Automatic data calibration method and system for air tightness tester

Abstract

The invention relates to the field of air tightness testers, and discloses an automatic data calibration method and system for an air tightness tester, and the method comprises the steps: obtaining a temperature sensor value T, a pressure sensor value P and a humidity sensor value H from the air tightness tester; performing three-dimensional coordinate positioning in the three-dimensional coordinate space to obtain a coordinate position C1; mapping the coordinate position C1 to a distribution model for coordinate mapping by using a coordinate transformation matrix M; obtaining a compensation parameter set S1 based on the error characteristic value E and the compensation parameter combination S; adjusting the compensation weight of each environment parameter according to the compensation parameter set S1; fusing the adjusted compensation weight with the error characteristic value E and the compensation parameter combination S to obtain a comprehensive compensation coefficient Kc; and performing correction processing on the original measurement data by using the comprehensive compensation coefficient Kc. The method has the advantages that automatic extraction and dynamic adjustment of calibration parameters are achieved when the test environment changes.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of air tightness testers, in particular to an automatic data calibration method and system for air tightness testers. BACKGROUND

[0002] As a key quality control link in modern industrial manufacturing, air tightness testing plays a crucial role in aerospace, automotive manufacturing, electronic equipment and other fields. With the continuous improvement of product precision requirements and the increasing complexity of application environment, the measurement accuracy of air tightness testers directly affects product quality and safety performance. Current air tightness testing equipment generally uses single parameter compensation or simple linear calibration methods, which have obvious shortcomings when facing complex and variable test environments. Traditional calibration methods usually only consider the influence of a single environmental factor, lack comprehensive consideration of the interaction of multiple environmental parameters, resulting in calibration accuracy that is difficult to meet high-precision testing requirements in actual application.

[0003] During the air tightness test, multiple environmental parameters such as temperature, pressure, and humidity will simultaneously affect the measurement results. There are complex nonlinear relationships and mutual coupling effects between these parameters. Temperature changes not only directly affect the gas density and viscosity, but also cause the drift of pressure sensor characteristics, thereby changing the response characteristics of the entire test system. Pressure fluctuations can affect the effectiveness of temperature compensation algorithms, while humidity changes can add interference to the measurement accuracy of pressure and temperature. This multi-parameter coupling effect makes it impossible for traditional single-dimensional calibration methods to accurately capture the comprehensive influence of environmental changes on test accuracy.

[0004] How to solve the above technical problems is a technical problem that needs to be overcome by those skilled in the art. SUMMARY

[0005] The present application provides an automatic data calibration method and system for air tightness testers to at least partially solve the above technical problems.

[0006] In a first aspect, to solve the above technical problems, the present application provides an automatic data calibration method for air tightness testers, comprising: obtaining temperature sensor value T, pressure sensor value P, and humidity sensor value H from the air tightness tester; if at least one of T, P, and H is not located within the corresponding preset range, generating an environmental parameter data set D containing T, P, and H; performing three-dimensional coordinate positioning in a three-dimensional coordinate space to obtain coordinate position C1 according to the environmental parameter data set D; the three-dimensional coordinate space is composed of temperature, pressure, and humidity; mapping the coordinate position C1 to a distribution model by using a coordinate transformation matrix M, querying an error characteristic value E and a compensation parameter combination S corresponding to the coordinate position C1; obtaining a compensation parameter set S1 based on the error characteristic value E and the compensation parameter combination S; adjusting compensation weights of each environmental parameter according to the compensation parameter set S1; fusing the adjusted compensation weights with the error characteristic value E and the compensation parameter combination S to obtain a comprehensive compensation coefficient Kc; applying the comprehensive compensation coefficient Kc to correct the original measurement data to generate corrected measurement data R1; if the corrected measurement data R1 is located in a corresponding preset range, the correction process is completed.

[0007] In an optional embodiment, according to the environmental parameter data set D, three-dimensional coordinate positioning is performed in a three-dimensional coordinate space to obtain a coordinate position C1, including: constructing a three-dimensional coordinate space and generating a preset three-dimensional grid division G; performing three-dimensional coordinate positioning on an input point C(T, P, H) in the three-dimensional coordinate space to obtain a preliminary coordinate position of the positioning point C; if the preliminary coordinate position of the positioning point C is located in a known distribution area, the preliminary coordinate position of the positioning point C is taken as the coordinate position C1; if the positioning point C is located in an unknown area, distances between the positioning point C and adjacent known points are calculated, weights W of the adjacent known points are determined based on the reciprocals of the distances, and an interpolation coefficient K is obtained through three-dimensional interpolation calculation according to the interpolation grid division G; the coordinate position C1 is generated according to the interpolation coefficient K and the preliminary coordinate position.

[0008] In an optional embodiment, mapping the coordinate position C1 to a distribution model by using a coordinate transformation matrix M, querying an error characteristic value E and a compensation parameter combination S corresponding to the coordinate position C1; obtaining a compensation parameter set S1 based on the error characteristic value E and the compensation parameter combination S, including: performing coordinate mapping operation on the coordinate position C1 in the three-dimensional coordinate space based on the coordinate transformation matrix M to generate a mapping coordinate C2; querying an error characteristic value E corresponding to the mapping coordinate C2 from a distribution database based on a database index structure; obtaining a compensation parameter combination S matched with the mapping coordinate C2, the compensation parameter combination S including a scaling factor and an offset; wherein if the error characteristic value E exceeds a preset threshold range, adjusting the compensation parameter combination S through an interpolation algorithm according to an adaptation relationship between a current environmental condition and historical compensation parameters; based on the compensation parameter combination S and the error characteristic value E, performing optimization iterative calculation on the compensation parameters by using a least square method to obtain a compensation parameter set S1.

