Self-calibration method and system for micro-water sensor based on dynamic adsorption and desorption curve

Abstract

The present application relates to the technical field of sensor calibration, and discloses a micro-water sensor self-calibration method and system based on a dynamic adsorption and desorption curve, which comprises the following steps: applying a controllable heating pulse to the micro-water sensor to drive the desorption of adsorbed water molecules during the normal operation gap of the device; recording a dynamic recovery curve of the capacitance value of the micro-water sensor changing with time continuously after the heating pulse ends; and inversely calculating the current real water content based on the mapping relationship between the characteristic value of the slope of the dynamic recovery curve and the current real water content; correcting the measurement reference of the micro-water sensor according to the real micro-water content value; and the micro-water sensor continues to perform an online monitoring task according to the corrected measurement reference. The present application can improve the consistency of the micro-water sensor in long-term measurement.

Description

Technical Field

[0001] This invention relates to the field of sensor calibration technology, and in particular to a self-calibration method and system for micro water sensors based on dynamic adsorption-desorption curves. Background Technology

[0002] Capacitive micro-moisture sensors measure trace amounts of moisture based on factory calibration curves and rely on a built-in heating element to perform desorption-cooling-single-point reading calibration at fixed time intervals or simple state triggering methods. This static reading corrects for zero-point offset. The core of this method relies on a preset triggering timing and a single steady-state correction value. However, the actual drift of the sensor is dynamically coupled with multiple factors such as temperature, aging, and contaminants. Fixed triggering cannot synchronize with the actual drift changes, and single-point correction cannot describe the characteristic evolution of the sensitive film throughout the adsorption and recovery process. As a result, the effectiveness of calibration is fundamentally limited by the inherent defects of the triggering conditions and correction methods.

[0003] In dynamic operating conditions such as power GIS and chemical process gas analysis, fluctuations in ambient temperature and gas composition can cause drift rates to change constantly. Fixed-period calibration may introduce additional errors due to being too early or too late, and may also cause unnecessary heating aging. When the sensitive membrane undergoes irreversible fouling or structural aging, the rate and equilibrium trajectory of its capacitance recovery process deviate from the initial state, but corrections based solely on single-point readings after cooling cannot capture these dynamic deviations. Therefore, even after calibration, the sensor may still have uncorrected nonlinear deviations across its entire range, making it difficult to effectively guarantee the reliability and consistency of long-term online monitoring data. Summary of the Invention

[0004] This invention provides a self-calibration method and system for micro water sensors based on dynamic adsorption-desorption curves to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides a self-calibration method for a micro-water sensor based on dynamic adsorption-desorption curves, comprising: Step 110: During normal equipment operation intervals, apply a controllable heating pulse to the micro water sensor to drive the desorption of adsorbed water molecules; Step 120: After the heating pulse ends, continuously record the dynamic recovery curve of the capacitance value of the micro-water sensor changing over time, and based on the mapping relationship between the slope characteristic value of the dynamic recovery curve and the current true water content, reversely calculate the current true water content. Step 130: Calibrate the measurement benchmark of the micro-moisture sensor based on the actual micro-moisture content value; Step 140: The micro-water sensor continues to perform online monitoring tasks according to the calibrated measurement benchmark.

[0006] To address the aforementioned problems, this invention also provides a self-calibration system for a micro-water sensor based on dynamic adsorption-desorption curves, the system comprising: The heating module is used to apply controllable heating pulses to the micro water sensor during normal equipment operation intervals to drive the desorption of adsorbed water molecules; The water content estimation module is used to continuously record the dynamic recovery curve of the capacitance value of the micro water sensor changing over time after the heating pulse ends, and to reverse estimate the current true water content based on the mapping relationship between the slope characteristic value of the dynamic recovery curve and the current true water content. The measurement reference calibration module is used to calibrate the measurement reference of the micro-moisture sensor according to the actual micro-moisture content value; The re-monitoring module is used by the micro-water sensor to continue performing online monitoring tasks based on the calibrated measurement benchmark.

[0007] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention drives the desorption of adsorbed water molecules by applying a controllable heating pulse, and continuously records the dynamic recovery curve of the capacitance value over time after heating. The slope characteristic value of this curve is used as the basis for estimating the current true water content. This slope characteristic value characterizes the dynamic response behavior of the entire adsorption recovery process, rather than relying on a single static reading. Therefore, even when the characteristic curve of the sensor deviates due to aging or contamination, the true water content matching the actual adsorption characteristics can still be obtained through the slope-water content mapping relationship, and the measurement benchmark can be corrected accordingly. Thus, the effectiveness of calibration is freed from dependence on preset triggering timing and static correction values, and the measurement benchmark can be updated synchronously with changes in the actual characteristics of the sensor, fundamentally solving the technical problem of full-range nonlinear deviation after calibration as described in the background art.

[0008] 2. This invention acquires the current ambient temperature value before applying the heating pulse, compares the current capacitance value with the factory reference capacitance value at the same temperature, and generates a calibration trigger signal only when the difference exceeds the drift threshold. This scheme uses the quantitative detection result of the sensor's current actual drift as the basis for determining calibration initiation, avoiding redundant calibration operations caused by fixed-cycle triggering, reducing the cumulative aging effects of unnecessary heating cycles on the sensitive membrane, and ensuring that calibration is only performed when drift has actually occurred to the point of requiring intervention, thus improving the accuracy of calibration triggering and resource utilization efficiency. Attached Figure Description

[0009] Figure 1 This is a schematic flowchart of a self-calibration method for a micro water sensor based on dynamic adsorption-desorption curves provided in an embodiment of the present invention. Figure 2This is a functional block diagram of a micro water sensor self-calibration system based on dynamic adsorption-desorption curves provided in an embodiment of the present invention. Figure 3 This is a diagram showing the determination of normal operating intervals and idle confirmation times for a micro water sensor self-calibration system based on dynamic adsorption-desorption curves, provided in an embodiment of the present invention. Figure 4 A diagram illustrating the dual triggering mechanism based on drift exceedance and cumulative operating cycle of a micro water sensor self-calibration system based on dynamic adsorption-desorption curves, provided in an embodiment of the present invention. Figure 5 A mapping diagram showing the slope characteristic value of the recovery curve and the actual micro-water content of a micro-water sensor self-calibration system based on dynamic adsorption-desorption curves provided in an embodiment of the present invention. Figure 6 This is a comparison chart of the measurement accuracy and linearity of a micro water sensor before and after self-calibration in a micro water sensor self-calibration system based on dynamic adsorption-desorption curves, provided in an embodiment of the present invention.

