Passive nucleon level gage and level output control method and monitoring system thereof during nuclear radiation flaw detection

Abstract

The invention discloses a passive nucleon level gage and a level output control method and a monitoring system thereof in a nuclear radiation flaw detection period. The level output method comprises the following steps: judging that nuclear radiation flaw detection interference exists in a measurement environment in which the passive nucleon level gage is positioned; executing at least one of the following output guarantee operations: setting an output switch of the passive nucleon level gage to be in a preset guarantee state, wherein the preset guarantee state is a switch-on state or a switch-off state; or the output switch of the passive nucleon level gage is maintained in the on-off state before it is judged that nuclear radiation flaw detection interference exists; or the normal material level value of the passive nucleon level gauge is replaced with the alternative material level value, or the on-off state of an output switch of the passive nucleon level gauge is determined based on the alternative material level value, and the alternative material level value is used for replacing the normal material level value when flaw detection interference is detected. The normal material level value is a material level value generated on the basis of the real-time measurement data in the state that the guarantee operation is not executed. The device realizes material level output guarantee control during nuclear radiation flaw detection.

Description

Technical Field

[0001] This invention relates to the field of passive nuclear level measurement, and in particular to a passive nuclear level gauge and its level output control method and monitoring system during nuclear radiation flaw detection. Background Technology

[0002] Passive nuclear level gauges are sensors that measure the level or quantity of materials by utilizing the nuclear radiation released from radioactive substances contained in the material. Their core principle is based on measuring the material level by measuring the nuclear radiation in the measurement space. Currently, they are widely used in industrial fields such as thermal power plants, steel mills, and waste incineration.

[0003] In actual measurements, passive nuclear level gauges operate in environments containing not only nuclear radiation released from materials but also background nuclear radiation from the cosmic environment. Furthermore, in industrial applications, nuclear radiation testing of equipment is sometimes necessary, and this testing process releases extremely strong nuclear radiation.

[0004] When a passive nuclear level gauge is measured in an environment, especially at close range (e.g., tens of meters), nuclear radiation testing can severely affect the nuclear radiation data measured by the gauge. Even at distances of hundreds of meters, nuclear radiation testing can significantly impact the data. When nuclear radiation testing is strong, very close, or lacks obstruction, the actual nuclear radiation emitted by the material is much less than that emitted by the testing equipment, severely masking the true material signal. In industrial settings, the presence of nuclear radiation testing can trigger a high-level alarm on the passive nuclear level gauge or result in a falsely high-level reading.

[0005] During nuclear radiation testing, the impact of nuclear radiation testing on passive nuclear level gauges varies greatly due to differences in testing distance, operational radiation, type of radiation source, and equipment environment. Sometimes it highly overlaps with normal material measurement data; sometimes it causes severe vibrations; sometimes it fluctuates at low frequencies; and sometimes it causes a sharp increase in the measured value. Furthermore, because the effects of nuclear radiation vary at different locations, passive nuclear level gauges located far from the testing source or with multiple obstacles making it difficult to identify the nuclear radiation released from the material within the radiation coverage of the testing. The interference caused by nuclear radiation testing is highly complex, unlike easily identifiable high-frequency oscillation interference such as power supply interference or device quality degradation.

[0006] In existing technologies, such as the method for improving the signal-to-noise ratio of passive nuclear level gauges disclosed in Chinese invention patent application CN102706409A, the main idea is to use data from external radiation sensors or multiple passive nuclear level gauges to extract environmental radiation, and then correct the measurement data of the passive nuclear level gauges for environmental radiation changes. This patent application proposes three typical methods: Method 1: Reference method, using a gamma-ray detector to specifically measure the background radiation outside the container and using it to correct other measuring level gauges; Method 2: Multi-point balance calculation method; Method 3: Characteristic ray method. From the perspective of technical application scenarios, the above three methods mainly address the problem of environmental radiation correction, rather than specifically for nuclear radiation flaw detection scenarios. Under the complex working conditions of nuclear radiation flaw detection, due to the diversity of flaw detection radiation and the spatial differences of the passive nuclear level gauges, the above methods have certain limitations in practical applications and cannot directly meet the complex anti-interference requirements of nuclear radiation flaw detection scenarios.

[0007] In addition, Chinese invention patent application CN104359527A discloses a power management method for a passive nuclear level gauge. Its main purpose is to solve the problem of faults in the passive nuclear level gauge itself. This method reduces the problem of internal temperature rise in the passive nuclear level gauge by cutting off or connecting the power supply to the passive nuclear level gauge.

[0008] For integrated passive nuclear level gauges, even with analog-to-digital conversion (ADC) sampling technology, it is virtually impossible to distinguish between nuclear radiation emitted by the material and nuclear radiation emitted during flaw detection, especially when the nuclear radiation flaw detection distance is far from the passive nuclear level gauge, the radiation intensity of the flaw detection is low, and the nuclear radiation emitted by the flaw detection is similar to that emitted by the material.

[0009] For changes in background nuclear radiation caused by the presence of radioactive sources in the cosmic environment, material accumulation at the site, and the measurement environment, existing technologies can effectively obtain accurate material levels through technical means, such as correction or compensation. Interference from nuclear radiation flaw detection is fundamentally different from background radiation changes caused by the presence of radioactive sources in the environment, changes in cosmic background radiation, and material accumulation at the site.

[0010] In real-world industrial scenarios, passive nuclear level gauges are often used as control signals rather than warning signals in many processes. However, due to interference from nuclear radiation during flaw detection, passive nuclear level gauges can cause measurement errors, leading to chaotic control commands. This directly affects the operation of the entire process system and can even cause the entire process system to shut down, resulting in significant impacts.

[0011] It should be noted that in routine flaw detection inspections in the industrial field, nuclear radiation flaw detection is usually carried out at night and ends in the early morning because it causes nuclear pollution to the environment. This interference period increases the difficulty of detecting abnormalities in level gauges. Once the equipment loses normal control as a result, it will seriously affect the orderly operation of industrial production.

[0012] The disclosure of the above background technical content is only for the purpose of assisting in understanding the concept and technical solution of this application, and does not necessarily provide technical instruction. Summary of the Invention

[0013] The purpose of this invention is to solve the production safety problems caused by interference with passive nuclear level gauges during nuclear radiation flaw detection operations in industrial sites, which prevents them from accurately detecting material levels, and the material level data from being effectively corrected or compensated, leading to false alarms, equipment malfunctions, or material level monitoring failures.

[0014] To achieve the above objectives, the technical solution adopted by the present invention is as follows: A method for controlling the level output of a passive nuclear level gauge during nuclear radiation flaw detection includes the following steps: It was determined that nuclear radiation flaw detection interference existed in the measurement environment where the passive nuclear level gauge was located. Perform at least one of the following output guarantee operations: Set the output switch of the passive nuclear level gauge to a preset protection state, wherein the preset protection state is the switch-on state; Alternatively, the output switch of the passive nuclear level gauge can be set to a preset protection state, wherein the preset protection state is the switch open state; Alternatively, the output switch of the passive nuclear level gauge can be maintained in the switching state before the presence of nuclear radiation flaw detection interference is detected. Alternatively, the normal level value of the passive nuclear level gauge can be replaced with a substitute level value, wherein the normal level value is the level value generated by the passive nuclear level gauge based on real-time measurement data when the output protection operation is not performed; and the substitute level value is the level value used to replace the normal level value when nuclear radiation flaw detection interference is detected. Alternatively, the switching state of the output switch of the passive nuclear level gauge can be determined based on the alternative level value, wherein the alternative level value is the level value used to replace the normal level value when nuclear radiation flaw detection interference is detected, and the normal level value is the level value generated by the passive nuclear level gauge based on real-time measurement data when the output protection operation is not performed.

[0015] Furthermore, following any one or a combination of the aforementioned technical solutions, the passive nuclear level gauge level output control method provided by this invention during nuclear radiation flaw detection further includes: The nuclear radiation flaw detection interference in the measurement environment of the passive nuclear level gauge has ended. Cancel the output guarantee operation in progress.

[0016] Furthermore, if any or a combination of the aforementioned technical solutions meets the preset first condition, it is determined that there is nuclear radiation flaw detection interference in the measurement environment where the passive nuclear level gauge is located. If the preset second condition is met, it is determined that the nuclear radiation flaw detection interference in the measurement environment where the passive nuclear level gauge is located has ended.

[0017] Furthermore, based on any one or a combination of the aforementioned technical solutions, the presence of nuclear radiation flaw detection interference in the measurement environment of the passive nuclear level gauge is determined by at least one of the following first conditions: The passive nuclear level gauge acquires nuclear radiation measurement values ​​within one or more preset time windows, and determines statistical characteristics and / or changes in statistical characteristics based on the nuclear radiation measurement values. When the statistical characteristics and changes in statistical characteristics meet the corresponding preset third condition; And / or, the nuclear radiation measurement value of the passive nuclear level gauge reaches the preset first radiation threshold; And / or, the operating time of the passive nuclear level gauge experiences a preset first duration; And / or, the current time reaches the preset time to start flaw detection; And / or, the passive nuclear level gauge or its control system receives a confirmation signal from an external device indicating that flaw detection has been initiated.

[0018] Furthermore, following any one or a combination of the aforementioned technical solutions, the statistical feature is at least one of the following: measured value, average value, extreme value, extreme value difference, standard deviation, variance, quantile, interquartile range, or mode.

[0019] Furthermore, based on any one or a combination of the aforementioned technical solutions, the preset third condition is at least one of the following: The statistical characteristics of one or more radiation measurements of the passive nuclear level gauge exceed the preset first characteristic range within a preset time window; And / or, the change in the statistical characteristics of the nuclear radiation measurement values ​​of the passive nuclear level gauge in the current time window compared to the statistical characteristics of the nuclear radiation measurement values ​​in the previous time window exceeds the preset second characteristic range; And / or, the absolute value of the statistical characteristics of the nuclear radiation measurement values ​​of the passive nuclear level gauge exceeding the preset third characteristic range the number of times; And / or, the variation between the maximum and minimum values ​​of the nuclear radiation measurements corresponding to multiple preset time windows by the passive nuclear level gauge exceeds the preset fourth characteristic range; And / or, based on the statistical characteristics of the nuclear radiation measurements corresponding to multiple preset time windows of the passive nuclear level gauge, calculate the change in the statistical characteristics corresponding to each two adjacent time windows, and the number of times the change exceeds the preset fifth characteristic range meets the standard.

[0020] Furthermore, based on any one or a combination of the aforementioned technical solutions, the nuclear radiation flaw detection interference in the measurement environment of the passive nuclear level gauge is determined to have ended by at least one of the following second conditions: From the moment nuclear radiation flaw detection interference is detected or output protection operation is executed, a preset second duration will elapse; And / or, the current time reaches the preset stop flaw detection time; And / or, acquire the nuclear radiation measurement values ​​of the passive nuclear level gauge within one or more preset time windows, determine statistical characteristics and / or changes in statistical characteristics based on the nuclear radiation measurement values, and when the statistical characteristics and changes in statistical characteristics meet the corresponding preset fourth condition; And / or, the passive nuclear level gauge or its control system receives a confirmation signal from an external device indicating the end of flaw detection; And / or, the alternative material level value generated based on the real-time generated real-time alternative material level calculation formula reaches the preset material level threshold, wherein the real-time alternative material level calculation formula is a real-time prediction model established based on historical material level and time data within a preset time before the determination of nuclear radiation flaw detection interference; And / or, the alternative material level value generated based on the pre-built prediction model reaches the preset material level threshold, wherein the prediction model is a prediction model pre-built based on historical real material level and time data within at least one process cycle.

[0021] Furthermore, based on any one or a combination of the aforementioned technical solutions, the preset fourth condition is at least one of the following: The statistical characteristics of the radiation measurement values ​​of the passive nuclear level gauge in one or more consecutive sampling cycles are restored to the preset first characteristic range; And / or, the number of times the statistical characteristics of the radiation measurement values ​​of the passive nuclear level gauge recover to the preset characteristic range in multiple sampling cycles meets the standard.

[0022] Furthermore, following any one or a combination of the aforementioned technical solutions, the output protection operation in progress is to replace the normal level value of the passive nuclear level gauge with a substitute level value, wherein the substitute level value is configured to be determined according to a preset protection strategy; After the nuclear radiation flaw detection interference ends and the ongoing output assurance operation is canceled, the following also applies: The actual level data measured by the passive nuclear level gauge is compared and analyzed with the alternative level values ​​used in the output support operation. The preset protection strategy is adaptively optimized based on the comparison analysis results.

[0023] Furthermore, based on any one or a combination of the aforementioned technical solutions, the alternative material level value is determined through at least one of the following safeguard strategies: The corresponding material level value is preset for the alternative material level value and stored; The normal material level value at the last moment before the determination of nuclear radiation flaw detection interference was used as the alternative material level value. The statistical characteristic value within a preset time period before the determination of nuclear radiation flaw detection interference is used as the alternative material level value; After determining that there is nuclear radiation flaw detection interference, at least one unit time window of radiation measurement values ​​are collected as a candidate sample set. Based on the historical measurement data before the nuclear radiation flaw detection interference, specific measurement values ​​or local feature values ​​that meet the preset deviation conditions are selected from the candidate sample set and used as the alternative material level values. The real-time replacement material level calculation formula is used to calculate the replacement material level value. The real-time replacement material level calculation formula is a real-time prediction model established based on historical material level and time data within a preset time before the determination of nuclear radiation flaw detection interference. The predicted material level value generated by the prediction model pre-built based on historical real material level and time data within at least one process cycle is used as the alternative material level value.

[0024] Furthermore, following any one or a combination of the aforementioned technical solutions, the pre-built prediction model includes any of the following methods: The prediction model is a fitting function obtained by fitting or regression analysis of historical material level and time data; The prediction model calls upon historical material level and time data classified according to process in the absence of nuclear radiation flaw detection interference to determine the material level value that matches the current process and time period as the alternative material level value; The prediction model is a function model constructed based on historical operating data according to process classification; The artificial intelligence model is trained using historical operating data to obtain the prediction model used to predict material level.

[0025] Furthermore, following any one or a combination of the aforementioned technical solutions, the pre-built prediction model is integrated inside the passive nuclear level gauge, or the pre-built prediction model is set in an external device that communicates with the passive nuclear level gauge.

[0026] Furthermore, based on any one or a combination of the aforementioned technical solutions, the real-time replacement material level calculation formula is a fitted function relationship: L=aT2 +bT+c, where L is the material level, T is the time, and a, b, c are fitting constants.

[0027] Furthermore, following any one or a combination of the aforementioned technical solutions, the starting time point of the historical material level and time data within the preset time period is determined by any of the following methods: The starting time point is defined as the trigger time of the start signal for the current process cycle. The starting time point is the moment when the material level reaches the preset material level threshold. The starting time point is the moment when the trigger signal sent by the external device is received. Based on the preset time period based on the patterns of historical flaw detection events, the moment when nuclear radiation flaw detection interference is determined to be earlier is moved forward to obtain the starting time point.

