A nuclear radiation surface contamination detection simulation training device and a control method thereof

Abstract

The application belongs to the technical field of simulation training equipment, and mainly provides a nuclear radiation surface contamination detection simulation training equipment and a control method thereof.The equipment comprises an excitation light source module, a photoelectric detection module, a multi-point distance measurement module and a processor, the multi-point distance measurement module detects distance values of multiple positions between a detection window and a measured surface, the processor calculates an average distance D avg , and locks an effective detection interval as 0.5cm≤D avg ≤2.5cm, and when the value exceeds the range, an invalid identification is forcedly output, and a simulation radioactivity value is not output; the equipment further comprises a hardware comparison circuit independent of the processor, the hardware comparison circuit comprises an SR latch, the hardware comparison circuit is connected to a driving circuit of the excitation light source module, when any distance is less than 0.3cm, the power is reduced to 30%, and the detection is not interrupted, when all distances are less than 0.2cm and last for more than or equal to 200ms, the SR latch cuts off the driving circuit of the excitation light source module, and when the distance is more than or equal to 0.6cm and keeps for 1s, the SR latch is unlocked.The application binds the detection output and the standard detection distance, and the hardware redundant interlocking considers safety and training fluency.

Description

Technical Field

[0001] This invention belongs to the technical field of simulation training equipment, and in particular relates to a simulation training device for detecting nuclear radiation surface contamination and its control method. Background Technology

[0002] Surface contamination detection is a crucial part of nuclear radiation reconnaissance and emergency response. Personnel must be proficient in operating radioactive surface contamination detectors, including proper detection distance, scanning speed, and detection posture, to ensure accurate results and operational safety. The CoMo170 surface contamination detector is a widely used standard device in the nuclear industry, with a detection area of ​​approximately 170 cm². 2 Weighing approximately 750g, it features a specific ergonomic shape and a bottom detection window structure, with a standard detection distance of 0.5~2.5cm. To enable trainees to develop operating habits transferable to real-world testing scenarios, simulation training equipment matching the operating parameters of the actual instrument is required.

[0003] Currently, nuclear radiation surface contamination detection simulation training equipment mainly adopts the fluorescence-induced detection method. It uses an excitation light source to irradiate a fluorescent tracer coated on the surface to be tested, and the photoelectric detection module collects the fluorescence signal and converts it into a simulated radioactivity value. There are three main technical approaches currently available: The first approach (such as Chinese patent CN202311370740.X) uses 365nm ultraviolet light to induce visible fluorescence in inorganic fluorescent substances such as SrO and CaO for simulated detection. It only uses a single ranging module, and when the device is tilted, the single-point ranging cannot reflect the actual parallelism between the detection window and the measured surface, which can easily lead to misjudgment of distance. The second approach (such as Chinese patent CN102519927A) uses 488nm blue LEDs to induce fluorescence in sodium fluorescein and uses a fixed probe array to detect the concentration field in the water area. The distance is physically limited to about 1cm by an isolation cap. There is no active distance sensor, so it cannot simulate the dynamic detection distance in real detection. The third approach (such as Chinese patent CN214705037U) uses a split structure where the main unit connects the handle and the probe via a wire. The front end of the probe is equipped with a photoelectric transmitter and a photoelectric receiver. There is no distance sensor or distance control logic, so it cannot cultivate correct detection distance habits.

[0004] The existing technology has the following shortcomings: First, the detection output of the simulation training equipment is independent of the detection distance. Regardless of whether the distance between the probe and the surface being measured is within the standard range, the equipment outputs a simulated activity reading. This makes it impossible for trainees to develop standardized operating habits for detection distance, and may even lead to bad muscle memory of placing the probe too close to the surface being measured. However, the actual nuclear radiation detection standard requires the probe to maintain a detection distance of 0.5~2.5cm. Placing the probe too close will increase the risk of the detector being contaminated with radioactive pollutants, while raising the probe too high will reduce the detection sensitivity. Traditional training can be understood in terms of specifications, but the trainees cannot perform them correctly, and the training results cannot be transferred to the operation of real instruments. Second, in terms of safety protection, the existing technology has a simple safety interlock mechanism, which usually relies solely on software logic for close-range protection. Once the processor crashes or malfunctions, the safety protection fails. There is no hardware redundancy, and the reliability is insufficient. Meanwhile, the close-range protection threshold is set unreasonably, and the human-machine safety threshold lacks hysteresis buffer. When scanning steps, gaps, or protruding bolts, the light source is directly cut off as soon as the distance is close, frequently triggering the light source to shut down, causing frequent training interruptions, poor trainee experience, and failing to balance safety protection and training smoothness. Third, some solutions use ultraviolet light as the excitation source, which poses a risk of skin erythema and photochemical damage to the eyes, making them unsuitable for multi-person rotation training and training on detecting contamination on human skin surfaces. Fourth, the split-structure equipment differs from the integrated handheld structure of real radioactive detection instruments. The operating parameters such as the position of the detection window, detection area, and weight do not match those of the real instrument, making it impossible to transfer the training feel.

[0005] Therefore, there is an urgent need for a nuclear radiation surface contamination detection simulation training device that can bind the effective detection range with the standard detection distance, has hardware redundancy and safety interlocks, and takes into account the safety and smoothness of trainees. Summary of the Invention

[0006] In view of the shortcomings of the prior art described above, the purpose of this invention is to provide a nuclear radiation surface contamination detection simulation training device and its control method, which solves the problems of the detection output of the nuclear radiation surface contamination detection simulation training device being unrelated to the detection distance, making it impossible to develop standardized operating habits, and the safety interlock being single and unreliable, and the training being easily interrupted by close-range protection.

[0007] To achieve the above and other related objectives, the first aspect of the present invention provides a nuclear radiation surface contamination detection simulation training device. The device as a whole adopts a blue light-induced fluorescence detection method for simulation training of nuclear radiation surface contamination detection. It includes a shell, a detection window, an excitation light source module, a photoelectric detection module, a multi-point ranging module, and a processor. The excitation light source module is used to generate excitation light, the photoelectric detection module is used to detect fluorescence signals, and the multi-point ranging module detects the distance values ​​between the detection window and multiple positions on the surface being measured.

[0008] The processor calculates the average distance D based on the distance value. avg The effective detection range is defined as 0.5cm≤D. avg ≤2.5cm, D avg When the value exceeds this range, an invalid flag will be forcibly output and no simulated radioactivity value will be output.

[0009] It also includes a hardware comparison circuit independent of the processor, which synchronously receives the distance signal output by the multi-point ranging module and is connected to the drive circuit of the excitation light source module. The hardware comparison circuit includes an RC delay network and an SR latch whose reset terminal is connected to the processor. The set terminal of the SR latch receives the distance comparison signal through the delay network.

[0010] When any distance value is less than 0.3cm, the processor will reduce the driving power of the excitation light source to 30% of the driving power corresponding to the current motion state without interrupting the detection. After the distance is restored to ≥0.3cm and stabilized for 200ms, the driving power will be automatically restored.

[0011] When all distance values ​​are less than 0.2cm and last for ≥200ms, the SR latch activates to cut off the drive circuit of the excitation light source module and sends an interrupt signal to the processor;

[0012] After the entire length is restored to ≥0.6cm and held for 1 second, the processor sends a reset pulse to unlock.

[0013] In some embodiments of the present invention, the housing is a one-piece molded structure, including a handheld part and a main body;

[0014] The detection window is located at the bottom of the main body, and the area of ​​the detection window is 150~200 cm². 2 The detection window is equipped with an optical protective cover.

[0015] In some embodiments of the present invention, the excitation light source module includes an LED excitation light source array with an emission wavelength of 450~500nm. The LED excitation light source array is arranged around the inner side of the detection window and tilts to illuminate the central area. A diffuser plate is provided below the LED excitation light source array.

[0016] The excitation light source module also includes a constant current drive circuit and a frequency modulator, the frequency modulator outputting a high-frequency square wave modulation signal of 4~5kHz.

[0017] In some embodiments of the present invention, a reflection correction light source and a timing control circuit are also included;

[0018] The reflection correction light source is a low-power LED with an emission wavelength of 470nm, and the driving circuit of the reflection correction light source is connected to an extremely short pulse generation circuit with a pulse width of <1μs; the photoelectric detection module is connected to a high-speed sample-and-hold circuit with a response time of <500ns.

[0019] The timing control circuit is connected to the driving end of the excitation light source module and the driving end of the reflection correction light source, respectively, and outputs a time-division staggered driving signal.

[0020] In some embodiments of the present invention, the multi-point ranging module includes four ToF infrared distance sensors and corresponding infrared light-shielding compartments. The infrared light-shielding compartments are respectively disposed at the four corners of the detection window, and each ToF infrared distance sensor is assembled in an independent infrared light-shielding compartment.

[0021] The four ToF infrared distance sensors are connected to a hardware synchronization triggering circuit, which outputs a synchronization triggering signal with a 10ns micro-timing stagger.

[0022] In some embodiments of the present invention, a miniature calibration cavity with a built-in standard reference fluorescent sheet is provided on one side of the detection window.

[0023] A second aspect of the present invention provides a control method for the above-mentioned device, comprising the following steps:

[0024] S1. Distance Detection: Distance values ​​between the detection window and the measured surface at multiple locations are detected using multi-point ranging, and the average distance D is calculated. avg ;

[0025] S2. Distance validity determination: The effective detection range is locked as 0.5cm≤D avg ≤2.5cm, when D avg When the value is less than 0.5cm or greater than 2.5cm, an invalid label will be forcibly output, and the simulated radioactivity value will not be output.

[0026] S3. Hierarchical Safety Control: When any distance value is less than 0.3cm, the driving power of the excitation light source is reduced to 30% of the driving power corresponding to the current motion state without interrupting detection. When all distance values ​​less than 0.3cm recover to greater than or equal to 0.3cm and remain stable for 200ms, the power reduction protection is automatically released. When all distance values ​​are less than 0.2cm and the duration is greater than or equal to 200ms, the SR latch in the hardware comparison circuit, independent of the processor, is activated to directly cut off the driving circuit of the excitation light source module and send an interrupt signal to the processor. When all distance values ​​recover to greater than or equal to 0.6cm and remain stable for 1s, the processor sends a reset pulse to the SR latch of the hardware comparison circuit to unlock it.

[0027] S4, Activity Output: 0.5cm≤D avg Within an effective detection range of ≤2.5cm, the simulated radioactivity value is calculated and output based on the output signal of the photoelectric detection module.

[0028] In some embodiments of the present invention, in step S1, the distance values ​​of the four corners of the detection window are synchronously acquired at a sampling frequency of ≥100Hz. Based on the preset coordinates of the four corners and the corresponding distance values, the spatial plane of the measured surface is fitted using the least squares method to calculate the average distance D. avg The tilt angles α and β in the two orthogonal directions, and the flatness deviation ΔD;

[0029] Gradient-discriminatory abrupt change filtering is performed on the four-point distance values: when the distance between the four points shows a monotonically increasing or decreasing linear gradient and the residuals of each measurement point relative to the fitting plane all fall within the sensor noise range of ±2mm, all data participate in the plane fitting. When the distance difference between a single data point and two adjacent measurement points on the side is >5mm and remains stable for more than 3 frames, and the difference between the maximum and minimum values ​​of the other three points is ≤0.3mm, the single point is removed and recalculated. When two or more distance anomalies occur and there is no linear gradient, the two middle sets of valid distance measurement data are retained after sorting by distance value for local fitting.