[0010] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0011] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0012] This application provides a self-calibration method for a micro-water sensor based on dynamic adsorption-desorption curves. The execution subject of this self-calibration method includes, but is not limited to, at least one of the following electronic devices that can be configured to execute the method provided in this application: a server, a terminal, etc. In other words, the self-calibration method for a micro-water sensor based on dynamic adsorption-desorption curves can be executed by software or hardware installed on a terminal device or a server device. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cluster of cloud servers. The server can be an independent server or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDN), and big data and artificial intelligence platforms.

[0013] Reference Figure 1 The diagram shown is a schematic flowchart of a micro-water sensor self-calibration method based on dynamic adsorption-desorption curves according to an embodiment of the present invention. In this embodiment, the micro-water sensor self-calibration method based on dynamic adsorption-desorption curves includes: Step 110: During normal equipment operation intervals, apply a controllable heating pulse to the micro water sensor to drive the desorption of adsorbed water molecules; In this embodiment, the "controllable heating pulse" refers to a finite-duration heating energy output generated within a controlled time window by a heating element built into the micro-water sensor, such as a thin-film resistance heater integrated into the sensor substrate, a ceramic heating element, or a resistance heating wire wound around the sensor housing. The core function of this heating pulse is to rapidly raise the temperature of the sensor's sensitive film, such as a polyimide film or alumina porous medium, above the water molecule desorption activation energy threshold. This breaks the van der Waals forces or hydrogen bonds between adsorbed water molecules and the active sites on the sensitive film surface, driving the adsorbed water molecules to desorb from the sensitive film surface and leave the sensing area. This physical process establishes a unified initial state for subsequent dynamic recovery curve acquisition—a "desorption baseline state" where the water molecule coverage on the sensitive film surface approaches zero.

[0014] In one feasible implementation, the heating duration threshold of the heating pulse is set to 2 to 10 seconds. Preferably, the heating duration threshold can be set to 5 seconds.

[0015] The determination of this value range is based on the following: If the heating time is too short, such as less than 2 seconds, the internal temperature of the sensitive membrane may not rise uniformly to the effective desorption temperature before heating stops, resulting in incomplete desorption. Residual water molecules will still occupy some active sites, thus affecting the consistency of the initial state of the subsequent recovery curve. If the heating time is too long, such as more than 10 seconds, the temperature of the sensitive membrane may be too high, accelerating the thermal oxidation aging of the polymer material and shortening the service life of the sensor.

[0016] Those skilled in the art can determine the specific heating time threshold through prior experiments based on the material type of the sensor's sensitive film, such as the thermal decomposition temperature of polyimide, which is typically higher than 300°C, the power density of the heating element, and the thermal capacity of the sensor package.

[0017] In another feasible implementation, the heating duration threshold of the heating pulse can also be dynamically adjusted according to the ambient temperature of the sensor.

[0018] For example, when the ambient temperature is below 0°C, the heating time threshold can be extended to 8 to 12 seconds to compensate for the reduced heating rate of the heating element in low-temperature environments; when the ambient temperature is above 40°C, the heating time threshold can be shortened to 2 to 4 seconds to avoid overheating of the sensitive membrane.

[0019] In another feasible implementation, the heating temperature of the heating pulse, i.e. the highest temperature reached by the sensitive membrane at the end of heating, is controlled between 90°C and 150°C.

[0020] Preferably, the heating temperature can be controlled between 110°C and 130°C.

[0021] The selection of this temperature range is based on the following criteria: below 90℃, the desorption rate of water molecules is slow, making it difficult to achieve sufficient desorption within a limited time; above 150℃, although the desorption rate increases, it may adversely affect the long-term stability of the sensitive membrane material. The actual heating temperature can be indirectly achieved by controlling the power of the heating pulse, such as the voltage or current amplitude, and its duration, or by implementing closed-loop feedback control through a temperature sensor installed inside or near the sensor.

[0022] In this embodiment, by applying a controllable heating pulse during the normal operation interval of the device, a uniform initial desorption state can be established for the sensor's sensitive membrane without interfering with the online monitoring task, thereby laying the foundation for the accurate acquisition of the subsequent dynamic recovery curve.

[0023] Step 120: After the heating pulse ends, continuously record the dynamic recovery curve of the capacitance value of the micro-water sensor changing over time, and based on the mapping relationship between the slope characteristic value of the dynamic recovery curve and the current true water content, reversely calculate the current true water content. In this embodiment of the application, the "dynamic recovery curve" refers to the trajectory of the capacitance value changing continuously over time as the sensor's sensitive membrane, starting from the desorption state after the heating pulse ends, re-adsorbs water molecules in the ambient atmosphere until a new adsorption-desorption equilibrium is reached.

[0024] The physical mechanism is as follows: after the heating pulse ends, the temperature of the sensitive membrane gradually decreases to the ambient temperature, and water molecules begin to re-adsorb onto the active sites on the surface of the sensitive membrane. Since the adsorption rate is related to factors such as the current water molecule coverage on the surface of the sensitive membrane, the partial pressure of water vapor in the environment, and the temperature, the increase in capacitance over time actually reflects the dynamic response characteristics of the sensitive membrane to trace amounts of moisture in the environment. There is a one-to-one mapping relationship between the slope characteristic value of this dynamic response characteristic and the actual water content in the current environment.

