Air tightness detection method and related equipment

Abstract

The invention discloses an air tightness detection method and related equipment, and the method comprises the steps: controlling an air inflation device to carry out N times of air inflation processing on a test cavity, the air inflation amount of the first time of air inflation processing is determined according to the air amount capable of enabling the air pressure of the test cavity to reach the target air pressure, and the test cavity is internally provided with a tested object and an air pressure sensor, n is an integer greater than 1; acquiring N parts of sensor data which are acquired by an air pressure sensor and are in one-to-one correspondence with the N times of inflation treatment; based on the target air pressure and the sensor data corresponding to the first inflation processing, adjusting the N parts of sensor data to obtain N parts of first sensor data; determining N parts of second sensor data based on the respective control information of the N times of inflation processing and the N parts of first sensor data; and determining the air tightness detection result of the detected object at least based on the respective inflation volumes of the N times of inflation processing and the N parts of second sensor data, so that the air tightness detection precision can be improved.

Description

Technical Field

[0001] This application relates to the field of airtightness testing technology, and in particular to an airtightness testing method and related equipment. Background Technology

[0002] In modern industrial systems, many industrial sectors have certain requirements for the airtightness of the workpieces used. If the airtightness of the workpiece does not meet the standards, it may affect the normal use of the workpiece under certain working conditions, and may even cause safety problems due to poor airtightness of the workpiece.

[0003] In related technologies, a professional airtightness tester (such as a differential pressure leak detector) is usually used to test the airtightness of the workpiece. During the test, a sensor adapted to the tester is used to collect relevant data, and the test results are obtained by analyzing the collected data.

[0004] However, the airtightness detection accuracy of this method is limited by the accuracy of the collected data. Summary of the Invention

[0005] To address the aforementioned technical problems, this application provides an airtightness testing method, apparatus, storage medium, and computer equipment, which can improve the accuracy of airtightness testing.

[0006] In a first aspect, embodiments of this application provide an airtightness detection method, including: The inflation device is controlled to inflate the test chamber N times. The inflation amount of the first inflation is determined based on the amount of gas that can make the air pressure of the test chamber reach the target air pressure. The test chamber is equipped with the test object and an air pressure sensor. N is an integer greater than 1. Obtain N sets of sensor data collected by the air pressure sensor, each corresponding to one of the N inflation processes; Based on the target air pressure and the sensor data corresponding to the first inflation process, the N sets of sensor data are adjusted to obtain N sets of first sensor data; Based on the control information of each of the N inflation processes and the N sets of first sensor data, N sets of second sensor data are determined; The air tightness test result of the object under test is determined based at least on the inflation volume of each of the N inflation processes and the N sets of second sensor data.

[0007] Optionally, determining the airtightness test result of the object under test based at least on the inflation volume of each of the N inflation processes and the N sets of second sensor data includes: Based on the N sets of second sensor data and the obtained covariate sequence, N sets of third sensor data are determined, wherein the covariate sequence includes N sets of covariate information corresponding one-to-one with the N inflation processes; Based on the inflation volume of each of the N inflation processes and the N sets of data from the third sensor, the airtightness test result of the tested object is determined.

[0008] Optionally, determining N sets of third sensor data based on the N sets of second sensor data and the acquired covariate sequence includes: For each of the N sets of second sensor data, Determine the data characteristics of the second sensor data and the variable characteristics of the covariate information corresponding to the second sensor data; The data features and the variable features are combined to obtain a first combined feature; Based at least on the first merging feature, a third sensor data corresponding to the second sensor data is obtained.

[0009] Optionally, the first merging feature includes at least one merging feature vector, and obtaining a third sensor data corresponding to the second sensor data based at least on the first merging feature includes: The preset attention model is configured using the variable features of the covariate information corresponding to the second sensor data. The configured attention model is invoked to assign corresponding weights to the at least one merged feature vector; Based on the assigned weights, the first merged feature is weighted and fused to obtain the first fused feature; Based on the first fusion feature, a third sensor data corresponding to the second sensor data is obtained.

[0010] Optionally, the control information corresponds to at least one control factor; The determination of N sets of second sensor data based on the control information of each of the N inflation processes and the N sets of first sensor data includes: For each of the N inflation processes, Based on the control information corresponding to this inflation process, determine the reference sensor data; Extract the first data feature and the at least one first factor feature corresponding to the at least one control factor from the reference sensor data; Extract the second data feature and the at least one second factor feature corresponding to the at least one control factor from the first sensor data corresponding to the inflation process; According to the at least one control factor, the at least one first factor feature and the at least one second factor feature are combined into at least one factor feature pair, wherein the first factor feature and the second factor feature corresponding to the same control factor are combined into one factor feature pair; Based on the first data feature, the second data feature, and the at least one factor feature pair, a second set of sensor data corresponding to the inflation process is determined.

[0011] Optionally, determining a set of second sensor data corresponding to the inflation process based on the first data feature, the second data feature, and the at least one factor feature pair includes: For each of the aforementioned factor characteristic pairs Based on the factor feature pair, the first data feature, and the second data feature, determine the intermediate feature of the control factor corresponding to the factor feature pair; Based on the control factors corresponding to the factor feature pair, the weights of the factor feature pair, the first data feature, and the second data feature are determined, wherein the sum of the weights of the first data feature and the second data feature is less than the weight of the factor feature pair. Based on their respective weights, intermediate features, first data features, and second data features, a set of second sensor data corresponding to this inflation process is determined.