[0009] In an optional embodiment, the compensation weight of each environmental parameter is adjusted according to the compensation parameter set S1; the adjusted compensation weight is fused with the error characteristic value E and the compensation parameter set S to obtain the comprehensive compensation coefficient Kc, including: based on the initial compensation weight in the compensation parameter set S1 , , ), the compensation weight of each environmental parameter is adjusted according to the absolute value of the deviation between the current environmental parameters T, P, H and the reference values T0, P0, H0; if |T-T0|>|P-P0|, the temperature compensation adjustment value At is calculated based on the temperature deviation value |T-T0|, and the temperature compensation weight is updated based on the temperature compensation adjustment value At ; if |P-P0|>|H-H0|, the pressure compensation adjustment value Ap is calculated based on the pressure deviation value |P-P0|, and the pressure compensation weight is updated based on the pressure compensation adjustment value Ap ; the humidity compensation weight remains unchanged; For the temperature compensation adjustment value At, if At is greater than a preset threshold Th1, At is weighted and fused with the error characteristic value E to obtain the first fused parameter F1; For the pressure compensation adjustment value Ap, if Ap is greater than a preset threshold Th2, Ap is weighted and fused with the error characteristic value E to obtain the second fused parameter F2; Common features are extracted from the first fused parameter F1 and the second fused parameter F2, and the comprehensive compensation coefficient Kc is calculated using the weighted average method.

[0010] In an optional embodiment, the comprehensive compensation coefficient Kc is applied to correct the original measurement data to generate corrected measurement data R1, including: The original measurement data R collected by the air tightness tester and the comprehensive compensation coefficient Kc are obtained; The original measurement data R and the comprehensive compensation coefficient Kc are fused and operated using a preset correction algorithm to generate corrected measurement data R1; wherein R1=R×Kc.

[0011] In an optional embodiment, the method further includes: S601, calculating the error residual Er based on the corrected measurement data R1 and the standard reference value R0; S602, comparing the error residual absolute value |Er| with the preset residual threshold Th3; if the error residual absolute value does not exceed the preset residual threshold, it is determined that the current calibration accuracy meets the requirements, and jumping to S605; if the error residual absolute value is greater than the preset residual threshold, jumping to S603; S603, taking the current error residual Er as a feedback parameter together with the temperature T, the pressure P and the humidity H in the environment parameter set D to form a new input parameter vector D' = [T, P, H, Er]; updating the coordinate transformation matrix M based on the parameter vector D' to make the mapping relationship contain error feedback information; S604, regenerating the mapping coordinates C2 according to the updated coordinate transformation matrix M, and querying the corresponding error eigenvalue E and the compensation parameter combination S in the distribution model; based on the new error eigenvalue E and the compensation parameter combination S, repeatedly performing the optimization calculation of the compensation parameter set S1, the update of the comprehensive compensation coefficient Kc and the re-correction of the original measurement data R; returning to S601; S605, outputting the final corrected measurement data R1.

[0012] In an optional implementation, the method further comprises: acquiring the current environment parameter set D = (T, P, H) and evaluating the data sparsity degree of the local area based on the historical data distribution density; adjusting the granularity of the three-dimensional grid division G according to the data sparsity degree; wherein the area with a data sparsity degree less than or equal to a preset density is recorded as a sparse area; the area with a data sparsity degree greater than the preset density is recorded as a dense area; for the sparse area, the grid cell edge length is increased to reduce the grid density; for the dense area, the grid cell edge length is appropriately reduced to improve the interpolation accuracy; When the positioning point C is located in an unknown area or a sparse area, the initial search radius is a standard reference value R0, and if a preset number of adjacent points cannot be found within the R0 range, the search radius is gradually expanded until the preset number is met.

[0013] In a second aspect, the application provides an automatic data calibration system for a gas tightness tester, comprising: a first processing module configured to acquire a temperature sensor value T, a pressure sensor value P and a humidity sensor value H from the gas tightness tester; if at least one of T, P and H is not located within a corresponding preset range, an environment parameter data set D containing T, P and H is generated; a second processing module configured to perform three-dimensional coordinate positioning to obtain a coordinate position C1 in a three-dimensional coordinate space according to the environment parameter data set D; the three-dimensional coordinate space is composed of temperature, pressure and humidity; a third processing module configured to map the coordinate position C1 to a distribution model using a coordinate transformation matrix M to perform coordinate mapping, query an error eigenvalue E and a compensation parameter combination S corresponding to the coordinate position C1, and obtain a compensation parameter set S1 based on the error eigenvalue E and the compensation parameter combination S; The fourth processing module is configured to adjust the compensation weight of each environmental parameter according to the compensation parameter set S1, and fuse the adjusted compensation weight, the error characteristic value E and the compensation parameter combination S to obtain a comprehensive compensation coefficient Kc. The fifth processing module is configured to apply the comprehensive compensation coefficient Kc to the original measurement data to perform correction processing and generate corrected measurement data R1, and if the corrected measurement data R1 is located in the corresponding preset range, the correction process is completed.