[0030] In some embodiments of the present invention, step S4 includes:

[0031] Simultaneously acquire signals in the first channel (510-550nm band) and the second channel (570-590nm band);

[0032] The background signal is collected every 200ms in the dark segment and then subtracted from the original dual-channel signal after being filtered by a 5th-order moving average.

[0033] When the signal values ​​of both channels after subtracting the background are both lower than three times the standard deviation of their respective background signals, they are determined to be undetectable and the activity is forced to be zero.

[0034] In some embodiments of the present invention, in step S4, the activity calculation uses a piecewise standard curve:

[0035] Equivalent fluorescence signal corresponding concentration ≤5μg / cm 2 When a forced linear fit is used, the formula is A=k×I. net Where k is the slope coefficient, I net Net fluorescence signal;

[0036] Corresponding concentration >5μg / cm 2 A second-order polynomial fitting is used, with the formula A = a × I. net 2 +b×I net , where a and b are fitting coefficients;

[0037] Between 4.5 and 5.5 μg / cm 2 The weights for linear fitting and polynomial fitting change by linear interpolation within the transition interval.

[0038] As described above, the present invention has the following beneficial effects:

[0039] 1. The processor calculates the average distance D based on the multi-point distance measurement values. avg The effective detection range is strictly limited to 0.5cm ≤ D. avg ≤2.5cm, D avg When the range is exceeded, an invalid label is forcibly output and no simulated radioactivity value is output, so that the detection output is directly bound to the standard detection distance; trainees can only obtain effective detection results by keeping the probe within the standard distance range. This mechanism forces trainees to master the standard detection distance and avoids developing bad operating habits such as placing the probe too close or raising it too high. It ensures that the training results can be fully transferred to the 0.5~2.5cm operating specifications of the real radioactive surface contamination detector.

[0040] 2. A hardware comparison circuit independent of the processor is set up, including an SR latch with the reset terminal connected to the processor, which constitutes a hardware redundancy safety design. Even if the processor crashes or the program runs away, the hardware comparison circuit can still directly cut off the drive circuit of the excitation light source module by the SR latch when the protection conditions are met. The safety protection does not depend on the software operation, avoids the safety hazards caused by the failure of a single software interlock, and meets the safety and reliability requirements of multiple people taking turns to operate in the training field.

[0041] 3. A two-level protection mechanism is adopted. The first level is software power reduction protection. When any distance value is less than 0.3cm, the driving power of the excitation light source is reduced to 30% of the driving power corresponding to the current motion state without interrupting the detection. This reduces the risk of local over-irradiation in lightly close working conditions while ensuring scanning continuity. The second level is firmware hard latch protection. When all distance values ​​are less than 0.2cm and last for ≥200ms, the SR latch cuts off the driving circuit of the excitation light source module, achieving absolute safety protection in heavily close working conditions. The processor sends a reset pulse to unlock only after all distances recover to ≥0.6cm and remain at that distance for 1s. The hysteresis buffer effectively avoids frequent switching of the light source in conditions such as steps, gaps, and protruding bolts, ensuring the smoothness of continuous scanning training.

[0042] 4. The multi-point ranging module detects the distance values ​​at multiple locations between the probe window and the measured surface, and calculates the average distance D based on the multi-point distances. avg It can more accurately reflect the distance between the entire detection window and the measured surface, avoid distance misjudgment caused by single-point ranging when the equipment is tilted, and provide a reliable data basis for distance validity determination and hierarchical safety control. Attached Figure Description

[0043] Figure 1 The diagram shown is a structural schematic from a first perspective of the present invention.

[0044] Figure 2 The diagram shown is a structural schematic from a second perspective of the present invention.

[0045] Figure 3 Displayed as Figure 2 Enlarged view of point A in the middle;

[0046] Figure 4 The diagram shows the hardware comparator circuit interlocking principle block diagram.

[0047] Figure 5 Displayed as the overall flowchart of the control method;

[0048] Figure 6 The diagram shows the transition between different levels of safety protection status.

[0049] Explanation of icon numbers:

[0050] 1. Housing; 2. Audio-visual prompt module; 3. Display unit; 4. Optical cavity; 5. Multi-point ranging module; 6. Photoelectric detection module; 7. NTC temperature sensor; 8. Humidity acquisition sensor; 9. Excitation light source module;

[0051] 11. Main body; 12. Handheld part;

[0052] 21. LED indicator light; 22. Buzzer;

[0053] 51. ToF infrared distance sensor; 511. Infrared light-shielding compartment;

[0054] 61. Photoelectric detection cavity;

[0055] 91. Black light-absorbing and light-blocking barrier; 92. Excitation light source cavity. Detailed Implementation

[0056] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention.

[0057] like Figures 1-3As shown, the first aspect of this invention provides a nuclear radiation surface contamination detection simulation training device. The device includes a housing 1 manufactured using a one-piece molding process. The housing 1 includes a handheld part 12 and a main body 11, with no seams between the handheld part 12 and the main body 11. The surface of the handheld part 12 is provided with an anti-slip texture. The main body 11 contains a main board cavity and an optical cavity 4. The optical cavity 4 is located inside the front end of the main body 11, and a detection window for the optical cavity 4, connecting to the outside, is located at the bottom of the front end of the main body 11. The entire device employs blue light-induced fluorescence detection for simulation training of nuclear radiation surface contamination detection.

[0058] The optical cavity 4 houses an excitation light source module 9, a photoelectric detection module 6, a multi-point ranging module 5, a miniature calibration cavity, an NTC temperature sensor 7, and a humidity acquisition sensor 8. The motherboard cavity houses a processor, hardware comparator circuitry, non-volatile memory, an independent power-down cache chip, an attitude sensor, a battery grading protection circuit, an extremely short pulse generation circuit, a high-speed sample-and-hold circuit, and a timing control circuit. All components within the optical cavity 4 are electrically connected to the processor and hardware comparator circuitry within the motherboard cavity via photoelectric analog signal lines.

[0059] The hardware comparison circuit is independent of the processor and specifically includes four analog voltage comparators, four-input AND gates, an RC delay network, and an SR latch. The multi-point ranging module 5 includes four ToF infrared distance sensors 51 and a hardware synchronous triggering circuit. Each ToF infrared distance sensor 51 outputs an analog voltage signal inversely proportional to the measured distance (the smaller the distance, the higher the voltage). This analog voltage signal is simultaneously connected to the processor's ADC input and the non-inverting input of the corresponding analog voltage comparator in the hardware comparison circuit. The inverting input of each analog voltage comparator is connected to a preset reference voltage V. ref-0.2 The reference voltage corresponds to a distance threshold of 0.2cm. The outputs of the four analog voltage comparators are connected to a four-input AND gate. The AND gate outputs a high level only when all four distance measurements are less than 0.2cm. The output of the AND gate is connected to an RC delay network (delay time constant set to 200ms). The output of the RC delay network is connected to the set terminal of the SR latch. The reset terminal of the SR latch is connected to the GPIO unlock control pin of the processor. The output of the SR latch is connected to the power switch path control terminal of the drive circuit of the excitation light source module 9, which is used to cut off the drive circuit of the excitation light source module 9 when the distance protection condition is met. The other output of the SR latch is connected to the external interrupt pin of the processor, which is used to report the latch trigger event to the processor.

[0060] In the above hardware comparison circuit, the four analog voltage comparators use the LMV339 four-channel voltage comparator or equivalent devices, with a reference voltage V. ref-0.2The output is generated by a high-precision voltage divider resistor network and buffered; the RC delay network consists of a 100kΩ resistor and a 2.2μF capacitor with a time constant of about 220ms, ensuring that instantaneous proximity does not trigger latching; the SR latch uses a CD4043 or equivalent device, and even if the analog voltage comparator input is restored after latching, the output remains cut off until the processor sends an unlock pulse to reset it via GPIO.

[0061] The detection window area is 150~200cm² 2 The shape is rectangular. In this embodiment, the detection window measures 12cm × 14cm, with an area of ​​168cm². 2 (approximately 170cm) 2 An optical protective cover is installed on the detection window, protruding slightly 0.2mm from the bottom surface of the housing 1, with silicone sealing gaskets at its seams. The optical protective cover is made of double-sided anti-reflective optical glass with a hard, wear-resistant coating. This optical glass has a transmittance of ≥92% in the 450~500nm wavelength range, ≥90% in the 510~550nm wavelength range, ≥85% in the 570~590nm wavelength range, and ≥85% in the 800~1000nm wavelength range. A black matte frame is provided on the inner side of the optical glass to suppress stray light reflected from the interface.

[0062] A low-power, temperature-controlled heating wire is installed inside the optical protective cover. The rated power of the heating wire is 0.5~1.0W, and its on / off state is controlled by the processor via a MOSFET switch controlled by a GPIO pin. The temperature feedback signal of the heating wire comes from the NTC temperature sensor 7 inside the optical cavity 4. The processor sets a heating start-stop hysteresis interval: when the temperature of the optical cavity 4 collected by the NTC temperature sensor 7 is below 8°C, the processor starts the heating wire for intermittent heating (duty cycle of 50%, i.e., heating for 5 seconds and stopping for 5 seconds in a cycle); when the temperature rises to 12°C, the processor turns off the heating wire. Through the above hysteresis control, the decrease in light transmittance caused by condensation on the optical protective cover in low-temperature environments can be eliminated, ensuring the stability of the optical path and adapting to winter outdoor training conditions.

[0063] A metal counterweight is also installed inside the optical cavity 4. This counterweight keeps the total weight of the device within the range of 600-900g, with the center of gravity located at the connection area between the handheld part 12 and the main body 11. The housing 1 is entirely IP54 dustproof and waterproof sealed, and silicone sealing gaskets are installed at the seams of the buttons and display screen on the housing 1. The aforementioned shape, weight, and center of gravity distribution are similar to a standard radioactive surface contamination detector, allowing trainees to transfer their grip and scanning posture to actual instrument operation. The detection window is located at the bottom rather than the front, guiding operators to face the bottom of the main body 11 towards the surface being measured, and to conduct scanning training in a parallel posture close to the surface.

[0064] The photoelectric analog signal line adopts a layered shielding structure. The inner layer is a copper foil shielding layer to suppress the 50Hz low-frequency magnetic field. The outer layer is a metal braided mesh shielding layer to shield against high-frequency radio frequency interference. The power input terminal is equipped with a common-mode and differential-mode composite LC filter circuit to ensure stable readings in training scenarios where multiple electromagnetic devices such as walkie-talkies and flaw detectors coexist, and to reduce activity fluctuations caused by electromagnetic interference.

