels, dual constraints are imposed on the predicted values. The first constraint is a boundary constraint based on historical extreme values. This involves statistically analyzing the maximum and minimum values ​​of each spectral feature parameter in the historical state vector sequence and limiting the predicted value to a certain proportion of these historical extreme values. For example, the predicted value is limited to no more than 1.1 times the historical maximum and no less than 0.9 times the historical minimum, thus preventing discontinuous abrupt changes in the prediction results. The second constraint is a physical nonnegativity constraint, explicitly limiting the spectral centroid energy, event count rate, and high-energy event proportion to no negative values, with the high-energy event proportion further limited to between 0 and 1. The predicted spectral feature state value after the above constraint processing is used as the final prediction result output to characterize the evolution trend of an individual's exposure state within the future monitoring time window.

[0068] In the dose monitoring module, the exposure risk assessment for individuals is performed based on the spectral feature prediction results of the intermediate energy deposition spectrum.

[0069] The predicted values ​​of the spectral centroid energy and the total count rate of single-pulse events corresponding to the future monitoring time window are read from the spectral feature prediction results output by the spectral prediction module. To ensure the stability and resistance to random fluctuations in risk assessment, the predicted values ​​are analyzed using a continuous time window comparison method. Specifically, the predicted value of the spectral centroid energy in the current monitoring time window is compared with the predicted value of the spectral centroid energy in the immediately preceding prediction window. When the predicted value of the spectral centroid energy shows an increasing trend within a consecutive preset number of prediction windows, it is determined that the overall energy deposition distribution is migrating towards higher energy. At the same time, the same continuous window comparison is performed on the predicted value of the total count rate of single-pulse events. When the predicted count rate continues to rise within the same continuous window, it is determined that the density of radiation interaction events per unit time is increasing. Only when both the predicted value of the spectral centroid energy and the predicted count rate simultaneously meet the above-mentioned continuous increase condition is it determined that the individual is in an increased exposure risk situation within the future monitoring time window. By adopting a judgment rule that allows for the simultaneous increase of two features, misjudgments caused by short-term fluctuations in a single feature are avoided. This ensures that the risk situation assessment reflects both changes in radiation energy levels and changes in the frequency of events, thus ensuring that the judgment results are consistent with the actual evolution characteristics of the radiation field.

[0070] When an individual is determined to be at increased exposure risk, a quantitative estimate of the expected cumulative dose increment within the future monitoring window is further performed. The specific calculation method is as follows: based on the predicted spectral centroid energy, the average energy deposition level of a single radiation interaction event within the future time window is determined in terms of tissue equivalence; combined with the predicted total count rate of single-pulse events, the number of interaction events expected to occur per unit time is calculated; under the defined equivalent tissue mass condition, the average energy deposition level and the number of events are cumulatively converted to obtain the estimated incremental value of the individual's expected cumulative absorbed dose within the future monitoring window. The equivalent tissue mass is uniformly set during the initialization phase, for example, by taking a fixed value according to a standard human soft tissue mass model, thereby ensuring the consistency of dose conversion. Subsequently, the estimated dose increment is compared with a preset baseline increment, which is obtained by statistically analyzing the dose increments of an individual within multiple consecutive historical monitoring time windows under normal background radiation conditions, specifically using the average of historical dose increments as the baseline reference. Based on the degree of deviation of the estimated dose increment from the baseline increment, the individual's exposure risk is graded and assessed. For example, when the estimated increment is significantly higher than the baseline increment, a high-risk exposure assessment result is output; when the estimated increment is slightly higher than the baseline, a medium-risk assessment result is output; and when the estimated increment is close to the baseline, a low-risk assessment result is output. In this way, a quantitative assessment and clear output of an individual's exposure risk level in the near future can be achieved.

[0071] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. The semiconductor medium can be a solid-state drive.

[0072] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0073] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and modules described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0074] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or modules may be electrical, mechanical, or other forms.

[0075] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical modules; they may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0076] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0077] If the aforementioned functions are implemented as software functional modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0078] 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.

[0079] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A smart system for acquiring and monitoring personal dose of nuclear radiation, characterized in that, include: The acquisition module is used to acquire raw pulse signals through the personal detection unit and convert them into digital waveform data, simultaneously acquire real-time personal body position data, and jointly establish a raw data stream; The waveform analysis module is used to dynamically configure pulse waveform separation parameters based on the noise statistics characteristics and body position data of the original data stream, and to parse discrete single-pulse event sequences from the digitized waveform data. The correction module is used to extract the waveform feature parameters of each single pulse event and perform position-related Compton scattering correction on the feature parameters based on the body position data and the ratio of the pulse leading edge to the trailing edge time. The spectrum construction module is used to input the corrected pulse characteristic parameters into the preset energy deposition response model to generate an intermediate energy deposition spectrum that reflects the interaction between radiation and human tissue. The spectral prediction module is used to predict the spectral characteristics of intermediate energy deposition spectra within a future monitoring time window based on the characteristic evolution sequence of intermediate energy deposition spectra. The dose monitoring module is used to assess an individual's exposure risk based on the spectral characteristics prediction results of the intermediate energy deposition spectrum; The generation of intermediate energy deposition spectra reflecting the interaction between radiation and human tissue in the spectrum construction module specifically includes: The corrected pulse characteristic parameters are input into the energy deposition response model. The distribution of all input single pulse events in the discrete equivalent deposition energy range is counted. The count value of each energy range is normalized to generate an intermediate energy deposition spectrum with equivalent deposition energy as the horizontal axis and normalized count rate as the vertical axis. The spectral prediction module, based on the characteristic evolution sequence of intermediate energy deposition spectra, specifically includes predicting the spectral characteristics of intermediate energy deposition spectra within a future monitoring time window, including: Calculate the spectral centroid energy of the intermediate energy deposition spectrum within the current monitoring time window to obtain the first spectral characteristic parameter; Calculate the total count rate of single-pulse events within the same time window to obtain the second spectral characteristic parameters; The proportion of event counts in the interval above the average energy in the intermediate energy deposition spectrum is calculated to obtain the third spectral characteristic parameters; The first, second and third spectral feature parameters are combined into the current spectral feature state vector. The current spectral feature state vector and the historical state vector sequence stored in chronological order are input into the recursive state predictor. The recursive state predictor fits a first-order dynamic model of the spectral feature state changing over time using a recursive least squares algorithm with an embedded forgetting factor. The first-order dynamic model is used to perform forward prediction, calculate the predicted value of the spectral feature state of the future monitoring time window, apply boundary constraints and physical non-negativity constraints based on historical extreme values ​​to the predicted value, and output the final prediction result.