[0012] Optionally, determining a set of second sensor data corresponding to the inflation process based on the respective weights, the intermediate features, the first data feature, and the second data feature includes: The first feature filtering network is invoked to process the first data feature to obtain the first filtering feature, process the intermediate feature to obtain the intermediate filtering feature, and process the second data feature to obtain the second filtering feature. The first feature filtering network includes at least two deep feature extraction layers, a pooling layer and an activation layer that are electrically connected in sequence. The first filtering feature and the intermediate filtering feature are upsampled to obtain their respective upsampled features. Based on their respective weights, the upsampled feature and the second filtering feature are weighted, and the weighted upsampled feature and the weighted second filtering feature are concatenated to obtain the concatenated feature. The second feature filtering network is invoked to process the spliced ​​features to obtain the processed spliced ​​features. The second feature filtering network includes a convolutional layer, a pooling layer, an activation layer and a convolutional layer that are electrically connected in sequence. Based on their respective weights, the intermediate feature, the first data feature, and the second data feature are weighted; The weighted intermediate features, the weighted first data features, the weighted second data features, and the processed spliced ​​features are fused together to obtain the second fused feature. Based on the second fusion feature, a second set of sensor data corresponding to this inflation process is determined.

[0013] Secondly, embodiments of this application provide an airtightness detection device, comprising: An inflation control module is used to control the inflation device to perform N inflation processes on the test chamber. The inflation amount of the first inflation process is determined based on the amount of gas that can make the air pressure of the test chamber reach the target air pressure. The test chamber is equipped with the test object and an air pressure sensor. N is an integer greater than 1. The data acquisition module is used to acquire N sets of sensor data collected by the air pressure sensor, which correspond one-to-one with the N inflation processes; The first processing module is used to adjust the N sets of sensor data based on the target air pressure and the sensor data corresponding to the first inflation process to obtain N sets of first sensor data. The second processing module is used to determine N sets of second sensor data based on the control information of each of the N inflation processes and the N sets of first sensor data. The detection result determination module is used to determine the airtightness detection result of the test object based at least on the inflation volume of each of the N inflation processes and the N sets of second sensor data.

[0014] Thirdly, embodiments of this application provide a non-transitory computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the method described in any of the above-mentioned embodiments.

[0015] Fourthly, embodiments of this application provide a computer device including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor executes the computer program to implement the steps of the method described in any of the preceding claims.

[0016] In summary, the embodiments of this application have at least the following beneficial effects: In this embodiment, the inflation device is controlled to perform N inflation processes on the test chamber. The inflation amount in the first inflation process is determined based on the amount of gas required to bring the test chamber to a target pressure. The test chamber contains the test object and a pressure sensor, where N is an integer greater than 1. N sets of sensor data, each corresponding to one of the N inflation processes, are acquired from the pressure sensor. Based on the target pressure and the sensor data corresponding to the first inflation process, the N sets of sensor data are adjusted to obtain N sets of first sensor data. Based on the control information of each of the N inflation processes and the N sets of first sensor data, N sets of second sensor data are determined. At least based on… Based on the inflation volume of each of the N inflation processes and the N sets of second sensor data, the airtightness test result of the tested object is determined. In this way, the N sets of sensor data can be coarsely adjusted by analyzing the target air pressure and the sensor data corresponding to the first inflation process to initially improve the errors in the data. Then, the control information of each inflation process is used to further eliminate the data noise caused by the inflation process in each set of first sensor data to generate second sensor data with higher accuracy. Finally, the accurate airtightness test result can be obtained by analyzing the high-precision second sensor data and the inflation volume of each inflation, thereby improving the airtightness test accuracy. Attached Figure Description

[0017] Figure 1 This is a schematic flowchart of the airtightness testing method provided in the embodiments of this application; Figure 2 This is a schematic diagram of the network structure of the feature filtering network provided in the embodiments of this application; Figure 3 This is a schematic diagram of the airtightness testing device provided in the embodiments of this application; Figure 4 This is a schematic diagram of the structure of the computer device provided in the embodiments of this application. Detailed Implementation

[0018] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments / examples are only a part of the embodiments / examples of this application, and not all of the embodiments / examples. Based on the embodiments / examples in this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0019] In the description of this application, the terms "first," "second," "third," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Therefore, a feature defined with "first," "second," "third," etc., may explicitly or implicitly include one or more of that feature. In the description of this application, unless otherwise stated, "multiple" means two or more. In the description of this application, the term "comprising" and its variations are open-ended, meaning "including but not limited to." The term "based on" means "at least partially based on." The term "according to" means "at least partially according to." The term "one embodiment / example" means "at least one embodiment / example"; the term "another embodiment / example" means "at least one additional embodiment / example"; the term "some embodiments / examples" means "at least some embodiments / examples."

[0020] In the description of this application, it should be noted that, unless otherwise expressly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection between two components. Those skilled in the art can understand the specific meaning of the above terms in this application based on the specific circumstances.

[0021] In the description of this application, it should be noted that, unless otherwise defined, all technical and scientific terms used in this application have the same meaning as commonly understood by one of ordinary skill in the art. The terminology used in this application is for the purpose of describing specific embodiments only and is not intended to limit the application. Those skilled in the art can understand the specific meaning of the above terms in this application according to the specific circumstances.

[0022] Firstly, see [the following] Figure 1 The diagram shows a flowchart of an airtightness testing method provided in an embodiment of this application. The airtightness testing method can be applied to a computer device that can be communicatively connected to an inflation device to control the inflation device. The method includes steps S101-S105, as detailed below.