[0025] In one feasible implementation, the "continuous recording" specifically includes the following operations: While heating is stopped, the capacitance value of the micro-water sensor is continuously acquired at a high sampling frequency. Preferably, the sampling frequency is set to 10Hz to 100Hz, and more preferably to 20Hz.

[0026] The sampling frequency was selected based on the fact that the rate of change is relatively fast in the initial stage of the capacitance recovery process. A lower sampling frequency, such as below 5Hz, may lead to the loss of key inflection point information. On the other hand, an excessively high sampling frequency, such as above 200Hz, will generate a large amount of redundant data, increasing the burden of storage and processing, while having limited improvement on the accuracy of slope feature value calculation.

[0027] At each capacitance value acquisition moment, the temperature value of the environment where the micro-water sensor is located is read synchronously, and the temperature value, the currently acquired capacitance value, and the corresponding timestamp are combined into a sampling data point.

[0028] The necessity of synchronous temperature acquisition lies in the fact that the capacitance value itself is temperature-dependent. Synchronously recording the temperature value allows for temperature compensation of the capacitance value during subsequent data processing, eliminating the interference of temperature fluctuations on the calculation of slope characteristic values.

[0029] All the continuously generated sampling data points are stored sequentially as capacitance-temperature-time series data in chronological order.

[0030] The acquisition, reading, and storage operations are continuously performed until the number of sampling data points contained in the capacitance-temperature-time series data reaches the sampling point threshold, at which point the acquisition stops, and the stored complete capacitance-temperature-time series data is used as a dynamic recovery curve.

[0031] Preferably, the threshold for the number of sampling points is set to 100 to 1000 points.

[0032] More preferably, the threshold is set to 500 points. Combined with the aforementioned sampling frequency of 20Hz, 500 sampling points correspond to a continuous acquisition time of 25 seconds. This duration is usually sufficient to cover the complete recovery process of the capacitance value rising from the lowest point after desorption to a stable equilibrium.

[0033] In another feasible implementation, the threshold number of sampling points can also be adjusted according to the sensor type and application scenario.

[0034] For example, for thin-film capacitive sensors with a fast response time of less than 5 seconds, the sampling point threshold can be set to 200 points; for sensors with a slower response time, the sampling point threshold can be set to 800 to 1000 points.

[0035] Those skilled in the art can determine the threshold number of sampling points based on the sensor's response time constant, which is usually provided by the sensor manufacturer or obtained through actual measurement, according to the principle that the acquisition time covers 3 to 5 times the response time constant.

[0036] Through the above method, the embodiments of this application can obtain a complete dynamic recovery curve containing temperature compensation information. This curve not only records the recovery trajectory of the capacitance value over time, but also records the changes in ambient temperature, providing a high-quality data foundation for the accurate extraction of slope feature values ​​in the future.

[0037] Step 130: Calibrate the measurement benchmark of the micro-moisture sensor based on the actual micro-moisture content value; In the embodiments of this application, the "slope characteristic value" refers to the rate of change of capacitance value per unit time during the capacitance value rising phase of the dynamic recovery curve. Its physical meaning is to characterize the dynamic rate at which the sensitive membrane re-adsorbs water molecules from the desorbed state to achieve equilibrium with the partial pressure of water vapor in the environment.

[0038] According to the theory of adsorption kinetics, at the same ambient temperature, the higher the partial pressure of water vapor in the environment, that is, the higher the water content, the greater the driving force for water molecules to occupy the active sites on the surface of the sensitive membrane, the faster the adsorption rate, which is reflected in the recovery curve as the greater the slope of the increase in capacitance value.

[0039] Therefore, there is a monotonic mapping relationship between the slope characteristic value and the current true water content, which can be reversed through a pre-calibrated mapping table.

[0040] In one feasible implementation, the "reverse calculation" specifically includes the following operations: Based on the capacitance value and timestamp of each sampling data point in the dynamic recovery curve, the rising phase of the capacitance value over time in the dynamic recovery curve is determined.

[0041] Specifically, the rising phase can be identified as follows: starting from the first sampling point after the heating pulse ends, calculate the change in capacitance between adjacent sampling points in sequence. When a sequence of sampling points with continuously increasing capacitance values ​​appears, the starting point to the ending point of the sequence is defined as the rising phase.

[0042] During the rising phase, consecutive sampled data points are selected as the slope calculation interval.

[0043] Preferably, the slope calculation range is selected as the range from 10% to 90% of the capacitance value rising from the lowest value after desorption to the stable equilibrium value during the rising phase, so as to avoid the nonlinear transient response in the initial stage and the asymptotic saturation region when approaching equilibrium, and to ensure the best linearity of the slope calculation.

[0044] The ratio of the difference between the initial and final capacitance values ​​to the difference between the initial and final times within the slope calculation interval is taken as the capacitance change rate, and this capacitance change rate is output as the slope characteristic value. The calculation formula is as follows: ; in, This is the slope characteristic value, in units of pF / s; and These are the capacitance values ​​at the start and end times of the slope calculation interval, respectively, in pF; and These are the start and end times of the slope calculation interval, in seconds.

[0045] Call the pre-stored table of correspondence between slope feature values ​​and micro-water content.

[0046] This correspondence table is established during the sensor's factory calibration stage by placing the sensor in a standard atmosphere with known trace moisture content, collecting dynamic recovery curves and calculating slope characteristic values ​​using the same method as in this application, thereby establishing a one-to-one correspondence between slope characteristic values ​​and trace moisture content. The correspondence table records the range of slope characteristic values ​​and the trace moisture content value corresponding to each range.

[0047] The slope feature value is matched with each slope feature value interval in the corresponding relationship table to determine the target slope feature value interval to which the slope feature value belongs; according to the target slope feature value interval, the corresponding micro water content value is read from the corresponding relationship table and the micro water content value is used as the current true water content.

[0048] In another feasible implementation, the mapping relationship between the slope feature value and the micro-water content can also be achieved by function fitting.

[0049] For example, the calibration data can be fitted into a polynomial or exponential function using the least squares method, and the slope characteristic value can be directly substituted into the function to calculate the current true water content during actual calculation.