[0065] The excitation light source module 9 includes a drive circuit and an LED excitation light source array. The drive circuit of the excitation light source module 9 comprises a constant current drive circuit, a frequency modulator, and a power switch path. The output of the constant current drive circuit is connected to the LED excitation light source array, and the output of the frequency modulator is connected to the control terminal of the constant current drive circuit for high-frequency square wave modulation. The power switch path is connected in series in the power supply path between the constant current drive circuit and the LED excitation light source array. The LED excitation light source array, as the load of the drive circuit of the excitation light source module 9, is electrically connected to the output of the power switch path.

[0066] The LED excitation light source array emits light with a wavelength of 450-500nm; in this embodiment, LEDs with a peak wavelength of 470nm are selected. The LED excitation light source array is arranged around the inner perimeter of the detection window and tilted towards the central area. The luminous power of a single LED is 25-40mW, and the total peak light power is controlled within the range of 200-250mW. The 450-500nm blue light belongs to the visible light band, avoiding the cumulative damage risk to the skin and eyes from ultraviolet light, making it suitable for long-term training scenarios such as multi-person rotation training and human body simulated contamination training.

[0067] A diffuser plate, made of milky-white optical-grade PC material, is positioned below the LED excitation light array and facing the detection window. This plate spatially homogenizes the excitation light, ensuring a uniformity of excitation light intensity ≥90% within the measured area corresponding to the detection window. A constant current drive circuit provides a stable current to the LED excitation light array, with current fluctuations ≤±1%. A frequency modulator modulates the constant current drive circuit with a high-frequency square wave, at a frequency of 4~5kHz; 4kHz is selected in this embodiment. This modulation frequency is significantly higher than the critical flicker frequency of the human eye and far from the 50Hz power frequency and its lower harmonics, facilitating the filtering of ambient light interference by the back-end lock-in amplifier.

[0068] The excitation light source module 9 is also equipped with a dynamic duty cycle control circuit. The input terminal of the dynamic duty cycle control circuit is connected to the motion state detection signal from the processor, and the output terminal of the dynamic duty cycle control circuit is electrically connected to the control terminal of the power switch path of the drive circuit of the excitation light source module 9, which is used to adjust the duty cycle output of the drive circuit of the excitation light source module 9 according to the motion state.

[0069] When the device is stationary or in a low-speed scanning state, the drive duty cycle is reduced to 8%. This value is a compromise in engineering: under the condition of a 4kHz light source modulation period of 250μs, a 20μs conduction pulse is obtained, which can significantly reduce the average excitation light power, suppress static photobleaching of the fluorescent tracer, and reduce heat accumulation in the cavity; at the same time, it can still ensure that the fluorescence signal meets the detection criterion of 3 times the standard deviation (3σ), avoiding missed detection of low-concentration simulated contamination. When the device is in a high-speed moving scanning state, the drive duty cycle is restored to 12%, using higher average light power to compensate for insufficient signal integration caused by the short scanning dwell time.

[0070] The motion state detection signal is generated by the processor based on the triaxial acceleration data output by the attitude sensor. Specifically, the processor reads the raw triaxial accelerometer data ax, ay, and az at a sampling frequency of ≥100Hz and calculates the resultant acceleration magnitude a. total =√(ax 2 +ay 2 +az 2 ), and after subtracting the gravitational acceleration g, we obtain the acceleration a. motion =|a total -g|. When a motion <0.15m / s 2 And when this continues for more than 500ms, the processor determines that the device is in a static or low-speed scanning state, and outputs a low-level signal to the dynamic duty cycle control circuit, reducing the drive duty cycle to 8%; when a motion ≥0.15m / s 2 When this occurs, the device is determined to be in a high-speed scanning state, and a high-level signal is output to restore the drive duty cycle to 12%. The aforementioned acceleration threshold is 0.15 m / s². 2 The duration of 500ms can be adjusted via processor firmware parameters. The output of the hardware comparator circuit is also electrically connected to the control terminal of the power switch path of the excitation light source module 9's drive circuit, used to cut off the drive circuit of the excitation light source module 9 when the distance protection condition is met.

[0071] The photoelectric detection module 6 includes a dual-channel photodiode array, a fluorescence collection light guide layer, a dual bandpass filter assembly, and two signal conditioning links. The dual-channel photodiode array is arranged in a 4×4 matrix, with a total of 16 photodiodes. Each photodiode has a dual-sensitive-region stacked structure, specifically a silicon-based monolithically integrated vertically stacked double PN junction device: a first N-type epitaxial layer and a second N-type epitaxial layer are sequentially grown on an N-type silicon substrate, and a first P-type doped region and a second P-type doped region are formed by ion implantation, thereby vertically stacking from top to bottom along the light incident direction to form the first sensitive region (shallow junction) and the second sensitive region (deep junction). The first sensitive region has a junction depth controlled between 0.3 and 0.8 μm, exhibiting high quantum efficiency for short-wavelength visible light in the 510–550 nm band, with a center response wavelength of 525 nm, and is used to output the first photocurrent signal corresponding to the main peak fluorescence. The second sensitive region has a junction depth controlled between 2 and 5 μm, exhibiting high quantum efficiency for long-wavelength visible light in the 570–590 nm band, with a center response wavelength of 580 nm, and is used to output the second photocurrent signal corresponding to the secondary peak fluorescence. The first and second sensitive regions are led out through independent metal electrodes, and the two signals can be read out synchronously. This device can be made using existing commercial silicon-based double-junction photodiodes (such as the Hamamatsu S6795 series or equivalent devices), or it can be custom-fabricated by a semiconductor foundry according to the above parameters.

[0072] A fluorescence collection light guide layer and a dual bandpass filter assembly are sequentially arranged in front of the light-receiving surface of the dual-channel photodiode array along the incident direction of the light. The fluorescence collection light guide layer is located near the detection window and is responsible for uniformly focusing the fluorescence generated within the detection window range. The dual bandpass filter assembly is located close to the light-receiving surface of the photodiode and includes two independent transmission bands of 510~550nm and 570~590nm. It also has an optical density OD≥4 for 470nm excitation light to fully filter out reflected excitation light.

[0073] In this embodiment, a black light-absorbing and light-blocking barrier 91 is installed between the excitation light source module 9 and the photoelectric detection module 6 inside the optical cavity 4. This barrier divides the internal space of the optical cavity 4 into two sub-cavities: the excitation light source cavity 92 and the photoelectric detection cavity 61, allowing for independent arrangement of the excitation light source module 9 and the photoelectric detection module 6. The installation height of the black light-absorbing and light-blocking barrier 91 is higher than the light-emitting surface of the LED and also higher than the light-receiving surface of the photodiode, preventing excitation light from directly entering the photoelectric detection side inside the cavity. The inner walls of both the excitation light source cavity 92 and the photoelectric detection cavity 61 are blackened and matte-finished, and a textured light-absorbing layer is added. The black light-absorbing and light-blocking barrier 91, the blackened and matte-finished inner walls, and the textured light-absorbing layer together form an internal matte-finishing structure, which absorbs stray reflected light within the cavity, suppresses the direct coupling of lateral scattered light from the excitation light source to the photodiode, and reduces baseline drift caused by stray light. The excitation light source cavity 92 and the photoelectric detection cavity 61 are not independently sealed to the outside. They share a set of optical protective covers, and the light uniformly interacts with the external object being measured through the optical protective cover at the detection window position.

[0074] Two signal conditioning links are connected to the independent output electrodes of the two channels of the dual-channel photodiode array, respectively. Each signal conditioning link includes a differential amplifier circuit, an AC amplifier circuit, a demodulation circuit, a DC amplifier circuit, and an ADC acquisition circuit connected in series. The demodulation frequency of the demodulation circuit is consistent with the modulation frequency of the frequency modulator, which is 4kHz in this embodiment. The two ADC acquisition circuits acquire data synchronously at a frequency ≥100Hz to ensure the time consistency of the dual-channel signals.

[0075] The device also includes a reflection correction light source located in optical cavity 4, and an extremely short pulse generation circuit, a high-speed sample-and-hold circuit, and a timing control circuit located in the mainboard cavity. The reflection correction light source is a low-power LED emitting a wavelength of 470nm. A beam-expanding optical element, a microlens array, is placed on the light-emitting side of the reflection correction light source to expand and homogenize the emitted light, forming a surface light source covering the entire detection window area. The output terminal of the driving circuit of the reflection correction light source is connected to the reflection correction light source; the output terminal of the extremely short pulse generation circuit is electrically connected to the control terminal of the driving circuit of the reflection correction light source, used to output driving pulses with a pulse width of less than 1μs to the driving circuit of the reflection correction light source. The input terminal of the high-speed sample-and-hold circuit is connected to the output terminal of the photoelectric detection module 6, with a response time of <500ns, used to complete reflectivity signal sampling during the extremely short pulse period; the output terminal of the high-speed sample-and-hold circuit is connected to the processor.

[0076] The specific calculation method for reflectivity is as follows: During the calibration phase, a standard white board (reflectivity R) is used. white =98%) The high-speed sample-and-hold circuit output voltage V was acquired at a standard distance of 1.5cm. whiteIt is then permanently stored in the factory reference area; during actual testing, the voltage collected during the second 470nm extremely short pulse segment is V. sample The surface reflectance R of the measured surface meas =(V sample / V white )×R white Reflectivity normalization coefficient k ref According to formula k ref =R std / R meas Calculate, where R std The reflectance of the calibration carrier used for calibrating the standard curve (in this embodiment, the calibration carrier is 170 cm²) is determined. 2 Area, 30μm film thickness standard white board, R std =98%). When R meas Exceeding the preset reasonable range (e.g., R) meas <5% or R meas When the reflectance is >100%, the reflectance acquisition is considered abnormal, the device will display a prompt and pause the activity output. The above R... white R std and k ref The computational logic is all embedded in the processor firmware.

[0077] The timing control circuit is located in the motherboard cavity. The output terminal of the timing control circuit is electrically connected to the driving terminal of the excitation light source module 9 and the driving terminal of the reflection correction light source, respectively. It outputs time-division staggered driving signals to the two to perform timing isolation control, dividing one acquisition cycle into a backgroundless acquisition stage, a 470nm ultra-short pulse reflectivity acquisition stage, and a blue light excitation fluorescence acquisition stage.

[0078] A complete acquisition cycle is divided into three time segments, with each segment only activating its corresponding light source and performing corresponding sampling: The first segment is the background light acquisition stage, where all light sources are turned off, and the dual-channel ADC acquisition circuit of the photoelectric detection module 6 acquires ambient light and circuit dark current background signals as subsequent baseline subtraction; the second segment is the 470nm ultra-short pulse reflection correction acquisition (pulse < 1μs), where only the 470nm reflection correction light source is activated, outputting sub-microsecond ultra-short pulses; the high-speed sample-and-hold circuit completes the surface blue light reflectance sampling before the fluorescent tracer produces fluorescence afterglow (fluorescence lifetime 3-5ns); the third segment is the blue light excitation fluorescence signal acquisition, where the reflection correction light source is turned off, the 470nm main excitation light source array is activated, illuminating the tested surface to generate fluorescence, and the dual-channel photoelectric detection module 6 acquires I 515 I 580 Fluorescent signal.