[0023] S101, control the inflation device to perform N inflation processes on the test chamber, wherein the inflation amount of the first inflation process is determined based on the amount of gas that can make the air pressure of the test chamber reach the target air pressure, the test chamber is equipped with the test object and the air pressure sensor, and N is an integer greater than 1.

[0024] It should be noted that this barometric pressure sensor belongs to a type of mechanical sensor. Mechanical sensors are devices used to measure mechanically related physical quantities, such as pressure, force, displacement, vibration, and acceleration—parameters related to the force or motion state of an object. Mechanical sensors convert mechanical signals (such as pressure and stress) into readable electrical signals (such as voltage and resistance changes). Correspondingly, the physical essence of barometric pressure is the collision pressure of gas molecules on the surface of an object. Therefore, the core working principle of a barometric pressure sensor is to sense gas pressure (a mechanical quantity) and convert it into an electrical signal.

[0025] In some examples, the test chamber may also be equipped with a temperature sensor, which can be used to measure the temperature as part of the environmental factor values ​​included in subsequent control information.

[0026] It is understandable that there will be a time interval between two adjacent inflation processes; in other words, the execution time points corresponding to the above N inflation processes are different. In some examples, the same inflation device can be used to perform an inflation process on the test chamber at each of the N different time points. Of course, multiple inflation devices can also be used to perform the N inflation processes, in which case one inflation device can be used to perform at least one inflation process. Furthermore, N inflation devices can be used to perform the N inflation processes separately.

[0027] It is understandable that the test chamber is sealed when not inflated. In some examples, the inflation volume of the first inflation process refers to the total amount of gas injected into the test chamber by the inflation device during the first inflation process. Theoretically, the inflation volume of the first inflation process should be sufficient to raise the pressure in the test chamber to the target pressure. This can be easily understood by using a commonly used pressure calculation formula in physics, based on the pre-measured available volume of the test chamber (which is obtained by subtracting the volume of all objects placed within the test chamber from its total volume), to calculate the theoretical amount of gas required to raise the pressure in the test chamber from the current atmospheric pressure to the target pressure. The total amount of gas required to inject into the test chamber during the first inflation process is then configured according to this theoretical amount of gas. The current atmospheric pressure can be measured in real-time in the external environment of the test chamber.

[0028] In some examples, the object under test may include at least one of the following: a battery (battery module, battery pack, etc.), a gas control valve, etc.

[0029] S102, acquire N sets of sensor data collected by the air pressure sensor that correspond one-to-one with the N inflation processes.

[0030] In some examples, each of the N sensor data sets can be sensor data collected by a pressure sensor at the same moment (or at a later moment) after the completion of its corresponding inflation process.

[0031] S103, based on the target air pressure and the sensor data corresponding to the first inflation process, adjust the N sets of sensor data to obtain N sets of first sensor data.

[0032] In some examples, the first air pressure indicated by the sensor data corresponding to the first inflation process can be compared with the target air pressure. Based on the comparison result, N sets of sensor data can be adjusted (i.e., coarse adjustment) to obtain N sets of first sensor data that correspond one-to-one with the N sets of sensor data (this can be understood as compensating for the difference represented by the comparison result in the sensor data to form the first sensor data, so that the difference is roughly eliminated in the first sensor data). Since the inflation volume of the first inflation process (referred to as the first inflation volume) is known, the air pressure that the test chamber should theoretically reach after being inflated with the first inflation volume (i.e., the target air pressure) is also known. Therefore, the comparison result can be used to indicate the difference between the first air pressure and the target air pressure. That is, the comparison result can be used to represent the difference between the theoretical air pressure change of the test chamber after being inflated and the measured air pressure change measured by the air pressure sensor. This discrepancy mainly stems from the unavoidable error between theory and practice. The causes of this discrepancy are numerous, including the difficulty in achieving a completely sealed test chamber, measurement errors in the pressure sensor, and the difficulty in ensuring the amount of gas injected by the inflation device matches the set amount perfectly. Therefore, in this embodiment, the approximate impact of this discrepancy can be estimated by comparing the theoretical and actual measurements of the first inflation process, thereby roughly eliminating this discrepancy in the N sets of sensor data.

[0033] S104, based on the control information of each of the N inflation processes and the N sets of first sensor data, determine N sets of second sensor data.

[0034] In some cases, a barometric pressure sensor detects air pressure via a sensor diaphragm (which can be a pressure-sensitive element). When the external air pressure changes, the diaphragm undergoes a slight deformation (such as stretching or bending), and the barometric pressure sensor converts this deformation into an electrical signal (such as changes in resistance or capacitance). The signal is then processed to output quantified data corresponding to the changed air pressure, thus achieving air pressure detection. However, when the test chamber is rapidly filled with gas, the gas flows in at high speed, creating pressure waves and / or turbulence. This causes the sensor diaphragm to experience excessive instantaneous force (pressure caused by the pressure wave and / or turbulence). In this case, the barometric pressure sensor is prone to overshoot and / or ringing. Furthermore, the measured value may lag behind the actual pressure, resulting in time-related errors. Consequently, the N data points collected by the barometric pressure sensor may contain errors due to overshoot and / or ringing and / or time-related errors. In some examples, the control factor values ​​contained in the control information can be used to adjust / correct N sets of first sensor data to form N sets of second sensor data that correspond one-to-one with the N sets of first sensor data, thereby eliminating errors caused by the overshoot phenomenon and / or the ringing phenomenon and / or the time-related error. The control factor values ​​may include at least one of the following: inflation rate, inflation duration, and valve opening / closing delay between the inflation device and the test chamber. Specifically, different pressure sensors can have their rate thresholds (or rates with no significant overshoot) pre-determined by the respective manufacturers. When the inflation rate exceeds this threshold, the first sensor data can be smoothed (e.g., using a low-pass filtering algorithm, such as a moving average filter) to reduce the impact of instantaneous spikes and / or oscillations in the pressure information, making the pressure information in the first sensor data closer to the true value of the stable pressure. Furthermore, since the inflation time is strongly correlated with the stabilization time of the gas in the test chamber, if the inflation time is too short, the gas will not be completely uniformly distributed, and the pressure sensor will output a value deviating from the actual stable pressure due to measurement lag. Therefore, when the inflation time is less than the stabilization time threshold (which can be used to indicate the shortest time for the gas to reach a preset uniform and stable state in the chamber), the first sensor data can be corrected for time-dimension delay.