[0028] Furthermore, if any one or a combination of the aforementioned technical solutions meets any of the following preset conditions, the output guarantee operation will be prohibited: The measurement results of the passive nuclear level gauge are within the preset safety range; The measurement results of the passive nuclear level gauge are within the preset fault range; The passive nuclear level gauge displays and / or outputs a warning level for the material level. The output protection function of the passive nuclear level gauge or its control system is turned off. The current time is within the preset normal working hours.

[0029] According to another aspect of the present invention, a passive nuclear level gauge is provided, comprising a radiation detector, a processor, and a signal output unit, wherein: The radiation detector is configured to collect nuclear radiation in the measurement environment; The processor is configured to determine whether there is nuclear radiation detection interference in the measurement environment based on the nuclear radiation collected by the radiation detector or based on a preset first condition. If so, the processor executes a preset output protection operation to generate corresponding switch signals, digital signals, and / or analog signals; otherwise, the processor generates corresponding switch signals, digital signals, and / or analog signals based on the nuclear radiation collected by the radiation detector; and the signal result generated based on the output protection operation is different from the signal result generated based on the collected nuclear radiation. The signal output unit is configured to output switch signals, digital signals, and / or analog signals generated by the processor.

[0030] Furthermore, following any one or a combination of the aforementioned technical solutions, the processor is configured to perform the steps of the material level output control method during nuclear radiation testing as described above.

[0031] Furthermore, following any or a combination of the aforementioned technical solutions, the processor is further configured to: when performing the output protection operation, the processor generates and stores a status flag bit, which is used to identify that the signal currently output by the signal output unit is an output based on the protection strategy under nuclear radiation flaw detection interference; And / or, the signal output unit is further configured to output the status flag bit; And / or, the passive nuclear level gauge further includes a communication interface configured to communicate with the signal output unit and an external device.

[0032] Furthermore, following any one or a combination of the aforementioned technical solutions, the signal output unit includes one or more of the following: Isolated digital output channels; Analog output circuit for outputting 4-20mA or 0-10V analog signals; Digital output circuit used to output digital level signals.

[0033] According to another aspect of the present invention, an industrial material level monitoring system is provided, comprising: The passive nuclear level gauge described above; An external control device, communicatively connected to the passive nuclear level gauge, is configured to perform the steps of the level output control method described above during nuclear radiation testing, or the external control device is configured to send at least one of the following signals to the passive nuclear level gauge: A confirmation signal used to determine the presence of nuclear radiation flaw detection interference in the measurement environment of a passive nuclear level gauge; A confirmation signal used to determine the end of interference with nuclear radiation flaw detection; A process synchronization signal used to synchronize historical material level and time data at the start time point.

[0034] Furthermore, based on any or a combination of the aforementioned technical solutions, the external control device is a distributed control system or a programmable logic controller; And / or, the passive nuclear level gauge is configured to upload a status flag bit and / or output mode information to identify the current output guarantee operation state via a communication interface, and the external control device is configured to display the information differently on the monitoring screen or generate corresponding operation logs based on the uploaded information.

[0035] The beneficial effects of the technical solution provided by this invention are as follows: a. Passive nuclear level gauges can still provide safe, continuous, and meaningful system outputs even when nuclear radiation testing interference prevents accurate material level measurement; ensuring the continuity of equipment operation in the process flow. This prevents equipment from malfunctioning and stopping due to alarm signals triggered by interference with the passive nuclear level gauge during nuclear radiation testing. b. During nuclear radiation flaw detection operations, the passive nuclear level gauge automatically enters the output protection state, enabling it to autonomously enter a controllable output protection state when interference occurs during flaw detection; at the end of nuclear radiation flaw detection, the passive nuclear level gauge can automatically return to normal measurement state, thus automatically resuming normal measurement after the flaw detection operation is completed, thereby improving the stability of level detection in industrial production processes and ensuring production safety; it significantly avoids personnel being harmed by nuclear radiation by adjusting settings on-site during nuclear radiation flaw detection. c. After a flaw detection occurs, maintain the output protection status, keep the original alarm status, automatically cancel or enable the alarm function to avoid the forced shutdown of process equipment due to nuclear radiation flaw detection; or generate alternative material levels based on prediction functions and time parameters to replace the actual measured material levels to avoid production safety issues. d. Ensure system safety. Prevent safety issues arising during nuclear radiation testing due to passive nuclear level gauges failing to measure material radiation or generating incorrect level readings, thus failing to determine if the material has reached a high level. e. By utilizing the planned and scheduled time characteristics of nuclear radiation flaw detection, the automatic activation and deactivation of fault-tolerant measurements during nuclear radiation flaw detection can be achieved, greatly reducing the workload of staff; and information manageability can be achieved, making the status transparent and the output knowable and controllable. Attached Figure Description

[0036] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0037] Figure 1 A schematic flowchart of a method for controlling the level output of a passive nuclear level gauge during nuclear radiation flaw detection, provided as an exemplary embodiment of the present invention; Figure 2 A schematic diagram of the operation process during and after nuclear radiation flaw detection interference is provided as an exemplary embodiment of the present invention; Figure 3A schematic flowchart of a material level output method with an output protection operation that is prohibited from being performed is provided as an exemplary embodiment of the present invention; Figure 4 A schematic block diagram of a passive nuclear level gauge provided as an exemplary embodiment of the present invention; Figure 5 A schematic block diagram of an industrial material level monitoring system provided as an exemplary embodiment of the present invention. Detailed Implementation

[0038] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0039] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, apparatus, product, or device that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or devices.

[0040] Passive nuclear level gauges are sensors that use the nuclear radiation released by radioactive substances contained in materials to measure the material level or quantity. Their core principle is to convert the nuclear radiation measurement value of the measurement space into the material level measurement result.

[0041] In actual measurements, passive nuclear level gauges operate in environments containing not only nuclear radiation released from the material itself, but also background radiation from the cosmic environment and adjacent containers. Furthermore, in industrial applications, nuclear radiation testing of equipment can sometimes be performed, which also releases very strong nuclear radiation.

[0042] Changes in background nuclear radiation caused by changes in the cosmic environment, material accumulation at the site, or the addition of radioactive sources in the measurement environment are all considered steady-state interferences. Existing technologies can effectively correct or compensate for these changes to obtain accurate material levels. The background radiation changes caused by nuclear radiation testing are entirely different from those caused by adding a radioactive source to the measurement environment, as the latter can be corrected using existing techniques for addressing background radiation variations. Interference caused by nuclear radiation testing arises from the addition of a strong radioactive source in the environment. This interference is caused by factors such as irregular opening and closing of the source, movement leading to changes in obstructions in the radiation path of the source and the passive nuclear level gauge, frequent changes in the emission direction of the source, and exchanges in the type, intensity, and type of the source. Nuclear radiation interference caused by nuclear radiation testing is an irregular dynamic interference, fundamentally different from background radiation changes caused by the presence of radioactive sources in the environment, changes in cosmic background radiation, and material accumulation at the site.

[0043] The impact of industrial flaw detection on passive nuclear level gauges in operation depends primarily on the type and intensity of the radioactive source used during flaw detection, the straight-line distance between the source and the gauge, and any obstructions between them. Typically, nuclear radiation emitted during flaw detection within a hundred meters of the gauge significantly affects its operation. For passive nuclear level gauges located in the unobstructed path of the radiation from flaw detection, nuclear radiation interference will prevent normal measurement, hindering the accurate calculation of the level or the generation of an accurate alarm signal indicating the level has reached the warning position. However, when the distance from the flaw detection location is far, or when there are numerous obstructions between the radiation source and the gauge, the impact of flaw detection on the measured values ​​is relatively small. Because the nuclear radiation signal from flaw detection is mixed with the nuclear radiation information released by the measured material, the normal level measurement operation of passive nuclear level gauges is affected, making it impossible to accurately measure the material level. Even if existing passive nuclear level gauges use ADC sampling technology to sample and analyze the obtained nuclear radiation pulses, it is difficult to distinguish between the nuclear radiation information from flaw detection and the nuclear radiation information released by normal material because the nuclear radiation released by the material is emitted by various radioactive substances contained in coal, and the nuclear radiation released by the material and the nuclear radiation from flaw detection highly overlap.

[0044] Meanwhile, in actual nuclear radiation flaw detection processes, frequent switching on and off of radioactive sources, changes in detection direction, movement of personnel within the flaw-detected equipment, or changes in shielding objects in the radiation path due to shifts in the flaw detection location can all lead to more irregular changes in nuclear radiation data caused by nuclear radiation flaw detection, further increasing the difficulty for passive nuclear level gauges to identify and eliminate nuclear radiation flaw detection signals.

[0045] It should be noted that due to the complexity of the impact of nuclear radiation testing on passive nuclear level gauge measurements, it is incorrect to simply understand the measurement impact of nuclear radiation testing as an extremely severe vibration similar to power supply interference, grounding interference, or interference caused by other surrounding electrical equipment. The changes in measurement data caused by nuclear radiation testing interference may be rapid oscillations, slow oscillations, or simply a stable increase in background radiation. Using methods for eliminating electrical interference such as power supply and grounding interference will, in most cases, fail to solve the problem.

[0046] At the same time, it should be noted that due to the complexity of nuclear radiation testing, it is often impossible to distinguish whether the measured data is caused by nuclear radiation released from the material or by the radioactive source used in the testing. Therefore, when nuclear radiation testing occurs, existing techniques such as characterization of measured values, such as simply statistically analyzing the minimum value within a time window or extracting data from the measurement data that closely approximate the nuclear radiation released from the material and meet statistical characteristics, are usually meaningless. Existing techniques struggle to make the measurement results approximate the actual condition of the material through compensation and correction.

[0047] Under current technology, when passive nuclear level gauges are used for nuclear radiation flaw detection, users can typically only watch as the level data generated by the passive nuclear level gauge fluctuates randomly, or helplessly witness frequent alarm triggers. Since many enterprises, such as thermal power plants, conduct flaw detection on boilers and other pipelines at night, and the detection area is inaccessible to non-detection personnel, operating and maintenance personnel far from the detection site can only watch helplessly as the passive nuclear level gauge loses control during nuclear radiation flaw detection, unable to take any action.

[0048] In industrial applications, passive nuclear level gauges typically output digital alarm signals and / or analog continuous level signals. Continuous level signals assist operators in continuously monitoring changes in material levels within containers, while digital alarm signals usually indicate that the material level has reached a dangerous level. However, in many processes, the digital signals output by passive nuclear level gauges are used as control signals for starting and stopping process equipment, thus controlling the operation of equipment within the process system. For example, in a thermal power plant's dry slag silo, if a passive nuclear level gauge is used as the full-sludge control signal to control the opening and closing of the associated slag removal machine, nuclear radiation testing could directly trigger a false full-sludge alarm, causing the system to incorrectly stop the slag removal machine. Incorrect control signals severely impact the safe operation of the unit.

[0049] To address the production safety issues arising from interference during nuclear radiation flaw detection operations in industrial settings, which can lead to inaccurate level detection, ineffective correction or compensation of level data by passive nuclear level gauges, resulting in false alarms, equipment malfunctions, or level monitoring failures, this paper proposes a method for output assurance and recovery of passive nuclear level gauges. This method enables the level gauge to autonomously enter a controllable output assurance state during flaw detection interference and automatically resume normal measurement after the flaw detection operation is completed. This improves the stability of level detection in industrial production processes and ensures production safety.

[0050] It is important to emphasize that the passive nuclear level gauge solution provided by this invention for ensuring level output during nuclear radiation testing is fundamentally incapable of achieving measurement effectiveness against interference from nuclear radiation testing. It cannot identify and eliminate the nuclear radiation detected by the passive nuclear level gauge during nuclear radiation testing, thus failing to accurately measure the nuclear radiation released from the material. When nuclear radiation testing occurs, the method of this invention does not compensate for or correct the measurement results of the passive nuclear level gauge, and it does not solely rely on the measurement data of the passive nuclear level gauge during the testing operation.

[0051] In one embodiment of the present invention, a method for controlling the level output of a passive nuclear level gauge during nuclear radiation flaw detection is provided, such as... Figure 1 As shown, the material level output method includes the following steps: S100: Determine whether there is nuclear radiation flaw detection interference in the measurement environment where the passive nuclear level gauge is located; specifically, determine whether the preset first condition is met. If so, determine whether there is nuclear radiation flaw detection interference in the measurement environment where the passive nuclear level gauge is located, and then execute step S200. If the first condition is not met, then perform normal measurement of the passive nuclear level gauge, that is, generate corresponding switch signals, digital signals and / or analog signals based on the collected nuclear radiation.

[0052] S200: Perform output assurance operation. After the passive nuclear level gauge determines that it is currently in a nuclear radiation flaw detection interference environment and enters the "output assurance phase," its core task is to stop outputting invalid real-time data affected by interference and instead output a stable and relatively reliable alternative level value or a controllable switching signal to prevent erroneous level or switching signals from causing malfunctions in the control system. The output assurance operation includes at least one of the following: T1: Set the output switch of the passive nuclear level gauge to a preset protection state, where the switch is in the on state. The output switch here is the physical signal switch that outputs the switching signal from the passive nuclear level gauge, typically a relay. In practical applications, the relay closing usually means the switch is closed, triggering an alarm signal. However, the relay can also open to trigger an alarm signal. For example, the output switch might be connected to an alarm circuit; when the output switch is on, the alarm circuit is activated; when the output switch is off, the alarm circuit is not activated. In other words, in this T1 output protection operation, if nuclear radiation detection interference is detected, an alarm will be triggered. In actual industrial settings, the inability to accurately measure the material level in a container can lead to extremely serious safety consequences. Therefore, if nuclear radiation detection interference in the industrial setting cannot be eliminated or corrected, it is preferable to shut down the equipment and system to ensure the safety-first production objective. Upon encountering nuclear radiation detection interference, an alarm will be triggered, equipment operation will be stopped, and production safety will be ensured.

[0053] T2: Set the output switch of the passive nuclear level gauge to a preset protection state, which is the switch-off state. In this output protection operation, once nuclear radiation testing interference is detected, even if the actual nuclear radiation of the measured material (excluding radiation generated by the testing process) indicates that it has reached the alarm threshold, no alarm will be triggered. In some industrial sites, the equipment monitored by the passive nuclear level gauge is only an auxiliary or secondary device in the system, and the entire system will not completely stop due to malfunction of the monitored auxiliary device. Therefore, to ensure the safe operation of the overall system, during nuclear radiation testing operations, no alarm actions are forcibly triggered to avoid system shutdown due to malfunctions.

[0054] T3: Maintain the output switch of the passive nuclear level gauge in the state it was in before the detection of nuclear radiation testing interference. In this output protection operation, once nuclear radiation testing interference is detected, if the switch was previously on, it remains on; if it was previously off, it remains off. In actual industrial settings, for large containers where the material inside is expected to change slowly, maintaining the state before nuclear radiation testing often more closely reflects the actual operating conditions.

[0055] In practical applications, the switching unit of a passive nuclear level gauge may indicate an alarm when it is open or when it is closed; the specific choice should be determined based on the actual system configuration.