[0079] In the second segment, the reflection correction light source is immediately turned off after the pulse ends. At this time, the main excitation blue LED is not turned on, and the pulse is short enough that the fluorescent material does not have time to respond. Therefore, the correction light will not excite fluorescence, thus eliminating crosstalk of fluorescence signals caused by the correction light. The ToF infrared distance sensor 51 uses an 850nm near-infrared LED as the ranging light source. This 850nm near-infrared LED is only used for ToF distance measurement and does not participate in surface reflectivity correction.

[0080] Temporal isolation can prevent crosstalk between reflected correction light and fluorescence signal, ensure the normal operation of material correction logic, and make the simulated radioactivity values ​​of the same simulated pollution output on training carriers of different materials tend to be consistent. The readings mainly reflect the simulated pollution level and are less affected by carrier color and reflective properties.

[0081] The processor has a built-in ratio calibration unit that automatically calculates the measured fluorescence intensity ratio R. 实测 , where R 实测 =I 580 / I 515 I 515 I is the signal value of the first channel after subtracting the background. 580 This is the signal value of the second channel after subtracting the background. The background (bg) is the signal output by the photodiode channel under no-light excitation conditions, which is the superposition of the circuit's dark current background and the ambient light background. The ratio calibration calculation unit will calculate R... 实测 Compare with the pre-entered baseline ratio R0 of the current batch, according to Formula I 515-校准 =I 515 ×(R0 / R 实测 The fluorescence intensity of the first channel was normalized and calibrated. The calibrated I... 515-校准 Substitute the pre-stored standard curve to convert it into a simulated radioactivity value. Ratio calibration is a common-mode suppression technique that can compensate for common-mode errors caused by slow fluctuations in the light source and aging of optical components, making it suitable for scenarios involving continuous training of large numbers of students. When the measured ratio R... 实测 An alarm is issued when the deviation from the baseline R0 exceeds ±30%, indicating that the consumable batch is incompatible or the consumable is invalid, thus preventing the use of incorrect parameters for training without the knowledge of the problem.

[0082] The equipment is equipped with a batch parameter input interface for the composition. This interface includes at least one of three methods: a barcode scanning interface, a manual input interface, and a wireless communication interface, all of which are connected to the processor. The barcode scanning interface scans the QR code or barcode on the composition packaging to automatically read the R0 value. The manual input interface allows manual input of the R0 value via the equipment menu, with an accuracy of 0.001. The wireless communication interface downloads the R0 value from the upper-level management platform. The processor performs range verification on the input R0 value; a reasonable range is 0.100 to 0.500. Values ​​outside this range are flagged as R0 abnormal. The processor also verifies the R0's expiration date; if the expiration date indicated on the composition label has expired, re-entry is prompted. Detection is directly locked when parameters are missing or abnormal, preventing operators from forgetting to update batch parameters and using incorrect parameters for training and assessment.

[0083] Each of the four corners of the detection window has an independent infrared light-shielding compartment 511. Four ToF infrared distance sensors 51 are each housed within an independent infrared light-shielding compartment 511, forming a quadrilateral ranging layout. The four ToF infrared distance sensors 51 correspond to the four vertices of the rectangular detection window, with preset two-dimensional coordinates P1(0,0), P2(0,W), P3(H,0), and P4(H,W). The infrared light-shielding compartments 511 and the optical protective cover are separately packaged, and each infrared light-shielding compartment 511 is equipped with a light-transmitting protective sheet. The ranging accuracy of the ToF infrared distance sensors 51 is ≤±2mm, and the detection range is 0.2~15cm.

[0084] The output of the hardware synchronous trigger circuit is connected to the trigger input of four ToF infrared distance sensors 51, respectively, and outputs a synchronous trigger signal with a 10ns micro-timing stagger. This 10ns micro-timing stagger avoids instantaneous saturation crosstalk at the receiver caused by the overlapping of the four infrared light pulse emission times. The total sampling time for a single round of four points is ≤50ms.

[0085] A distance correction table for highly reflective materials is pre-stored in the non-volatile memory connected to the processor. Highly reflective materials include stainless steel and aluminum alloy. The specific identification criteria for highly reflective materials are: the processor uses the average distance measurement value D from four points in the current frame... avg The expected distance D compared to a standard non-metallic surface (such as an ABS plastic sheet) at the same distance. expected Comparison, when 5 consecutive frames satisfy D expected -D avg When the thickness is ≥5mm, the surface being measured is determined to be a highly reflective material, and the corresponding compensation coefficient is automatically retrieved based on the actual measured distance D. avg The correction amount ΔD is obtained by looking up the table. corr The corrected ranging value D corr =D avg +ΔD corrSubstitute these values ​​into subsequent plane fitting and tilt angle calculations to eliminate ranging errors caused by the superposition of infrared diffuse reflection from the metal surface. The distance correction table for high-reflectivity materials was obtained during the factory calibration stage using stainless steel and aluminum alloy plates, calibrated point by point within the range of 0.5~2.5cm, and then stored in non-volatile memory.

[0086] A miniature calibration chamber is located on one side of the probe window. This chamber contains a standard reference fluorescent sheet, which is used to detect the measured value I during the power-on self-test. meas .

[0087] The standard reference fluorescent sheet is prepared using the liquid fluorescent tracer composition of this invention: An optical-grade quartz glass sheet with a thickness of 1.0±0.05 mm is selected as the substrate. The composition is uniformly coated onto the substrate surface using a spin-coating method (800~1200 rpm, 30~60 s, wet film thickness 30±5 μm). It is then horizontally dried for 2~4 hours in a light-proof, dust-free environment at 25℃ to form a light pink translucent fluorescent film layer. A 0.5 mm optical-grade quartz glass cover is then placed on top, and the edges are sealed with UV-curable adhesive to form a sandwich encapsulation structure to isolate oxygen and moisture. Under 470 nm excitation light, the emission peaks of this fluorescent sheet are located at 515~520 nm (main peak) and 575~585 nm (secondary peak), matching the two detection bands of the dual-channel photoelectric detection module 6. The first channel reference signal value I under the conditions of 25℃, 1.5 cm standard distance, and full voltage is... std The solidified material is stored in the factory reference area (approximately 120 ADC code value in this example). Under normal light-protected storage conditions, the annual fluorescence intensity decay rate of this fluorescent sheet is <2%. If I meas Compare I std If the degradation exceeds 10%, the processor prompts for replacement and recalibration.

[0088] An independent monitoring photodiode is provided on the side or back of the excitation light source module 9. The output of the monitoring photodiode is connected to the closed-loop feedback signal input of the processor for changes in LED drive intensity and backlight leakage.

[0089] The device also includes non-volatile memory and a separate power-down cache chip located inside the motherboard cavity. The non-volatile memory is divided into two independent partitions. The first partition is the factory reference area, which permanently stores the original standard curve and the reference phosphor reference value I. std The first partition is a distance correction table for high-reflectivity materials, and the factory reference area cannot be overwritten. The second partition is a dynamic correction area, which stores the gain coefficient, temperature and humidity correction coefficient, and the current batch reference ratio R0 generated each time the device is powered on or at a set time. After R0 is entered, it is stored in the dynamic correction area of ​​non-volatile memory and bound to the current batch composition. An independent power-down cache chip is connected to the processor to save training log data when power is off.

[0090] In addition to the NTC temperature sensor 7 and humidity sensor 8 located in the optical cavity 4, the device also includes an attitude sensor and a battery grading protection circuit located in the mainboard cavity. The attitude sensor includes a three-axis accelerometer and a three-axis gyroscope. The output of the attitude sensor is connected to the processor to record raw data such as scanning trajectory, jitter frequency, pitch angle, roll angle, and yaw angle. The NTC temperature sensor 7 is used to acquire the temperature of the optical cavity 4 in real time. The humidity sensor 8 is used to monitor the internal humidity. Tilt determination is based solely on the angle between the detection window and the relative normal to the measured surface, regardless of whether the measured surface is on a horizontal, vertical wall, or inclined slope, making it suitable for diverse detection objects such as the outer surfaces of buildings, tanks, and equipment.

[0091] The device also includes a display unit 3, which is an LCD screen mounted on the housing 1 with a resolution of 128×64 pixels. The device also features a 3-level adjustable backlight located on the upper part of the main body 11. The signal input terminal of the display unit 3 is connected to the processor. The display unit 3 displays simulated radioactivity activity values, real-time detection distance, battery level, current operating mode, tilt status indication, and device temperature and humidity status. The display format is consistent with standard radioactive surface contamination detectors. The device menu includes a built-in continuous training mode switch; activating continuous training mode extends the sleep time to 300 seconds. A long press of the power button temporarily disables sleep mode.

[0092] The device also includes an audio-visual prompt module 2, which comprises a buzzer 22 and an LED indicator 21, mounted on the housing 1. The control terminal of the audio-visual prompt module 2 is connected to the processor. The buzzer 22 supports two output modes: training and evaluation mode and real operation simulation mode. In training and evaluation mode, it emits different frequency prompts when the distance exceeds the threshold, the device tilts excessively, or simulated contamination is detected, suitable for beginners to correct distance and posture issues in real time through differentiated prompts. In real operation simulation mode, it outputs a pulse ticking sound proportional to the count rate of the simulated radioactivity value, replicating the feedback of the real machine, suitable for advanced and assessment stages. The LED indicator 21 is used to indicate the device's operating status, such as standby, detection in progress, and alarm.

[0093] The battery grading protection circuit includes a battery voltage detection unit. The output of the battery voltage detection unit is connected to the processor. Within the high voltage range, all functions of the device operate normally. Within the medium and low voltage range, the processor automatically applies the voltage compensation coefficient k. V This is to compensate for the attenuation of LED light power. When the undervoltage threshold is reached, the processor locks and saves all training data, automatically shuts down, and prohibits further detection.

[0094] The equipment is also equipped with high and low temperature protection thresholds. When the temperature of the optical cavity 4 is higher than 50℃ or lower than -10℃, the LED power will be automatically reduced by 50%. If the temperature exceeds the limit for 5 minutes, the equipment will automatically shut down for protection.

[0095] like Figures 4-6 As shown, the second aspect of the present invention provides a control method for the above-mentioned nuclear radiation surface contamination detection simulation training device, the method comprising four steps: distance detection, distance validity determination, graded safety control, and activity output.

[0096] In the distance detection step, the processor synchronously acquires the distance values ​​d1, d2, d3, and d4 at the four corners of the detection window at a sampling frequency of ≥100Hz. Based on the preset coordinates of the four corners and the corresponding distance values, the processor fits the spatial plane of the measured surface using the least squares method to calculate the average distance D. avg The tilt angles α and β in two orthogonal directions, and the flatness deviation ΔD. Where D... avg = (d1+d2+d3+d4) / 4. ΔD is the difference between the maximum and minimum distance values ​​among the four points.

