Method for determining the measurement period of a palladium alloy hydrogen sensor in a closed environment

Abstract

The application discloses a kind of closed environment in palladium alloy hydrogen sensor measurement period determination method, comprising the following steps: 1) determine the basic parameters of hydrogen sensor and establish accuracy index requirement;2) build hydrogen sensor accelerated thermal aging test platform;3) test hydrogen sensor resistance value response curve and recovery curve;4) calculate the characteristic index reflecting the performance of hydrogen sensor;5) identify hydrogen sensor thermal aging acceleration factor;6) sensor degradation trajectory analysis and reliability model construction;7) according to the preset reliability threshold, the sensor failure time is calculated and the measurement period is determined.The measurement period determination method of palladium alloy hydrogen sensor in closed environment provided by the application can obtain the expected service life of hydrogen sensor in working environment, to optimize the measurement period and mode of hydrogen sensor.

Description

Technical Field

[0001] This invention relates to the field of metrology, and more specifically to a method for determining the metrology cycle of a palladium alloy hydrogen sensor in a closed environment. Background Technology

[0002] Because hydrogen exhibits hydrogenation corrosion and hydrogen embrittlement, and oxygen possesses strong oxidizing properties, it can significantly accelerate equipment damage and performance degradation by directly participating in or inducing physicochemical reactions. Furthermore, from a safety perspective, hydrogen has a very low ignition energy: 0.019 mJ in air and 0.007 mJ in oxygen. Ordinary impacts, friction, electrical discharges, and hot gas flows can ignite hydrogen-oxygen mixtures, potentially leading to explosions and causing significant property damage and personal injury. Therefore, using hydrogen sensors to monitor hydrogen levels is beneficial for reducing the probability of accidents and ensuring production safety.

[0003] During long-term operation under service conditions, the performance of components in hydrogen sensors gradually deteriorates, leading to "drift" in hydrogen concentration detection data. Metrology is a crucial technical means to maintain and guarantee the quality and performance of hydrogen sensors throughout their service life. To ensure the accuracy of hydrogen sensor readings, the detection system must be periodically traced. In particular, hydrogen sensors installed in confined spaces are often required to operate continuously for many years due to the inconvenience of removal for calibration. Therefore, the metrological cycle of hydrogen sensors is an extremely important metrological parameter. The measurement performance of hydrogen sensors gradually decreases during operation in the working environment; stability within the calibration-free period is a prerequisite for the long-term normal operation of hydrogen sensors.

[0004] Since the working conditions of hydrogen sensors are relatively stable under closed conditions, such as constant air pressure, humidity, and vibration, the main factor affecting the stability of hydrogen sensors is temperature. Therefore, a measurement cycle calculation model for hydrogen sensors can be established based on thermal aging test and analysis to determine whether the measurement uncertainty of hydrogen sensors under long-term operation can meet the actual requirements. Summary of the Invention

[0005] The purpose of this invention is to provide a method for determining the metrology cycle of a hydrogen sensor to assess whether the measurement uncertainty of the hydrogen sensor meets practical requirements under long-term operation. First, the basic parameters of the hydrogen sensor and accuracy requirements are determined. Then, a thermal aging test is conducted on this type of hydrogen sensor, and the resistance response curve and recovery curve of this type of hydrogen sensor are tested. Next, characteristic indicators reflecting the performance of the hydrogen sensor are calculated, the thermal aging acceleration factor of the hydrogen sensor is identified, and the degradation trajectory analysis and reliability model construction of the hydrogen sensor are performed. Based on a preset reliability threshold, the sensor failure time is calculated, and the metrology cycle is determined. This invention can obtain the expected lifespan of the hydrogen sensor under operating conditions, thereby optimizing the metrology cycle and method of the hydrogen sensor.

[0006] The technical solution for achieving the objective of this invention is as follows:

[0007] A method for determining the metering cycle of a palladium alloy hydrogen sensor in a confined environment, comprising the following steps:

[0008] S1. The operating limit temperature of the hydrogen sensor under test is found by using a high-temperature step stress test.

[0009] S2. Using the hydrogen sensor under test as the test sample, multiple test samples are subjected to accelerated thermal aging tests at different aging temperatures and for different aging times; the maximum value of the aging temperature is taken as the nearest integer value that is lower than the working limit temperature of the hydrogen sensor.