[0056] Furthermore, in practical applications and designs, during flaw detection interference, whether the output switch of the passive nuclear level gauge should be forced to a preset open or closed state, or forced to the switch state before the flaw detection interference occurred, should be comprehensively considered in conjunction with factors such as the duration of the flaw detection interference, the operating characteristics of the equipment, or which method is the safest.

[0057] Currently, passive nuclear level gauges are widely used as continuous level measurement sensors due to their ability to continuously measure changes in material level within a certain range. However, during nuclear radiation testing, the nuclear radiation data measured by passive nuclear level gauges is not entirely from the material itself; in fact, much of it may be interference radiation from the testing process. Therefore, calculating the material level using normal methods becomes meaningless and seriously misleading. To address this, in this application, after the passive nuclear level gauge enters the output assurance phase following nuclear radiation testing, a substitute level value is used instead of the normal level value. Specifically: T4: Replace the normal level value of the passive nuclear level gauge with a substitute level value. The normal level value is the level value generated by the passive nuclear level gauge under normal measurement conditions based on real-time measurement data. Normal measurement conditions refer to the state before the output safeguard operation is performed. The level value generated based on real-time measurement data follows the existing level measurement principle of passive nuclear level gauges: using a radiation detector to collect nuclear radiation in the measurement environment and using this as measurement data to estimate the level. The substitute level value is used to replace the normal level value when nuclear radiation flaw detection interference is detected. The substitute level value is configured to be determined according to a preset safeguard strategy. In this output safeguard operation, once nuclear radiation flaw detection interference is detected, the substitute level value determined based on safeguard measurement is used as the analog output of the passive nuclear level gauge. As mentioned above, the normal level value usually loses its ability to represent the true level of the material at this time.

[0058] T5: Based on the aforementioned alternative level value, determine the on / off state of the passive nuclear level gauge's output switch. This output assurance operation is based on the T4 assurance operation. Unlike T4, which directly outputs the alternative level value, T5 outputs a switch state quantity calculated based on the alternative level value. In many applications, the on / off level signal of the passive nuclear level gauge is an important warning or control signal. The switch signal generated based on the alternative level value is more accurate than the forced switch signals preset by T1, T2, and T3, and can better ensure the accurate operation of the system.

[0059] While performing the output protection operations of T1 to T5, a specific "protection mode activated" status word is sent to the host computer via the display screen of the equipment's human-machine interface or through the communication protocol. However, this invention does not limit the display of the status word to be mandatory. In practical applications, if the passive nuclear level gauge outputs a continuous level signal, a specific level value can be output to alert the operator. For example, if the operator sees the level value remaining stable at 50% for an extended period on the monitoring screen (the level fluctuates during normal production), they can immediately realize that "the level gauge is in anti-interference protection mode, and the current displayed value is not the actual measured value." This method is simple to implement and the alarm intent is clear.

[0060] In the T4 passive nuclear level gauge, during the flaw detection response phase, the analog level signal output is switched to output protection mode, using a substitute level value to replace the normal level value. Alternatively, in T5, the output switch state is calculated based on the substitute level value, which is determined according to a preset output protection strategy. In other words, when flaw detection interference occurs, the passive nuclear level gauge will no longer calculate the material level according to the normal measurement data, but will instead use a substitute level value to forcibly replace the level data generated during normal measurement. In this application, the substitute level value can be a fixed value or a dynamic value.

[0061] After forcibly replacing the normal material level value with a substitute material level value, the method of this application can also generate an output switch state based on the substitute material level value. It should be noted that the switch output judgment threshold can be either the judgment threshold used during normal material level measurement or a judgment threshold specifically preset for the flaw detection response stage.

[0062] Continue to refer to Figure 1 In one embodiment of the present invention, the method for controlling the level output of a passive nuclear level gauge during nuclear radiation flaw detection further includes: S300: Determine that the nuclear radiation flaw detection interference in the measurement environment where the passive nuclear level gauge is located has ended; specifically, determine whether the preset second condition is met. If it is met, determine that the nuclear radiation flaw detection interference in the measurement environment where the passive nuclear level gauge is located has ended, and execute step S400; if the second condition is not met, it means that the nuclear radiation flaw detection interference has not ended, and the output protection operation of step S200 still needs to be executed.

[0063] S400: Cancel the output protection operation in progress, that is, restore the normal measurement to generate corresponding switch signals, digital signals and / or analog signals based on the collected nuclear radiation.

[0064] The core steps of this application consist of three steps: flaw detection interference determination, flaw detection interference response, and recovery. After determining that nuclear radiation flaw detection is present in the measurement environment, the passive nuclear level gauge in this application does not stop normal measurement of the measurement environment, but continues normal nuclear radiation measurement. However, it will no longer calculate the material level or generate an alarm signal based on the measured nuclear radiation data. Therefore, measuring the material level or implementing a level alarm using normal methods is ineffective at this point, and it cannot truly monitor the material level inside the container.

[0065] It should be noted that how passive nuclear level gauges normally measure material level, calculate material level, or output alarm information is existing technology in this field and will not be elaborated further.

[0066] In the flaw detection interference response step of this application, absolute, transparent and predictable control of the level signal output of the passive nuclear level gauge can be achieved, avoiding various adverse consequences caused by level measurement errors due to flaw detection interference.

[0067] The final step in this embodiment is to exit the output protection state and restore the normal level measurement and signal output process when the flaw detection interference is determined to have ended. Clearly, a key technical feature of this application is that the passive nuclear level gauge can automatically return to normal measurement mode after the flaw detection interference ends, rather than remaining in the output protection state indefinitely.

[0068] The method in this application is essentially to avoid equipment malfunctions or system safety issues caused by false alarms or incorrect material levels due to flaw detection interference. However, it cannot replace normal material level measurement. Therefore, the passive nuclear level gauge must be able to automatically resume normal material level measurement after the flaw detection interference ends.

[0069] Since the recovery step in this application is automatically executed by the passive nuclear level gauge itself without user intervention, it can ensure that the passive nuclear level gauge can be restored to normal measurement status in a timely manner, while effectively reducing the workload of maintenance personnel.

[0070] The following explains the determination of whether nuclear radiation testing interference exists in the measurement environment of the passive nuclear level gauge in step S100: It is very important to determine whether nuclear radiation testing interference exists in the measurement environment of the passive nuclear level gauge. The existence of nuclear radiation testing interference in the measurement environment of the passive nuclear level gauge is determined by a combination of one or more of the following first conditions: The passive nuclear level gauge acquires nuclear radiation measurement values ​​within one or more preset time windows, and determines statistical characteristics and / or changes in statistical characteristics based on the nuclear radiation measurement values. When the statistical characteristics and changes in statistical characteristics meet the corresponding preset third condition; And / or, the nuclear radiation measurement value of the passive nuclear level gauge reaches the preset first radiation threshold; And / or, the operating time of the passive nuclear level gauge experiences a preset first duration; And / or, the current time reaches the preset time to start flaw detection; And / or, the passive nuclear level gauge or its control system receives a confirmation signal from an external device indicating that flaw detection has been initiated.

[0071] The statistical characteristics are at least one of the following: measured value, mean, extreme value, extreme value difference, standard deviation, variance, quantile, interquartile range, or mode. By analyzing the statistical characteristics of nuclear radiation data measured by a passive nuclear level gauge, it is possible to effectively determine whether nuclear radiation interference exists in the environment. In practical applications, the statistical characteristics can be one or more statistical quantities, such as mean, extreme value (maximum / minimum), extreme value difference, standard deviation, variance, quantile, or interquartile range; or they can be based on the original measured values ​​themselves. Furthermore, statistical characteristics can refer to either the specific numerical values ​​of these statistical quantities or the changes in these values.

[0072] The simplest method is to check if the nuclear radiation measurement value of the passive nuclear level gauge reaches the preset first radiation threshold. In a normal material level measurement process, the nuclear radiation generated by the material will generally not exceed 2900. The first radiation threshold can be set to 3000. If the nuclear radiation measurement value of the passive nuclear level gauge reaches 3000 or higher, it is determined that there is nuclear radiation flaw detection interference, and output protection operation is performed.

[0073] Since nuclear radiation testing in industrial settings exhibits significant temporal patterns, utilizing these patterns to determine whether nuclear radiation testing is present in the environment is a simple and efficient method.

[0074] And / or, acquire the nuclear radiation measurement values ​​of the passive nuclear level gauge within one or more preset time windows, determine statistical characteristics and / or changes in statistical characteristics based on the nuclear radiation measurement values, and when the statistical characteristics and changes in statistical characteristics meet the corresponding preset third condition. The preset condition is at least one of the following: A1: The passive nuclear level gauge shows that the variation between the maximum and minimum radiation measurements within a preset time window exceeds a preset radiation threshold. Specifically, the preset time window covers N consecutive measurements, where the maximum nuclear radiation measurement value is V. max The minimum nuclear radiation measurement value is V min If there is no interference from nuclear radiation testing, the continuous measurement values ​​will not show violent fluctuations, i.e., V max With V min The difference is not large; if V appears max With V min If the difference is greater than or equal to the second radiation threshold, it is determined that there is nuclear radiation flaw detection interference, and output protection operation is performed.

[0075] Numerical Example S1 (Nuclear Radiation Flaw Detection Identification Method Based on Extreme Difference of Count Values ​​within a Time Window), the determination of nuclear radiation flaw detection interference includes the following steps: S101: Data Acquisition and Buffering. The core detector of the passive nuclear level gauge (such as a scintillator detector and photomultiplier tube) converts the received gamma rays into electrical pulse signals. A counter counts the pulses and outputs a raw count value at a fixed sampling period (e.g., 100 milliseconds). The system runs continuously and stores the latest count value sequence in a circular data buffer. In this embodiment, a preset time window T is set, with a length of 10 seconds. Therefore, the buffer retains continuous count values ​​within the time window T, totaling 10 seconds / 0.1 seconds = 100 data points, forming a count value sequence {X1, X2, ..., X...}. 100}

[0076] S102: Calculate the extreme value difference using statistical features. Statistical analysis is performed on the sequence of count values ​​in the buffer at each sampling period, or according to a set calculation period (e.g., every 1 second). This embodiment selects the extreme value difference (Range) as the core statistical feature. The formula for calculating the extreme value difference R is: R = max(X1, X2, ..., X...). 100 )-min(X1,X2,...,X 100 The functions max() and min() are used to find the maximum and minimum count values ​​within the time window T, respectively, while the extreme value difference R intuitively reflects the absolute fluctuation range of the nuclear radiation intensity during that time period.

[0077] S103: Compare the calculated extreme value difference R with a preset second radiation threshold, which can be calibrated based on a large amount of experimental and field data. For example, when gamma-ray or X-ray flaw detection is performed in or near the measurement environment, high-intensity, directional flaw detection rays will be captured by the level gauge detector, causing the count value to show one or more spikes far exceeding the normal range in a short period of time, thus causing the extreme value difference R to increase sharply. Therefore, the second radiation threshold is set to 2000, and the corresponding judgment logic is as follows: If R is greater than 2000, it is determined that there is nuclear radiation flaw detection interference, and output protection operation is performed.

[0078] If R is less than or equal to 2000, the current environment is determined to be normal (there is no interference from nuclear radiation testing), and monitoring will continue.

[0079] S104: Outputs a flaw detection interference signal and triggers corresponding output protection measures.

[0080] A passive nuclear level gauge installed on the dust collector hopper of a thermal power plant was tested in the field. During five consecutive days of monitoring, three planned gamma-ray flaw detection operations occurred within the plant area. The method described in this embodiment successfully triggered an alarm within 3-5 seconds after the flaw detection began (i.e., after the time window was filled with abnormal data), with the measured extreme value difference R reaching 2500-4000. During the normal time without flaw detection, the extreme value difference R remained consistently below 600. This indicates that the technical approach combining a "10-second time window" with an "extreme value difference threshold of 2000" can quickly, accurately, and reliably identify this specific interference event of nuclear radiation flaw detection, effectively avoiding misjudgments and missed detections.

[0081] A2: The statistical characteristics of one or more radiation measurements of the passive nuclear level gauge within a preset time window exceed a preset first characteristic range. If it is for a single measurement result, the statistical characteristic is the radiation measurement value; if it is for multiple measurement results, the statistical characteristic can be one or more of the following statistical values: average, extreme values, standard deviation, variance, quantiles, interquartile range, or mode. The A1 embodiment can be a specific embodiment of the A2 method, and may also include the following specific embodiments: Numerical Example S2 (An Alternative Based on Joint Judgment of Standard Deviation and Mean): In order to further enhance the robustness of nuclear radiation flaw detection judgment and reduce false alarms caused by a single extreme noise pulse, this example provides a joint statistical feature judgment method.

[0082] The data acquisition steps are the same as step S101 in the numerical embodiment S1, that is, the count value sequence {X1,X2,...,X} is collected. 100}, calculate the average: M = (X1 + X2 + ... + X 100 ) / 100, calculate the standard deviation: S=sqrt([(X1-M)) / 100 2 +(X2-M) 2 +…(X 100 -M) 2 ] / (100-1)), or use the population standard deviation formula. Set the mean threshold θ. M and standard deviation threshold θ S The decision logic is as follows: If and only if M>θ M And S>θ S Upon simultaneous establishment, if nuclear radiation flaw detection interference is detected, output protection operation is executed. Where, θ M It can be set to 1.5-2 times the average value under normal operating conditions (for example, if the normal average value is about 1500, then θ can be set). M =2700); θ SIt can be set to 2-3 times the standard deviation under normal operating conditions (for example, if the normal standard deviation is approximately 250, then θ can be set). S =600).

[0083] This scheme combines two dimensions: an overall increase in radiation intensity (increased M) and drastic numerical fluctuations (increased S), which can capture flaw detection features more comprehensively and has stronger anti-interference ability against individual field values.

[0084] Numerical Example S3 (Rapid Initial Screening Method for Flaw Detection Based on Maximum Count Value (Extreme Value) within a Time Window): This example addresses flaw detection alarm scenarios requiring extremely fast responses. It utilizes the maximum value (Max) as an extreme statistical feature, aiming to achieve millisecond- to second-level anomaly initial screening. The specific steps are as follows: S301: Data Acquisition. Similar to numerical embodiment S1, the passive nuclear level gauge samples at a 100-millisecond cycle. The system maintains a short-time window buffer of 3 seconds, storing the most recent 30 count values ​​{X1, X2, ..., X...}. 30}

[0085] S302: Calculate the maximum value of statistical characteristics. Calculate the maximum value M of the count within this short time window in real time. max =max(X1,X2,...,X 30 This calculation requires very little computation and has an extremely fast response time.

[0086] S303: Fast threshold determination. Set a relatively high upper threshold θ. high (For example, 4500). When M max >θ high Upon activation, a primary alarm signal is immediately triggered. This threshold is set much higher than the normal material fluctuation range, specifically designed to capture the strongest pulse signal generated at the beginning of flaw detection.