[0097] The specific process of least squares fitting the plane and calculating the tilt angle is as follows: Let the preset two-dimensional coordinates of the four corners of the detection window be P1(0,0), P2(0,W), P3(H,0), and P4(H,W), corresponding to the measured distances d1, d2, d3, and d4, where W is the distance in the first direction and H is the distance in the second direction perpendicular to the first direction. Fit the plane equation d = a·x + b·y + c using the least squares method for the four data points (xi,yi,di), where a, b, and c are the fitting coefficients. The tilt angles in the two orthogonal directions are calculated using the following formulas: α = arctan(|b|), representing the tilt angle of the detection window around the x-axis (i.e., along the first direction); β = arctan(|a|), representing the tilt angle of the detection window around the y-axis (i.e., along the second direction).

[0098] The distance detection step also includes gradient-discriminatory abrupt change filtering of the four-point distance values. When the four-point distances are combined with the preset two-dimensional coordinates of the four corners of the detection window to perform least-squares spatial plane fitting, and the residuals of each measuring point relative to the fitting plane all fall within the ±2mm sensor noise range and show a monotonically increasing or decreasing linear gradient in two-dimensional space, it is determined that the device is generally gently tilted, and all data participate in the plane fitting. When only the distance difference between a single data point and two adjacent measuring points is ≥5mm and remains stable for more than 3 frames, and the difference between the maximum and minimum values ​​of the other three points is ≤0.3mm, it is determined that there is a local bulge at a single point, and the single point is removed and the ΔD and tilt angles α and β are recalculated. When there are two or more distance anomalies, and the residuals of multiple points after global fitting significantly exceed the sensor noise threshold without an effective two-dimensional linear gradient, it is determined that there are multiple points of concavity or convexity. After sorting by distance values, the middle two sets of distance measurement data are retained for local fitting, and all four-point data are not directly discarded. The 5mm threshold is used to filter out the ±2mm noise jitter of the ToF infrared distance sensor itself, and the 3-frame continuous condition is used to remove occasional interference.

[0099] The following are numerical examples of four sets of uneven filtering conditions (in the following conditions, "linear gradient" refers to the two-dimensional spatial gradient. The algorithm combines the preset two-dimensional coordinates of the four corner points of the detection window to perform least squares spatial plane fitting, and makes judgments based on the residuals of each measurement point relative to the fitted plane; if the residuals of each point fall within the inherent noise range of the ToF infrared distance sensor ≤ ±2mm, it is judged as overall device tilt; if there are multiple points where the residuals are significantly greater than the sensor noise threshold, and the measured surface cannot be described by a single plane, it is judged as multi-point unevenness):

[0100] Condition 1 is a simple large incline, with four-point distance measurements of 0.7cm, 2.4cm, 0.8cm, and 2.5cm, and an average distance D. avg =1.6cm is within the range; after fitting the spatial plane with the four-point corner coordinates, the residual of each point is ≤2mm, which is within the sensor noise range, and it is determined that the whole machine is gently tilted and there is no local protrusion; the flatness deviation ΔD=18mm. According to the detection window width W=12cm and height H=14cm in this embodiment, after fitting the spatial plane with the four-point corner coordinates, the plane equation coefficients are b≈0.15 (distance change rate along the y direction), and the tilt angle α=arctan(|b|)=arctan(0.15)≈8.6°, which exceeds the 6° boundary. The screen prompts excessive tilt, stops tilt compensation, and only stores the original data.

[0101] Condition 2 is normal noise fluctuation, with the four-point distance measurements being 0.7cm, 2.4cm, 0.8cm, and 2.6cm respectively. Only the fourth-path distance measurement in the last frame has a difference of 2mm compared to its neighboring points, which does not reach the 5mm judgment threshold. Moreover, this jump only occurs in a single frame and does not meet the condition of continuous stability for 3 frames. The difference between the maximum and minimum values ​​of the other three points is ≤0.3mm. Therefore, it is judged as normal noise fluctuation, and no rejection is triggered. All data are used for plane fitting.

[0102] Condition 3 is a single-point protrusion, with four-point distance measurements of 0.7cm, 2.4cm, 0.8cm, and 3.1cm respectively; the difference between the fourth-path distance measurement and the adjacent point is 7mm (greater than the 5mm threshold), and this abnormal reading has been consistently stable for more than 3 frames. The difference between the maximum and minimum values ​​of the other three points is ≤0.3mm; it is determined to be a real single-point local protrusion. After removing the abnormal point, the plane is refitted, and the updated ΔD and tilt angles α and β are calculated.

[0103] Condition 4 is a multi-point concave-convex surface, with four-point distance measurements of 0.6cm, 2.6cm, 1.7cm, and 2.2cm respectively. When performing global plane fitting based on the four-corner coordinates, there are two measurement points with residuals significantly greater than 5mm, making it impossible to obtain a self-consistent single measured plane and resulting in no effective linear gradient. Therefore, it is determined to be a multi-point concave-convex surface. After sorting by distance values, the middle two sets of distance data (1.7cm and 2.2cm) are retained for local plane fitting, without directly discarding all four-point distance measurements.

[0104] The above filtering logic can distinguish between noise jitter, real single-point protrusions, multi-point concavity and convexity, and overall tilt. For non-flat training carriers such as welds, steps, and concave and convex components, it will not directly invalidate all ranging data due to local protrusions, and can still output reliable distance and attitude data.

[0105] In the distance validity determination step, the processor locks the detection range to 0.5cm≤D. avg ≤2.5cm. When D avg When the distance is less than 0.5cm, the reading is forcibly determined to be invalid due to excessive proximity. The display unit outputs an invalid indicator "---" accompanied by a rapid alarm sound, and no simulated radioactivity value is output. When D avg When the distance exceeds 2.5cm, the data is forcibly judged as invalid due to exceeding the distance limit, and the display unit outputs an invalid flag "---", without outputting the simulated radioactivity value.

[0106] The following are two examples of distance validity determination verification. In near-field verification, the trainer brings the device close to the test surface at an average distance D. avg =0.3cm, the device instantly determines the distance is too close, the display unit outputs an invalid mark accompanied by a rapid alarm sound, and does not output simulated radioactivity activity values. At the same time, the software power reduction protection automatically reduces the blue LED drive power to 30% of the current value. The trainee adjusts the distance to the average distance D. avg After reaching a distance of 1.5cm, the device automatically resumes normal drive power detection and outputs activity readings. During far-field validation, the trainee raises the device to an average distance D. avg =3.0cm, the device judges the distance to be out of limit, displays an invalid output flag, and does not output simulated radioactivity value. The trainee returns the device to the average distance D. avg After reaching 1.5cm, normal detection output resumes. This mechanism transforms the written requirements of the operating procedures into feedback that the device can perceive. During training playback, the percentage of time the detection window remains within the distance range throughout the training process can be statistically analyzed as an objective scoring criterion for operational compliance.

[0107] like Figure 6 As shown, the graded safety control procedure sets up two levels of hazardous condition protection, which are executed in parallel with the distance validity determination. The hardware execution priority, from high to low, is firmware hard latch protection, software power reduction protection, and distance validity determination.

[0108] The first level (mild proximity condition) is software-based power reduction protection. When any ToF infrared distance sensor detects a distance value less than 0.3cm, the processor reduces the drive power of the excitation light source module to 30% of the drive power corresponding to the current motion state and issues a low-frequency beep. This software-based power reduction protection logic takes precedence over the dynamic duty cycle switching of the device's motion state drive. The above-mentioned software-based power reduction protection is executed through a dynamic duty cycle control circuit: the processor outputs a power reduction control signal to the dynamic duty cycle control circuit, forcibly reducing the duty cycle of the power switch path of the excitation light source module's drive circuit to 30% of the duty cycle corresponding to the current motion state (e.g., from 8% to 2.4% in a stationary state, and from 12% to 3.6% in a high-speed scanning state). When all distance values ​​less than 0.3cm recover to greater than or equal to 0.3cm and remain stable for 200ms, the protection logic automatically releases the power reduction protection, resumes dynamic duty cycle control, and the detection process is uninterrupted.

[0109] The second level (heavy proximity condition) is firmware hardware latch protection. The hardware comparison circuit detects the distance between four points with a microsecond-level response speed. When all distance values ​​are less than 0.2cm and the duration is greater than or equal to 200ms, the hardware comparison circuit instantaneously triggers an interrupt signal, the processor records the event, and the output of the SR latch in the hardware comparison circuit directly cuts off the power switch path of the excitation light source module's drive circuit, simultaneously cutting off the blue light and correction light; at the same time, the output of the SR latch sends an interrupt signal to the processor, the processor records the event and triggers a red audible and visual alarm. The condition of four points being simultaneously close can eliminate single-point false triggering caused by ±2mm noise jitter of the ToF infrared distance sensor and local micro-protrusions on the measured surface.

[0110] The firmware features a hysteresis buffer for hardware latch protection. When the distance drops to the 0.4~0.6cm range, the light source maintains its current latched state and does not trigger unlocking. Once all distance values ​​recover to ≥0.6cm and remain stable for 1 second, the device automatically unlocks the excitation light source module without requiring manual button restart. When a momentary close proximity occurs (0.2cm≤distance<0.5cm), only a low-frequency warning pops up, and the light source is not turned off.

[0111] In the activity output step, when 0.5cm≤D avg Within the range of ≤2.5cm, the processor calculates and outputs a simulated radioactivity value based on the output signal from the photoelectric detection module. The photoelectric detection module simultaneously acquires the 510~550nm band signal I from the first channel. 515-raw Second channel 570~590nm band signal I 580-raw Background signals are acquired every 200ms in the dark segment, and a 5th-order moving average filter is used to eliminate transient strong light jumps. The moving average background signal is subtracted from each of the two channels to obtain I. 515 =I 515-raw -I515-bg I 580 =I 580-raw -I 580-bg Moving average filtering can eliminate activity reading jumps caused by instantaneous strong light pulses and avoid false alarms caused by sunlight and flashlight shaking, thus interfering with training. A moving average background reference table established under illuminance conditions of 0, 100, 500, and 1000 lux is used to quickly estimate the background value during the equipment startup initialization phase. During normal testing, the background value collected in real-time every 200ms in a dark segment and filtered by a 5th-order moving average is used as the standard. The moving average background reference table is only used as an initialization aid.

[0112] The processor performs 3σ noise removal judgment. When I 515 <3σ1 and I 580 When the value is less than 3σ², it is determined as undetectable, and a simulated radioactivity activity A = 0.00 Bq / cm² is forcibly output. 2 This is to avoid misclassifying noise as a signal. Here, σ1 is the standard deviation of the background signal in the first channel, and σ2 is the standard deviation of the background signal in the second channel. This determination has higher priority than the activity fitting calculation.

[0113] The processor performs ratio calibration and calculates the measured fluorescence intensity ratio R. 实测 =I 580 / I 515 According to formula I 515-校准 =I 515 ×(R0 / R 实测 Ratio calibration is performed. 实测 When the deviation from R0 exceeds ±30%, the equipment indicates that the composition batch is mismatched or has deteriorated, and suspends the test output.