[0010] S3. During the test in step S2, test the resistance response curve and resistance recovery curve of the hydrogen sensor under test at different aging temperatures and different aging times; the resistance response curve refers to the resistance curve of the hydrogen sensor from the moment the standard hydrogen gas is introduced into the test tank where the hydrogen sensor is located to the moment the data stabilizes during the accelerated thermal aging test.

[0011] The resistance value recovery curve refers to the resistance value curve of the hydrogen sensor from the moment pure nitrogen gas is introduced into the test tank where the hydrogen sensor is located to the moment when the data stabilizes after the resistance value response curve measurement is completed in the accelerated thermal aging test.

[0012] S4. Based on the resistance response curves and resistance recovery curves obtained in step S3 at different aging temperatures and aging times, calculate the response value and recovery value of the hydrogen sensor at different aging temperatures and aging times; where the response value is the difference between the stable value and the initial value of the response curve, and the recovery value is the difference between the initial value and the stable value of the recovery curve.

[0013] S5. Using the lowest aging temperature in the accelerated thermal aging test in step S2 as the reference temperature, calculate the acceleration factors of the response and recovery values ​​at other aging temperatures in the accelerated aging test relative to the reference temperature; after obtaining the acceleration factors, perform regression calculations using the following formula to obtain the activation energy of the reaction. The value,

[0014] ;

[0015] In the formula, Indicates the aging temperature in the accelerated aging test. Relative to reference temperature acceleration factor, R is the activation energy of the reaction; R is the ideal gas constant.

[0016] The activation energy of the reaction can be obtained from the above equation. Fitted values Then, from the reference temperature Extrapolation to obtain room temperature Acceleration factor ,

[0017] ;

[0018] S6. Sensor degradation trajectory analysis and reliability model construction; Indicates the mean parameter, The shape parameter represents the performance degradation threshold of the hydrogen sensor failure, and D is the performance degradation threshold of the hydrogen sensor failure. The mean parameter and shape parameter are the model parameter vectors to be estimated. It is a non-linear power exponent;

[0019] The relationship between the reliability of the hydrogen sensor's lifespan and time is as follows:

[0020] ;

[0021] In the formula, t represents the aging time;

[0022] The probability density distribution function of the hydrogen sensor lifetime prediction is:

[0023] ;

[0024] In the formula, Represents a nonlinear power exponent The derivative function, and These represent the probability density function and distribution function of the standard normal distribution, respectively.

[0025] Using the maximization of the cumulative probability density distribution function of all sample points as the objective function L, the mean and shape parameters of the model are obtained using the Bayesian method, thereby yielding the probability density distribution function of the predicted lifetime value, and then the baseline temperature is obtained. The reliability function R(t) of the hydrogen sensor;

[0026] S7. Calculate the sensor failure time based on the preset reliability threshold and determine the metering cycle; determine the probability R of the overall performance of the hydrogen sensor remaining within the expected acceptable range during reconfirmation or subsequent testing. default When R(t0) = R default When the response termination time t0 is obtained, the metering cycle of the hydrogen sensor is: .

[0027] Furthermore, in step S2, the maximum value of the aging temperature is T. n Set temperature interval Set n aging temperatures, with the i-th aging temperature being... Set the aging time interval. If the total aging time is m days, then the aging time at the i-th aging temperature is expressed as: .

[0028] Furthermore, in step S2, at least three test samples are selected for each aging temperature. Before the accelerated thermal aging test begins, the test container used to store the test samples is placed in a 50℃ / 50Pa vacuum drying oven for 2 hours. The test samples are connected to the aviation plug, and then the test container is sealed. The response curves and recovery curves of the hydrogen sensor resistance values ​​at different aging temperatures and different aging times are tested.

[0029] Furthermore, in step S3, the specific method for testing the resistance response curves of the hydrogen sensor at different aging temperatures and times is as follows:

[0030] Place the test container containing the test sample in a constant temperature chamber and adjust the temperature inside the chamber to set the aging temperature. Once the temperature in the test container containing the test sample stabilizes at the set aging temperature, collect and save the data.

[0031] Air is extracted from the test chamber. Once the vacuum level inside the test chamber reaches the preset vacuum level, extraction is stopped.