[0014] Compared with the prior art, the present application has at least the following beneficial effects: by monitoring the temperature, pressure and humidity in real time, performing positioning and interpolation operations in a preset three-dimensional coordinate space, and obtaining a special compensation parameter set, the compensation weight is dynamically adjusted according to the change of the environmental parameters, the error characteristic value and the compensation parameter are fused to obtain a comprehensive compensation coefficient, and the original measurement data is corrected in precision. BRIEF DESCRIPTION OF DRAWINGS

[0015] Figure 1 is a flowchart of an automatic data calibration method of a gas tightness tester provided by the first embodiment of the present application; Figure 2 is a block diagram of an automatic data calibration system of a gas tightness tester provided by the second embodiment of the present application. DETAILED DESCRIPTION

[0016] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0017] With reference to Figure 1 The first embodiment of the present application provides an automatic data calibration method of a gas tightness tester, which comprises the following steps: S101, obtaining a temperature sensor value T, a pressure sensor value P and a humidity sensor value H from a gas tightness tester, and if at least one of T, P and H is not located in the corresponding preset range, generating an environmental parameter data set D containing T, P and H.

[0018] Specifically, the temperature sensor value T, the pressure sensor value P and the humidity sensor value H are obtained from the air tightness tester by the data acquisition module. The air tightness tester is built-in with high-precision temperature sensor, pressure sensor and humidity sensor, which are used to collect the temperature, humidity and air pressure information in real time. The collection frequency can be set to once per second, and the collected data is transmitted to the central processing unit for storage and analysis through a serial communication protocol (such as RS485 or Modbus). The collected environmental parameters are monitored, and whether the current collected T, P and H is within the normal range is judged by setting the preset range [Tmin, Tmax], [Pmin, Pmax], [Hmin, Hmax], for example, the temperature range is 20.0-30.0℃, the pressure range is 100.0-105.0kPa, and the humidity range is 50.0%-70.0%. If any parameter exceeds the corresponding preset range (for example, temperature T=31.2℃>Tmax=30.0℃), the precision compensation process is triggered, the data point is marked, and the environmental parameter data set D is generated.

[0019] S102, according to the environmental parameter data set D, three-dimensional coordinate positioning is performed in a three-dimensional coordinate space to obtain a coordinate position C1; the three-dimensional coordinate space is composed of temperature, pressure and humidity; S103, the coordinate position C1 is mapped to a distribution model by using a coordinate transformation matrix M for coordinate mapping, and the error characteristic value E and the compensation parameter combination S corresponding to the coordinate position C1 are queried; the compensation parameter set S1 is obtained based on the error characteristic value E and the compensation parameter combination S. Specifically, the distribution model is a database established in advance through a large amount of experimental data, which records the influence of different environmental conditions (temperature, pressure, humidity) on the measurement accuracy.

[0020] S104, the compensation weights of each environmental parameter are adjusted according to the compensation parameter set S1; the adjusted compensation weights, the error characteristic value E and the compensation parameter combination S are fused to obtain a comprehensive compensation coefficient Kc; S105, the comprehensive compensation coefficient Kc is applied to correct the original measurement data to generate corrected measurement data R1; if the corrected measurement data R1 is located within the corresponding preset range, the correction process is completed.

[0021] In an embodiment, according to the environmental parameter data set D, three-dimensional coordinate positioning is performed in a three-dimensional coordinate space to obtain a coordinate position C1, which includes: constructing a three-dimensional coordinate space and generating a preset three-dimensional grid division G; performing three-dimensional coordinate positioning on the input point C(T, P, H) in the three-dimensional coordinate space to obtain the preliminary coordinate position of the positioning point C; If the preliminary coordinate position of the positioning point C is located in the known distribution area, the preliminary coordinate position of the positioning point C is taken as the coordinate position C1; If the positioning point C is located in the unknown area, the distance between the positioning point C and the adjacent known points is calculated, the weight W of each adjacent known point is determined based on the reciprocal of the distance, and the interpolation coefficient K is calculated through three-dimensional interpolation according to the interpolation grid division G; the coordinate position C1 is generated according to the interpolation coefficient K and the preliminary coordinate position.