[0087] S304: Correlation Confirmation. Upon triggering the primary alarm, a secondary confirmation process based on more complex statistical characteristics (such as extreme value difference and standard deviation), as described in numerical embodiment S1 or S2, is automatically initiated. A comprehensive analysis is performed within a subsequent 10-second window. If the secondary confirmation also determines an anomaly, it is confirmed as a nuclear radiation flaw detection event; otherwise, the primary alarm is deactivated to avoid false alarms caused by single, instantaneous high-energy particles such as cosmic rays or electrical interference.

[0088] This embodiment adopts a two-level judgment mechanism of "rapid initial screening of maximum value + secondary confirmation of complex features", which takes into account both the immediacy of alarm (the first response can be completed within <3 seconds) and reliability, and is suitable for occasions with extremely high security requirements and the need for immediate warning.

[0089] Numerical Example S4 (Stability Monitoring Method Based on Variance of Count Values ​​within a Time Window): This example focuses on identifying flaw detection interference that may not be the strongest but has good persistence, or on assessing the overall stability of the measurement environment, using variance as the core feature. The specific steps are as follows: S401: Data Acquisition. Similar to numerical example S1, the time window T is set to 15 seconds, and the count value sequence {X1, X2, ..., X} is acquired. 150}

[0090] S402: Calculate the variance using statistical characteristics. Calculate the variance V of the sequence: V=[(X1-M) 2 +(X2-M) 2 +...+(X 150 -M) 2 ] / 150, where M is the average count for this window. Variance V reflects the square of the dispersion of the data around the mean and is extremely sensitive to fluctuations.

[0091] S403: Threshold determination. Set the variance threshold θ. V When V>θ V If interference causing drastic and continuous fluctuations in the readings is detected, it is highly likely to be nuclear radiation testing.

[0092] S404: Auxiliary Application. The calculation result V in this embodiment is not only used for alarms, but also as a long-term monitoring indicator, which can be used to plot the trend of variance over time. During normal maintenance without flaw detection, a stable variance trend indicates that the detector is working normally; if the variance baseline shows an unexplained gradual increase, it may indicate that the detector performance is deteriorating or that a new and persistent background radiation source has appeared on site, providing data support for equipment pre-maintenance.

[0093] Variance features can effectively capture persistent and fluctuating interference signals, avoiding missed detections caused by the decline in statistical characteristics after individual spikes. At the same time, it provides a quantitative indicator for measuring the long-term stability of the system, thus expanding the application value of this method.

[0094] Numerical Example S5 (Robust Identification Method Based on Quantiles to Combat Outlier Interference): This example addresses scenarios where occasional, non-flaw-detection-induced electrical noise or transient interference ("outliers") may exist in the field. It employs quantile features to improve the robustness of the algorithm. The specific steps are as follows: S501: Data acquisition. Same as the numerical example S1, but with a time window of 10 seconds.

[0095] S502: Calculate the difference between the upper and lower quartiles using statistical features. First, arrange the count sequence within the time window in ascending order. Then, find the 75th percentile (upper quartile, denoted as Q3) and the 25th percentile (lower quartile, denoted as Q1) of the sequence. The statistical feature used in this embodiment is Q3-Q1.

[0096] S503: Threshold determination. Set the quantile difference threshold θ. Q When (Q3-Q1)>θ Q At that time, it was determined that there was nuclear radiation flaw detection.

[0097] Unlike extreme value differences (affected by both the maximum and minimum values), Q3-Q1 only focuses on the distribution range of the middle 50% of the data. During nuclear radiation testing, a large number of count values ​​increase overall and become more dispersed, causing both Q3 and Q1 to shift significantly upwards and the difference to widen. Occasional single outliers, whether extremely high or low, typically only affect the maximum or minimum value, with negligible impact on Q3 and Q1. Therefore, this method effectively filters outlier interference.

[0098] In industrial environments with occasional random noise, this embodiment exhibits higher robustness and a lower false alarm rate than methods based on extreme values ​​or extreme differences.

[0099] Numerical Example S6 (Standardized Volatility Measurement Method Based on Interquartile Range): This example is an extension and optimization of numerical example S5. It uses the interquartile range (IQR) combined with the median to achieve a standardized volatility measurement that is independent of the absolute count level. The specific steps are as follows: S601: Data acquisition. Same as numerical example S1.

[0100] S602: Calculate the standardized interquartile range (IQR) for statistical characteristics. Calculate the median (Med) of the sequence and the IQR = Q3 - Q1. Then, calculate a standardized ratio feature: R0 IQR =IQR / Med.

[0101] S603: Threshold determination. Set a ratio threshold θratio; when R... IQR When the value is greater than θratio, it is determined that nuclear radiation flaw detection is present.

[0102] The IQR value alone is affected by material density and material level (which affect the average count rate). IQR=IQR / Med eliminates the influence of absolute count levels, measuring instead the relative intensity of fluctuations. Regardless of whether the background count is 300 or 1000, any relative fluctuation intensity caused by flaw detection exceeding the threshold will be detected. This makes the algorithm adaptive under different material levels and conditions, without requiring significant threshold adjustments for each condition.

[0103] The standardized measurement method provided in this embodiment reduces the system's dependence on different application scenarios and the workload of parameter tuning, while improving the algorithm's universality and ease of engineering use. It is particularly suitable for storage tanks with frequent material changes or large material level ranges.

[0104] The measured values, average values, and extreme values ​​are basically similar. Taking the average value as an example: when the nuclear radiation value collected by the passive nuclear level gauge during flaw detection is far greater than the nuclear radiation of the material being measured in the container, for example, if the average nuclear radiation value collected over five consecutive sampling cycles suddenly rises to 800 to 2000 (within the first characteristic range), then it is determined that there is nuclear radiation flaw detection interference, and an output protection operation is executed. However, in terms of reflecting the measurement effect, it is weaker than the six numerical examples mentioned above.

[0105] A3: The change in the statistical characteristics of the nuclear radiation measurement values ​​in the current time window compared to the statistical characteristics of the nuclear radiation measurement values ​​in the previous time window exceeds a preset second characteristic range. Specifically, if each time window covers one measurement, and the difference between the current nuclear radiation measurement value and the previous measurement value exceeds the second characteristic range, then nuclear radiation flaw detection interference is determined to exist, and output protection operation is executed. If each time window covers multiple measurements, then the statistical characteristics of the measurement values ​​corresponding to the current time window and the measurement values ​​corresponding to the previous window are calculated separately. If the difference between the statistical characteristics corresponding to the two windows exceeds the second characteristic range, then nuclear radiation flaw detection interference is determined to exist, and output protection operation is executed.

[0106] Numerical Example S7 (Flaw Detection Start Identification Method Based on Abrupt Changes in Statistical Characteristics of Adjacent Time Windows): This example aims to solve the problem of accurately identifying the start time of nuclear radiation flaw detection operations and effectively distinguishing between flaw detection signals and slow changes caused by material movement. Its core lies in comparing the variation amplitude of the same statistical characteristic within multiple consecutive preset time windows. The specific steps are as follows: S701: Data Segmentation and Buffer Setup. The system sets up two First-In-First-Out (FIFO) data buffers, defined as the first window buffer W1 and the second window buffer W2. Each buffer stores the raw count values ​​within a preset time window length L (e.g., 10 seconds). As data continuously enters, the system ensures that W1 stores the data of the immediately preceding complete L-window (e.g., from time t-20 seconds to t-10 seconds), and W2 stores the data of the most recent complete L-window (e.g., from time t-10 seconds to the current time t). As time progresses, the two windows move forward synchronously like a "sliding window."

[0107] S702: Dual-window feature calculation. At the end of each calculation cycle (e.g., every 10 seconds), the same preset statistical feature is calculated in parallel for two windows. This embodiment uses the maximum value (Max) of the count within the window as an example: Calculate the maximum value of the first window W1: Max1 = max(data sequence W1); Calculate the maximum value of the second window W2: Max2 = max(data sequence W2); S703: Calculation and Determination of Feature Abrupt Change. Calculate the difference (abrupt change) of statistical features between two adjacent time windows: ΔMax = Max2 - Max1. Then, compare this difference ΔMax with the abrupt change threshold θ. Δ Compare. Mutation threshold θ Δ Based on experiments, the boundary line used to characterize the "maximum fluctuation caused by normal material changes" and the "sudden increase in X-ray intensity at the start of flaw detection" is defined. For example, θ is set... Δ =1500 (counting units).

[0108] Decision logic: If ΔMax > θ Δ If so, it is determined that a nuclear radiation flaw detection initiation event occurred within the time period of the second window W2.

[0109] S704: Output and Marking. Once the determination is established, the system will not only output a general alarm for "nuclear radiation testing is present", but also additionally mark the start time of the testing (which can be marked as the start time of W2 or the time when ΔMax exceeds the threshold), and freeze / mark the level measurement values ​​from that time.

[0110] This method focuses on the characteristic transitions between adjacent windows, making it extremely sensitive to the instantaneous nature of the start of flaw detection. Changes in the count rate caused by variations in material height are typically slow and continuous, with small ΔMax values ​​for adjacent windows. However, at the start of flaw detection, the ray intensity changes in a "step" manner, resulting in a very significant ΔMax value. Therefore, this method can achieve high-precision capture of the event initiation point and effectively reduce false alarms caused by material movement.

[0111] A4: The absolute value of the statistical characteristics of the nuclear radiation measurement values ​​of the passive nuclear level gauge exceeding the preset third characteristic range multiple times within multiple preset time windows meets the standard. In other words, the number of times the change or absolute value of the statistical characteristics of the radiation measurement values ​​of the passive nuclear level gauge within multiple consecutive preset time windows exceeds the corresponding mutation threshold reaches the preset number condition or preset proportion condition.

[0112] For example, the standard for compliance is that within 20 time windows, the absolute value of the corresponding statistical characteristic exceeds the preset third characteristic range more than 12 times, or the percentage of times exceeding the preset third characteristic range exceeds 60%. If the standard is met, it is determined that there is nuclear radiation flaw detection interference, and output protection operations are performed.

[0113] Specifically, numerical embodiment S8 (a method for confirming flaw detection events based on the number of times statistical characteristics continuously exceed the limit): This embodiment aims to solve the problem of identifying intermittent, periodic, or weak but long-lasting flaw detection interference, and to improve immunity to occasional spike noise. Its core is to statistically analyze the frequency of single-window statistical characteristics exceeding the primary threshold over a relatively long observation period. The specific steps are as follows: S801: Long-period observation window setting. Sets a long observation time range T. long For example, 100 seconds. Divide this long range into N consecutive, non-overlapping basic time windows, each with a length of L (for example, 10 seconds). Therefore, N = T_long / L = 10. The system maintains a queue to store the data for these 10 basic windows.

[0114] S802: Single-window feature calculation and preliminary judgment. For long observation range T long For each base window i (i=1 to 10) within the range, its preset statistical characteristics are calculated. This embodiment uses the range as an example to calculate the range R for each 10-second window. i . Each R i With threshold θ single (For example, 1000) are compared, and a primary binary decision is performed: if R i >θ single If the window is not found, mark it as an "exceeded limit window" and record an excessive limit event.

[0115] S803: Frequency Statistics and Advanced Judgment. In long observation range T long At the end, count the total number of times the windows were marked as "exceeding the limit" across these 10 basic windows. Then, multiply the count by a preset count condition C. TH Compare them.

[0116] Decision logic: If Count ≥ C THThen it is determined that within the long observation range T long There was a nuclear radiation testing incident inside.

[0117] Among them, the degree condition C TH The setting can be adjusted based on the stringency of the judgment on the continuity of flaw detection, for example, setting C. TH =5 (meaning that more than half of the 10 windows exceeded the limit).

[0118] The strategy in this embodiment is very flexible: It can be used for confirmation: It can be used in conjunction with numerical examples S1 to S7. For example, when the mutation alarm of numerical example S7 is triggered, the continuous monitoring of this example is immediately started (100 seconds of observation starting from the mutation time). If the frequency condition is also met, the flaw detection event is double-confirmed, and the reliability is extremely high.

[0119] It can be used for independent detection: it can be directly used to detect flaw detection modes that have frequent fluctuations and continuous low intensity but whose intensity does not reach the high threshold of a single window (such as the threshold of 2000 in numerical example S1).

[0120] It can be used to determine when the flaw detection operation has ended: when multiple consecutive windows no longer exceed the limit (the count is very low).

[0121] This method expands the judgment from a "point" to a "line" by introducing frequency statistics in the time dimension. It does not rely on a single strong signal, but rather looks for anomaly patterns. This gives it a strong ability to identify weak but continuous flaw detection signals or flaw detection signals contaminated by short-term noise, while effectively filtering out truly occasional and isolated interference pulses, significantly improving the overall robustness and detection rate of the system.

[0122] Compared with numerical embodiment S8, the core of numerical embodiment S7 is to capture the "rate of change" or "gradient", which detects the transition of the state, while the core of numerical embodiment S8 is to statistically measure the "abnormal density" or "duration", which detects the maintenance of the state.

[0123] A5: The variation between the maximum and minimum values ​​of the statistical characteristics of nuclear radiation measurements corresponding to multiple preset time windows of the passive nuclear level gauge exceeds the preset fourth characteristic range. Specifically, each preset time window covers one or more nuclear radiation measurements, and the statistical characteristic F corresponding to each time window is calculated; the maximum statistical characteristic among the preset multiple time windows is F. max The smallest statistical characteristic is F min If F appears max With F min If the difference exceeds the preset fourth feature range, it is determined that there is nuclear radiation flaw detection interference, and output protection operation is performed.

[0124] A6: Calculate the change in statistical characteristics of the nuclear radiation measurements corresponding to multiple preset time windows of the passive nuclear level gauge. The number of times this change exceeds the preset fifth characteristic range is considered satisfactory. Similar to method A5, calculate the statistical characteristic F corresponding to each time window. If the number of preset time windows is 21, then there should be 20 instances of change in statistical characteristics ΔF corresponding to two adjacent time windows. The satisfactory standard is, for example, that among the 20 instances of change ΔF, the number exceeding the preset fifth characteristic range reaches 12 or more, or that the percentage exceeding the preset fifth characteristic range reaches 60% or more. If the standard is met, it is determined that there is nuclear radiation flaw detection interference, and output protection operation is performed.

[0125] A7: The passive nuclear level gauge operates for a preset first duration. Specifically, assuming the preset flaw detection time is 12 hours after the passive nuclear level gauge starts working, a 12-hour countdown begins when the level gauge starts working. When the countdown reaches its end, nuclear radiation flaw detection interference is detected, and output protection operation is performed.