2. The intelligent personal dose acquisition and monitoring system for nuclear radiation according to claim 1, characterized in that, The acquisition module specifically includes the following steps for establishing the raw data stream: The personal detection unit includes a plastic scintillator detector and a photomultiplier tube; The scintillation light signal induced by radiation particles is collected and converted into an analog electrical pulse signal. The analog electrical pulse signal is sampled at equal intervals to convert the continuous voltage change into a discrete digital sequence, forming the original digital waveform data. Simultaneously, real-time body position data of an individual's body axis relative to a set reference coordinate system is acquired, and a unified timestamp is assigned to the original digital waveform data and the real-time body position data. Based on this timestamp, the two types of data are aligned and spliced ​​to generate the original data stream.

3. The intelligent personal dose acquisition and monitoring system for nuclear radiation according to claim 1, characterized in that, The waveform analysis module dynamically configures pulse waveform separation parameters to parse discrete single-pulse event sequences from digitized waveform data, specifically including: From the digitized waveform data of the original data stream, select a local waveform of the non-pulse signal segment, and calculate the amplitude standard deviation of the waveform sample as the real-time baseline noise level; Extract the real-time angle between the axis of the personal detection unit and the direction of gravity from the raw data stream. Calculate the equivalent solid angle correction coefficient of the radiation incident direction relative to the sensitive volume of the detector based on the angle. Set the reference time window for pulse width discrimination based on the reciprocal of the equivalent solid angle correction coefficient. The real-time baseline noise level is used as the reference threshold for pulse detection, and the equivalent solid angle correction coefficient is used as the position-related response scaling factor to proportionally adjust the reference threshold. Using a reference threshold and a reference time window as pulse waveform separation parameters, the digitized waveform data is scanned point by point. When the amplitude of consecutive data points exceeds the reference threshold and falls within the reference time window, a complete single-pulse event waveform is captured and arranged according to the timestamp to form the single-pulse event sequence.

4. The intelligent personal dose acquisition and monitoring system for nuclear radiation according to claim 1, characterized in that, In the correction module, the waveform feature parameters of each single-pulse event are extracted, and position-related Compton scattering correction is performed on the feature parameters based on the body position data and the ratio of the pulse leading edge to the trailing edge time. Specifically, this includes: The specific feature parameters are the peak amplitude, waveform integral area, and the ratio of the leading edge time to the trailing edge time of the pulse, which are calculated from the digital waveform of each single pulse event read from the single pulse event sequence. Based on the real-time sine value of the angle between the axis of the personal detection unit and the direction of gravity, as well as the ratio of the leading and trailing edges, the classification of the current pulse event and the corresponding Compton scattering correction weight coefficient are determined according to the set event classification mapping relationship. Then, scalar multiplication is performed on the waveform integral area to obtain the corrected pulse integral area.

5. The intelligent personal dose acquisition and monitoring system for nuclear radiation according to claim 1, characterized in that, In the spectrum construction module, the energy deposition response model establishes a quantitative correspondence between the pulse signal amplitude and the actual deposited energy in human soft tissue by testing the response pulse amplitude of the personal detection unit to different known energy monoenergetic radiation sources, and combining the energy deposition share in the tissue equivalent material calculated by Monte Carlo simulation. The integral area of ​​each pulse is mapped to a discrete equivalent deposition energy range.

6. The intelligent personal dose acquisition and monitoring system for nuclear radiation according to claim 1, characterized in that, In the dose monitoring module, the exposure risk assessment for individuals based on the spectral characteristic prediction results of the intermediate energy deposition spectrum specifically includes: Obtain the characteristic predicted values ​​of the intermediate energy deposition spectrum. If both the predicted value of the spectral centroid energy and the predicted count rate show a monotonically increasing trend, it is determined that the individual is in a state of increased exposure risk within the future monitoring time window. When the risk of exposure is rising, the incremental estimate of the individual's expected cumulative dose within the future monitoring time window is calculated based on the predicted spectral centroid energy and total count rate. The incremental estimate is compared with the incremental estimate based on the preset baseline, and the individual's exposure risk assessment result is output.

Images