[0022] Specifically, the temperature-pressure-humidity three-dimensional coordinate space is constructed by using the environmental parameter data set D, and a preset three-dimensional grid division G is generated; wherein the three-dimensional grid division G: divides the three-dimensional coordinate space into a plurality of small unit cells; the three-dimensional coordinate space is composed of three dimensions of temperature, pressure and humidity. In the above constructed three-dimensional coordinate space, the input point C(T, P, H) is located in three-dimensional coordinates. Assuming that the current environmental parameters are T=21℃, P=101.5kPa, H=52%, the corresponding preliminary coordinate position in the three-dimensional coordinate space is determined, and this process is realized by searching for the closest known sampling point to the input point. These known sampling points usually include the pre-stored temperature, pressure, humidity and the corresponding compensation coefficient K. If the positioning point C(T, P, H) falls within the known distribution area (i.e. there is a direct matching sampling point), the compensation coefficient K corresponding to the point is directly extracted as the final coordinate position C1. For example, the known point A1(20℃, 101kPa, 50%) corresponds to the compensation coefficient K=1.05, and if the distance d1 between C and A1 satisfies the preset accuracy condition (such as d≤3), the compensation coefficient K=1.05 of A1 is directly used as the compensation coefficient of C, so as to determine the accurate coordinate position C1 of C. If the positioning point C(T, P, H) is not in the known distribution area (i.e. there is no direct matching sampling point), the compensation coefficient needs to be estimated by the adjacent point interpolation method. First, the distance between the positioning point C and all adjacent known points is calculated, and the formula is: For example, for the target positioning point C(21℃, 101.5kPa, 52%), the distances between C and the known points A1(20℃, 101kPa, 50%), A2(25℃, 100kPa, 60%) and A3(22℃, 102kPa, 55%) are d1≈2.24, d2≈9.06 and d3≈3.20, respectively. Assuming that the known distribution area is d≤3, C and A1 and A3 are in the accuracy area, and A2 is not. The weight W of each adjacent known point is determined based on the reciprocal of the distance. The weight calculation formula is: The normalized weight represents the influence degree of each adjacent point on the interpolation result; for example, for points A1 and A3 in the above example, W1=0.35 and W3=0.25 are obtained after normalization. Then, the interpolation coefficient K is calculated through three-dimensional interpolation according to the interpolation grid division G. The interpolation formula is: ; select the adjacent 8 grid points, such as A1, A2, A3 and A4 (24℃, 101kPa, 56%, K=1.09), the interpolation coefficient K is calculated as follows: K=0.35×1.05+0.15×1.10+0.25×1.08+0.20×1.09=1.0205K=0.35×1.05+0.15×1.10+0.25×1.08+0.20×1.09=1.0205; adjust the coordinates of the positioning point C according to the interpolation coefficient K and the preliminary coordinate position to generate the accurate coordinate position C1. For example, C'(23℃, 100.5kPa, 58%) obtains the interpolation coefficient K=1.0205 through the above interpolation calculation, and then generates the accurate coordinate position C1. When the preliminary positioning point is not in the known distribution area, a more accurate coordinate position is calculated through the adjacent point interpolation method, and this new position is C'.

[0023] By using the above method, compared with the traditional single parameter compensation method, the scheme utilizes three-dimensional coordinate space and multi-dimensional interpolation algorithm, which can more accurately capture the influence of environmental changes on measurement results. Especially when dealing with unknown area data, through the adjacent point interpolation method and weight distribution mechanism, it is ensured that even in the case of lack of direct matching data, reliable compensation coefficients can be obtained, thereby improving the accuracy of air tightness test.

[0024] In an embodiment, the coordinate position C1 is mapped to the distribution model using the coordinate transformation matrix M for coordinate mapping, and the error characteristic value E and the compensation parameter combination S corresponding to the coordinate position C1 are queried; based on the error characteristic value E and the compensation parameter combination S, a compensation parameter set S1 is obtained, including: Perform coordinate mapping operation on the coordinate position C1 in the three-dimensional coordinate space based on the coordinate transformation matrix M to generate a mapping coordinate C2; Based on the database index structure, the error characteristic value E corresponding to the mapping coordinate C2 is queried from the distribution database; Obtain the compensation parameter combination S matched with the mapping coordinate C2, the compensation parameter combination S includes a scaling factor and an offset; wherein, if the error characteristic value E exceeds the preset threshold range, the compensation parameter combination S is adjusted through the interpolation algorithm according to the adaptation relationship between the current environmental condition and the historical compensation parameter; Based on the compensation parameter combination S and the error characteristic value E, the least square method is used to optimize and iteratively calculate the compensation parameter to obtain the compensation parameter set S1.

[0025] Specifically, a coordinate transformation matrix M is used to perform a mapping operation on the input coordinates C1 to generate mapped coordinates C2. The coordinate transformation matrix M is used to convert points in a three-dimensional coordinate space to another coordinate system. Based on the database index structure, the error characteristic value E corresponding to the mapped coordinates C2 is queried from the pre-set precision distribution database. The error characteristic value E represents the degree of deviation between the measurement result and the true value at a specific coordinate point. For example, assuming that C2 falls into the grid (10.5-11.0, 18.5-19.0), the query returns an error characteristic value E = 0.12, representing the average coordinate deviation in this region. The compensation parameter combination S matching the mapped coordinates C2 is obtained, which usually includes a scaling factor s1 and an offset s2. The scaling factor s1 is used to adjust the scaling factor of the measurement value, and the offset s2 is used to adjust the fixed offset of the measurement value.

[0026] Suppose the database provides compensation parameters for temperatures of 20°C and 30°C as S20 = (1.01, (0.10, -0.10)) and S30 = (1.03, (0.20, -0.05)), respectively, and the current environmental temperature is 25°C. By using a linear interpolation algorithm to adjust S, the calculation gives s1 = 1.01 + (1.03 - 1.01) * 0.5 = 1.02, the offset x = 0.10 + (0.20 - 0.10) * 0.5 = 0.15, y = -0.10 + (-0.05 - (-0.10)) * 0.5 = -0.075, and the special compensation parameter set S1 = (1.02, (0.15, -0.075)) is obtained. If the error characteristic value E exceeds the pre-set threshold range, the compensation parameter combination S is adjusted by an interpolation algorithm according to the adaptive relationship between the current environmental conditions and the historical compensation parameters. For example, if E = 0.12 exceeds the pre-set threshold range, a new compensation parameter combination needs to be recalculated to adapt to the current environmental conditions.

[0027] Based on the compensation parameter combination S and the error characteristic value E, the least squares method is used to optimize and iteratively calculate the compensation parameters to obtain the final compensation parameter set S1. For example, the preliminary compensation parameter set S2 = (1.02, (0.15, -0.075)), and the error characteristic value E = 0.12. After optimization using the least squares method, the final compensation parameter set S1 = (1.02, (0.15, -0.075)) is obtained. At this time, the compensation parameter set S1 has been optimized and can be more accurately applied to subsequent data correction processes.