[0032] Hydrogen standard gas is introduced into the test tank, and the flow rate of the hydrogen standard gas is controlled to be no less than the set flow rate.

[0033] Once the gas pressure inside the test tank reaches standard atmospheric pressure, open the exhaust port of the test tank and continue to introduce standard hydrogen gas for a set time. Then close the inlet and outlet ports of the test tank. Observe the collected data until the data remains stable.

[0034] The resistance curve from the moment the standard hydrogen gas is input to the moment the data stabilizes is the response curve.

[0035] Furthermore, the specific test method for the resistance recovery curve of the hydrogen sensor under different aging temperatures and aging times is as follows:

[0036] After ensuring that the temperature in the test tank is stable at the set aging temperature, data is collected and saved.

[0037] Introduce pure nitrogen into the air inlet of the test tank and open the exhaust port. Control the flow rate of pure nitrogen to be no less than the set flow rate. Observe the collected data until the data stabilizes. The resistance curve from the moment of pure nitrogen input to the moment of data stabilization is the resistance recovery curve.

[0038] Furthermore, in step S5, the process of solving for the acceleration factor is as follows:

[0039] The regression equation for the degradation law of the hydrogen sensor response value at the reference temperature is as follows: ;

[0040] The regression equation for the degradation law of the recovery value of the hydrogen sensor at the reference temperature is as follows: ;

[0041] In the formula, t represents the aging time, a continuous value starting from 0; , It is a regression function;

[0042] At other aging temperatures The aging time below The following will be accelerated by the factor Converted to reference temperature based on the time-temperature equivalence principle The aging time is as follows X represents the number of response values ​​and the number of recovery values ​​corresponding to different aging times at the i-th aging temperature. ;

[0043] Accelerator The value of is determined by the objective function. Find the best option. In the formula, Aging temperature T i The aging time is converted to the aging time at the reference temperature. Substitute into and In the process, the aging temperature is obtained. Equivalent response value at reference temperature and recovery value equivalent value Using the average relative dispersion coefficient as the objective function, for each aging temperature T i The objective function is as follows In the form of

[0044] -1;

[0045] A parameter optimization algorithm is used to iteratively optimize each acceleration factor to achieve the desired result at each aging temperature T. i Below Approaching 0, we get The optimal solution is to equivalently shift the accelerated performance degradation data points at n-1 different temperatures to the distribution characteristics at the baseline temperature, perform regression calculations on the equivalent data sample points, and calculate the absolute dispersion coefficient of the regression prediction value of each sample point relative to the original value. When the predicted value of the regression model is within 1.5 times the dispersion line, the acceleration factor has high accuracy; otherwise, the acceleration factor needs to be calculated again.

[0046] Furthermore, in step S6, the prior distribution in the Bayesian method is selected using the Markov chain Monte Carlo method.

[0047] In the formula, y represents the number of sample points.

[0048] This invention obtains the metering cycle of the hydrogen sensor through experimental testing, which can be used to optimize the metering scheme of the hydrogen sensor. Attached Figure Description

[0049] Figure 1 This is a flowchart of the present invention.

[0050] Figure 2 This is a schematic diagram of the structure of the hydrogen sensor accelerated thermal aging test platform of the present invention. Detailed Implementation

[0051] The technical solution of the present invention will be clearly and completely described below with reference to specific embodiments.

[0052] like Figure 1 As shown, a method for determining the metering cycle of a palladium alloy hydrogen sensor in a closed environment includes the following steps:

[0053] S1: Determine the basic parameters and accuracy requirements of the hydrogen sensor.

[0054] Determine the basic parameters and accuracy requirements of the hydrogen sensor. Based on the sensor's design parameters, identify the voltage, temperature, humidity, and pressure parameters that allow the sensor to operate normally.