[0114] A single acquisition cycle is executed in three time-division segments: background light acquisition, 470nm ultra-short pulse reflectance acquisition, and blue light excitation acquisition. Based on the full-area surface reflectance acquired in the second segment, the processor retrieves the reflectance normalization coefficient k. ref The calibrated fluorescence signal is then subjected to reflectivity correction. The reflectivity correction light source is expanded to cover the entire detection window area, and the collected reflectivity signal is matched with the combined fluorescence detection area of ​​the 16 photodiodes.

[0115] Processor access is only available within 0.5cm ≤ D. avg The global cubic polynomial distance compensation function g(D) effective within the interval ≤2.5cm. avg ), perform tilt compensation determination: when ΔD≤8mm and α≤6° and β≤6°, introduce tilt compensation coefficient k tilt =cos 2 α×cos 2β. This formula represents the Lambert biaxial cosine square attenuation: when the detector window is tilted, there are two independent attenuations: the excitation light incident irradiance attenuation (cosθ) and the fluorescence collection solid angle loss (cosθ). The total uniaxial attenuation is cosθ. 2 θ, the two orthogonal directions of inclination are independent, so their product is taken. ΔD constrains the measured surface to be nearly flat, and a 6° tilt angle threshold limits the small tilt angle range to ensure the Lambertian plane assumption holds. Exceeding the boundary avoids model failure and overcompensation. When any of the above conditions exceeds the boundary, tilt compensation stops, and only the original distance and attitude data are recorded.

[0116] The processor then combines the temperature compensation coefficient k obtained from the optical cavity temperature collected by the NTC temperature sensor. T And the voltage compensation coefficient k obtained from the battery voltage collected by the battery voltage detection unit. V Finally, according to the formula A=f(I) 515-校准 ×k ref )×g(D avg )×k tilt ×k T ×k V Calculate the simulated radioactivity. Here, f is a piecewise standard curve function, and the input to f is the equivalent fluorescence signal quantity after reflectance correction.

[0117] The segmented standard curve is further illustrated below: the equivalent fluorescence signal corresponds to a concentration ≤ 5 μg / cm³. 2 When a forced linear fit through the origin is used, the formula is A=k×I. net The linearity R of this interval 2 >0.999, where k is the slope coefficient of the forced least-squares fit through the origin. The equivalent fluorescence signal corresponds to a concentration >5 μg / cm³. 2 At that time, due to the downward bend in the curve caused by high concentration of fluorescence self-absorption, a second-order polynomial was used for fitting, with the formula A=a×I. net 2 +b×I net , where 'a' is the coefficient of the quadratic term obtained from the fitting. (The value is in the range of 4.5–5.5 μg / cm³.) 2 Within the transition interval, the linear fitting weights and polynomial fitting weights change by linear interpolation at 4.5 μg / cm³. 2 The linear fitting weight is 100%, and the polynomial fitting weight is 0%, at 5.5 μg / cm³. 2 The linear fitting weight is 0%, and the polynomial fitting weight is 100%, in order to achieve continuous smoothness of the simulated radioactivity values ​​near the segmentation point.

[0118] The method for establishing the standard curve is explained below: Gradient samples include 0.1, 0.2, 0.5, 1.0, 2.0, 5.0, 8.0, and 10.0 μg / cm³. 2A total of 8 concentration points were used, with the optimized formulation as the standard substance for calibration, and the concentration was calculated based on the equivalent concentration of sodium fluorescein. A uniform 170cm² measurement was adopted. 2 A standard white plate with a surface area and a film thickness of 30 μm was used as the calibration carrier. The calibration environment was 25℃, 50%RH, and 300 lux illuminance. The conversion coefficient was locked when the deviation of three readings from the same plate was less than 1%. During the calibration process, the first channel signal I at each concentration point was recorded simultaneously. 515 Second channel signal I 580 Calculate the ratio R=I at each concentration point. 580 / I 515 The average value of all concentration points is taken as the reference ratio R0 for this batch. The relative standard deviation (RSD) of the R values ​​at each concentration point should be ≤2%; otherwise, the composition is considered inhomogeneous and the calibration is invalid. The signal I after the moving average background is collected at a standard distance of 1.5 cm. eff Calculate I net =I eff -I bg I net Represents net fluorescence signal, I bg The background signal includes the circuit dark current background and the ambient light background.

[0119] A global cubic polynomial distance compensation function was calibrated at six points (0.5, 0.8, 1.2, 1.6, 2.0, and 2.5 cm), within the range of 0.5 to 2.5 cm. A moving average background reference table was established under illuminance conditions of 0, 100, 500, and 1000 lux. Multi-grayscale standard plates (with reflectivities of 10%, 30%, 50%, 70%, 90%, and 98%) were used to collect V values ​​corresponding to each grayscale plate at a standard distance of 1.5 cm. sample Calculate the R value of each grayscale plate. meas A linear regression was performed with the nominal reflectivity to obtain the reflectivity-voltage calibration curve. The slope and intercept of this curve were then stored in the factory reference area. The temperature compensation coefficient k was calibrated in a gradient manner within the range of 0~40℃. T The voltage compensation coefficient k calibrated under multiple voltage conditions V The high-reflectivity material distance correction table established using high-reflectivity metal plates is used to solidify the parameters obtained from the factory calibration, along with the distance compensation function parameters and the moving average background comparison table, into the factory reference area of ​​the non-volatile memory, and R0 is stored in the dynamic correction area of ​​the non-volatile memory.

[0120] The device performs a power-on calibration upon startup. The processor reads the reference phosphor reference value I stored in the factory reference area. std Measured value I of standard reference fluorescent sheet in miniature calibration cavity meas Calculate the calibration coefficient k cal =I std / I meas, will k cal The data is stored in the dynamic correction area. During device operation, temperature-triggered calibration is performed. When the NTC temperature sensor detects a temperature change of ≥±2℃ in the optical cavity, the processor automatically performs rapid segmented calibration during the detection interval (approximately 200ms idle time), updating the temperature compensation coefficient. If the device is currently in continuous detection mode, the calibration process can be interrupted momentarily, and the calibration data will be written during the next idle time. The device also performs dynamic voltage calibration; the processor reads the battery voltage in real time and automatically loads the corresponding voltage correction coefficient. All original factory references are permanently stored in the factory reference area of ​​non-volatile memory. When replacing optical components such as LEDs and photodiodes, only the dynamic correction parameters are remeasured; complete gradient calibration is not required.

[0121] Temperature compensation coefficient k T Calibration Example: Calibration temperature points of 0℃, 10℃, 20℃, 25℃, 30℃, and 40℃ were selected within the 0~40℃ temperature range. The equipment was placed in a high and low temperature test chamber, and the optical cavity temperature was allowed to stabilize at the target calibration temperature point and maintained at that temperature for at least 15 minutes. Throughout the calibration process, a standard reference fluorescent sheet within the miniature calibration cavity was used as a stable signal source. The distance between the probe window and the reference fluorescent sheet was fixed at a standard distance of 1.5cm, and the battery was maintained at a fully charged high voltage. The fluorescence signal acquired at the nominal operating point of 25℃ was used as the reference signal I. ref-T Effective fluorescence signals I after background subtraction were collected sequentially at each temperature gradient point. T According to formula k T =I ref-T / I T Calculate the temperature compensation coefficient k corresponding to each calibration temperature point. T , construct temperature -k T The lookup table uses linear interpolation for intermediate temperatures and is then fixed to the factory reference area of ​​the non-volatile memory.

[0122] Voltage compensation coefficient k V Calibration Example: Typical supply voltages at multiple levels within the complete discharge lifespan of the lithium battery were selected as calibration voltage points. During the calibration process, the optical cavity temperature was kept stable at 25°C, and the detection distance was fixed at a standard distance of 1.5 cm. A standard reference fluorescent sheet inside the miniature calibration cavity was used as the signal source. The fluorescence signal collected under the high voltage setting of a fully charged lithium battery was used as the reference signal I. ref-V Each power supply voltage was applied sequentially, and after the voltage stabilized, the effective fluorescence signal I after background subtraction was collected. V According to formula k V =I ref-V / I V Calculate the voltage compensation coefficient k corresponding to each voltage level. V Establish voltage -k V The lookup table uses linear interpolation for the intermediate voltage, which is then permanently stored in the factory reference area of ​​the non-volatile memory.

[0123] Distance compensation function g(D) avg Calibration Example: Six calibration distances were selected: 0.5cm, 0.8cm, 1.2cm, 1.6cm, 2.0cm, and 2.5cm. A 170cm calibration carrier was used. 2 A standard white plate with a surface area and a film thickness of 30 μm was used, coated with a fixed equivalent of standard fluorescent material. The calibration environment was fixed at 25℃, 50%RH, and 300 lux illuminance, with the device battery fully charged. The device was positioned perpendicularly to the calibration sample, and adjusted sequentially to various average calibration distances. Multiple samples were taken at each distance point, and background subtraction was performed. The average value was taken to obtain the effective fluorescence signal I at that distance. D The fluorescence signal measured at a standard distance of 1.5 cm was used as the distance reference signal I. ref-D Calculate the original distance compensation factor g for each distance point. point (D)=I ref-D / I D The global cubic polynomial distance compensation function g(D) is obtained by fitting six sets of discrete sample points. avg )=c0+c1×D avg +c2×D avg 2 +c3×D avg 3 Where c0, c1, c2, and c3 are the polynomial coefficients obtained from the fitting; this distance compensation function is only effective within the range of 0.5cm to 2.5cm, and the coefficients are permanently stored in the factory reference area of ​​the non-volatile memory.

[0124] The processor stores the timestamp, simulated radioactivity, four-point distance values, flatness deviation ΔD, tilt angles α and β, raw accelerometer data, temperature, battery voltage, and the current batch baseline ratio R0 in non-volatile memory and an independent power-down cache chip. Training logs are saved even when power is off, and can be exported as training and evaluation reports via wired or wireless methods.

[0125] Training evaluation is based on four objective quantitative indicators of operational behavior dimensions, statistically analyzed from stored raw detection data, combined with "whether simulated contamination was detected" to complete a comprehensive score: ① Distance interval percentage: D avg① The percentage of the total scan time within the 0.5~2.5cm effective range reflects the standardization of distance control; a higher percentage results in a higher score. ② Total time of excessive tilt of the detection window: the cumulative duration when any of the following conditions are met: ΔD>8mm, α>6°, β>6°; a higher value results in a higher deduction. ③ Number of dangerous proximity events to the detection window: each time the software power reduction protection is triggered when any ToF distance is less than 0.3cm is counted as one event (distinct from firmware hard latch protection triggered by simultaneous proximity of four points); a higher number of events results in a higher deduction. ④ Scanning process jitter variance: the statistical variance calculated based on the raw triaxial acceleration data, characterizing the severity of handheld device jitter; a higher value results in a higher deduction. The device can completely retain the raw data and supports playback of the entire scanning process.

[0126] The default standby / sleep duration is 90 seconds. When continuous training mode is enabled, the sleep duration is extended to 300 seconds. The device enters low-power sleep mode when there is no fluorescent signal and no displacement. When the multi-point ranging module detects an approaching object or a button press, the device quickly wakes up and resumes standby within ≤50ms.