[0028] Through the above method, the system can perform coordinate mapping in different reference frames by using the coordinate transformation matrix M, ensuring the consistency of measurement results under different environmental conditions. By querying the error characteristic value E from the distributed database and combining the adaptive relationship between the current environmental conditions and the historical compensation parameters, the compensation parameter combination S is dynamically adjusted, so that accurate and reliable compensation parameters can be obtained even if the error characteristic value exceeds the preset range. The application of the least square method further improves the accuracy of the compensation parameters, ensuring the robustness and stability of the system.

[0029] In an embodiment, the compensation weight of each environmental parameter is adjusted according to the compensation parameter set S1; the adjusted compensation weight, the error characteristic value E, and the compensation parameter combination S are fused to obtain a comprehensive compensation coefficient Kc, including: based on the initial compensation weight in the compensation parameter set S1 、 、 ), the compensation weight of each environmental parameter is adjusted according to the absolute value of the deviation between the current environmental parameters T, P, H and the reference values T0, P0, H0; if |T-T0|>|P-P0|, the temperature compensation adjustment value At is calculated based on the temperature deviation value |T-T0|, and the temperature compensation weight is updated based on the temperature compensation adjustment value At ; if |P-P0|>|H-H0|, the pressure compensation adjustment value Ap is calculated based on the pressure deviation value |P-P0|, and the pressure compensation weight is updated based on the pressure compensation adjustment value Ap ; the humidity compensation weight remains unchanged; For the temperature compensation adjustment value At, if At is greater than a preset threshold Th1, At and the error characteristic value E are weighted and fused to obtain a first fused parameter F1; For the pressure compensation adjustment value Ap, if Ap is greater than a preset threshold Th2, Ap and the error characteristic value E are weighted and fused to obtain a second fused parameter F2; Common features are extracted from the first fused parameter F1 and the second fused parameter F2, and a weighted average method is used to calculate the comprehensive compensation coefficient Kc.

[0030] Specifically, the initial compensation weight is 、 、 The initial compensation weight is set based on historical data analysis and the relationship between environmental parameters and measurement results. In one implementation, a large amount of historical data is analyzed to understand the impact of different environmental conditions (temperature T, pressure P, humidity H) on measurement accuracy. For example, analyzing measurement errors under different temperature, pressure, and humidity conditions over a past period helps determine which environmental factors have a greater impact on the measurement results. Based on the above analysis results, an initial compensation weight is pre-set, for example, a temperature compensation weight. =0.5; Pressure compensation weight =0.3; Humidity compensation weight =0.2; These weight values ​​represent the degree of importance of each environmental factor on the measurement results by default, and the specific values ​​can be adjusted according to the actual situation.

[0031] Based on the initial compensation weights in the compensation parameter set S1, the compensation weights of each environmental parameter are adjusted according to the absolute value of the deviation between the current environmental parameters T, P, and H and the reference values ​​T0, P0, and H0.

[0032] If |T-T0|>|P-P0|, then the temperature compensation weights are obtained from the dedicated parameter set S1. A linear regression algorithm is used to process the temperature deviation value |T-T0| to obtain the temperature compensation adjustment value At. For example, the current environmental parameters are T=25℃, P=101kPa, H=60%RH, and the reference parameters are T0=20℃, P0=100kPa, H0=50%RH. The calculated deviations are |T-T0|=|25-20|=5℃, |P-P0|=|101-100|=1kPa, and |H-H0|=|60-50|=10%RH. Since 5>|T-T0|>|P-P0|=1, the temperature compensation weight is adjusted first. , = ×(1+|T-T0| / T0)=0.5×(1+5 / 20)=0.625.

[0033] If |P-P0|>|H-H0|, then the pressure compensation weight is obtained from the dedicated parameter set S1. The pressure deviation value |P-P0| is processed using a support vector machine algorithm to obtain the pressure compensation adjustment value Ap. Comparing |P-P0|=1<|H-H0|=10, the pressure compensation weight is adjusted first. , = ×(1+|P-P0| / P0)=0.3×(1+1 / 100)=0.303. The humidity compensation weight Wh remains unchanged or is adjusted according to the humidity deviation. Assume that the humidity compensation weight Wh remains at 0.2.

[0034] For the temperature compensation adjustment value At, if At is greater than a preset threshold Th1 (for example, Th1 = 0.5), At is fused with the error characteristic value E by weighting to obtain a first fusion parameter F1.

[0035] For the pressure compensation adjustment value Ap, if Ap is greater than a preset threshold Th2 (for example, Th2 = 0.3), Ap is fused with the error characteristic value E by weighting to obtain a second fusion parameter F2. For example, assuming that Ap = 0.303 and Ap > Th2, Ap is fused with the error characteristic value E by weighting to obtain the second fusion parameter F2. Common features are extracted from the first fusion parameter F1 and the second fusion parameter F2, which means that the first fusion parameter F1 and the second fusion parameter F2 are integrated to generate a comprehensive compensation coefficient Kc that can comprehensively reflect the influence of measurement error under current environmental conditions, for example, the comprehensive compensation coefficient Kc is calculated by using a weighted average method. For example, assuming that F1 = 0.078125 and F2 = 0.007575, the comprehensive compensation coefficient Kc is calculated by using a weighted average method, Kc = (F1 + F2) / 2 = 0.04285.

[0036] In one of the implementation modes, , is a weight coefficient.