[0035] In addition, the control information may also include the control factor value and / or environmental factor value, which may include temperature and / or humidity.

[0036] In some examples, a pre-trained error elimination model can be used to determine the second sensor data corresponding to each inflation process based on the control information and first sensor data. This error elimination model can be a trained model capable of predicting using control information and first sensor data as input and second sensor data as output. During training, sample control information and sample sensor data (which also carry corresponding expected data labels representing the expected sensor data) can be used to obtain the predicted sensor data generated by the model based on these sample data. Then, based on the difference between the predicted sensor data and the expected sensor data represented by the label, a general loss function is used to calculate the loss value. Finally, a general training algorithm (e.g., gradient descent) is used to train the model based on this loss value, enabling the trained model to possess the aforementioned capabilities. For example, the model may include an input representation layer, a feature fusion and representation layer, and a prediction output layer. The input representation layer can receive data from the input model and convert the data into the desired feature vector form. For example, the input representation layer can use word embeddings or pre-trained language models (such as bidirectional language representation models based on the Transformer architecture) to generate semantic vectors, and / or use embedding layers to generate dense vectors. The feature fusion and representation layer can be used to fuse the feature vectors converted by the input representation layer. For example, the feature fusion and representation layer can implement the fusion through fully connected layers or attention mechanism layers. The prediction output layer can be used to generate prediction results based on the fused features. For example, the prediction output layer can use a softmax layer to output the probability distribution for different categories, and then output the prediction result based on the probability (for example, it can output the top one or more classification results with the highest probability as the prediction result).

[0037] It is easy to understand that the sample data in this embodiment can be experimental data obtained in advance through multiple tests related to the above-mentioned "rapid filling of the test chamber with gas".

[0038] S105, based at least on the inflation volume of each of the N inflation processes and the N sets of second sensor data, determine the airtightness test result of the object under test.

[0039] In this embodiment, since the second sensor data has undergone two rounds of error correction, random noise, systematic errors, operational interference, and / or environmental interference have been largely eliminated, retaining the effective signals related to airtightness. Finally, by combining N inflation volumes (equivalent to known inputs) with N sets of second sensor data (equivalent to corrected outputs), and through cross-validation of multiple sets of data (the inflation volume corresponding to the same inflation process and the second sensor data constitute one set), the true gas leakage characteristics of the tested object can be more accurately identified, avoiding misjudgments caused by the random errors of a single data point. This results in an accurate airtightness detection result for the tested object. Here, this cross-validation can be achieved by analyzing the consistency of gas pressure change trends and the calculation stability of gas leakage amounts.

[0040] In some other examples, for each inflation process, the theoretical pressure corresponding to the inflation volume of that inflation process can be calculated, and the measured pressure value represented by the second sensor data corresponding to that inflation process can be determined. The airtightness information corresponding to that inflation process can be analyzed by calculating the difference between the theoretical pressure value and the measured pressure value. Finally, the airtightness test result of the tested object can be generated based on the airtightness information of each of the N inflation processes. For example, the airtightness test result can be determined by the average of the airtightness information of each of the N inflation processes.

[0041] In one optional implementation, determining the airtightness test result of the object under test based at least on the inflation volume of each of the N inflation processes and the N sets of second sensor data includes: Based on the N sets of second sensor data and the obtained covariate sequence, N sets of third sensor data are determined, wherein the covariate sequence includes N sets of covariate information corresponding one-to-one with the N inflation processes; Based on the inflation volume of each of the N inflation processes and the N sets of data from the third sensor, the airtightness test result of the tested object is determined.

[0042] It is understandable that the N inflation processes are performed sequentially, so the pressure changes in the test chamber will have certain temporal characteristics. Correspondingly, each of the N sets of sensor data will also be affected by the previous sets of sensor data. At this time, if uncontrollable interference (i.e., sudden covariates) occurs between two inflation processes, it will easily affect the accuracy of the sensor data corresponding to the subsequent inflation processes. Therefore, in this embodiment, it is equivalent to using the covariate sequence to further perform error elimination processing on the second sensor data to eliminate the disturbance signal caused by the covariates, thereby forming N sets of third sensor data that correspond one-to-one with the N sets of second sensor data.