[0127] The following are several specific numerical examples (in each example, the baseline ratio R0 of the current batch composition is 0.280, and the temperature compensation coefficient k is given under full charge at 25℃). T and voltage compensation coefficient k V Both are 1):

[0128] Example 1 (Standard normal testing conditions):

[0129] The four-point ToF distance measurements were 1.2cm, 1.3cm, 1.25cm, and 1.35cm, with an average distance D. avg =1.275cm, within the range of 0.5~2.5cm. The distance between the four points shows a gentle gradient, the flatness deviation ΔD=1.5mm, the tilt angles α and β are close to 0°, and the light source is normally turned on. The first channel acquires the raw signal I. 515-raw =150, background signal I 515-bg =5, I after deducting the background 515 =145; Second channel acquires raw signal I 580-raw =45.6, background signal I 580-bg =5, I after deducting the background 580 =40.6. Measured ratio R 实测 =40.6 / 145=0.280, matching R0, after ratio calibration I 515-校准 =145, this value is higher than the 3σ detection limit. Distance compensation function g(D) avg =0.99, tilt compensation coefficient k tilt=0.95, the slope coefficient of the piecewise standard curve for the low concentration segment is k=0.0705. Simulated radioactivity A=0.0705×145×0.99×0.95≈9.70 Bq / cm³ 2 The activity is less than 10 Bq / cm³. 2 Alarm threshold, output pulse ticking sound.

[0130] Example 2 (Dark carrier causes ratio shift, not exceeding 30% threshold, normal ratio calibration output alarm activity):

[0131] The four-point ToF distance measurements were 1.10cm, 1.18cm, 1.06cm, and 1.22cm, with an average distance D. avg =1.14cm, within the range. The distance between the four points shows a gentle gradient, the flatness deviation ΔD=1.6mm, and the tilt angles α≈2.2° and β≈1.5°, satisfying the tilt compensation activation conditions. First Channel I 515 =137, Second Channel I 580 =29.25, measured ratio R 实测 ≈0.213, relative deviation from R0 is approximately 23.9%, which does not exceed the ±30% alarm threshold. After execution ratio calibration, I 515-校准 ≈180.14. Distance compensation g=0.98, tilt compensation coefficient k tilt =cos22.2°×cos21.5°≈0.997, k=0.0705 in the low concentration range. Simulated radioactivity A=0.0705×180.14×0.98×0.997≈12.43Bq / cm³ 2 The activity is higher than 10 Bq / cm 2 When the alarm threshold is reached, the device triggers a high-frequency audible and visual alarm.

[0132] Example 3 (Detector window tilt exceeds limit, tilt compensation disabled, distance valid, no alarm activity output):

[0133] The four-point ToF distance measurements were 0.80cm, 1.85cm, 0.86cm, and 1.91cm, with an average distance D. avg =1.355cm, within the range. Flatness deviation ΔD=11.1mm, tilt angle α≈7.4°, exceeding the tilt compensation judgment boundary, tilt compensation is stopped, only the original attitude data is recorded. First channel I 515 =119, Second Channel I 580 =28.32, measured ratio R 实测 ≈0.280 matches R0, I after ratio calibration 515-校准 =119. Distance compensation g=1.02, no tilt compensation coefficient k is introduced due to excessive tilt. tiltThe simulated radioactivity A = 0.0705 × 119 × 1.02 ≈ 8.56 Bq / cm³ 2 If the activity level is below the alarm threshold, the screen will display an excessive tilt status warning, and this duration will be included in the total duration statistics of excessive tilt of the detection window.

[0134] Example 4 (Local proximity triggers software power reduction protection; average distance less than 0.5cm, distance invalid, no activity output):

[0135] The four-point ToF distance measurements were 0.22cm, 0.42cm, 0.38cm, and 0.46cm, respectively, with an average distance D. avg =0.37cm, less than the effective lower limit of 0.5cm. For single-point distance measurement of 0.22cm, which is less than 0.3cm, software power reduction protection is triggered, the excitation light source power is reduced to 30% of the drive power corresponding to the current motion state, a low-frequency buzzer is emitted, and the count of dangerous close-in events on the detection window is incremented by 1; if not all four points are less than 0.2cm, firmware hard latch cutoff is not triggered. Distance validity determination D avg <0.5cm, the distance is considered too close, the reading is invalid, the display unit outputs an invalid flag "---", and no simulated radioactivity value is output. This time period is not included in the effective duration of the distance interval. When the detection window is raised, and the distance between the four points is ≥0.3cm and remains stable for 200ms, the power reduction protection is automatically released; D avg Once the reading returns to the 0.5~2.5cm range, the device resumes normal detection output.

[0136] Example 5 (Fluorescent tracer batch failure, R) 实测 (If the value exceeds ±30% threshold, an alarm will be triggered and detection output will be paused.)

[0137] The four-point ToF distance measurements were 1.22cm, 1.28cm, 1.24cm, and 1.30cm, with an average distance D. avg =1.26cm, within the range. The distance between the four points shows a gentle gradient, the flatness deviation ΔD=8mm, and the tilt angles α and β are both less than 6°, meeting the tilt compensation activation conditions. First Channel I 515 =125, Second Channel I 580 =17.0, measured ratio R 实测 =17.0 / 125=0.136, the relative R0 deviation reaches 51.4%, exceeding the ±30% threshold. The equipment outputs an alarm indicating that the composition batch is incompatible or the consumable is invalid, pausing the activity test output. The test can only be resumed after the correct batch R0 is re-entered.

[0138] Example 6 (The tested surface has no simulated contamination, meets the 3σ noise rejection criterion, and the forced output A = 0.00 Bq / cm) 2 ):

[0139] The four-point ToF distance measurements were 1.15cm, 1.20cm, 1.12cm, and 1.23cm, with an average distance D. avg =1.175cm, within the range. The distance between the four points shows a gentle gradient, the flatness deviation ΔD=11mm, and the tilt angles α≈3.1° and β≈2.4°. First channel I 515-raw =8 minus the background I 515 =3, Second Channel I 580-raw =7 minus the background I 580 =2, the signals after background subtraction from both channels are simultaneously below their respective 3σ noise detection thresholds. A 3σ noise cancellation judgment is performed, forcibly outputting a simulated radioactivity activity A = 0.00 Bq / cm². 2 Without triggering pollution-related audio and visual prompts, the raw data is stored in the training log.

[0140] Example 7 (High-concentration simulated pollution triggers second-order polynomial piecewise standard curve calculation of activity):

[0141] The four-point ToF distance measurements were 1.21cm, 1.29cm, 1.24cm, and 1.32cm, respectively, with an average distance D. avg =1.265cm, within the range. The distance between the four points shows a gentle gradient, the flatness deviation ΔD=11mm, and the tilt angles α≈2.8° and β≈2.1°, satisfying the tilt compensation activation conditions. First Channel I 515 =255, Second Channel I 580 =66.4, measured ratio R 实测 ≈0.260, relative R0 deviation is about 7.1%, which does not exceed the ±30% alarm threshold. After ratio calibration, I 515-校准 ≈274.62. Distance compensation g=0.99, tilt compensation coefficient k tilt =cos 2 2.8°×cos 2 2.1°≈0.996. The equivalent net fluorescence signal after conversion corresponds to a concentration greater than 5 μg / cm³. 2 Entering the high-concentration second-order polynomial fitting interval, substituting the polynomial coefficients a=-0.00012, b=0.082, we first calculate the equivalent net signal after distance and tilt compensation: I net =274.62 × 0.99 × 0.996 ≈ 270.40. Simulated radioactivity A = a × I net 2 +b×I net =-0.00012×270.40 2 +0.082×270.40≈13.40Bq / cm 2 If the activity level exceeds the alarm threshold, the device will trigger a high-frequency audible and visual alarm.

[0142] The device in this embodiment is equipped with a liquid fluorescent tracer composition. This composition, by weight percentage, comprises 0.005wt%~0.05wt% sodium fluorescein, 0.05wt%~0.5wt% hydroxypropyl-β-cyclodextrin, 0.0005wt%~0.005wt% rhodamine B, 0.01wt%~0.1wt% Tween-80, 0.12wt%~0.60wt% phosphate buffer pair, 0.005wt%~0.02wt% vitamin C, 0.01wt%~0.025wt% methylparaben, with the balance being deionized water. The phosphate buffer pair comprises sodium dihydrogen phosphate and disodium hydrogen phosphate in a molar ratio of 1:4 to 1:16.

[0143] The composition is a nearly colorless or very pale pink transparent liquid at 25°C, with a pH of 7.0–8.0, a viscosity of 1.0–5.0 mPa·s, an excitation peak of 490–495 nm, and emission peaks of 515–520 nm and 575–585 nm. After being sprayed onto an object surface and dried, the composition forms a very pale pink, semi-transparent fluorescent film layer, making the contamination boundary difficult to discern with the naked eye under normal training lighting conditions. This composition is a preferred fluorescent tracer provided in this embodiment. By inputting the baseline ratio R0 of different batches of the composition, this device can be adapted to various liquid fluorescent tracer compositions with similar fluorescence emission characteristics, and is not limited to the specific formulation described above.

[0144] The preparation process of this composition is as follows: Hydroxypropyl-β-cyclodextrin and one-third of the total water volume of deionized water are weighed according to the formulation amount, and stirred at 300-500 rpm at room temperature until dissolved; sodium fluorescein is added, and the mixture is ultrasonically treated at 38-42℃, 40kHz, and 200W for 20-40 min, followed by stirring at 200-300 rpm for 10-20 min to obtain an inclusion complex solution. Rhodamine B is weighed according to the formulation amount, and after wetting with 20wt% of Tween-80, one-tenth of the total water volume of deionized water is added, and the mixture is ultrasonically treated at 40kHz and 100W for 3-8 min to form a drug-loaded micelle dispersion. Phosphate buffer is weighed according to the formulation amount and dissolved in one-fifth of the total water volume of deionized water, and vitamin C is added and stirred until dissolved to obtain a buffer solution. Under light-protected and room temperature conditions, the inclusion complex solution, drug-loaded micelle dispersion, and buffer solution were mixed sequentially, and the remaining Tween-80 solution was added and stirred. Methylparaben dissolved in 0.8wt%~1.0wt% anhydrous ethanol was slowly added dropwise to the mixture while stirring. The volume was brought to a final volume with the remaining deionized water, and stirring was continued at 200~300 rpm for 10~15 min. The mixture was then ultrasonically degassed for 3~5 min to obtain the crude product. The crude product was transferred to a brown glass container and aged at 2~6℃ in the dark for 18~30 hours. After returning to room temperature, visual inspection revealed no crystallization (trace amounts of precipitate were removed by pre-filtration using a 0.45μm filter membrane). The mixture was then sterilized using a 0.22μm hydrophilic polyvinylidene fluoride filter membrane, filled into brown glass bottles under nitrogen protection, headspace purged with nitrogen to replace oxygen, sealed, and stored at 2~8℃ in the dark.