[0050] Another feasible implementation is to use interpolation algorithms, such as linear interpolation or spline interpolation, to perform fine interpolation between calibration data points in order to obtain higher resolution water content estimation results.

[0051] In this embodiment of the application, by using the slope characteristic value of the dynamic recovery curve instead of a single static reading to estimate the water content, the dynamic characteristic information of the sensitive membrane throughout the adsorption recovery process can be captured. Thus, even when the static reading shifts due to sensor aging or contamination, the true water content value that matches the current actual adsorption characteristics can still be obtained.

[0052] Step 140: The micro-water sensor continues to perform online monitoring tasks according to the calibrated measurement benchmark.

[0053] In this embodiment of the invention, the "measurement benchmark" refers to the reference mapping relationship used by the micro-moisture sensor to convert the original capacitance measurement value into a moisture content output value. In capacitive micro-moisture sensors, this mapping relationship is usually reflected in the factory calibration curve or calibration equation, that is, the correspondence between capacitance value and moisture content. When the sensor drifts (caused by factors such as sensitive membrane aging, contaminant adhesion, temperature cycling, etc.), the capacitance value corresponding to the same true moisture content will change, causing the moisture content calculated according to the original calibration curve to deviate from the true value. The purpose of this step is to use the current true moisture content calculated in step three to correct the measurement benchmark online, so that it is matched again with the current actual characteristics of the sensor.

[0054] In one feasible implementation, the "correction" specifically includes the following operations: First, based on the current measurement benchmark of the micro-water sensor, i.e. the current calibration curve or calibration equation, the original capacitance measurement value at the current moment is substituted into the calculation to obtain the original measured water content.

[0055] Secondly, the difference between the original measured moisture content and the current true moisture content calculated in step three is used to characterize the degree of drift of the measurement benchmark. This difference can be positive or negative; a positive value indicates that the reading under the current measurement benchmark is too high, that is, the sensor has drifted positively, while a negative value indicates that the reading is too low, that is, the sensor has drifted negatively.

[0056] Based on the drift degree, a correction compensation parameter is generated, and the measurement reference is corrected according to the correction compensation parameter. In one specific implementation, the correction compensation parameter can be generated as follows: the capacitance reading when the capacitance value in the dynamic recovery curve drops to its minimum value after the heating pulse ends is taken as the lowest capacitance value after desorption, denoted as... The stable capacitance value reached after the capacitance value in the dynamic recovery curve increases and tends to stabilize over time is recorded as the recovery equilibrium capacitance value. The difference between the two values ​​yields the adsorption recovery amplitude. The adsorption recovery amplitude reflects the actual adsorption capacity of the sensitive membrane under current environmental conditions. This adsorption recovery amplitude and the standard adsorption recovery amplitude are recorded as follows: This refers to the adsorption recovery deviation between the adsorption recovery amplitudes of the sensor under factory calibration conditions. Simultaneously, the slope feature value is obtained. The standard slope eigenvalue is denoted as... The slope deviation between the slope characteristic values ​​corresponding to the sensor in its factory calibration state is... .

[0057] The adsorption recovery deviation and the offset correction coefficient are used. The product of the slope deviation and the time constant, Adding the products together, we obtain the correction compensation parameters. : ; Among them, the offset correction coefficient The dimensions are water content / capacitance, for example, ppm / pF, and time constant. The units are water content·second / pF, for example, ppm·s / pF, determined by the factory calibration experiment of the micro-water sensor. Their physical meanings respectively characterize the contribution weights of adsorption recovery amplitude deviation and slope deviation to the measurement reference drift. Finally, according to the aforementioned correction compensation parameters... The measurement reference is corrected, for example by shifting the original calibration curve along the capacitance axis. The corresponding capacitance offset, or directly superimposed on the original measured moisture content. As the corrected measurement value.

[0058] In another feasible implementation, the correction compensation parameter can be generated in a simpler way. For example, the difference between the original measured moisture content and the current true moisture content can be directly used as the correction compensation parameter to adjust subsequent measurements.

[0059] In another feasible implementation, a recursive filtering method can be used to perform a weighted average of the current correction compensation parameters and the historical correction compensation parameters to smooth the correction process and avoid excessive disturbance to the measurement reference caused by the random error of a single correction.

[0060] During the next online monitoring, the original measured moisture content is calculated using the corrected measurement benchmark, thereby outputting the corrected trace moisture content value.

[0061] In another feasible implementation, the correction can also be performed in a segmented manner.

[0062] Preferably, the full range of the sensor is divided into multiple sub-intervals, and correction compensation parameters are calculated independently for each sub-interval to achieve fine correction of nonlinear drift. This method is particularly suitable for situations where the calibration curve shape undergoes nonlinear distortion due to the aging of the sensitive membrane.

[0063] In this embodiment, by using the information of two dimensions, namely the adsorption recovery amplitude and slope characteristic value extracted from the dynamic recovery curve, to generate correction compensation parameters, a multi-dimensional comprehensive correction of the measurement benchmark is achieved. Compared with the traditional method that relies only on single-point readings, it can more comprehensively compensate for the drift of the sensor caused by factors such as aging and contamination, thereby improving the measurement accuracy across the entire range after calibration.

[0064] In one feasible implementation, in step 110 of the above embodiments, the calibration start-up method based on cumulative running time can also be triggered by a fixed time period. The preset time period is a fixed time length, which is preset by the operator according to the sensor type, the historical drift pattern of the application scenario, and the measurement accuracy requirements.

[0065] Preferably, for the capacitive sensor used for monitoring the trace moisture in insulating gas in power GIS equipment, the preset time period is set to 7 to 30 days; more preferably, it is set to 15 days.