[0043] In some examples, each set of covariate information may include at least one of the following: gas temperature change information within the test chamber, air pressure fluctuation information of the external environment where the test chamber is located, and mechanical vibration information inside and / or outside the test chamber. Specifically, the gas temperature change information may be used to indicate sudden changes in gas temperature caused by the start-up of the compressor in the inflation device; the air pressure fluctuation information may be used to indicate sudden changes in air pressure in the external environment caused by changes in environmental factors; and the mechanical vibration information may be used to indicate mechanical vibrations caused by high-frequency noise, which may be caused by the operation of the inflation device and / or equipment related to the inflation device.

[0044] In some examples, when determining the airtightness test result, this embodiment can refer to the relevant embodiment of step S105 above. It is only necessary to replace the second sensor data in step S105 above with the third sensor data in this embodiment, which will not be repeated here.

[0045] In one optional implementation, determining N sets of third sensor data based on the N sets of second sensor data and the acquired covariate sequence includes: For each of the N sets of second sensor data, Determine the data characteristics of the second sensor data and the variable characteristics of the covariate information corresponding to the second sensor data; The data features and the variable features are combined to obtain a first combined feature; Based at least on the first merging feature, a third sensor data corresponding to the second sensor data is obtained.

[0046] In some examples, the "features" described in any embodiment of this application can be obtained by a general feature extraction model. For example, the feature extraction model may include an encoder, which can be used to encode the input information / data to generate corresponding features.

[0047] In some examples, the first merged feature can be formed by fusing (e.g., splicing) data features and variable features.

[0048] In some examples, a corresponding set of third sensor data can be generated by decoding the first merged feature. This decoding can be implemented by a decoder that can be used to perform decoding corresponding to the target encoding process (i.e., the inverse operation of the target encoding process), which refers to the encoding operation used to extract data features from the second sensor data.

[0049] In one optional implementation, the first merging feature includes at least one merging feature vector, and obtaining a third sensor data corresponding to the second sensor data based at least on the first merging feature includes: The preset attention model is configured using the variable features of the covariate information corresponding to the second sensor data. The configured attention model is invoked to assign corresponding weights to the at least one merged feature vector; Based on the assigned weights, the first merged feature is weighted and fused to obtain the first fused feature; Based on the first fusion feature, a third sensor data corresponding to the second sensor data is obtained.

[0050] In some examples, the attention model can be configured by using the variable feature as a weight allocation condition (i.e., the variable feature can be configured into the attention model). This allows the configured attention model to assign weights to at least one merged feature vector based on the variable feature. For example, the configured attention model can perform weight allocation by analyzing the correlation strength (e.g., correlation and / or importance) between the variable feature and each merged feature vector. Merged feature vectors with higher correlation strength can be assigned higher weights, thereby strengthening effective features, weakening irrelevant features, and improving the accuracy of subsequent processing.

[0051] In some examples, at least one merged feature vector in the first merged feature can be weighted and fused based on the assigned weights to transform the first merged feature into a first fused feature.

[0052] In some examples, a corresponding set of third sensor data can be generated by decoding the first fused feature. This decoding can be implemented by a decoder that can be used to perform decoding corresponding to the target encoding process described above.

[0053] In one alternative implementation, the control information corresponds to at least one control factor; The determination of N sets of second sensor data based on the control information of each of the N inflation processes and the N sets of first sensor data includes: For each of the N inflation processes, Based on the control information corresponding to this inflation process, determine the reference sensor data; Extract the first data feature and the at least one first factor feature corresponding to the at least one control factor from the reference sensor data; Extract the second data feature and the at least one second factor feature corresponding to the at least one control factor from the first sensor data corresponding to the inflation process; According to the at least one control factor, the at least one first factor feature and the at least one second factor feature are combined into at least one factor feature pair, wherein the first factor feature and the second factor feature corresponding to the same control factor are combined into one factor feature pair; Based on the first data feature, the second data feature, and the at least one factor feature pair, a second set of sensor data corresponding to the inflation process is determined.

[0054] In this embodiment, the factor feature pair can be used to indicate the mapping relationship between the first factor feature and the second factor feature under the corresponding control factor. Thus, the factor feature pair can be further used to indicate the transformation mapping relationship of the second factor feature in the reference sensor data relative to the first sensor data. In this way, the source of interference error corresponding to each type of control factor can be accurately determined by using the factor feature pair, and the interference caused by various control factors can be reversed and eliminated from the first sensor data by screening.

[0055] In some examples, based on the first data feature and the at least one factor feature pair, the second data feature in the first sensor data can be subjected to interference correction, and based on the corrected second data feature, a second sensor data corresponding to the inflation process can be decoded and generated.

[0056] In some examples, the reference sensor data may be information that matches the control information. The reference sensor data may be obtained by querying a preset reference database based on the control information. The preset reference database may include pressure response data collected by high-precision barometric pressure sensors under various standardized scenarios. The standardized scenario may refer to a test scenario configured according to at least one control factor.

[0057] In one optional implementation, determining a set of second sensor data corresponding to the inflation process based on the first data feature, the second data feature, and the at least one factor feature pair includes: For each of the aforementioned factor characteristic pairs Based on the factor feature pair, the first data feature, and the second data feature, determine the intermediate feature of the control factor corresponding to the factor feature pair; Based on the control factors corresponding to the factor feature pair, the weights of the factor feature pair, the first data feature, and the second data feature are determined, wherein the sum of the weights of the first data feature and the second data feature is less than the weight of the factor feature pair. Based on their respective weights, intermediate features, first data features, and second data features, a set of second sensor data corresponding to this inflation process is determined.

[0058] In this embodiment, the sum of the weights of the first and second data features is also less than the weight of the factor feature pair. This allows the correction of interference related to control factors to be placed at the core of this data processing step, thereby increasing the likelihood that interference caused by various control factors will be accurately removed.