[0037] By the above method, the compensation weight of each environmental parameter can be dynamically adjusted according to the absolute value of the deviation between the current environmental parameters T, P, H and the reference values T0, P0, H0, and the accuracy of the compensation process is ensured. Especially when the temperature or pressure deviation is large, the first fusion parameter F1 and the second fusion parameter F2 are generated by fusing the special compensation adjustment value (such as At and Ap) with the error characteristic value E by weighting, which further improves the compensation accuracy. The comprehensive compensation coefficient Kc is calculated by using a weighted average method, so that the whole compensation process not only considers the influence of a single environmental parameter, but also combines the comprehensive action of multiple factors, ensuring the accuracy of the measurement results.

[0038] In one of the implementation modes, the comprehensive compensation coefficient Kc is applied to correct the original measurement data to generate corrected measurement data R1, including: The original measurement data R collected by the air tightness tester and the comprehensive compensation coefficient Kc are obtained. The original measurement data R and the comprehensive compensation coefficient Kc are fused by using a preset correction algorithm to generate corrected measurement data R1; wherein R1 = R x Kc.

[0039] In one of the implementation modes, the method further includes: S601, calculating an error residual Er based on the corrected measurement data R1 and a standard reference value R0. S602, compare the error residual absolute value |Er| with a preset residual threshold Th3; if the error residual absolute value does not exceed the preset residual threshold, it is determined that the current calibration accuracy meets the requirements, and jump to S605; if the error residual absolute value is greater than the preset residual threshold, jump to S603; S603, the current error residual Er is taken as a feedback parameter to form a new input parameter vector D' = [T, P, H, Er] together with the temperature T, pressure P, and humidity H in the environmental parameter set D; update the coordinate transformation matrix M based on the parameter vector D' to make the mapping relationship contain error feedback information; S604, regenerate the mapping coordinates C2 according to the updated coordinate transformation matrix M, and query the corresponding error characteristic value E and compensation parameter combination S in the distribution model; based on the new error characteristic value E and the compensation parameter combination S, repeat the optimization calculation of the compensation parameter set S1, the update of the comprehensive compensation coefficient Kc, and the re-correction of the original measurement data R; return to S601; S605, output the final corrected measurement data R1.

[0040] Specifically, the air tightness tester calculates the error residual Er based on the corrected measurement data R1 and the standard reference value R0; compare the error residual absolute value |Er| with a preset residual threshold Th3. If the error residual absolute value does not exceed the preset residual threshold, it is determined that the current calibration accuracy meets the requirements, and jump to S605 to output the final corrected measurement data R1. For example, assuming that the preset residual threshold Th3 is 3.0, since |Er| = 2.71 < 3.0, it is considered that the current calibration accuracy meets the requirements, and directly jump to S605. If the error residual absolute value is greater than the preset residual threshold, the calibration parameters need to be further adjusted and optimized, and jump to S603. When the error residual absolute value is greater than the preset residual threshold, the system takes the current error residual Er as a feedback parameter, and forms a new input parameter vector D' = [T, P, H, Er] together with the temperature T, pressure P, and humidity H in the environmental parameter set D, and updates the coordinate transformation matrix M to ensure that the mapping relationship can contain error feedback information. The effect of updating the coordinate transformation matrix M to contain error feedback information is to dynamically adjust the calibration model, so that the system can more accurately reflect the actual measurement error under the current environmental conditions and perform corresponding compensation. According to the updated coordinate transformation matrix M, the mapping coordinates C2 are regenerated, and the error characteristic value E corresponding to the mapping coordinates C2 is queried from the distribution database. The compensation parameter combination S matching the mapping coordinates C2 is obtained. Based on the new error characteristic value E and the compensation parameter combination S, the optimization calculation of the compensation parameter set S1, the update of the comprehensive compensation coefficient Kc, and the re-correction of the original measurement data R are repeated, and the step S601 is returned to continue to evaluate the correction effect.

[0041] In an embodiment, assuming that the measurement data obtained after the initial correction is R1=52.71, and the standard reference value R0=50.0, the error residual Er is calculated: Er=R1−R0=2.71; if the preset residual threshold Th3 is 3.0, since |Er|=2.71<3.0, it is considered that the current calibration accuracy meets the requirements, and the final corrected measurement data R1 is directly output. But if the absolute value of the error residual is greater than the preset residual threshold, for example, Er=3.5, further adjustment is needed: the current error residual Er=3.5 is taken as a feedback parameter, together with the temperature T=25℃, the pressure P=101kPa, and the humidity H=60% in the environmental parameter set D to form a new input parameter vector D'=[25, 101, 60, 3.5]. The coordinate transformation matrix M is updated, the mapping coordinates C2 are regenerated, and the corresponding error characteristic value E and the compensation parameter combination S in the distribution model are queried. Based on the new error characteristic value E and the compensation parameter combination S, the optimization calculation of the compensation parameter set S1, the update of the comprehensive compensation coefficient Kc, and the re-correction of the original measurement data R are repeatedly performed.

[0042] Through the above method, the error residual Er is calculated after the initial correction, which is combined with the environmental parameter set D to form a new input parameter vector D', and the error feedback information is included by dynamically updating the coordinate transformation matrix M, so that closed-loop calibration control is realized. The accuracy and stability of the calibration process are improved, and especially in the case where the error exceeds the preset threshold, the compensation parameter set S1 and the comprehensive compensation coefficient Kc are optimized through multiple iterations to ensure the accuracy of the final correction result.