[0066] The determination of this value range is based on the following: A cycle that is too short, less than 7 days, will lead to excessively frequent calibration operations, accelerating the cumulative thermal aging effect of the heating cycle on the sensitive membrane, and simultaneously occupying a significant amount of normal equipment downtime resources; a cycle that is too long, greater than 30 days, may result in drift accumulating to the point of affecting measurement accuracy between calibrations. For online micro-water analyzers operating continuously in chemical processes, due to relatively stable operating conditions, this preset time period can be set to 30 to 90 days. In one specific implementation, the cumulative operating time is recorded using a non-volatile counter. This counter increments from the previously saved value each time the sensor is powered on, and after each calibration is completed, the processor resets the counter to zero and restarts the timing.

[0067] In another feasible implementation, in step 110 of the above embodiments, the calibration start-up method based on cumulative running time can also be triggered by a dynamically adjusted time period.

[0068] The preset time period is not a fixed value, but is dynamically adjusted according to the actual drift rate of the sensor in the most recent calibration period.

[0069] Specifically, the difference between the original measured water content and the current actual water content calculated at each calibration is recorded, i.e., the drift amount. This drift amount is then divided by the time interval between the previous calibration and the current calibration to obtain the average drift rate. The average drift rate is compared with a preset rate threshold: when the average drift rate is higher than a first rate threshold, such as 0.5 ppm / day, the time period for the next calibration is shortened to the first period value, such as 7 days; when the average drift rate is lower than a second rate threshold, such as 0.1 ppm / day, the time period for the next calibration is extended to the second period value, such as 60 days; when the average drift rate is between the two, a default period value, such as 30 days, is used.

[0070] This method is applicable to application environments where the sensor's operating conditions fluctuate in stages and the drift rate changes significantly over time. It can adaptively adjust the calibration frequency according to the actual state of the sensor, minimizing redundant calibration operations while ensuring measurement accuracy.

[0071] In one feasible implementation, in step 110 of the above embodiments, the calibration start-up method based on cumulative running time can also be that the preset time period and the cumulative number of heating pulses applied are combined as the calibration triggering condition.

[0072] Specifically, the cumulative running time since the last calibration is recorded, along with the number of heating pulses executed by the micro-water sensor since the last calibration. A calibration trigger signal is generated when the cumulative running time reaches a first time threshold (e.g., 20 days) and the cumulative number of heating pulses reaches a first number threshold (e.g., 50 times), or when the cumulative running time has not reached the first time threshold but the cumulative number of heating pulses has exceeded a second number threshold (e.g., 100 times). The physical basis of this method is that each heating pulse application causes a certain degree of temperature shock to the sensitive membrane, accelerating the aging process of the material. Therefore, incorporating the number of heating pulses into the trigger determination factor allows for a more comprehensive assessment of the sensor's actual aging state. This method is particularly suitable for applications with high calibration frequencies and sensors subjected to numerous thermal cycles.

[0073] Regardless of which method is used, the embodiments of this application can ensure that the sensor undergoes periodic calibration during long-term continuous operation, preventing excessive drift accumulation due to time accumulation, thereby ensuring the long-term reliability of online monitoring data.

[0074] The first fixed time period method is simple to implement and consumes the least amount of computing resources, making it suitable for scenarios with stable operating conditions and known drift patterns. The second dynamic adjustment method can adaptively adjust the calibration frequency according to the actual drift state of the sensor, improving resource utilization efficiency while ensuring accuracy. The third dual-factor triggering method can more comprehensively consider the impact of thermal cycling aging on the sensitive membrane, making it suitable for applications with high calibration frequency or dense heating pulses.

[0075] In one feasible implementation, in step 120 of the above embodiments, the slope calculation interval can also be selected based on a fixed percentage interval.

[0076] Specifically, the dynamic recovery curve is defined as the complete upward phase between the lowest value after desorption (i.e., the minimum value to which the capacitance first drops after the heating pulse ends) and the stable equilibrium value (i.e., the capacitance reading corresponding to the capacitance value changing less than a preset threshold over time).

[0077] During this rising phase, the sampling point corresponding to the point where the capacitance value rises from the lowest value after desorption to 10% of the stable equilibrium value is taken as the starting point of the interval, and the sampling point corresponding to the point where the capacitance value rises to 90% of the stable equilibrium value is taken as the ending point of the interval; the continuous sampling data point sequence defined by the starting point and the ending point is taken as the slope calculation interval.

[0078] In this method, the selection of the range of 10% to 90% is based on the following: avoiding the nonlinear transient response region caused by thermal residual effect and non-uniform temperature distribution in the initial stage of rise, which is usually located in the range of 0% to 10%, and avoiding the asymptotic saturation region caused by the weakening of adsorption driving force when approaching equilibrium, which is usually located in the range of 90% to 100%, thereby ensuring that the capacitance-time relationship in the selected range has good linearity, making the calculation of slope characteristic values ​​more stable and reliable.

[0079] In another feasible implementation, in step 120 of the above embodiments, the slope calculation interval can also be selected based on an adaptive interval selection of the rate of change threshold.

[0080] Specifically, after identifying the rising phase, the capacitance change rate between adjacent sampling points within that phase is calculated sequentially, i.e., the capacitance increment per unit time. Starting from the beginning of the rising phase and scanning backwards, when the capacitance change rate of three or more consecutive sampling points is greater than a preset lower limit threshold (e.g., 0.5 pF / s), the first sampling point meeting this condition is taken as the starting point of the interval. Starting from the end of the rising phase and scanning forwards, when the capacitance change rate of three or more consecutive sampling points is greater than the lower limit threshold, the last sampling point meeting this condition is taken as the ending point of the interval. The sequence of continuous sampling data points defined by the starting and ending points is used as the slope calculation interval.

[0081] This method is applicable when the sensor response speed is slow or the rising segment of the dynamic recovery curve exhibits obvious nonlinear characteristics. Its advantage lies in its ability to exclude the regions with low change rates at both ends of the rising segment and extract the middle linear segment with the most active dynamic response, thereby improving the sensitivity of the slope feature value to changes in water content.

[0082] In one feasible implementation, in step 120 of the above embodiment, the slope feature value can also be calculated by least squares linear fitting slope calculation.