[0126] In practice, nuclear radiation flaw detection exhibits a strong temporal regularity. Because industrial flaw detection processes aim to minimize radiation exposure to personnel in the work environment, detection is typically scheduled for late at night, with fixed and clearly defined start and end times, such as between 10 PM and 5 AM the following morning. Therefore, based on this temporal regularity, the first time-based method for determining entry into nuclear radiation flaw detection provided in this embodiment is as follows: Based on the user-preset fixed time parameters in the passive nuclear level gauge, if the user knows the operation time of nuclear radiation testing, the user can set the working time parameters for the passive nuclear level gauge through external devices, such as a wireless remote control, a remote wired control system, or the human-machine interface built into the passive nuclear level gauge. This time parameter is a cumulative duration. For example, after the passive nuclear level gauge is activated, it enters the nuclear radiation testing mode after 17 hours, continues for 7 hours, then continues working for another 17 hours, then enters the nuclear radiation testing mode again, continues working for another 7 hours, and so on in a regular cycle.

[0127] A8: The current time has reached the preset start time for flaw detection. Specifically, assuming the preset flaw detection time is 1:00 AM on February 20, 2026, if the current time reaches that point, it is determined that there is nuclear radiation flaw detection interference, and output protection operation is executed.

[0128] In addition, other methods for determining entry into nuclear radiation flaw detection based on time include the following: The passive nuclear level gauge in this application automatically generates time parameters based on historical data and patterns in flaw detection events. If the user cannot obtain a nuclear radiation flaw detection work schedule from other departments, the passive nuclear level gauge in this application can also automatically mark the working time parameters based on historical data and confirmation of flaw detection events. For example, based on historical data from the previous day and various nuclear radiation flaw detection events triggered by statistical characteristics, the passive nuclear level gauge marks the trigger time as the start time of nuclear radiation flaw detection and automatically uses this time as the start time of nuclear radiation flaw detection, or it can obtain a user's instruction to confirm the marked time as the start time of nuclear radiation flaw detection and set that time as the start time of nuclear radiation flaw detection. In subsequent work, the passive nuclear level gauge no longer needs to rely on the statistical characteristics of the measured value to determine whether to enter the nuclear radiation flaw detection stage, but can directly determine it based on time.

[0129] A9: The passive nuclear level gauge or its control system receives a confirmation signal from an external device indicating the start of nuclear radiation testing. Using instructions from external devices to determine whether nuclear radiation testing is present in the measurement environment can greatly reduce the workload of the passive nuclear level gauge itself, while eliminating the need for complex calculations.

[0130] In practical implementation, an external device can be installed to monitor the measurement environment of the passive nuclear level gauge. This external device can detect the presence of nuclear radiation flaw detection operations in the measurement environment through its own methods and send a signal indicating the presence of nuclear radiation flaw detection to the passive nuclear level gauge via wired or wireless means. Alternatively, after the operator learns that nuclear radiation flaw detection has begun in the measurement environment of the passive nuclear level gauge, they can input a command indicating the presence of nuclear radiation flaw detection in the environment through the human-machine interface configured on the passive nuclear level gauge. The external device can be a signal transmitting device connected to the passive nuclear level gauge via a wired connection. In practice, the external device can be a nuclear radiation detection device with environmental nuclear radiation detection capabilities, similar to the passive nuclear level gauge, capable of detecting nuclear radiation in the environment. Of course, as a dedicated nuclear radiation flaw detection detection device, it should be more sensitive and have a stronger nuclear radiation measurement method than the passive nuclear level gauge. The external device with nuclear radiation measurement capabilities can determine the presence of nuclear radiation flaw detection in the measurement environment according to the aforementioned determination method of the passive nuclear level gauge itself, and then send a signal to the passive nuclear level gauge to begin nuclear radiation flaw detection. External devices significantly reduce the workload of passive nuclear level gauges. These external devices can also be smart mobile terminals that send confirmation signals indicating the start of flaw detection via an app.

[0131] In practice, external equipment can determine the presence of nuclear radiation testing events in the measurement environment based on data analysis of multiple passive nuclear level gauges within the same geographical area. These multiple passive nuclear level gauges can refer to multiple gauges on a single container or multiple gauges on multiple containers. In industrial settings, for example, a dust collector often has 2-3 passive nuclear level gauges installed on a single ash hopper, while a dust collector in a thermal power plant or steel plant typically has dozens of ash hoppers, each equipped with a passive nuclear level gauge. Once nuclear radiation flaw detection begins, the nuclear radiation from the radioactive source can affect many passive nuclear level gauges within a radius of several hundred meters. A significant characteristic is that although the intensity of nuclear radiation may vary among multiple passive nuclear level gauges due to differences in location and shielding, their temporal characteristics are completely consistent. For example, multiple passive nuclear level gauges within the measurement space may simultaneously measure a significant increase in nuclear radiation values. Furthermore, the increase in nuclear radiation at the same time will differ significantly, and there will be a certain degree of decreasing or increasing increase in the rate of increase based on geographical location or spatial layout. This is clearly different from many electrical interferences faced by passive nuclear level gauges. External equipment can collect nuclear radiation data measured by all passive nuclear level gauges in the measurement space, and then, based on the above statistical characteristics, determine that nuclear radiation flaw detection is present in the measurement environment and send the data to the passive nuclear level gauges, informing them that flaw detection work is underway.

[0132] The following explains the determination of whether the nuclear radiation flaw detection interference has ended in step S300: The nuclear radiation flaw detection interference in the measurement environment of the passive nuclear level gauge is determined to have ended by at least one of the following second conditions: B1: From the moment nuclear radiation interference is detected or an output support operation is executed, a preset second time period will elapse. For example, if the predetermined detection time is 5 hours, a 5-hour countdown will begin from the moment nuclear radiation interference is detected, and the ongoing output support operation will be canceled after 5 hours.

[0133] B2: The current time has reached the preset stop time for flaw detection. For example, if the flaw detection time is predetermined to end at 2:00 AM, then the output protection operation in progress will be canceled at that time.

[0134] The passive nuclear level gauge acquires nuclear radiation measurement values ​​within one or more preset time windows, and determines statistical characteristics and / or changes in statistical characteristics based on these measurements. The statistical characteristics and changes in statistical characteristics satisfy a corresponding preset fourth condition. The preset fourth condition is at least one of B3 or B4 below: B3: The statistical characteristics of the radiation measurements from the passive nuclear level gauge over one or more consecutive sampling periods recover to the preset sixth characteristic range. The statistical characteristics described here, and those used below, are used to describe the measured value or the characteristics of its variation. The statistical characteristics can be at least one of the following: measured value, average value, extreme value, extreme value difference, standard deviation, variance, quantile, or interquartile range. Specifically, for a single measurement result, the statistical characteristic is the measured value; for multiple measurement results, the statistical characteristic is the average value, extreme value, standard deviation, variance, quantile, or interquartile range.

[0135] The specific numerical implementation is conceptually the same as numerical implementations S1 to S8, the difference being that the corresponding threshold is the threshold / feature range restored to a non-flaw detection environment. For example, corresponding to numerical implementation S1, the sixth feature range, determined based on a large number of experiments and field data, is that under normal operating conditions without nuclear radiation flaw detection interference (affected only by natural background radiation and stable radiation from the material being tested), the extreme difference of the level gauge's count values ​​within a 10-second time window is typically distributed at a low level (e.g., between 200 and 800). Based on this, this embodiment sets the sixth feature range to 200 to 800 (count units). The determination logic is: If R is between 200 and 800, the flaw detection is considered to have ended, and the output protection operation in progress is cancelled. If R is greater than 800, the flaw detection is considered not to have ended.

[0136] B4: The passive nuclear level gauge meets the standard if the statistical characteristics of the radiation measurement values ​​over multiple sampling periods return to the preset seventh characteristic range the number of times. For example, if the proportion threshold is 80% or the number threshold is 8 times, and 8 or more of the 10 statistical characteristics reach the preset seventh characteristic range, then condition B4 is met, that is, the flaw detection is determined to be completed, and the output protection operation in progress is canceled.

[0137] This method is essentially the same as the method for determining nuclear radiation flaws through statistical characteristics, the only difference being the required conditions. B2 and B3 are two specific methods; others will not be elaborated upon.

[0138] B5: The passive nuclear level gauge or its control system receives a confirmation signal from an external device indicating the end of flaw detection. Similarly, the external device can be a signal transmitting device connected to the passive nuclear level gauge via a wired connection, or it can send the confirmation signal indicating the end of flaw detection via an app integrated into a smart mobile terminal.

[0139] B6: The substitute material level value generated based on the real-time substitute material level calculation formula reaches the preset material level threshold. The real-time substitute material level calculation formula is a real-time prediction model established based on historical material level and time data within a preset time period before the determination of nuclear radiation flaw detection interference.

[0140] Numerical Example S9 (Dynamic Extrapolation Output Method Based on Quadratic Curve Fitting Model): The core of this example is to use a quadratic polynomial to fit the material level change trend before the flaw detection occurs, and to dynamically generate a smooth and continuous predicted material level value as a substitute material level value based on this fitting function during the output assurance stage. The specific steps are as follows: S901: Historical Data Acquisition and Preprocessing. During normal operation and without interference from nuclear radiation flaw detection, continuously acquire and cache valid measurement data from the level gauge. When it is determined, based on any of the aforementioned flaw detection identification methods (such as numerical examples S1-S8), that the system is about to enter the "output assurance stage," immediately lock and extract a preset fitting time T preceding that determination moment. fit Historical effective material level-time data series.

[0141] For example, set T fit =10 minutes. Assuming the system determines to enter the protection phase at time t0, the extracted data sequence is: (t1,L1),(t2,L2),...(t m ,L m ), where t1 to t m Located within the interval [t0-10 minutes, t0), L i For the corresponding time t i The effective material level value (unit: %), where m is the number of effective sampling points within this time period.

[0142] S902: Construct a quadratic curve fitting prediction model. Perform quadratic polynomial least squares fitting on the historical data sequence extracted in step S901 to construct the functional relationship between the predicted material level L and time T. Normalize or serialize the time variable; for example, define a new time variable T, let T=0 corresponding to the moment t0 when the protection phase is determined, and set the historical time point t... i Convert to an offset relative to t0 (in seconds or sampling periods).

[0143] By fitting, the time-level fitting function in the following form is obtained: Fitting function L(T) = aT 2 +bT+c, where L(T) represents the predicted material level value at relative time T (unit: %), T is the relative time variable, with the moment of entering the protection stage as the origin; a, b, c are fitting constants, where a is the quadratic term correction coefficient, reflecting the acceleration of material level change (for example, the acceleration a≈0 due to linear change caused by uniform feeding / discharging; if the feeding speed increases, a may be positive); b is the linear term correction coefficient, reflecting the instantaneous speed (slope) of material level change; c is the correction constant, when T=0, L(0)=c, that is, the material level reference value predicted by the model at the instant of entering the protection stage.

[0144] S903: Model Application and Dynamic Prediction Output. After the system enters the output guarantee phase (T>0), the following loop is executed in real time: 1. Calculate the current relative time: Obtain the time elapsed since entering the protection phase and update the relative time variable T. current .

[0145] 2. Substitute into the model for calculation: Substitute T current Substituting the fitted function L(T)=aT obtained in step S312 2 In +bT+c, the predicted material level L at the current moment is calculated. predicted =a×(T current ) 2 +b×(T current )+c.

[0146] If the predicted material level L at the current moment predicted If the level exceeds the preset threshold, it can be assumed that the material level inside the container has reached a very high level according to the preset model. At this point, the output protection mode should be terminated, and the system should be restored to normal measurement. The fundamental logic behind this is that production safety is more important than interference from flaw detection. Restoring to normal operation, even with the risk of interference from nuclear radiation flaw detection, will encourage operators to strengthen monitoring of whether the material level inside the container has reached a high level through other means, thus ensuring production safety.

[0147] B7: The alternative material level value generated based on the pre-built prediction model reaches the preset material level threshold. The prediction model is a prediction model pre-built based on historical real material level and time data within at least one process cycle.

[0148] The prediction model obtains the predicted material level value through any of the following methods: B7.1: This prediction model is a fitting function obtained by fitting or regression analysis of historical material level and time data, that is, the relationship between material level value and time is obtained. Substituting the current time into the function relationship, the corresponding predicted material level value is obtained. B7.2: The prediction model calls upon historical material level and time data classified by process under the condition of no nuclear radiation flaw detection interference, and determines the material level value that matches the current process and time period as the predicted material level value; B7.3: This prediction model is a function model constructed based on historical operating data according to process classification. This method is based on method B7.1. It obtains historical material level data of different categories according to process classification, and then obtains the function model (relationship between material level value and time) corresponding to each process through fitting or regression analysis. Substituting the current time into the function model corresponding to the current process, the corresponding predicted material level value is obtained. B7.4: Train the artificial intelligence model using historical operating data to obtain the prediction model used to predict material level.

[0149] This method shares the same underlying logic as B6. At least in this method, the pre-built prediction model is not generated ad-hoc but pre-set and stored in the passive nuclear level gauge. During nuclear radiation testing, i.e., when the passive nuclear level gauge enters the output assurance mode, a prediction function is used to generate alternative level values. As time progresses, the generated alternative level values ​​gradually increase. When the alternative level reaches a constraint threshold, such as 3 meters or 80% of the level value, it means that, based on trend estimates, the material inside the container is now in a very dangerous state. Continuing to maintain the output assurance mode at this point might be dangerous. To give the user a better chance to see the actual measurement situation, it is best to exit the output assurance mode. If the actual nuclear radiation testing continues after exiting the assurance mode, the user will see the testing interference and avoid complacency. If the nuclear radiation testing has indeed ended, then the normal measurement mode is entered.

[0150] Further specific implementation examples Figure 2 As shown, the output protection operation in progress is to replace the normal level value of the passive nuclear level gauge with a substitute level value; In this embodiment, the alternative material level value is determined using any of the following safeguard strategies: C1: A corresponding level value is preset for the alternative level value and stored. The user or the passive nuclear level gauge has a preset level value at the factory, which serves as the fixed output level value after entering the output assurance phase. When the user sees the level remaining at a fixed value for a long time during operation, for example, consistently outputting 50% of the level, they can easily understand that they are currently in the output assurance phase for nuclear radiation flaw detection operations.

[0151] To make the preset fixed material level value more meaningful, passive nuclear level gauges can use the average value calculated from recorded historical data as the fixed material level value, or use the actual material level value recorded at the same historical moment as the alternative material level value.

[0152] C2: The material level value at the last moment before the presence of nuclear radiation flaw detection interference is determined is used as the alternative material level value. The valid material level value recorded at the last moment before entering the output support state is used as the alternative material level value, or the alternative material level value is calculated based on the valid material level values ​​recorded during the last period before entering the output support state. The specific steps are as follows: C201. Historical Data Caching. The system continuously caches valid historical material level values ​​within a recent short period, for example, the material level data sequence {L1, L2, ..., L...} from the 5 minutes prior to entering the output guarantee phase. nThe system ensures that this data is valid data collected before it is determined to be interference with flaw detection; C202. Calculate the replacement value. At the instant of entering the protection phase, calculate the average value of the cached historical valid data sequence: L replace =(L1+L2+...+L n ) / n.

[0153] C203, Output and Hold. The calculated L... replace This is used as an alternative level value for locking the output. Throughout the entire output assurance phase, this fixed locked value is continuously output.