[0059] In some examples, the feature pair, the first data feature, and the second data feature can be feature aligned (e.g., through an embedding layer), and then the aligned features can be mapped into the latent space through a latent space transformation model to form the intermediate feature.

[0060] It is understood that the specific method of obtaining the weights determined based on the control factors is not specifically limited in this application embodiment. It can be a fixed weight pre-configured for different control factors, or it can be another method.

[0061] In one optional implementation, determining a set of second sensor data corresponding to the inflation process based on the respective weights, the intermediate features, the first data feature, and the second data feature includes: The first feature filtering network is invoked to process the first data feature to obtain the first filtering feature, process the intermediate feature to obtain the intermediate filtering feature, and process the second data feature to obtain the second filtering feature. The first feature filtering network includes at least two deep feature extraction layers, a pooling layer and an activation layer that are electrically connected in sequence. The first filtering feature and the intermediate filtering feature are upsampled to obtain their respective upsampled features. Based on their respective weights, the upsampled feature and the second filtering feature are weighted, and the weighted upsampled feature and the weighted second filtering feature are concatenated to obtain the concatenated feature. The second feature filtering network is invoked to process the spliced ​​features to obtain the processed spliced ​​features. The second feature filtering network includes a convolutional layer, a pooling layer, an activation layer and a convolutional layer that are electrically connected in sequence. Based on their respective weights, the intermediate feature, the first data feature, and the second data feature are weighted; The weighted intermediate features, the weighted first data features, the weighted second data features, and the processed spliced ​​features are fused together to obtain the second fused feature. Based on the second fusion feature, a second set of sensor data corresponding to this inflation process is determined.

[0062] In this embodiment, a two-stage feature selection network is provided. The first feature selection network can achieve in-depth feature extraction. Through the first feature selection network composed of multiple deep feature extraction layers, pooling layers, and activation layers, it can mine deep effective information from the first data features, intermediate features, and second data features, while filtering shallow noise (such as thermal noise from sensor circuits). Then, the features after the initial selection are upsampled to ensure feature dimension matching. Then, weighted concatenation is performed by combining weights, so that the interference correction related features and the original effective features (i.e., the second selected features processed from the second data features in the first sensor data) can be deeply integrated to improve the targeting of the features. Finally, the second feature selection network is used for optimization. The fusion effect, through a network structure of "convolution-pooling-activation-convolution", further refines the spliced ​​features, eliminates redundant information that may be generated during the feature fusion process, and makes the processed spliced ​​features more focused on the effective features of airtightness detection. Furthermore, the fusion process is used to supplement the processed spliced ​​features with weighted intermediate features, first data features and second data features, so as to achieve the fusion and complementarity of the basic signal (weighted intermediate features, first data features and second data features) and the corrected signal (the processed spliced ​​features). In the final decoding process, the basic signal and the corrected signal can use each other as references, thereby forming more reasonable and accurate second sensor data.

[0063] In some examples, the deep feature extraction layer described above can be composed of residual blocks.

[0064] In some examples, a corresponding set of second sensor data can be generated by decoding the second fused feature, where the decoding process can be implemented by a decoder.

[0065] See in some examples Figure 2 , Figure 2A schematic diagram of the network structure of the feature filtering network provided in this application embodiment is shown. Here, the first feature filtering network 201 may include at least two deep feature extraction layers (shown as two deep feature extraction layers in the figure), a pooling layer, and an activation layer connected in sequence. The first deep feature extraction layer can be used to receive input data input to the first feature filtering network 201, and the output data of the activation layer can be used as the output data of the first feature filtering network 201. The second feature filtering network 202 may include a convolutional layer, a pooling layer, an activation layer, and a convolutional layer connected in sequence. The first convolutional layer can be used to receive input data input to the second feature filtering network 202, and the output data of the last convolutional layer can be used as the output data of the second feature filtering network 202.

[0066] Secondly, correspondingly, the embodiments of this application also provide an airtightness testing device, which can implement all the processes of the airtightness testing method provided in the above embodiments.

[0067] See Figure 3 The diagram shows a schematic of the structure of an airtightness testing device 300 provided in an embodiment of this application. The airtightness testing device 300 includes: The inflation control module 301 is used to control the inflation device to perform N inflation processes on the test chamber. The inflation amount of the first inflation process is determined based on the amount of gas that can make the air pressure of the test chamber reach the target air pressure. The test chamber is equipped with the test object and an air pressure sensor. N is an integer greater than 1. The data acquisition module 302 is used to acquire N sets of sensor data collected by the air pressure sensor that correspond one-to-one with the N inflation processes; The first processing module 303 is used to adjust the N sets of sensor data based on the target air pressure and the sensor data corresponding to the first inflation process to obtain N sets of first sensor data. The second processing module 304 is used to determine N sets of second sensor data based on the control information of each of the N inflation processes and the N sets of first sensor data. The detection result determination module 305 is used to determine the air tightness detection result of the test object based at least on the inflation volume of each of the N inflation processes and the N sets of second sensor data.

[0068] In one optional implementation, determining the airtightness test result of the object under test based at least on the inflation volume of each of the N inflation processes and the N sets of second sensor data includes: Based on the N sets of second sensor data and the obtained covariate sequence, N sets of third sensor data are determined, wherein the covariate sequence includes N sets of covariate information corresponding one-to-one with the N inflation processes; Based on the inflation volume of each of the N inflation processes and the N sets of data from the third sensor, the airtightness test result of the tested object is determined.