[0083] Specifically, all sampled data points within the slope calculation interval are used as the fitting dataset, with timestamp as the independent variable and capacitance as the dependent variable, to construct a univariate linear regression model; the slope parameter of the regression model is solved by the least squares method to minimize the sum of squared residuals of all sampled data points to the fitted line, and the solved slope parameter is output as the slope feature value.

[0084] This method utilizes information from all sampling points within the interval, effectively suppressing the impact of random measurement noise on the slope calculation results. It is suitable for scenarios with high sampling frequencies but where the signal exhibits some noise jitter, resulting in higher calculation accuracy.

[0085] In another feasible implementation, in step 120 of the above embodiment, the slope feature value can also be calculated by weighted linear fitting of the slope.

[0086] Specifically, sampling points located in the middle region of the slope calculation interval are assigned higher weights, while sampling points at both ends of the interval are assigned lower weights. The weighting coefficients are distributed according to a Gaussian or triangular function. The physical basis for this approach is that the middle region of the dynamic recovery curve has the best linearity and best reflects the main controlling process of adsorption kinetics, while the regions at both ends are more affected by the residual effect of heat exchange and the adsorption equilibrium approximation effect. Weighted processing can further improve the accuracy of the slope characteristic values ​​in representing the true adsorption rate.

[0087] This method is suitable for applications where the dynamic response characteristics of the sensor are sensitive to changes in temperature or concentration and require higher estimation accuracy.

[0088] In one feasible implementation, in step 120 of the above embodiments, the matching method for the slope feature value and the mapping relationship of micro-water content can also be linear interpolation mapping.

[0089] The correspondence table records discrete calibration data points. Each data point contains a slope feature value calibration value and its corresponding micro-water content calibration value, rather than a slope feature value range.

[0090] When the calculated slope feature value is not exactly equal to the slope feature value of a certain calibration data point, two calibration data points adjacent to the slope feature value are searched from the correspondence table. One slope feature value is slightly less than the calculated value, and the other is slightly greater than the calculated value. Using a linear interpolation formula, the micro-moisture content value corresponding to the current slope feature value is calculated based on the slope feature values ​​and micro-moisture content values ​​of the two calibration data points.

[0091] This method is applicable to application environments where the density of calibration data points in the correspondence table is limited, but a high resolution of the output water content value is required. By interpolation, continuous and smooth output can be achieved between adjacent calibration points, reducing discretization errors.

[0092] In another feasible implementation, in step 120 of the above embodiments, the matching method for the slope feature value and the mapping relationship of micro-water content can also be a function fitting mapping.

[0093] Specifically, during the factory calibration stage, the slope characteristic values ​​measured at different trace moisture contents are used as input data and the corresponding trace moisture content values ​​are used as output data. The least squares method is used to fit a continuous function, which can be a polynomial function, an exponential function, or a logarithmic function, preferably a second-order or third-order polynomial function.

[0094] In actual calculations, the slope characteristic value obtained from the calculation can be directly substituted into the fitting function to calculate the corresponding micro-moisture content value as the current true moisture content.

[0095] The advantages of this method are: once the fitting function is determined, no table lookup or interpolation operation is required in actual operation. The result can be obtained by simply substituting the function into the calculation, which is the fastest calculation speed. At the same time, the fitting function can utilize the information of the full set of calibrated data points, avoiding the step-like error caused by discrete table lookup.

[0096] like Figure 2 The diagram shown is a functional block diagram of a micro water sensor self-calibration system based on dynamic adsorption-desorption curves provided in an embodiment of the present invention.

[0097] The micro-water sensor self-calibration system 100 based on dynamic adsorption-desorption curves described in this invention can be installed in an electronic device. Depending on the functions implemented, the micro-water sensor self-calibration system 100 may include a heating module 101, a water content calculation module 102, a measurement reference correction module 103, and a re-monitoring module 104. The module described in this invention can also be referred to as a unit, which refers to a series of computer program segments that can be executed by the processor of an electronic device and can perform a fixed function, and are stored in the memory of the electronic device.

[0098] In this embodiment, the functions of each module / unit are as follows: The heating module 101 is used to apply a controllable heating pulse to the micro water sensor during normal equipment operation intervals to drive the desorption of adsorbed water molecules. The water content estimation module 102 is used to continuously record the dynamic recovery curve of the capacitance value of the micro water sensor changing over time after the heating pulse ends, and to reverse-calculate the current true water content based on the mapping relationship between the slope characteristic value of the dynamic recovery curve and the current true water content. The measurement reference correction module 103 is used to correct the measurement reference of the micro-moisture sensor according to the actual micro-moisture content value; The re-monitoring module 104 is used for the micro-water sensor to continue performing online monitoring tasks according to the calibrated measurement benchmark.

[0099] Reference Figure 3 The figure shown is a determination diagram of the normal operation gap and idle confirmation time of a micro water sensor self-calibration system based on dynamic adsorption-desorption curves provided in an embodiment of the present invention.

[0100] In the engineering application of micro-moisture sensor online monitoring systems, avoiding thermal interference and data contamination caused by calibration pulses to routine monitoring tasks is the primary technical bottleneck to be addressed. This figure details the timing evolution of the idle state determination logic in this system. During the data acquisition cycle, the online monitoring main program sets the working status flag to a high level (1: acquisition state). At this time, the sensitive capacitance value exhibits transient changes (approximately...) along with fluctuations in external gas moisture content. Once a single data acquisition cycle ends, the flag is reset to low (0: idle state) by the main program, and the capacitance value enters a relatively stable baseline fluctuation phase. The adaptive controller begins to poll the flag's state in real time. When it detects that the flag has remained low for a preset idle acknowledgment time threshold (set to...), the controller will initiate a polling process. When the system determines that the device is currently in an absolute operational gap, it can execute the calibration command. The introduction of the idle confirmation interval serves as a thermal barrier buffer and state lock, ensuring that subsequent high-energy heating pulses will not introduce additional thermal noise and signal distortion during online monitoring in the main program. The setting of the judgment threshold comprehensively balances the system polling overhead and the thermodynamic equilibrium recovery time of the humidity measuring film. This logic path provides a safe hardware and software environment guarantee for the calibration process.