[0154] In a preferred embodiment, to increase practicality, a truncated mean can be used in step C202, that is, the highest and lowest 10% of data in the sequence are removed before averaging, in order to eliminate possible random fluctuations in historical data.

[0155] The alternative value output by this method reflects the average material level in the short period before the flaw detection occurs. It is closer to the actual process situation than a fixed value (such as 50%) and can provide operators with a relatively reliable reference.

[0156] C3: The statistical characteristic value within a preset time period before the determination of nuclear radiation flaw detection interference is used as the alternative material level value.

[0157] The specific implementation steps are as follows: C301. Long-term historical data caching. The system continuously caches valid material level data within a preset time period before determining nuclear radiation flaw detection interference, such as a preset time period of 10 minutes, 15 minutes, or 30 minutes, forming a continuous and stable historical material level dataset {L1, L2, ..., L...}. m The dataset removes abnormal jumps and invalid measurements, retaining only the normal process material level data before the occurrence of flaw detection interference.

[0158] C302. Extract statistical feature values. Based on the cached long-term historical dataset, calculate statistical feature values ​​to characterize the stable state of the material level as alternative material level values. Optional statistical feature values ​​include: weighted average, median, mode, moving steady-state mean, or trend-fitted value. The weighted average is weighted by the time period, with more recent data having a higher weight. The median and mode can further weaken the influence of extreme data, adapting to industrial conditions with relatively small material level fluctuations.

[0159] C303, Eigenvalue Locking and Output. The calculated statistical eigenvalue is locked as the replacement material level value Lreplace. During the entire nuclear radiation flaw detection interference output protection phase, this value is kept constant and does not change with the flaw detection radiation signal.

[0160] In a preferred embodiment, the system can simultaneously calculate the median value and the weighted average value, and use the average of the two as the final substitute material level value, taking into account both data stability and process trend, so that the output value is closer to the actual material level of the equipment, and provides a more valuable and stable material level signal for on-site operation.

[0161] C4: After determining that there is nuclear radiation flaw detection interference, collect radiation measurement values ​​within at least one unit time window as a candidate sample set. Based on the historical measurement data before the nuclear radiation flaw detection interference, select specific measurement values ​​or local feature values ​​that meet the preset deviation conditions from the candidate sample set and use them as the alternative material level values.

[0162] The specific implementation steps are as follows: C401. Real-time acquisition of radiation measurement values ​​from nuclear radiation level gauges, and determination of nuclear radiation flaw detection interference in the current level measurement scenario using the preset interference judgment algorithm of this application; C402. After determining that there is nuclear radiation flaw detection interference, the data acquisition mechanism is activated, and a 5-second time window is selected as a unit time window. Radiation measurement values ​​are continuously collected within 3 unit time windows, and a total of 15 sets of radiation measurement data are obtained to form a candidate sample set. C403. Retrieve stable historical measurement data within 10 minutes before the nuclear radiation flaw detection interference occurs, calculate the average value and standard deviation of the historical data, and set the historical measurement average value ± 1.5 times the standard deviation as the preset deviation condition. C404. Traverse the above candidate sample set, remove abnormal interference data that exceed the preset deviation conditions, select three consecutive stable measurement values ​​that fall within the preset deviation range, calculate the mean of the three measurement values ​​as the local feature value, and finally determine the mean as the alternative material level value under this interference.

[0163] By selecting sampling points whose values ​​remain close to historical benchmarks within a real-time window during interference periods, this method can restore the true material level information to the greatest extent possible under complex flaw detection interference environments. In practical implementation, the minimum value within a unit time window during nuclear radiation flaw detection can also be used as characteristic data for comparison with historical data. If the minimum value is within a preset fluctuation range, it is used as a substitute material level value. This method does not correct or compensate for measurement errors, but rather, when nuclear radiation flaw detection interference makes measurement data unreliable, it selects a value from the interfered sampling data that has a reasonable and reliable deviation from the historical stable state and uses it directly as a substitute material level value.

[0164] C5: Calculates the alternative material level value based on the real-time generated alternative material level calculation formula. The real-time alternative material level calculation formula is a real-time prediction model established based on historical material level and time data within a preset time period before the determination of nuclear radiation flaw detection interference.

[0165] Numerical Example S10 (Dynamic Extrapolation Output Method Based on Real-Time Prediction Model): This example provides a dynamic prediction scheme that can output a real-time "simulated" trend of material level change during the assurance phase. The specific steps are as follows: S1001: Data Preparation. Collect the level-time series data (t1, L1), (t2, L2), ... (t) of the system during a period T (e.g., 30 minutes) before entering the output assurance phase under normal operating conditions. m ,L m ).

[0166] S1002: Curve fitting to construct the prediction function. Using algorithms such as polynomial fitting or exponential smoothing, curve fitting is performed on the historical data to obtain an approximate function of material level with respect to time, L=f(t). For example, if the material level is decreasing, the fitting function might be a linear function L=a*t+b.

[0167] S1003: Model Application and Predicted Output. At the moment t0 when entering the output guarantee phase, the current time t (t>t0) is substituted into the fitting function f(t) to calculate the predicted material level value L. pre (t)=f(t).

[0168] S1004: Continuous predictive output. During the protection phase, the system continuously updates and calculates L based on real-time t. pre (t) is output as a substitute material level value, thereby simulating a dynamic change value that conforms to historical trends.

[0169] A more specific embodiment can be found in numerical embodiment S9.

[0170] C6: The predicted material level value generated by a prediction model pre-built based on historical real material level and time data within at least one process cycle is used as the alternative material level value.

[0171] Specifically, using the predicted material level value generated by the pre-built prediction model as the alternative material level value includes any of the following methods: C6.1: The prediction model is a fitting function obtained by fitting or regressing historical level and time data; for example, the passive nuclear level gauge of this application can generate a level-time fitting function based on historical events within one or more process cycles, as a subsequent prediction model. After entering the output assurance mode, an alternative level value is generated based on the prediction model and time parameters.

[0172] C6.2: The prediction model calls upon historical material level and time data categorized by process type under conditions where there is no interference from nuclear radiation flaw detection, and determines the material level value that matches the current process and time period as the alternative material level value. This method, based on method C6.1, obtains historical material level data for different types according to process classification, and then uses fitting or regression analysis to obtain the function model (relationship between material level value and time) corresponding to each process. Substituting the current time into the function model corresponding to the current process, the corresponding predicted material level value is obtained. C6.3: The prediction model is a function model constructed based on historical operating data according to process classification; it can be based on historical measurement data for a time period, such as data within each 24-hour period over 30 days, comprehensively analyzing the measurement data at each moment, for example, removing multiple maximum and minimum values, and then taking the average value as the material level value at that moment. Based on the above average value of all time periods within 24 hours, a material level time fitting function is generated through a fitting function method, which serves as the prediction model.

[0173] Numerical Example S11 (Periodic Process Matching Output Method Based on Historical Same-Time Data Templates): This example is applicable to processes where the production process is highly periodic and the material changes within each cycle have highly similar temporal characteristics. Its core idea is to use normal data from the same historical moment, such as the average of normal data from the same time in the past 7 or 30 days, as the replacement value for today's disturbance; alternatively, only the normal data from the same time in the past 24 hours can be used as the replacement value for today's disturbance. The specific steps are as follows: S1101: Template Library Construction (Learning Phase). During long-term normal system operation (without flaw detection), the system automatically learns and builds material level-time templates. For example, using a day (24 hours) as a cycle, the day is divided into 1440 minute-level time periods. For each time period (e.g., 10:00 AM - 10:01 AM every day), the system records all valid material level values ​​collected within that time period.

[0174] S1102: Template Value Calculation. For each time period, obvious outliers in the recorded data are removed (e.g., by using box plots), and then the median or robust average of the material level values ​​for that time period is calculated as the template material level value "L" for that time period. M (t)”, where t is the time of day.

[0175] S1103: Real-time matching output. At a certain time t... now1 When flaw detection is detected and the system enters the output assurance phase, the system immediately adjusts the output based on the current time t. now Search the template material level value L for the corresponding time period from the template library. M (t now ).

[0176] S1104: Output substitution value. (The L value will be replaced.) M (t now This is output as the current alternative material level value.

[0177] For production processes with strong regularity, this method can output a material level value that closely matches the expected process state at the current moment, with a high degree of intelligence and great reference value.

[0178] C6.4: Train the artificial intelligence model using historical operating data to obtain the prediction model used to predict material level.

[0179] Numerical Example S12 (Intelligent Prediction Output Method Based on Machine Learning Model): This example utilizes machine learning to handle more complex working conditions. The specific steps are as follows: S1201: Feature Engineering and Model Training. In historical normal operation data, not only the time series of the material level itself is used, but also potentially related process features, such as inlet / outlet valve status signals, pump start / stop signals, and timestamps (hours, days of the week), are added as features X, and the material level value is used as the label Y. Algorithms such as linear regression, support vector machines (SVR), or long short-term memory networks (LSTM) are used to train a prediction model M such that Y≈M(X).

[0180] S1202: Model Application. After entering the output assurance stage, although the system cannot obtain the actual real-time radiation measurement value Y, it can still acquire other process characteristics X (such as valve status). These real-time characteristics X... real By inputting a trained model M, a predicted material level value L based on multivariate inference can be output. AI =M(X_real).

[0181] S1203: Output and Update. Output L AI As an alternative value, model M can also perform online incremental learning and updates using new, valid data after each exit from the protection phase and return to normal, thus achieving self-optimization.

[0182] Numerical embodiments S10 and S12, especially the latter, provide a dynamically changing alternative level value that is logically consistent with other current process variables during the output assurance phase. This "simulates" the normal situation to the greatest extent possible, avoiding the impact that a fixed output value might have on advanced control loops (such as PID), and achieving a leap from "fail-safe" to "intelligent fault tolerance".

[0183] External devices can generate prediction models based on the data from the passive nuclear level gauge, or based on historical data from multiple passive nuclear level gauges, or based on machine learning or artificial intelligence algorithms.

[0184] The aforementioned prediction model is integrated inside the passive nuclear level gauge, or the prediction model is set in an external device that communicates with the passive nuclear level gauge.

[0185] The present invention acquires a preset model sent by an external device or a prediction model based on human-machine interface input. For passive nuclear level gauges, using the gauge's own processor as the carrier for machine learning or artificial intelligence algorithms would significantly increase the cost. Given the widespread use of passive nuclear level gauges in industrial settings, such high costs are unacceptable to users. To reduce costs, the present invention's passive nuclear level gauge can acquire a model from an external device or a model input by the user based on a human-machine interface as the prediction model.

[0186] It should be noted that in actual implementation, the C5 and C6 prediction model methods can be used as screening constraints for the C4 method, and the measured values ​​can be selected from the data set collected during nuclear radiation flaw detection as substitute values.

[0187] The accuracy of a prediction model heavily depends on the precise alignment of its time coordinate system with the starting point of the actual physical process. This application provides several flexible methods for determining the starting time point of the prediction model (i.e., the absolute time corresponding to the time variable T=0 in the model) to adapt to different application scenarios and information availability.

[0188] Specifically, the starting time point of the historical material level and time data within the preset time period is determined by any of the following methods: D1: The starting time is defined as the trigger time of the start signal for the current process cycle. Specifically, the start of each process cycle is essentially the starting time of material changes within the container, making it a suitable starting time for the preset model. For example, when the container closes its discharge valve, the material inside the container typically shows an increasing trend.

[0189] Numerical Example S13 (Method for Determining the Start Time Based on the Process Cycle Start Signal): This example is applicable to situations where the process flow has a clear and detectable periodic start point, such as batch feeding or batch production in a reactor. The specific steps are as follows: S1301: Signal Monitoring and Correlation. The passive nuclear level gauge is equipped with a digital input (DI) channel or communication interface (such as Modbus TCP) to monitor cycle start signals from the process control system. These signals can include a fully closed discharge valve signal (indicating the start of a feed cycle), a feed pump start signal, or a "new production cycle start" command from the batch controller.

[0190] S1302: Time point capture and setting. When the level gauge receives a valid cycle start signal, it immediately records the current system timestamp as the start time t of this process cycle. start_cycle .

[0191] S1303: Model Construction and Application. Thereafter, when it is necessary to construct a predictive model as described in numerical embodiment S9 (e.g., before flaw detection), the definition of the relative time variable T is modified to: T = t current -t start_cycle , where t current This is the current time. This means that the prediction model L(T) = aT 2 +bT+c describes the material level change pattern from the start of this process cycle to the present.

[0192] For example, in a batch feeding into a dry ash silo in a thermal power plant, each time the interlock signal of "closing the vent valve and opening the feed valve" is issued, the level gauge records the time as t. start_cycle When flaw detection occurs during the feeding process, the system utilizes its own... start_cycle The model, which fits historical data, will predict the material level progression "starting from the beginning of this feeding cycle," which is highly consistent with the process logic.

[0193] D2: The starting time point is defined as the moment when the material level reaches a preset threshold. Specifically, in actual operation, if it is difficult to obtain the time information of related equipment in the process flow, the moment when the material in the container reaches a preset value can be used as the starting time of the prediction model. For example, the starting time of the preset model can be the moment when the material level reaches 0.1 meters or 10%. In fact, below this material level, no matter how long it takes, it is meaningless for the prediction model.

[0194] Numerical Example S14 (Method for Determining the Start Time of Material Level Reaching a Preset Threshold): This example is applicable to situations where it is difficult to directly obtain external process signals, but the material level itself has clear characteristic points of change. The specific steps are as follows: S1401: Threshold setting. Based on process knowledge, a physically meaningful material level threshold L is preset. threshold For example, for dust collector hoppers, the level can be set to 10% (it is considered that predictions are meaningless below this level).

[0195] S1402: Status monitoring and triggering. The system continuously monitors the valid real-time material level value L. real When L is detected real From below L threshold Become the first time to reach or exceed L threshold At that time, the starting point record is triggered.

[0196] S1403: Time Point Recording and Setting. Record the moment when the above triggering event occurs as the model start time t. start_threshold .

[0197] S1404: Model Building. When building the prediction model later, define T=t. current -t start_threshold The model describes the pattern of change since the material level enters the effective / meaningful range.

[0198] The advantage of this numerical embodiment is that it relies entirely on the level gauge itself for measurement, requires no external signals, and has strong self-containment. It focuses the model's attention on the level range that is most important to the process operation.

[0199] D3: The start time point is the moment when the trigger signal sent by the external device is received. In application, it is preferable to obtain the start time data input by the external device. For example, the external device can send time information to the passive nuclear level gauge of this application based on the valve trigger signal in the process flow, which can be used as the start time.

[0200] Numerical Example S15 (Method for Determining Start Time Based on External Device Input Signal): This example aims to achieve deep integration with the factory information system and obtain the most authoritative timestamp of process events. The specific steps are as follows: S1501: Communication Protocol Configuration. The level gauge connects to the factory's Distributed Control System (DCS) or Manufacturing Execution System (MES) via industrial Ethernet or fieldbus (PROFINET, EtherNet / IP).