[0055] S2: Constructing a hydrogen sensor accelerated thermal aging test platform

[0056] Build a sensor accelerated thermal aging test platform, such as Figure 2 As shown, the system includes a temperature-adjustable constant temperature chamber 14, a shelf 13, a test container 12, a test sample 11, an aviation connector 10, a data acquisition system 9, a vacuum pump 8, a two-way valve 7, a temperature and pressure sensor 6, a gas dryer 5, a flow meter 4, a three-way valve 3, 2 standard hydrogen gases, and 1 pure nitrogen gas. The constant temperature chamber has an adjustable temperature, a circular opening at the top, and a metal shelf at the bottom. The chamber is filled with silicone oil, which completely submerges the test container. The test container, a cylindrical structure, is placed on the shelf. An aviation connector is located at the center of the top. The gas pipe on the left side is flush with the length of the top cover, and the gas pipe on the right side... The length is close to the bottom of the test tank, and the internal size of the test tank is no more than twice the volume of the test sample. Hydrogen standard gas and pure nitrogen are connected to the three-way valve, flow meter, gas dryer and test tank in sequence through gas pipelines. The gas pipeline on the left side has a preheating buffer bend in silicone oil before connecting to the test tank. The gas pipeline on the right side is connected to the two-way valve and vacuum pump in sequence. The data acquisition system is connected to the test sample through an aviation connector, and the temperature and pressure sensors monitor the temperature changes inside the test tank through the aviation connector. The concentration of hydrogen standard gas should be as high as possible higher than the hydrogen concentration in the working atmosphere, but not exceeding the maximum working concentration of the hydrogen sensor itself.

[0057] S3: Test the resistance response curve and recovery curve of the hydrogen sensor.

[0058] The working limit temperature of the test sample was determined by high-temperature step stress testing. The nearest integer temperature below the working limit temperature was taken as the highest accelerated aging temperature T5. Five aging temperatures were set at 10°C intervals, with the i-th aging temperature being T5. i Let i = 1, 2, 3, 4, 5. The aging interval is 10 days, and the total aging time is 180 days. The aging time t at the i-th aging temperature... ij That is, j=0, 10, 20, 30, ..., 180. Three qualified test samples were selected for each aging temperature. Before gas loading, the test tank was placed in a 50℃ / 50Pa vacuum drying oven for 2 hours. The test samples were connected to the aviation plug, and then the test tank was sealed. The recovery curve and response curve of the sensor resistance value at different aging times under five aging temperatures were tested.

[0059] The test method for the response curve of the sensor resistance value is as follows: Adjust the temperature of the constant temperature chamber to the aging temperature. After the temperature in the test tank stabilizes at the aging temperature using a temperature and pressure sensor, turn on the data acquisition system to test and save the data. Close the three-way valve and open the two-way valve. Turn on the vacuum pump to extract air from the test tank. When the vacuum degree reaches -0.9, close the vacuum pump and the two-way valve. Then, turn the three-way valve to the direction of hydrogen standard gas. Adjust the hydrogen standard gas pressure reducing valve so that the carrier gas flow rate in the gas pipeline is not less than 200 sccm. When the gas pressure in the test tank is the standard atmospheric pressure, open the two-way valve and continue to introduce hydrogen standard gas for 20 minutes. Then close the three-way valve and the two-way valve. Observe the data collected by the data acquisition system until the data remains stable. The resistance value curve from the moment of hydrogen input to the moment of data stabilization is the response curve.

[0060] The test method for the recovery curve of the sensor resistance value is as follows: keep the temperature of the constant temperature chamber at the aging temperature, turn on the data acquisition system to test and save the data, turn the three-way valve to the pure nitrogen direction and open the two-way valve, adjust the pure nitrogen pressure reducing valve so that the flow rate of the carrier gas in the gas pipeline is not less than 200 sccm, observe the data acquisition system to collect data until the data remains stable, and the resistance value curve from the time of pure nitrogen input to the time of data stabilization is the recovery curve.

[0061] S4: Calculate the characteristic indicators reflecting the performance of the hydrogen sensor

[0062] Response and recovery values ​​are calculated based on the response and recovery curves. The response value is the difference between the stable value and the initial value of the response curve, and the recovery value is the difference between the initial value and the stable value of the recovery curve. Different aging times (t) at different aging temperatures are obtained. ij The response value Res Ti, j and recovery value Rec Ti, j i = 1, 2, 3, 4, 5, j = 0, 10, 20, 30, ..., 180, where t represents different aging times at the i-th aging temperature. ij There are 19 response values ​​and 19 recovery values.

[0063] S5: Identify hydrogen sensor thermal aging acceleration factors

[0064] Using aging temperature T1 as the reference temperature, calculate the acceleration factor α of the response and recovery values ​​of T2, ..., T5 relative to the reference temperature. 21 a 31 a 41 a 51 The variable values ​​for each acceleration factor are in the range of [1, 100], and the variable dimension for the parameter identification problem is 4.