[0101] Reference Figure 4 The diagram shown is a schematic of a dual triggering mechanism based on drift exceeding limits and cumulative operating cycle of a micro water sensor self-calibration system based on dynamic adsorption-desorption curves provided in an embodiment of the present invention.

[0102] This figure illustrates the "event-time" dual-channel closed-loop triggering logic employed by the self-calibration system at the front-end control level. To avoid excessive thermal fatigue of sensitive materials caused by invalid calibration, the system abandons the single timed calibration mode and introduces a dual intersection judgment space consisting of cumulative drift detection (event channel) and maximum running time (time channel). In the figure, the blue curve depicts the trend of sensor cumulative zero-point drift after temperature compensation. As the sensitive porous film ages or contaminants adhere, its capacitance baseline in the moisture-free state slowly and monotonically increases. When the drift reaches the drift trigger threshold ( (approximately at the time shown in the figure) (Approximately one day), the event channel immediately alarms and generates a "drift over-limit trigger" signal. Simultaneously, the system's internal timer continuously records the cumulative working time since the last calibration. When the time reaches the set threshold... When the cycle threshold is reached, the time channel triggers and sends a "cycle timeout trigger" signal. These two signals are output as an "OR" logic in the trigger decision unit. If either signal exceeds the limit, the system, while ensuring safe operation, generates a trigger signal that allows the execution of a heating pulse during the adjacent equipment operation interval. This dual triggering criterion effectively balances the timeliness of calibration with the control of material aging, preventing data distortion caused by long-term online drift.

[0103] Reference Figure 5 The figure shown is a mapping relationship between the slope characteristic value of the recovery curve and the actual micro-water content of the micro-water sensor self-calibration system based on dynamic adsorption-desorption curve provided in an embodiment of the present invention.

[0104] The dynamic resorption rate of water molecules essentially depends on the water vapor partial pressure difference (adsorption driving force) at the sensitive membrane interface. This figure illustrates how the microscopic adsorption rate slope characteristic value can be determined through a calibration mechanism. Mapped to the macroscopic true micro water content The dark green continuous curve in the figure shows the result obtained after multi-point calibration and fitting at the factory. The function conforms to the nonlinear exponential growth law of typical kinetic models: under high micro-water content, the efficiency of active sites in capturing water molecules is greatly improved, and the corresponding upward slope becomes steeper and faster. The red circular standard points represent a stepped interval correspondence table discretely stored in read-only memory (ROM). The slope characteristic value calculated at the current measurement interval... When inputting into the mapping module, the system uses the orange test mapping line to perform multi-point weighted linear interpolation on the continuous curve, and then reverse-engineers the true micro-moisture content of the current environment with high precision. This "concentration by rate" method does not rely on the equilibrium absolute capacitance reading, which is prone to long-term drift after temperature drop. Therefore, it can still maintain ultra-high consistency of the mapping output when faced with membrane material aging and moderate surface contamination.

[0105] Reference Figure 6 The figure shown is a comparison of the measurement accuracy and linearity of a micro water sensor before and after self-calibration in a micro water sensor self-calibration system based on dynamic adsorption-desorption curves provided in an embodiment of the present invention.

[0106] This figure objectively evaluates the practical engineering correction performance of the self-calibration algorithm based on dynamic adsorption-desorption curves by assessing the accuracy reconstruction results across the entire measurement range. Within a large-span measurement space, due to sensitivity decay caused by dielectric aging and zero-point drift of background capacitance, the output value of the uncalibrated sensor (red dashed line) deviates significantly from the 1:1 absolute standard theoretical line, and... The checkpoints accumulated to a high level The positive nonlinear deviation. This self-calibration system utilizes a multidimensional compensation model to correct the compensation parameters. The sensor's measurement benchmark is corrected online. The corrected sensor output (solid blue line) closely matches the theoretical standard line, exhibiting extremely high consistency at various random test points (blue circles). This multi-dimensional fusion algorithm not only corrects the zero-point static drift through translation but also addresses the dynamic rate deviation. The curvature (linearity) of the calibration curve was corrected, reducing the system measurement deviation across the entire range from the original... Rapidly converged to Within this scope, the technical challenge of long-term measurement reliability degradation has been fundamentally overcome.

[0107] In the several embodiments provided by this invention, it should be understood that the disclosed methods and systems can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and other division methods may be used in actual implementation.

[0108] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0109] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in the form of hardware plus software functional modules.

[0110] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0111] This application embodiment can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence is the theory, method, technology, and application system that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0112] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A self-calibration method for micro-water sensors based on dynamic adsorption-desorption curves, characterized in that, The method includes: Step 110: During normal equipment operation intervals, apply a controllable heating pulse to the micro water sensor to drive the desorption of adsorbed water molecules; Step 120: After the heating pulse ends, continuously record the dynamic recovery curve of the capacitance value of the micro-water sensor changing over time, and based on the mapping relationship between the slope characteristic value of the dynamic recovery curve and the current true water content, reversely calculate the current true water content. Step 130: Calibrate the measurement benchmark of the micro-moisture sensor based on the actual micro-moisture content value; Step 140: The micro-water sensor continues to perform online monitoring tasks according to the calibrated measurement benchmark.

2. The self-calibration method for a micro-water sensor based on dynamic adsorption-desorption curves as described in claim 1, characterized in that, Before applying a controllable heating pulse to the micro-water sensor, the following steps are included: Obtain the current ambient temperature value of the environment where the micro-water sensor is located; The current capacitance value of the micro-water sensor is continuously collected in a periodic cyclic measurement mode. The current capacitance value is compared with the standard capacitance value, wherein the standard capacitance value is the reference capacitance value of the micro-water sensor in the factory calibration state, corresponding to the current ambient temperature value; A calibration trigger signal is generated when the difference between the current capacitance value and the standard capacitance value exceeds the drift threshold.