[0201] S1502: Receive authoritative time messages. When a critical process event occurs, the DCS / MES sends a message containing an event code and a precise timestamp to the level gauge.

[0202] S1503: Parsing and Storage. The level gauge parses this message and confirms the received timestamp as the officially authorized model start time t. start_external This timestamp is typically synchronized by the DCS server, ensuring high precision across the entire plant.

[0203] S1504: Model synchronization. (This is followed by a seemingly unrelated phrase: "with this t") start_external This serves as the absolute time origin (T=0) for the prediction model. This ensures that the model time base between multiple devices is fully synchronized with the process timing of the central control system, facilitating collaborative analysis and optimization across the entire process.

[0204] D4: Based on the preset time period according to the patterns of historical flaw detection events, the moment when nuclear radiation flaw detection interference is determined is moved forward to obtain the starting time point. Specifically, based on the material level data and prediction model at the moment before the nuclear radiation flaw detection is determined, the preset model start time is inferred.

[0205] Numerical Example S16 (Adaptive Start Time Determination Method Based on Model Back-Calculation): This example uses the actual material level value at the last moment before the flaw detection to "calibrate" the preset or candidate start time. The specific steps are as follows: S1601: Preset or obtain a candidate start time. First, through any of the aforementioned numerical embodiments (numerical embodiment S13, numerical embodiment S14, or numerical embodiment S15), a candidate start time t is obtained. candidate At the same time, the system caches from t candidate Then until the flaw detection determination time t disturbance All valid historical material level data.

[0206] S1602: Construct a preliminary prediction model. Using t candidate to t disturbance Based on historical data, a preliminary prediction model L is constructed. preliminary (T)=aT 2 +bT+c, where T=tt candidate .

[0207] S1603: Obtain key true values. Record the value at the last moment before the flaw detection judgment takes effect (t). disturbance -Δt, where Δt is one sampling period) is the actual effective material level value L collected. actual .

[0208] S1604: Back-dive and calibration. (L...) actual Substitute into the preliminary model L preliminary (T), and solve for the corresponding relative time T. actual That is, solving equation L actual =a*(T actual ) 2 +b*(T actual )+c. Because L preliminary (T) is monotonic (material level typically changes in one direction), and this equation has a unique solution. This T actual This means that if the model is completely accurate, then L will be generated. actual This true value corresponds to, from t candidate The starting point of the calculation.

[0209] S1605: Calculate the calibrated start time. The actual, calibrated model start time t. start_calibrated The following formula can be used to calculate: t start_calibrated =tdisturbance -Δt-T actual ; S1606: Reconstruct the final prediction model. Use the calibrated start time t. start_calibrated Using this as a new time origin, the same historical data is refitted (or the original model is directly shifted) to obtain the calibrated final prediction model, which is used to output dynamic predictions during the safeguard phase. At this point, the model is at T=0 (i.e., t... start_calibrated The theoretical value at time ) will be compared with t disturbance The true value L at time -Δt actual A perfect match.

[0210] Numerical implementation S16 forces the historical fitting curve of the prediction model to precisely pass through the last known true data point by "using the result to deduce the starting point". This is equivalent to performing a zero-point drift correction, which can significantly reduce the cumulative prediction error caused by inaccurate judgment of the starting time, making the dynamic extrapolation prediction after entering the protection phase more accurate and reliable.

[0211] Because the multiple methods for determining the start time mentioned above show little correlation between the material level at the time of flaw detection generated by the prediction model and the actual material level at the last moment before the flaw detection, the subsequent predicted alternative material level values ​​have significant deviations. However, having the actual material level value at the moment before the nuclear radiation flaw detection, querying its time in the prediction model, and then deriving the start time based on that time is a more scientific and reasonable approach.

[0212] In one embodiment of the present invention, such as Figure 3 As shown, if any of the following preset third conditions are met, the output guarantee operation will be prohibited (that is, the output guarantee operation will not be performed even if the first conditions described in B1 to B10 are met): E1: The measurement result of the passive nuclear level gauge is within the preset safety range. The radiation generated by conventional nuclear radiation testing is usually not less than 800, and the safety range can be set between 100-500. If the measurement result of the passive nuclear level gauge is less than 500 but greater than 100, then even if any of the conditions in B2 to B7 are met, the output protection operation is prohibited.

[0213] E2: The measurement result of the passive nuclear level gauge falls within the preset fault range. Although the radiation dose generated by conventional nuclear radiation flaw detection is large, it usually does not exceed 6000. The fault range can be set above 6000. If the measurement result of the passive nuclear level gauge is higher than 6000, it indicates that the current measurement is very likely to malfunction; or if the measured value is 0, 1, or other values ​​significantly lower than the conventional measurement, it also indicates that the passive nuclear level gauge is malfunctioning, and executing the output protection operation will hide the alarm signal. Therefore, if the measurement result exceeds the preset fault range, the output protection operation should not be executed. Therefore, in actual implementation, measured values ​​less than 50 or greater than 5000 can be considered as fault range values.

[0214] E3: The passive nuclear level gauge displays or outputs a warning level for the material level. In practice, if the material level in the container is already at a dangerous level before nuclear radiation testing, entering the protection output state at this time may more easily lead to overlooking potential production safety issues. Therefore, in this case, it is more reassuring for users to see the actual measurement feedback during nuclear radiation testing.

[0215] Numerical Example S17 (Protection Mode Disabled Based on Real-Time Material Level Safety Threshold): This example is applicable to preventing the true danger from being masked by activating the protection mode when the material is already at a dangerously high or low level. The specific steps are as follows: S1701: Preset safety thresholds. Users set two key thresholds based on container design safety specifications and process operating procedures: the high-level safety threshold H... high ; and the low-level safety threshold H low .

[0216] S1702: Real-time monitoring and condition determination. During real-time operation, the system continuously monitors the real-time material level value L after basic filtering. now Before executing the "Enter Output Assurance State" command, perform a security check: If L now ≥H high It was determined to be in a "high-risk state".

[0217] If L now ≤H low It is determined to be either a "low-level dangerous state" or a "low-level safe state".

[0218] S1703: Logical Decision. The "Third Condition" is satisfied if any of the above-mentioned dangerous conditions are met. The system will ignore the flaw detection result and will prohibit switching to the output protection mode.

[0219] S1704: Execute the security policy. The system maintains the original measurement output.

[0220] E4: The output protection function of the passive nuclear level gauge or its control system is disabled. In practical applications, to avoid errors in the automatic program that could cause the passive nuclear level gauge to erroneously enter the output protection state, a function switch is usually required. If the function of the method described in this application is enabled, the flaw detection interference will be determined according to the method of this application, and the output protection mode will be entered. If the function is disabled, the passive nuclear level gauge will not execute the determination steps of this application. If the user determines that nuclear radiation flaw detection operations are unlikely to occur within a certain period of time, the output protection function can be directly disabled. However, the user can enable the function after being troubled by nuclear radiation flaw detection interference or after learning that nuclear radiation flaw detection work is about to begin.

[0221] E5: The current time is within the preset normal working time period. For example, flaw detection is scheduled in the early morning, so users can preset the daytime period as the time period to exclude the presence of nuclear radiation flaw detection interference; if users have high trust in the nuclear radiation flaw detection schedule, they can set the normal working time period parameter. If the working time of the passive nuclear level gauge of this application is within the normal working time period, then the flaw detection interference judgment step will be forcibly not executed.

[0222] In the output safeguard operation based on T4 or T5 as described above, the alternative level value is configured to be determined according to a preset safeguard strategy. In a further embodiment, after the nuclear radiation inspection interference ends and the output safeguard operation is canceled, the method further includes: comparing and analyzing the level data currently measured by the passive nuclear level gauge with the alternative level value in the output safeguard operation; and adaptively optimizing the preset safeguard strategy based on the comparison and analysis results, so that the safeguard strategy is optimized with the goal of reducing the difference between the alternative level value determined by it and the actual measured value after the inspection interference ends and the output safeguard operation is canceled.

[0223] Numerical Example S18 (Self-learning Optimization Method for Predictive Model Based on Data Comparison): This example aims to achieve the self-evolution of the system. By comparing the "predicted value output during the protection phase" with the "actual value measured after the disturbance ends," the system can evaluate the accuracy of the predictive model and automatically optimize it, forming a complete intelligent closed loop of "perception-decision-execution-feedback-optimization." The specific steps are as follows: S1801: Data Recording and Alignment. During the output assurance phase, the system not only outputs the alternative material level value L... sub (t), and simultaneously using the timestamp t as an index, the sequence {t,L} is... sub The complete data (t) is recorded in non-volatile memory and marked as "predicted output data". Once the system determines that the nuclear radiation flaw detection interference has ended and returns to normal measurement mode, it continues to run and record the actual measured level value sequence {t,L} for a subsequent period (e.g., 30 minutes). real (t)}.

[0224] S1802: Data Synchronization and Comparison. On the timeline, align the "predicted output data" with the "actual measurement data" after recovery. Calculate the deviation ΔL(t) = L between the predicted and actual values ​​at the same or similar time points. real (t)-L sub (t). Statistical analysis is performed on these deviations to calculate indices such as mean absolute error (MAE) and root mean square error (RMSE).

[0225] S1803: Model parameter optimization. If the analysis in step S2102 shows that the prediction bias is systematically large (e.g., MAE consistently exceeds 3% of the total range), the model optimization procedure is triggered. The quadratic curve fitting model L(T)=aT from the numerical example S9 is used. 2 For example, +bT+c: Method A (Full Refit): Using historical data prior to the current disturbance event (i.e., the data used in the previous modeling), combined with bias analysis, a regression algorithm with a penalty term is used to refit the data to obtain optimized new parameters (a', b', c'), which replace the old parameters.

[0226] Method B (Incremental Fine-tuning): The original parameters are slightly modified based on the trend of the deviation ΔL(t). For example, if the predicted value is consistently lower than the actual value, the constant term c is increased appropriately; if the slope of the predicted trend does not match the actual value, b is adjusted.

[0227] S1804: Alternative Strategy Optimization. Besides optimizing specific model parameters, the system can also optimize the fault-tolerance strategy itself. For example, the system records the method of generating the alternative value (e.g., fixed value, historical average, prediction model, etc.) selected each time the system enters the protection phase, and its subsequent prediction deviation. Through machine learning algorithms (e.g., multi-armed slot machine model), the system can learn which alternative value generation method will yield smaller prediction errors under different operating conditions (e.g., material level change rate before disturbance, current time point, etc.). When encountering similar operating conditions in the future, the strategy with better historical performance will be prioritized.

[0228] S1805: Update and Store. Update the optimized model parameters or strategy selection rules to the system's configuration store for use in the next event.

[0229] This embodiment transforms the level gauge from a static device into an intelligent agent capable of learning from historical experience. Its predictive and fault-tolerant capabilities can continuously improve themselves over time, significantly enhancing long-term operational reliability and user trust.

[0230] In one embodiment of the present invention, a passive nuclear level gauge is provided, such as... Figure 4As shown, the passive nuclear level gauge includes a radiation detector, a processor, and a signal output unit, wherein: The radiation detector is configured to collect nuclear radiation in the measurement environment; The processor is configured to determine whether there is nuclear radiation detection interference in the measurement environment based on the nuclear radiation collected by the radiation detector or based on a preset first condition. If so, the processor executes a preset output protection operation to generate corresponding switch signals, digital signals, and / or analog signals; otherwise, the processor generates corresponding switch signals, digital signals, and / or analog signals based on the nuclear radiation collected by the radiation detector; and the signal result generated based on the output protection operation is different from the signal result generated based on the collected nuclear radiation. The signal output unit is configured to output switch signals, digital signals, and / or analog signals generated by the processor. Specifically, the signal output unit includes one or more of the following: an isolated switch output channel; an analog output circuit for outputting 4-20mA or 0-10V analog signals; and a digital output circuit for outputting digital level signals.

[0231] It should be noted that the passive nuclear level gauge provided in this embodiment belongs to the same inventive concept as the above-described embodiment of the level output control method during nuclear radiation testing, that is, the processor is configured to execute the steps of the level output control method during nuclear radiation testing as described above.

[0232] Further reference Figure 5 The processor is configured to generate and store a status flag bit when performing the output protection operation, which is used to identify that the signal currently output by the signal output unit is an output based on the protection strategy under nuclear radiation flaw detection interference. And / or, the signal output unit is further configured to output the status flag bit; And / or, the passive nuclear level gauge further includes a communication interface configured to communicate with the signal output unit and an external device.

[0233] This embodiment specifically describes the hardware device for implementing all the foregoing method embodiments. The specific device structure includes the following: Radiation detectors include scintillators (such as NaI(Tl) or plastic scintillators), photoelectric conversion devices (such as photomultiplier tubes or silicon photodiodes), and preamplifiers. Their function is to convert gamma rays in the measurement environment into electrical pulse signals proportional to their intensity. Signal conditioning and acquisition module: Receives the pulse signal from the detector, filters and shapes it, and uses a high-speed counter to count the pulses, converting the count value (count rate) per unit time into a digital signal.

[0234] The processor, acting as the "brain" of the passive nuclear level gauge device, is typically an industrial-grade microcontroller (MCU) or system-on-a-chip (SoC). It stores firmware configured to execute all or part of the method steps described in the preceding embodiments. Specifically, this includes: running a flaw detection algorithm (calculating statistical features and comparing thresholds), managing historical data cache, performing predictive model fitting and calculation, deciding on the output mode based on fault-tolerant strategies and safety conditions, and executing a self-learning optimization algorithm.

[0235] Signal output unit: Controlled by the processing unit, responsible for providing standard industrial signals, including: Isolated digital output channels: used to output status signals such as "flaw detection alarm", "safety mode activated", and "high / low level". Optocoupler or relay isolation is used to improve anti-interference capability.

[0236] Analog output circuit: Typically a 4-20mA current loop output circuit, used to output continuous material level percentage values ​​(whether actual measured values ​​or substitute values). This circuit features short-circuit protection and reverse connection protection.

[0237] Digital output circuit: integrates RS-485 or Ethernet communication interface (such as Modbus RTU / TCP protocol) for outputting high-resolution digital level values, status words, historical data, etc.

[0238] Human-machine interaction and communication interface: The local device can be equipped with an LCD display and buttons for status display and parameter setting, and a communication interface (such as RS-485 / Ethernet mentioned above) for receiving external commands (such as function switch signals, time synchronization signals) and uploading data.

[0239] One embodiment of the present invention is as follows: Figure 5 As shown, an industrial material level monitoring system is provided, comprising: The passive nuclear level gauge described above; An external control device, communicatively connected to the passive nuclear level gauge, is configured to perform the steps of the level output control method described above during nuclear radiation testing, or the external control device is configured to send at least one of the following signals to the passive nuclear level gauge: A confirmation signal used to determine the presence of nuclear radiation flaw detection interference in the measurement environment of a passive nuclear level gauge; A confirmation signal used to determine the end of interference with nuclear radiation flaw detection; A process synchronization signal used to synchronize historical material level and time data at the start time point.