[0069] In one optional implementation, determining N sets of third sensor data based on the N sets of second sensor data and the acquired covariate sequence includes: For each of the N sets of second sensor data, Determine the data characteristics of the second sensor data and the variable characteristics of the covariate information corresponding to the second sensor data; The data features and the variable features are combined to obtain a first combined feature; Based at least on the first merging feature, a third sensor data corresponding to the second sensor data is obtained.

[0070] In one optional implementation, the first merging feature includes at least one merging feature vector, and obtaining a third sensor data corresponding to the second sensor data based at least on the first merging feature includes: The preset attention model is configured using the variable features of the covariate information corresponding to the second sensor data. The configured attention model is invoked to assign corresponding weights to the at least one merged feature vector; Based on the assigned weights, the first merged feature is weighted and fused to obtain the first fused feature; Based on the first fusion feature, a third sensor data corresponding to the second sensor data is obtained.

[0071] In one alternative implementation, the control information corresponds to at least one control factor; The determination of N sets of second sensor data based on the control information of each of the N inflation processes and the N sets of first sensor data includes: For each of the N inflation processes, Based on the control information corresponding to this inflation process, determine the reference sensor data; Extract the first data feature and the at least one first factor feature corresponding to the at least one control factor from the reference sensor data; Extract the second data feature and the at least one second factor feature corresponding to the at least one control factor from the first sensor data corresponding to the inflation process; According to the at least one control factor, the at least one first factor feature and the at least one second factor feature are combined into at least one factor feature pair, wherein the first factor feature and the second factor feature corresponding to the same control factor are combined into one factor feature pair; Based on the first data feature, the second data feature, and the at least one factor feature pair, a second set of sensor data corresponding to the inflation process is determined.

[0072] In one optional implementation, determining a set of second sensor data corresponding to the inflation process based on the first data feature, the second data feature, and the at least one factor feature pair includes: For each of the aforementioned factor characteristic pairs Based on the factor feature pair, the first data feature, and the second data feature, determine the intermediate feature of the control factor corresponding to the factor feature pair; Based on the control factors corresponding to the factor feature pair, the weights of the factor feature pair, the first data feature, and the second data feature are determined, wherein the sum of the weights of the first data feature and the second data feature is less than the weight of the factor feature pair. Based on their respective weights, intermediate features, first data features, and second data features, a set of second sensor data corresponding to this inflation process is determined.

[0073] In one optional implementation, determining a set of second sensor data corresponding to the inflation process based on the respective weights, the intermediate features, the first data feature, and the second data feature includes: The first feature filtering network is invoked to process the first data feature to obtain the first filtering feature, process the intermediate feature to obtain the intermediate filtering feature, and process the second data feature to obtain the second filtering feature. The first feature filtering network includes at least two deep feature extraction layers, a pooling layer and an activation layer that are electrically connected in sequence. The first filtering feature and the intermediate filtering feature are upsampled to obtain their respective upsampled features. Based on their respective weights, the upsampled feature and the second filtering feature are weighted, and the weighted upsampled feature and the weighted second filtering feature are concatenated to obtain the concatenated feature. The second feature filtering network is invoked to process the spliced ​​features to obtain the processed spliced ​​features. The second feature filtering network includes a convolutional layer, a pooling layer, an activation layer and a convolutional layer that are electrically connected in sequence. Based on their respective weights, the intermediate feature, the first data feature, and the second data feature are weighted; The weighted intermediate features, the weighted first data features, the weighted second data features, and the processed spliced ​​features are fused together to obtain the second fused feature. Based on the second fusion feature, a second set of sensor data corresponding to this inflation process is determined.

[0074] Thirdly, embodiments of this application provide a non-transitory computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the method described in any of the above-mentioned embodiments.

[0075] Fourthly, embodiments of this application provide a computer device including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor executes the computer program to implement the steps of the method described in any of the preceding claims.

[0076] See Figure 4 The computer device in this embodiment includes a processor 401, a memory 402, and a computer program stored in the memory 402 and executable on the processor 401, such as an airtightness detection program. When the processor 401 executes the computer program, it implements the steps in the various airtightness detection method embodiments described above, for example... Figure 1 The steps S101-S105 are shown.

[0077] For example, the computer program may be divided into one or more modules / units, which are stored in the memory 402 and executed by the processor 401 to complete this application. The one or more modules / units may be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program in the computer device.

[0078] The computer device may be a desktop computer, laptop, handheld computer, or cloud server, etc. The computer device may include, but is not limited to, a processor 401 and a memory 402. Those skilled in the art will understand that the schematic diagram is merely an example of a computer device and does not constitute a limitation on the computer device. It may include more or fewer components than shown, or combine certain components, or different components. For example, the computer device may also include input / output devices, network access devices, buses, etc.

[0079] The processor 401 can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor, or the processor 401 can be any conventional processor. The processor 401 is the control center of the computer device, connecting various parts of the entire computer device through various interfaces and lines.

[0080] The memory 402 can be used to store the computer programs and / or modules. The processor 401 implements various functions of the computer device by running or executing the computer programs and / or modules stored in the memory 402 and calling the data stored in the memory 402. The memory 402 may mainly include a program storage area and a data storage area. The program storage area may store the operating system, at least one application program required for a function (such as sound playback function, image playback function, etc.), etc.; the data storage area may store data created according to the use of the mobile phone (such as audio data, phonebook, etc.). In addition, the memory 402 may include high-speed random access memory, and may also include non-volatile memory, such as hard disk, memory, plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, at least one disk storage device, flash memory device, or other volatile solid-state storage device.