[0043] In an embodiment, the method further comprises: obtaining the current environmental parameter set D=(T, P, H) and evaluating the data sparsity degree of the local area based on the historical data distribution density; adjusting the granularity of the three-dimensional grid division G according to the data sparsity degree; wherein the area with a data sparsity degree less than or equal to a preset density is recorded as a sparse area; the area with a data sparsity degree greater than the preset density is recorded as a dense area; for the sparse area, the grid cell edge length is increased to reduce the grid density; for the dense area, the grid cell edge length is appropriately reduced to improve the interpolation accuracy; When the positioning point C is located in an unknown area or a sparse area, the initial search radius is the standard reference value R0, and if a preset number of adjacent points are not found within the R0 range, the search radius is gradually expanded until the preset number is met.

[0044] Reference Figure 2 The second embodiment of the present application provides an automatic data calibration system of a gas tightness tester, comprising: The first processing module is configured to acquire a temperature sensor value T, a pressure sensor value P and a humidity sensor value H from the air tightness tester; and generate an environment parameter data set D containing T, P and H if at least one of T, P and H is not within a corresponding preset range. The second processing module is configured to perform three-dimensional coordinate positioning to obtain a coordinate position C1 in a three-dimensional coordinate space according to the environment parameter data set D; the three-dimensional coordinate space is constituted by temperature, pressure and humidity. The third processing module is configured to map the coordinate position C1 to a distribution model by using a coordinate transformation matrix M to perform coordinate mapping, query an error characteristic value E and a compensation parameter combination S corresponding to the coordinate position C1, and obtain a compensation parameter set S1 based on the error characteristic value E and the compensation parameter combination S. The fourth processing module is configured to adjust compensation weights of respective environment parameters according to the compensation parameter set S1, and fuse the adjusted compensation weights with the error characteristic value E and the compensation parameter combination S to obtain a comprehensive compensation coefficient Kc. The fifth processing module is configured to apply the comprehensive compensation coefficient Kc to perform correction processing on original measurement data to generate corrected measurement data R1; and complete the correction process if the corrected measurement data R1 is within a corresponding preset range.

[0045] It should be noted that the automatic data calibration system of the air tightness tester provided by the embodiments of the present application is used to perform all process steps of the automatic data calibration method of the air tightness tester of the above embodiments, and the working principles and beneficial effects of the two are one-to-one corresponding, so they will not be described again.

[0046] The above specific embodiments further illustrate the purpose, technical solutions and beneficial effects of the present application. It should be understood that the above description is only for specific embodiments of the present application and is not used to limit the protection scope of the present application. It is particularly pointed out that any modification, equivalent replacement, improvement, etc. made by those skilled in the art within the spirit and principles of the present application should be included in the protection scope of the present application.

Claims

1. An automated data calibration method for a gas tightness tester, the method comprising: The method comprises the following steps: obtaining temperature sensor value T, pressure sensor value P and humidity sensor value H from the air tightness tester; if at least one of T, P and H is not within the corresponding preset range, generating an environment parameter data set D containing T, P and H; performing three-dimensional coordinate positioning in a three-dimensional coordinate space according to the environment parameter data set D to obtain coordinate position C1; the three-dimensional coordinate space is composed of temperature, pressure and humidity; mapping the coordinate position C1 to a distribution model using a coordinate transformation matrix M to perform coordinate mapping, querying error characteristic value E and compensation parameter combination S corresponding to the coordinate position C1; and obtaining compensation parameter set S1 based on the error characteristic value E and the compensation parameter combination S; adjusting the compensation weight of each environment parameter according to the compensation parameter set S1; and fusing the adjusted compensation weight with the error characteristic value E and the compensation parameter combination S to obtain comprehensive compensation coefficient Kc; applying the comprehensive compensation coefficient Kc to correct the original measurement data to generate corrected measurement data R1; and if the corrected measurement data R1 is within the corresponding preset range, completing the correction process.

2. The method of claim 1, wherein, performing three-dimensional coordinate positioning in a three-dimensional coordinate space according to the environment parameter data set D to obtain coordinate position C1, which comprises the following steps: constructing a three-dimensional coordinate space and generating a preset three-dimensional grid division G; performing three-dimensional coordinate positioning on input point C(T, P, H) in the three-dimensional coordinate space to obtain the preliminary coordinate position of the positioning point C; if the preliminary coordinate position of the positioning point C is within the known distribution area, taking the preliminary coordinate position of the positioning point C as the coordinate position C1; if the positioning point C is in the unknown area, calculating the distance between the positioning point C and the adjacent known points, determining the weight W of each adjacent known point based on the reciprocal of the distance, and performing three-dimensional interpolation calculation according to the interpolation grid division G to obtain the interpolation coefficient K; and generating the coordinate position C1 according to the interpolation coefficient K and the preliminary coordinate position.

3. The method of claim 2, wherein, mapping the coordinate position C1 to a distribution model using a coordinate transformation matrix M to perform coordinate mapping, querying error characteristic value E and compensation parameter combination S corresponding to the coordinate position C1; obtaining compensation parameter set S1 based on the error characteristic value E and the compensation parameter combination S, which comprises the following steps: performing coordinate mapping operation on the coordinate position C1 in the three-dimensional coordinate space based on the coordinate transformation matrix M to generate mapping coordinate C2; querying error characteristic value E corresponding to the mapping coordinate C2 from the distribution database based on the database index structure; obtaining compensation parameter combination S matching the mapping coordinate C2, the compensation parameter combination S comprising a scaling factor and an offset; wherein if the error characteristic value E exceeds the preset threshold range, adjusting the compensation parameter combination S through an interpolation algorithm according to the adaptation relationship between the current environmental condition and the historical compensation parameters; based on the compensation parameter combination S and the error characteristic value E, performing optimization iterative calculation on the compensation parameters using the least square method to obtain the compensation parameter set S1.