[0240] Furthermore, based on any or a combination of the aforementioned technical solutions, the external control device is a distributed control system or a programmable logic controller; And / or, the passive nuclear level gauge is configured to upload a status flag bit and / or output mode information to identify the current output guarantee operation state via a communication interface, and the external control device is configured to display the information differently on the monitoring screen or generate corresponding operation logs based on the uploaded information.

[0241] An Example of an Industrial Material Level Monitoring System Integration: This example specifically describes a typical application of integrating the passive nuclear level gauge of the present invention into a modern industrial control system. The system configuration is as follows: Passive nuclear level gauge: installed on storage tanks or silos that require monitoring; External control equipment: typically the controllers of a factory's distributed control system (DCS) or safety instrumented system (SIS); Communication network: It adopts 4-20mA hard-wired connection for analog signals, and uses industrial Ethernet for high-speed data communication.

[0242] Advanced integration features: Signal input: The DCS can send planned flaw detection start / end signals to the level gauge via digital output (DO) cards or communication messages.

[0243] Signal Output and Display: The status flag bit uploaded by the level gauge via communication is received by the DCS's graphical station software. The software logic is configured as follows: when the flag bit is set, in addition to changing the display color, different colored curves are used in the historical trend graph to distinguish between the "actual value" and the "alternative value", and a record is automatically generated in the operation log: "The XX level gauge entered the intelligent protection mode due to flaw detection interference at [time] and output the predicted value."

[0244] Process synchronization: When the DCS starts a batch process, it sends a "batch start" synchronization signal to the level gauge, which uses this signal as the precise start time of the prediction model.

[0245] System workflow example: 1. The factory maintenance department plans to conduct flaw detection in area A from 14:00 to 15:00. At 13:55, the DCS operator remotely activated the "safety function" of the level gauge through the monitoring screen.

[0246] 2.14:00, flaw detection begins. The level gauge detects abnormal radiation, indicating that the system is entering safety mode.

[0247] 3. The level gauge immediately switches its output to a substitute value based on the prediction model and reports the "safety mode activated" flag via communication.

[0248] 4. The level box for silo A on the DCS screen turns yellow, and the level curve changes smoothly. The operator recognizes this as an intelligent output and can continue to monitor the general trend.

[0249] 5.15:00, flaw detection ends. DCS sends a "flaw detection end confirmation" signal. Upon receiving this signal, the level gauge, based on its own algorithm, resumes accurate measurement.

[0250] 6. The level gauge analyzed the data before and after this event and fine-tuned the parameters of the prediction model.

[0251] 7. All processes are recorded by DCS, forming traceable reports.

[0252] The industrial material level monitoring system provided in this embodiment and the material level output control method embodiment during nuclear radiation testing provided in the above embodiment belong to the same inventive concept. Here, by reference, the entire contents of the material level output control method embodiment during nuclear radiation testing are incorporated into this industrial material level monitoring system embodiment, and will not be repeated.

[0253] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0254] The above description is only a specific embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A method for controlling the level output of a passive nuclear level gauge during nuclear radiation flaw detection, characterized in that, Includes the following steps: It was determined that nuclear radiation flaw detection interference existed in the measurement environment where the passive nuclear level gauge was located. Perform at least one of the following output guarantee operations: Set the output switch of the passive nuclear level gauge to a preset protection state, wherein the preset protection state is the switch-on state; Alternatively, the output switch of the passive nuclear level gauge can be set to a preset protection state, wherein the preset protection state is the switch open state; Alternatively, the output switch of the passive nuclear level gauge can be maintained in the switching state before the presence of nuclear radiation flaw detection interference is detected. Alternatively, the normal level value of the passive nuclear level gauge can be replaced with a substitute level value, wherein the normal level value is the level value generated by the passive nuclear level gauge based on real-time measurement data when the output protection operation is not performed; and the substitute level value is the level value used to replace the normal level value when nuclear radiation flaw detection interference is detected. Alternatively, the switching state of the output switch of the passive nuclear level gauge can be determined based on the alternative level value, wherein the alternative level value is the level value used to replace the normal level value when nuclear radiation flaw detection interference is detected, and the normal level value is the level value generated by the passive nuclear level gauge based on real-time measurement data when the output protection operation is not performed.

2. The method for controlling the level output of a passive nuclear level gauge during nuclear radiation flaw detection according to claim 1, characterized in that, Also includes: The nuclear radiation flaw detection interference in the measurement environment of the passive nuclear level gauge has ended. Cancel the output protection operation in progress.

3. The method for controlling the level output of a passive nuclear level gauge during nuclear radiation flaw detection according to claim 1 or 2, characterized in that, If the preset first condition is met, it is determined that there is nuclear radiation flaw detection interference in the measurement environment where the passive nuclear level gauge is located; If the preset second condition is met, it is determined that the nuclear radiation flaw detection interference in the measurement environment where the passive nuclear level gauge is located has ended.

4. The method for controlling the level output of a passive nuclear level gauge during nuclear radiation flaw detection according to claim 1, characterized in that, The presence of nuclear radiation flaw detection interference in the measurement environment of the passive nuclear level gauge can be determined by at least one of the following first conditions: The passive nuclear level gauge acquires nuclear radiation measurement values ​​within one or more preset time windows, and determines statistical characteristics and / or changes in statistical characteristics based on the nuclear radiation measurement values. When the statistical characteristics and changes in statistical characteristics meet the corresponding preset third condition; And / or, the nuclear radiation measurement value of the passive nuclear level gauge reaches the preset first radiation threshold; And / or, the operating time of the passive nuclear level gauge experiences a preset first duration; And / or, the current time reaches the preset time to start flaw detection; And / or, the passive nuclear level gauge or its control system receives a confirmation signal from an external device indicating that flaw detection has been initiated.

5. The method for controlling the level output of a passive nuclear level gauge during nuclear radiation flaw detection according to claim 4, characterized in that, The statistical feature is at least one of the following: measured value, mean, extreme value, extreme value difference, standard deviation, variance, quantile, interquartile range, or mode.

6. The method for controlling the level output of a passive nuclear level gauge during nuclear radiation flaw detection according to claim 4, characterized in that, The preset third condition is at least one of the following: The statistical characteristics of one or more radiation measurements of the passive nuclear level gauge exceed the preset first characteristic range within a preset time window; And / or, the change in the statistical characteristics of the nuclear radiation measurement values ​​of the passive nuclear level gauge in the current time window compared to the statistical characteristics of the nuclear radiation measurement values ​​in the previous time window exceeds the preset second characteristic range; And / or, the absolute value of the statistical characteristics of the nuclear radiation measurement values ​​of the passive nuclear level gauge exceeding the preset third characteristic range the number of times; And / or, the variation between the maximum and minimum values ​​of the nuclear radiation measurements in the statistical characteristics of the passive nuclear level gauge in multiple preset time windows exceeds the preset fourth characteristic range; And / or, based on the statistical characteristics of the nuclear radiation measurements corresponding to multiple preset time windows of the passive nuclear level gauge, calculate the change in the statistical characteristics corresponding to each two adjacent time windows, and the number of times the change exceeds the preset fifth characteristic range meets the standard.

7. The method for controlling the level output of a passive nuclear level gauge during nuclear radiation flaw detection according to claim 2, characterized in that, The nuclear radiation flaw detection interference in the measurement environment of the passive nuclear level gauge is determined to have ended by at least one of the following second conditions: From the moment nuclear radiation flaw detection interference is detected or output protection operation is executed, a preset second duration will elapse; And / or, the current time reaches the preset stop flaw detection time; And / or, acquire the nuclear radiation measurement values ​​of the passive nuclear level gauge within one or more preset time windows, determine statistical characteristics and / or changes in statistical characteristics based on the nuclear radiation measurement values, and when the statistical characteristics and changes in statistical characteristics meet the corresponding preset fourth condition; And / or, the passive nuclear level gauge or its control system receives a confirmation signal from an external device indicating the end of flaw detection; And / or, the alternative material level value generated based on the real-time generated real-time alternative material level calculation formula reaches the preset material level threshold, wherein the real-time alternative material level calculation formula is a real-time prediction model established based on historical material level and time data within a preset time before the determination of nuclear radiation flaw detection interference; And / or, the alternative material level value generated based on the pre-built prediction model reaches the preset material level threshold, wherein the prediction model is a prediction model pre-built based on historical real material level and time data within at least one process cycle.

8. The method for controlling the level output of a passive nuclear level gauge during nuclear radiation flaw detection according to claim 7, characterized in that, The preset fourth condition is at least one of the following: The statistical characteristics of the radiation measurement values ​​of the passive nuclear level gauge in one or more consecutive sampling cycles are restored to the preset sixth characteristic range; And / or, the number of times the statistical characteristics of the radiation measurement values ​​of the passive nuclear level gauge recover to the preset seventh characteristic range in multiple sampling cycles meets the standard.

9. The method for controlling the level output of a passive nuclear level gauge during nuclear radiation flaw detection according to claim 2, characterized in that, The output safeguard operation in progress is to replace the normal level value of the passive nuclear level gauge with a substitute level value, which is configured to be determined according to a preset safeguard strategy; After the nuclear radiation flaw detection interference ends and the ongoing output assurance operation is canceled, the following also applies: The actual level data measured by the passive nuclear level gauge is compared and analyzed with the alternative level values ​​used in the output support operation. The preset protection strategy is adaptively optimized based on the comparison analysis results.

10. The method for controlling the level output of a passive nuclear level gauge during nuclear radiation flaw detection according to claim 1, characterized in that, The alternative material level value is determined using at least one of the following safeguard strategies: The corresponding material level value is preset for the alternative material level value and stored; The normal material level value at the last moment before the determination of nuclear radiation flaw detection interference was used as the alternative material level value. The statistical characteristic value within a preset time period before the determination of nuclear radiation flaw detection interference is used as the alternative material level value; After determining that there is nuclear radiation flaw detection interference, at least one unit time window of radiation measurement values ​​are collected as a candidate sample set. Based on the historical measurement data before the nuclear radiation flaw detection interference, specific measurement values ​​or local feature values ​​that meet the preset deviation conditions are selected from the candidate sample set and used as the alternative material level values. The real-time replacement material level calculation formula is used to calculate the replacement material level value. The real-time replacement material level calculation formula is a real-time prediction model established based on historical material level and time data within a preset time before the determination of nuclear radiation flaw detection interference. The predicted material level value generated by the prediction model pre-built based on historical real material level and time data within at least one process cycle is used as the alternative material level value.

11. The method for controlling the level output of a passive nuclear level gauge during nuclear radiation flaw detection according to claim 10, characterized in that, The pre-built prediction model includes any of the following: The prediction model is a fitting function obtained by fitting or regression analysis of historical material level and time data; The prediction model calls upon historical material level and time data classified according to process in the absence of nuclear radiation flaw detection interference to determine the material level value that matches the current process and time period as the alternative material level value; The prediction model is a function model constructed based on historical operating data according to process classification; The artificial intelligence model is trained using historical operating data to obtain the prediction model used to predict material level.

12. The method for controlling the level output of a passive nuclear level gauge during nuclear radiation flaw detection according to claim 11, characterized in that, The pre-built prediction model is integrated inside the passive nuclear level gauge, or the pre-built prediction model is set in an external device that communicates with the passive nuclear level gauge.

13. The method for controlling the level output of a passive nuclear level gauge during nuclear radiation flaw detection according to claim 7 or 10, characterized in that, The real-time replacement material level calculation formula is a fitting function relationship: L=aT 2 +bT+c, where L is the material level, T is the time, and a, b, c are fitting constants.

14. The method for controlling the level output of a passive nuclear level gauge during nuclear radiation flaw detection according to claim 7 or 10, characterized in that, The starting time point of the historical material level and time data within the preset time period is determined by any of the following methods: The starting time point is defined as the trigger time of the start signal for the current process cycle. The starting time point is the moment when the material level reaches the preset material level threshold. The starting time point is the moment when the trigger signal sent by the external device is received. Based on the preset time period based on the patterns of historical flaw detection events, the moment when nuclear radiation flaw detection interference is determined to be earlier is moved forward to obtain the starting time point.

15. The method for controlling the level output of a passive nuclear level gauge during nuclear radiation flaw detection according to claim 1, characterized in that, Output guarantee operations will be prohibited if any of the following preset conditions are met: The measurement results of the passive nuclear level gauge are within the preset safety range; The measurement results of the passive nuclear level gauge are within the preset fault range; The passive nuclear level gauge displays and / or outputs a warning level for the material level. The output protection function of the passive nuclear level gauge or its control system is turned off. The current time is within the preset normal working hours.

16. A passive nuclear level gauge, characterized in that, Includes a radiation detector, a processor, and a signal output unit, wherein: The radiation detector is configured to collect nuclear radiation in the measurement environment; The processor is configured to determine whether there is nuclear radiation detection interference in the measurement environment based on the nuclear radiation collected by the radiation detector or based on a preset first condition. If so, the processor executes a preset output protection operation to generate corresponding switch signals, digital signals, and / or analog signals; otherwise, the processor generates corresponding switch signals, digital signals, and / or analog signals based on the nuclear radiation collected by the radiation detector; and the signal result generated based on the output protection operation is different from the signal result generated based on the collected nuclear radiation. The signal output unit is configured to output switch signals, digital signals, and / or analog signals generated by the processor.

17. The passive nuclear level gauge according to claim 16, characterized in that, The processor is configured to perform the steps of the level output control method during nuclear radiation testing as described in any one of claims 1 to 15.

18. The passive nuclear level gauge according to claim 16, characterized in that, The processor is further configured to: when performing the output protection operation, generate and store a status flag bit, which is used to identify that the signal currently output by the signal output unit is an output based on the protection strategy under nuclear radiation flaw detection interference; And / or, the signal output unit is further configured to output the status flag bit; And / or, the passive nuclear level gauge further includes a communication interface configured to communicate with the signal output unit and an external device.

19. The passive nuclear level gauge according to claim 16, characterized in that, The signal output unit includes one or more of the following: Isolated digital output channels; Analog output circuit for outputting 4-20mA or 0-10V analog signals; Digital output circuit used to output digital level signals.

20. An industrial material level monitoring system, characterized in that, include: Passive nuclear level gauge as described in any one of claims 16 to 19; An external control device, communicatively connected to the passive nuclear level gauge, is configured to perform the steps of the level output control method during nuclear radiation testing as described in any one of claims 1 to 15, or the external control device is configured to send at least one of the following signals to the passive nuclear level gauge: A confirmation signal used to determine the presence of nuclear radiation flaw detection interference in the measurement environment of a passive nuclear level gauge; A confirmation signal used to determine the end of interference with nuclear radiation flaw detection; A process synchronization signal used to synchronize historical material level and time data at the start time point.

21. The industrial material level monitoring system according to claim 20, characterized in that, The external control device is a distributed control system or a programmable logic controller. And / or, the passive nuclear level gauge is configured to upload a status flag bit and / or output mode information to identify the current output guarantee operation state via a communication interface, and the external control device is configured to distinguish and display the uploaded information on the monitoring screen or generate corresponding operation logs.

Images