[0065] The regression equation for the degradation of the hydrogen sensor's response value at the reference temperature is as follows:

[0066] (1)

[0067] The regression equation for the degradation law of the sensor's recovery value at the reference temperature is as follows:

[0068] (2)

[0069] In the formula, t represents the aging time, a continuous value starting from 0; f s (t), f c (t) is the regression function, which can be in the form of an exponential function, a power function, etc., or a combination of multiple types of functions. The optimal form should be chosen based on the actual regression effect. For example, , .

[0070] The aging time t at the remaining aging temperatures T2, ..., T5 ij The following will be accelerated by factor a i1 The aging time converted to the reference aging temperature T1 based on the time-temperature equivalence principle is t. i1_k i=2, 3, 4, 5, k=1, 2, 3, ..., 19, acceleration factor a i1 The value of is obtained by optimizing the objective function of equation (4).

[0071] (3)

[0072] In the formula, t i1_k Aging temperature T i The aging time is converted to the aging time at the reference temperature, i=2, 3, 4, 5, k=1, 2, 3, ..., 19; t i1_k Substituting these values ​​into formulas (1) and (2) respectively, we obtain the equivalent values ​​of the response and recovery at the reference temperature under aging temperatures T2, ..., T5, denoted as f. s (t i1_k ) and f c (t i1_k Using the average relative dispersion coefficient as the objective function, T is calculated at each aging temperature. i objective function The format is:

[0073] (4)

[0074] A parameter optimization algorithm is used to iteratively optimize each acceleration factor to ensure that T is optimal at each aging temperature. i of To obtain the optimal solution a, we need to find a value as close to 0 as possible.21 a 31 a 41 a 51 The accelerated performance degradation data points at four different temperatures are equivalently shifted to the distribution characteristics at the baseline temperature. Regression calculations are performed on the equivalent data sample points, and the absolute dispersion coefficient of the regression prediction value relative to the original value is calculated. When the predicted value of the regression model is within 1.5 times the dispersion line, the acceleration factor has high accuracy; otherwise, the acceleration factor needs to be recalculated.

[0075] The acceleration factor a is obtained 21 a 31 a 41 a 51 Then, the value of the activation energy Ea can be obtained by regression calculation using equation (5):

[0076] (5)

[0077] In the formula, T i The aging temperature is i = 2, 3, 4, 5, in K; E a The value is the activation energy of the reaction, expressed in kJ / mol; R is the ideal gas constant, with a value of 8.314 J / (mol•K).

[0078] The activation energy E of the reaction is obtained. a The fitted value E a_fitted Then, the acceleration factor a at room temperature T0 was obtained by extrapolation from the aging temperature T1. 01 At room temperature, T0 is generally taken as 20℃;

[0079] (6)

[0080] S6: Sensor Degradation Trajectory Analysis and Reliability Model Construction

[0081] Let μ represent the mean parameter, λ represent the shape parameter, and D be the performance degradation threshold for hydrogen sensor failure (the threshold can be set according to actual needs, generally 90%. In JJF 1139-2005 "Principles and Methods for Determining the Verification Cycle," principle two states that when determining the verification cycle of a measuring instrument, the measurement reliability target R of the applicable measuring instrument should first be clarified; generally, the measurement reliability target R of a measuring instrument is ≥90%. Measurement reliability R refers to the probability that the overall performance of a measuring instrument remains within the expected acceptable range during reconfirmation or subsequent verification). The mean parameter and shape parameter are the model parameter vector to be estimated. Then, the hydrogen sensor lifespan L is L={t|Res}. Ti,t ≥D}; It is a non-linear power exponent.

[0082] The relationship between the reliability of the hydrogen sensor's lifespan and time is as follows:

[0083] (7)

[0084] The probability density distribution function of the hydrogen sensor lifetime is:

[0085] (8)

[0086] The objective function L is to maximize the cumulative probability density distribution function of all sample points. The sample points can be selected from the response values ​​of 19 different aging times at the reference temperature, or more response values ​​at different aging times can be selected by interpolation according to formulas (1) and (2). The mean parameter and shape parameter of the model are obtained by Bayesian method, and then the probability density distribution function of the lifetime prediction value is obtained. Then the reliability function R(t) of the hydrogen sensor at aging temperature T1 can be obtained. The prior distribution in the Bayesian method is selected by Markov chain Monte Carlo method:

[0087] (9)

[0088] In the formula, n is the number of sample points; The vector of model parameters to be estimated, i.e., the parameters and parameters ; Let represent the probability density distribution function, that is, the distribution function referred to in formula (8).