3. The self-calibration method for a micro-water sensor based on dynamic adsorption-desorption curves as described in claim 2, characterized in that, Before applying a controllable heating pulse to the micro-water sensor, the following is also included: Record the cumulative operating time of the micro-water sensor since the last calibration. When the cumulative running time reaches the preset time period, a calibration trigger signal is generated.

4. The self-calibration method for a micro-water sensor based on dynamic adsorption-desorption curves as described in claim 1, characterized in that, The normal operating intervals of the equipment include: The current working status flag of the micro-water sensor is read in real time. The working status flag is set by the online monitoring main program when data acquisition is performed and reset after data acquisition is completed. When the working status flag is detected to be in the reset state, it is determined that the micro water sensor is not currently performing online monitoring data acquisition operations; After the working status flag is continuously reset for an idle confirmation time, it is confirmed that the current moment is within the normal operation interval of the device, and a trigger command to allow the execution of the heating pulse is generated.

5. The self-calibration method for a micro-water sensor based on dynamic adsorption-desorption curves as described in claim 4, characterized in that, The process of applying a controllable heating pulse to the micro-water sensor to drive the desorption of adsorbed water molecules includes: Heating is initiated when the trigger command is received; Start the timer at the same time as starting heating to begin recording the heating duration; When the timer reaches the heating duration threshold, heating is terminated to complete the application of a single heating pulse.

6. The self-calibration method for a micro-water sensor based on dynamic adsorption-desorption curves as described in claim 1, characterized in that, The step of continuously recording the dynamic recovery curve of the capacitance value of the micro-water sensor over time after the heating pulse ends includes: While terminating heating, the capacitance value of the micro-water sensor is continuously acquired at a high sampling frequency; At each capacitance value acquisition moment, the temperature value of the environment where the micro water sensor is located is read synchronously, and the temperature value, the currently acquired capacitance value, and the corresponding timestamp are combined into a sampling data point; All continuously generated sampling data points are stored sequentially as capacitance-temperature-time series data in chronological order. The acquisition, reading, and storage operations are continuously performed until the number of sampling data points contained in the capacitance-temperature-time series data reaches the sampling point threshold, at which point the acquisition stops, and the stored complete capacitance-temperature-time series data is used as a dynamic recovery curve.

7. The self-calibration method for a micro-water sensor based on dynamic adsorption-desorption curves as described in claim 1, characterized in that, The method of reversely calculating the current true water content based on the mapping relationship between the slope characteristic value of the dynamic recovery curve and the current true water content includes: Based on the capacitance value and timestamp of each sampling data point in the dynamic recovery curve, the rising phase of the capacitance value over time in the dynamic recovery curve is determined, and continuous sampling data points are selected as the slope calculation interval within the rising phase. The ratio of the difference between the initial capacitance value and the final capacitance value to the difference between the initial time and the final time within the slope calculation interval is taken as the capacitance change rate, and the capacitance change rate is output as the slope characteristic value. Call the pre-stored correspondence table between slope feature values ​​and trace moisture content, wherein the correspondence table records the slope feature value intervals and the trace moisture content values ​​corresponding to each interval; The slope feature value is matched with each slope feature value interval in the corresponding relationship table to determine the target slope feature value interval to which the slope feature value belongs; Based on the target slope feature value range, the corresponding micro-moisture content value is read from the corresponding relationship table, and the micro-moisture content value is used as the current true moisture content.

8. The self-calibration method for a micro-water sensor based on dynamic adsorption-desorption curves as described in claim 1, characterized in that, The calibration of the measurement benchmark of the micro-moisture sensor based on the actual micro-moisture content value includes: Based on the measurement reference of the micro-water sensor, the original measured water content at the current moment is calculated; The difference between the original measured moisture content and the current actual moisture content is used to characterize the degree of drift of the measurement benchmark; Based on the degree of drift, correction and compensation parameters are generated, and the measurement reference is corrected according to the correction and compensation parameters; During the next online monitoring, the original measured moisture content will be calculated using the revised measurement baseline.

9. The self-calibration method for a micro-water sensor based on dynamic adsorption-desorption curves as described in claim 8, characterized in that, The generation of correction compensation parameters based on the drift degree includes: After the heating pulse ends, the capacitance reading at which the capacitance value in the dynamic recovery curve drops to the minimum value is taken as the lowest capacitance value after desorption. The stable capacitance reading reached after the capacitance value in the dynamic recovery curve increases over time and tends to stabilize is taken as the recovery balance capacitance value. The adsorption recovery amplitude of the micro-water sensor under the current environment is obtained by subtracting the minimum capacitance value after desorption from the recovery equilibrium capacitance value. The adsorption recovery deviation between the adsorption recovery amplitude and the standard adsorption recovery amplitude is obtained, wherein the standard adsorption recovery amplitude is the adsorption recovery amplitude of the micro water sensor under factory calibration conditions; Obtain the slope deviation between the slope feature value and the standard slope feature value, wherein the standard slope feature value is the slope feature value corresponding to the micro water sensor in the factory calibration state; The product of the adsorption recovery deviation and the offset correction coefficient is added to the product of the slope deviation and the time constant to obtain the correction compensation parameter, wherein the offset correction coefficient and the time constant are determined by the factory calibration experiment of the micro water sensor.

10. A self-calibration system for a micro-water sensor based on dynamic adsorption-desorption curves, used to implement the self-calibration method for a micro-water sensor based on dynamic adsorption-desorption curves as described in claim 1, the system comprising: The heating module is used to apply controllable heating pulses to the micro water sensor during normal equipment operation intervals to drive the desorption of adsorbed water molecules; The water content estimation module is used to continuously record the dynamic recovery curve of the capacitance value of the micro water sensor changing over time after the heating pulse ends, and to reverse estimate the current true water content based on the mapping relationship between the slope characteristic value of the dynamic recovery curve and the current true water content. The measurement reference calibration module is used to calibrate the measurement reference of the micro-moisture sensor according to the actual micro-moisture content value; The re-monitoring module is used by the micro-water sensor to continue performing online monitoring tasks based on the calibrated measurement benchmark.

Images