4. The method of claim 3, wherein, adjusting the compensation weight of each environment parameter according to the compensation parameter set S1. The adjusted compensation weight is fused with the error characteristic value E and the compensation parameter combination S to obtain a comprehensive compensation coefficient Kc, including: based on the initial compensation weight in the compensation parameter set S1 , , , adjusts the compensation weight of each environmental parameter according to the absolute value of the deviation between the current environmental parameters T, P, H and the reference values T0, P0, H0; if |T-T0|>|P-P0|, calculates the temperature compensation adjustment value At based on the temperature deviation value |T-T0|, and updates the temperature compensation weight based on the temperature compensation adjustment value At; if |P-P0|>|H-H0|, calculates the pressure compensation adjustment value Ap based on the pressure deviation value |P-P0|, and updates the pressure compensation weight based on the pressure compensation adjustment value Ap; the humidity compensation weight remains unchanged; For the temperature compensation adjustment value At, if At is greater than a preset threshold Th1, At is fused with the error characteristic value E by weighting to obtain a first fusion parameter F1; For the pressure compensation adjustment value Ap, if Ap is greater than a preset threshold Th2, Ap is fused with the error characteristic value E by weighting to obtain a second fusion parameter F2; Common features are extracted from the first fusion parameter F1 and the second fusion parameter F2, and a comprehensive compensation coefficient Kc is calculated by using a weighted average method.

5. The method of claim 4, wherein, The original measurement data is corrected by using the comprehensive compensation coefficient Kc to generate corrected measurement data R1, including: The original measurement data R collected by the air tightness tester and the comprehensive compensation coefficient Kc are obtained; The original measurement data R and the comprehensive compensation coefficient Kc are fused by using a preset correction algorithm to generate corrected measurement data R1; wherein R1=R*Kc.

6. The method of claim 5, wherein, The method further includes: S601, calculating an error residual Er based on the corrected measurement data R1 and a standard reference value R0; S602, comparing the error residual absolute value |Er| with a preset residual threshold Th3; if the error residual absolute value does not exceed the preset residual threshold, it is determined that the current calibration accuracy meets the requirements, and jumping to S605; if the error residual absolute value is greater than the preset residual threshold, jumping to S603; S603, taking the current error residual Er as a feedback parameter together with the temperature T, the pressure P and the humidity H in the environmental parameter set D to form a new input parameter vector D'=[T, P, H, Er]; updating the coordinate transformation matrix M based on the parameter vector D' to make the mapping relationship contain error feedback information; S604, regenerating the mapping coordinates C2 according to the updated coordinate transformation matrix M, and querying the corresponding error characteristic value E and the compensation parameter combination S in the distribution model; based on the new error characteristic value E and the compensation parameter combination S, repeating the optimization calculation of the compensation parameter set S1, the update of the comprehensive compensation coefficient Kc and the correction of the original measurement data R again; returning to S601; S605, outputting the final corrected measurement data R1.

7. The method of claim 6, wherein, The method further includes: The current environmental parameter set D=(T, P, H) is obtained and the data sparsity degree of the local area is evaluated based on the historical data distribution density; The granularity of the three-dimensional grid division G is adjusted according to the data sparsity degree; wherein the area with a data sparsity degree less than or equal to a preset density is recorded as a sparse area; the area with a data sparsity degree greater than the preset density is recorded as a dense area; for the sparse area, the grid cell edge length is increased to reduce the grid density; for the dense area, the grid cell edge length is appropriately reduced to improve the interpolation accuracy; When the positioning point C is located in an unknown area or a sparse area, the initial search radius is a standard reference value R0, and if a preset number of adjacent points cannot be found within R0, the search radius is gradually expanded until the preset number is met.

8. An automated data calibration system for a gas tightness tester, the system comprising: including: The first processing module is used for obtaining the temperature sensor value T, the pressure sensor value P and the humidity sensor value H from the air tightness tester; If at least one of T, P and H is not located in the corresponding preset range, an environmental parameter data set D containing T, P and H is generated; The second processing module is configured to perform three-dimensional coordinate positioning to obtain a coordinate position C1 in a three-dimensional coordinate space according to the environmental parameter data set D, wherein the three-dimensional coordinate space is constituted by temperature, pressure and humidity; The third processing module is configured to map the coordinate position C1 to a distribution model by using a coordinate transformation matrix M to perform coordinate mapping, query an error eigenvalue E and a compensation parameter combination S corresponding to the coordinate position C1, and obtain a compensation parameter set S1 based on the error eigenvalue E and the compensation parameter combination S; The fourth processing module is configured to adjust compensation weights of the respective environmental parameters according to the compensation parameter set S1, fuse the adjusted compensation weights with the error eigenvalue E and the compensation parameter combination S to obtain a comprehensive compensation coefficient Kc, and apply the comprehensive compensation coefficient Kc to correct the original measurement data to generate corrected measurement data R1. The fifth processing module is configured to complete the correction process if the corrected measurement data R1 is located in the corresponding preset range.

Images