[0089] S7: Calculate the sensor failure time based on the preset reliability threshold and determine the metering cycle.

[0090] The probability R of the overall performance of the hydrogen sensor remaining within the expected acceptable range during reconfirmation or subsequent testing. default (It can be set to 90%, or it can be set according to actual needs. For higher requirements, it can be set to 95% or 98%). When R(t0) = R default When the corresponding termination time t0 is obtained, the metering period T of the hydrogen sensor is determined. cal for:

[0091] (10).

Claims

1. A method for determining the measurement period of a palladium alloy hydrogen sensor in a closed environment, characterized in that: Including steps, S1. The operating limit temperature of the hydrogen sensor under test is found by using a high-temperature step stress test. S2. Using the hydrogen sensor under test as the test sample, multiple test samples are subjected to accelerated thermal aging tests at different aging temperatures and for different aging times; the maximum value of the aging temperature is taken as the nearest integer value that is lower than the working limit temperature of the hydrogen sensor. S3. During the test in step S2, test the resistance response curve and resistance recovery curve of the hydrogen sensor under test at different aging temperatures and different aging times; the resistance response curve refers to the resistance curve of the hydrogen sensor from the moment the standard hydrogen gas is introduced into the test tank where the hydrogen sensor is located to the moment the data stabilizes during the accelerated thermal aging test. The resistance value recovery curve refers to the resistance value curve of the hydrogen sensor from the moment pure nitrogen gas is introduced into the test tank where the hydrogen sensor is located to the moment when the data stabilizes after the resistance value response curve measurement is completed in the accelerated thermal aging test. S4. Based on the resistance response curves and resistance recovery curves obtained in step S3 at different aging temperatures and aging times, calculate the response value and recovery value of the hydrogen sensor at different aging temperatures and aging times; where the response value is the difference between the stable value and the initial value of the response curve, and the recovery value is the difference between the initial value and the stable value of the recovery curve. S5. Using the lowest aging temperature in the accelerated thermal aging test in step S2 as the reference temperature, calculate the acceleration factors of the response and recovery values ​​at other aging temperatures in the accelerated aging test relative to the reference temperature; after obtaining the acceleration factors, perform regression calculations using the following formula to obtain the activation energy of the reaction. The value, ； In the formula, Indicates the aging temperature in the accelerated aging test. Relative to reference temperature acceleration factor, R is the activation energy of the reaction; R is the ideal gas constant. The activation energy of the reaction can be obtained from the above equation. Fitted values Then, from the reference temperature Extrapolation to obtain room temperature Acceleration factor , ； S6. Sensor degradation trajectory analysis and reliability model construction; Indicates the mean parameter, The shape parameter represents the performance degradation threshold of the hydrogen sensor failure, and D is the performance degradation threshold of the hydrogen sensor failure. The mean parameter and shape parameter are the model parameter vectors to be estimated. It is a non-linear power exponent; The relationship between the reliability of the hydrogen sensor's lifespan and time is as follows: ； In the formula, t represents the aging time; The probability density distribution function of the hydrogen sensor lifetime prediction is: ； In the formula, Represents a nonlinear power exponent The derivative function, and These represent the probability density function and distribution function of the standard normal distribution, respectively. Using the maximization of the cumulative probability density distribution function of all sample points as the objective function L, the mean and shape parameters of the model are obtained using the Bayesian method, thereby yielding the probability density distribution function of the predicted lifetime value, and then the baseline temperature is obtained. The reliability function R(t) of the hydrogen sensor; S7. Calculate the sensor failure time based on the preset reliability threshold and determine the metering cycle; determine the probability R of the overall performance of the hydrogen sensor remaining within the expected acceptable range during reconfirmation or subsequent testing. default When R(t0) = R default When the response termination time t0 is obtained, the metering cycle of the hydrogen sensor is: .