[0145] The working process of this invention is as follows:

[0146] S1. Power-on Self-Test: After the device is powered on, the processor executes a self-test program, checking the operating status of the excitation light source module, photoelectric detection module, multi-point ranging module, attitude sensor, NTC temperature sensor, non-volatile memory, display unit, hardware comparison circuit, battery grading protection circuit, and heating wire. If any module malfunctions, the processor outputs the corresponding fault code through the display unit and audible / visual prompt module. After the self-test passes, the processor reads and loads the standard curve, distance compensation function, and temperature compensation coefficient k from the factory reference area of ​​the non-volatile memory. T The system includes a moving average background comparison table, a high-reflectivity material distance correction table, and a 470nm surface reflectivity correction table; simultaneously, it collects initial temperature and battery voltage, and loads the calibration coefficient k generated from the last power-on or intermittent calibration from the dynamic correction zone. cal The device then measures the dark current background and enters standby mode.

[0147] S2. Batch Parameter Input: The user inputs the baseline ratio R0 of the current batch composition via a barcode scanning interface, manual input interface, or wireless communication interface. The processor performs range and expiration date verification on the input R0, with a reasonable range of 0.100~0.500. After successful verification, R0 is stored in the dynamic correction area of ​​the non-volatile memory and bound to the current batch composition, and the device enters the detection state. The user sprays the liquid fluorescent tracer composition onto the surface of the training object, and after it dries to form a fluorescent film layer, the training operation begins.

[0148] S3. Scanning and Detection: The user holds the device by its handheld part and brings the detection window close to the surface to be measured for scanning. The multi-point ranging module synchronously acquires the distance values ​​d1, d2, d3, and d4 at the four corners of the detection window at a sampling frequency of ≥100Hz. The processor sequentially executes firmware hard latch protection judgment and software power reduction protection judgment, and then calculates the average distance D. avg And perform a distance validity check. When D avg When the light source is within the 0.5cm~2.5cm range, the excitation light source module is lit, and the photoelectric detection module acquires dual-channel fluorescence signals in a three-segment time-division sequence: a background light acquisition stage, a 470nm ultra-short pulse reflectance acquisition stage, and a blue light-excited fluorescence acquisition stage. When D avg When the value is less than 0.5cm or greater than 2.5cm, the display unit outputs an invalid flag and does not output simulated radioactivity value.

[0149] S4. Signal Processing and Activity Output: The processor sequentially performs the following processing on the dual-channel signals acquired by the photoelectric detection module: background subtraction, 3σ noise removal judgment, ratio calibration, reflectivity correction, distance compensation, tilt compensation judgment and calculation, temperature compensation, and voltage compensation. Finally, the simulated radioactivity value is calculated and displayed in real time through the display unit. Specifically, background subtraction uses the background signal acquired every 200ms in the dark segment, which is then filtered using a 5th-order moving average and subtracted from the original dual-channel signals. In the 3σ noise removal judgment, when the signal values ​​after background subtraction in both channels are lower than three times their respective background standard deviations, the simulated radioactivity activity is forcibly output as 0.00 Bq / cm². 2 Ratio calibration is performed according to formula I. 515-校准 =I 515 ×(R0 / R 实测 ) Execution, when R 实测 When the deviation from R0 exceeds ±30%, the detection output is paused and an alarm is issued. The attitude sensor simultaneously records the raw data from the triaxial acceleration and triaxial gyroscope during the scanning process. When the simulated radioactivity exceeds 10 Bq / cm²... 2 At that time, the processor triggers the audio-visual prompt module to output a high-frequency audio-visual alarm.

[0150] S5. Data Storage and Export: All detection data is stored in real-time in a non-volatile memory and an independent power-off cache chip. The detection data includes timestamps, simulated radioactivity, four-point distance values, flatness deviation ΔD, tilt angles α and β, raw accelerometer data, temperature, battery voltage, and the current batch baseline ratio R0. After training, the stored data can be exported as a training assessment report via wired or wireless means.

[0151] The above embodiments are merely illustrative of the principles and effects of the present invention and are not intended to limit the invention. Any person skilled in the art can modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or alterations made by those skilled in the art without departing from the spirit and technical concept disclosed in the present invention should still be covered by the claims of the present invention.

Claims

1. A nuclear radiation surface contamination detection simulation training device, wherein the device as a whole adopts blue light induced fluorescence detection method for simulation training of nuclear radiation surface contamination detection, comprising a shell (1), a detection window, an excitation light source module (9), a photoelectric detection module (6), a multi-point ranging module (5), and a processor, wherein the excitation light source module (9) is used to generate excitation light, and the photoelectric detection module (6) is used to detect fluorescence signals, characterized in that, The multi-point ranging module (5) detects the distance values ​​at multiple locations between the detection window and the surface being measured; The processor calculates the average distance D based on the distance value. avg The effective detection range is defined as 0.5cm≤D. avg ≤2.5cm, D avg When the value exceeds this range, an invalid flag will be forcibly output and no simulated radioactivity value will be output. It also includes a hardware comparison circuit independent of the processor, which synchronously receives the distance signal output by the multi-point ranging module (5) and is connected to the drive circuit of the excitation light source module (9). The hardware comparison circuit includes an RC delay network and an SR latch with the reset terminal connected to the processor. The set terminal of the SR latch receives the distance comparison signal through the delay network. When any distance value is less than 0.3cm, the processor will reduce the driving power of the excitation light source to 30% of the driving power corresponding to the current motion state without interrupting the detection. After the distance is restored to ≥0.3cm and stabilized for 200ms, the driving power will be automatically restored. When all distance values ​​are less than 0.2cm and last for ≥200ms, the SR latch cuts off the drive circuit of the excitation light source module (9) and sends an interrupt signal to the processor. After the entire length is restored to ≥0.6cm and held for 1 second, the processor sends a reset pulse to unlock.

2. The nuclear radiation surface contamination detection simulation training equipment according to claim 1, characterized in that, The housing (1) is an integrally molded structure, including a handheld part (12) and a main body (11). The detection window is located at the bottom of the main body (11), and the area of ​​the detection window is 150~200cm². 2 The detection window is equipped with an optical protective cover.

3. The nuclear radiation surface contamination detection simulation training equipment according to claim 1, characterized in that, The excitation light source module (9) includes an LED excitation light source array with an emission wavelength of 450~500nm. The LED excitation light source array is arranged around the inner side of the detection window and tilts to illuminate the central area. A diffuser plate is provided below the LED excitation light source array. The excitation light source module (9) also includes a constant current drive circuit and a frequency modulator, wherein the frequency modulator outputs a high-frequency square wave modulation signal of 4~5kHz.

4. The nuclear radiation surface contamination detection simulation training equipment according to claim 1, characterized in that, It also includes a reflection correction light source and timing control circuit; The reflection correction light source is a low-power LED with an emission wavelength of 470nm. The driving circuit of the reflection correction light source is connected to an extremely short pulse generation circuit with a pulse width of <1μs. The photoelectric detection module (6) is connected to a high-speed sample and hold circuit with a response time of <500ns. The timing control circuit is connected to the driving end of the excitation light source module (9) and the driving end of the reflection correction light source, respectively, and outputs a time-division staggered driving signal.

5. The nuclear radiation surface contamination detection simulation training equipment according to claim 1, characterized in that, The multi-point ranging module (5) includes four ToF infrared distance sensors (51) and corresponding infrared light-shielding compartments (511). The infrared light-shielding compartments (511) are respectively set at the four corners of the detection window, and each ToF infrared distance sensor (51) is assembled in an independent infrared light-shielding compartment (511). The four ToF infrared distance sensors (51) are connected to a hardware synchronization trigger circuit, which outputs a synchronization trigger signal with a 10ns micro-timing misalignment.

6. The nuclear radiation surface contamination detection simulation training equipment according to claim 1, characterized in that, A miniature calibration cavity with a built-in standard reference fluorescent sheet is provided on one side of the detection window.

7. A control method for the device as described in any one of claims 1-6, characterized in that, Includes the following steps: S1. Distance Detection: Distance values ​​between the detection window and the measured surface at multiple locations are detected using multi-point ranging, and the average distance D is calculated. avg ; S2. Distance validity determination: The effective detection range is locked as 0.5cm≤D avg ≤2.5cm, when D avg When the value is less than 0.5cm or greater than 2.5cm, an invalid label will be forcibly output, and the simulated radioactivity value will not be output. S3. Hierarchical Safety Control: When any distance value is less than 0.3cm, the driving power of the excitation light source is reduced to 30% of the driving power corresponding to the current motion state without interrupting detection. When all distance values ​​less than 0.3cm recover to greater than or equal to 0.3cm and remain stable for 200ms, the power reduction protection is automatically released. When all distance values ​​are less than 0.2cm and the duration is greater than or equal to 200ms, the SR latch in the hardware comparison circuit, independent of the processor, is activated to directly cut off the driving circuit of the excitation light source module and send an interrupt signal to the processor. When all distance values ​​recover to greater than or equal to 0.6cm and remain stable for 1s, the processor sends a reset pulse to the SR latch of the hardware comparison circuit to unlock it. S4, Activity Output: 0.5cm≤D avg Within an effective detection range of ≤2.5cm, the simulated radioactivity value is calculated and output based on the output signal of the photoelectric detection module.

8. The control method according to claim 7, characterized in that, In step S1, distance values ​​at the four corners of the detection window are synchronously acquired at a sampling frequency of ≥100Hz. Based on the preset coordinates of the four corners and the corresponding distance values, the spatial plane of the measured surface is fitted using the least squares method to calculate the average distance D. avg The tilt angles α and β in the two orthogonal directions, and the flatness deviation ΔD; Gradient-discriminatory abrupt change filtering is performed on the four-point distance values: when the distance between the four points shows a monotonically increasing or decreasing linear gradient and the residuals of each measurement point relative to the fitting plane all fall within the sensor noise range of ±2mm, all data participate in the plane fitting. When the distance difference between a single data point and two adjacent measurement points on the side is >5mm and remains stable for more than 3 frames, and the difference between the maximum and minimum values ​​of the other three points is ≤0.3mm, the single point is removed and recalculated. When two or more distance anomalies occur and there is no linear gradient, the two middle sets of valid distance measurement data are retained after sorting by distance value for local fitting.

9. The control method according to claim 7, characterized in that, Step S4 includes: Simultaneously acquire signals in the first channel (510-550nm band) and the second channel (570-590nm band); The background signal is collected every 200ms in the dark segment and then subtracted from the original dual-channel signal after being filtered by a 5th-order moving average. When the signal values ​​of both channels after subtracting the background are both lower than three times the standard deviation of their respective background signals, they are determined to be undetectable and the activity is forced to be zero.

10. The control method according to claim 7, characterized in that, In step S4, the activity calculation uses a piecewise standard curve: Equivalent fluorescence signal corresponding concentration ≤5μg / cm 2 When a forced linear fit is used, the formula is A=k×I. net Where k is the slope coefficient, I net Net fluorescence signal; Corresponding concentration >5μg / cm 2 A second-order polynomial fitting is used, with the formula A = a × I. net 2 +b×I net , where a and b are fitting coefficients; Between 4.5 and 5.5 μg / cm 2 The weights for linear fitting and polynomial fitting change by linear interpolation within the transition interval.

Images