2. The method for determining the metering cycle of a palladium alloy hydrogen sensor in a closed environment as described in claim 1, characterized in that: In step S2, the maximum value of the aging temperature is T. n Set temperature interval Set n aging temperatures, with the i-th aging temperature being... Set the aging time interval. If the total aging time is m days, then the aging time at the i-th aging temperature is expressed as: .

3. The method for determining the metering cycle of a palladium alloy hydrogen sensor in a closed environment as described in claim 2, characterized in that: In step S2, at least three test samples are selected for each aging temperature. Before the accelerated thermal aging test begins, the test container used to store the test samples is placed in a 50℃ / 50Pa vacuum drying oven and dried for 2 hours. The test samples are connected to the aviation plug, and then the test container is sealed. The response curves and recovery curves of the hydrogen sensor resistance values ​​at different aging temperatures and different aging times are tested.

4. The method for determining the metering cycle of a palladium alloy hydrogen sensor in a closed environment as described in claim 1, characterized in that: In step S3, the specific method for testing the resistance response curves of the hydrogen sensor at different aging temperatures and aging times is as follows: Place the test container containing the test sample in a constant temperature chamber and adjust the temperature inside the chamber to set the aging temperature. Once the temperature in the test container containing the test sample stabilizes at the set aging temperature, collect and save the data. Air is extracted from the test chamber. Once the vacuum level inside the test chamber reaches the preset vacuum level, extraction is stopped. Hydrogen standard gas is introduced into the test tank, and the flow rate of the hydrogen standard gas is controlled to be no less than the set flow rate. Once the gas pressure inside the test tank reaches standard atmospheric pressure, open the exhaust port of the test tank and continue to introduce standard hydrogen gas for a set time. Then close the inlet and outlet ports of the test tank. Observe the collected data until the data remains stable. The resistance curve from the moment the standard hydrogen gas is input to the moment the data stabilizes is the response curve.

5. The method for determining the metering cycle of a palladium alloy hydrogen sensor in a closed environment as described in claim 4, characterized in that: The specific test method for the resistance recovery curve of the hydrogen sensor at different aging temperatures and aging times is as follows: After ensuring that the temperature in the test tank is stable at the set aging temperature, data is collected and saved. Introduce pure nitrogen into the air inlet of the test tank and open the exhaust port. Control the flow rate of pure nitrogen to be no less than the set flow rate. Observe the collected data until the data stabilizes. The resistance curve from the moment of pure nitrogen input to the moment of data stabilization is the resistance recovery curve.

6. The method for determining the metering cycle of a palladium alloy hydrogen sensor in a closed environment as described in claim 1, characterized in that: In step S5, the process of solving for the acceleration factor is as follows: The regression equation for the degradation law of the hydrogen sensor response value at the reference temperature is as follows: ; The regression equation for the degradation law of the recovery value of the hydrogen sensor at the reference temperature is as follows: ; In the formula, t represents the aging time, a continuous value starting from 0; , For regression functions; At other aging temperatures The aging time below The following will be accelerated by the factor Converted to reference temperature based on the time-temperature equivalence principle The aging time is as follows X represents the number of response values ​​and the number of recovery values ​​corresponding to different aging times at the i-th aging temperature. ; Accelerator The value of is determined by the objective function. Find the best option. In the formula, Aging temperature T i The aging time is converted to the aging time at the reference temperature. Substitute into and In the process, the aging temperature is obtained. Equivalent response value at reference temperature and recovery value equivalent value Using the average relative dispersion coefficient as the objective function, for each aging temperature T i The objective function is as follows In the form of -1； A parameter optimization algorithm is used to iteratively optimize each acceleration factor to achieve the desired result at each aging temperature T. i Below Approaching 0, we get The optimal solution is to equivalently shift the accelerated performance degradation data points at n-1 different temperatures to the distribution characteristics at the baseline temperature, perform regression calculations on the equivalent data sample points, and calculate the absolute dispersion coefficient of the regression prediction value of each sample point relative to the original value. When the predicted value of the regression model is within 1.5 times the dispersion line, the acceleration factor has high accuracy; otherwise, the acceleration factor needs to be calculated again.

7. The method for determining the metering cycle of a palladium alloy hydrogen sensor in a closed environment as described in claim 1, characterized in that: In step S6, the prior distribution in the Bayesian method is selected using the Markov chain Monte Carlo method. In the formula, y represents the number of sample points.

Images