[0081] Wherein, if the modules / units integrated into the computer device are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments can also be implemented by a computer program instructing related hardware. The computer program can be stored in a non-transitory computer-readable storage medium. When the computer program is executed by the processor 401, it can implement the steps of the various method embodiments described above. Wherein, the computer program includes computer program code, which can be in the form of source code, object code, executable file, or some intermediate form, etc. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, portable hard drive, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal, and software distribution medium, etc.

[0082] In summary, the embodiments of this application have at least the following beneficial effects: In this embodiment, the inflation device is controlled to perform N inflation processes on the test chamber. The inflation amount in the first inflation process is determined based on the amount of gas required to bring the test chamber to a target pressure. The test chamber contains the test object and a pressure sensor, where N is an integer greater than 1. N sets of sensor data, each corresponding to one of the N inflation processes, are acquired from the pressure sensor. Based on the target pressure and the sensor data corresponding to the first inflation process, the N sets of sensor data are adjusted to obtain N sets of first sensor data. Based on the control information of each of the N inflation processes and the N sets of first sensor data, N sets of second sensor data are determined. At least based on… Based on the inflation volume of each of the N inflation processes and the N sets of second sensor data, the airtightness test result of the tested object is determined. In this way, the N sets of sensor data can be coarsely adjusted by analyzing the target air pressure and the sensor data corresponding to the first inflation process to initially improve the errors in the data. Then, the control information of each inflation process is used to further eliminate the data noise caused by the inflation process in each set of first sensor data to generate second sensor data with higher accuracy. Finally, the accurate airtightness test result can be obtained by analyzing the high-precision second sensor data and the inflation volume of each inflation, thereby improving the airtightness test accuracy.

[0083] Through the above description of the embodiments, those skilled in the art can clearly understand that this application can be implemented by means of software plus necessary hardware platforms, or it can be implemented entirely by hardware. Based on this understanding, all or part of the technical solutions of this application that contribute to the background technology can be embodied in the form of a software product. This computer software product can be stored in a storage medium, such as ROM (Read-Only Memory) / RAM (Random Access Memory), magnetic disk, optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in various embodiments or some parts of the embodiments of this application.

[0084] The above description is the preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications are also considered to be within the scope of protection of this application.

Claims

1. A method for detecting airtightness, characterized in that, include: The inflation device is controlled to inflate the test chamber N times. The inflation amount of the first inflation is determined based on the amount of gas that can make the air pressure of the test chamber reach the target air pressure. The test chamber is equipped with the test object and an air pressure sensor. N is an integer greater than 1. Obtain N sets of sensor data collected by the air pressure sensor, each corresponding to one of the N inflation processes; Based on the target air pressure and the sensor data corresponding to the first inflation process, the N sets of sensor data are adjusted to obtain N sets of first sensor data; Based on the control information of each of the N inflation processes and the N sets of first sensor data, N sets of second sensor data are determined; The air tightness test result of the object under test is determined based at least on the inflation volume of each of the N inflation processes and the N sets of second sensor data.

2. The method according to claim 1, characterized in that, The determination of the airtightness test result of the object under test, based at least on the inflation volume of each of the N inflation processes and the N sets of second sensor data, includes: Based on the N sets of second sensor data and the obtained covariate sequence, N sets of third sensor data are determined, wherein the covariate sequence includes N sets of covariate information corresponding one-to-one with the N inflation processes; Based on the inflation volume of each of the N inflation processes and the N sets of data from the third sensor, the airtightness test result of the tested object is determined.

3. The method according to claim 2, characterized in that, The step of determining N sets of third sensor data based on the N sets of second sensor data and the acquired covariate sequence includes: For each of the N sets of second sensor data, Determine the data characteristics of the second sensor data and the variable characteristics of the covariate information corresponding to the second sensor data; The data features and the variable features are combined to obtain a first combined feature; Based at least on the first merging feature, a third sensor data corresponding to the second sensor data is obtained.

4. The method according to claim 3, characterized in that, The first merging feature includes at least one merging feature vector, and obtaining a third sensor data corresponding to the second sensor data based at least on the first merging feature includes: The preset attention model is configured using the variable features of the covariate information corresponding to the second sensor data. The configured attention model is invoked to assign corresponding weights to the at least one merged feature vector; Based on the assigned weights, the first merged feature is weighted and fused to obtain the first fused feature; Based on the first fusion feature, a third sensor data corresponding to the second sensor data is obtained.

5. The method according to claim 1, characterized in that, The control information corresponds to at least one control factor; The determination of N sets of second sensor data based on the control information of each of the N inflation processes and the N sets of first sensor data includes: For each of the N inflation processes, Based on the control information corresponding to this inflation process, determine the reference sensor data; Extract the first data feature and the at least one first factor feature corresponding to the at least one control factor from the reference sensor data; Extract the second data feature and the at least one second factor feature corresponding to the at least one control factor from the first sensor data corresponding to the inflation process; According to the at least one control factor, the at least one first factor feature and the at least one second factor feature are combined into at least one factor feature pair, wherein the first factor feature and the second factor feature corresponding to the same control factor are combined into one factor feature pair; Based on the first data feature, the second data feature, and the at least one factor feature pair, a second set of sensor data corresponding to the inflation process is determined.

6. The method according to claim 5, characterized in that, The step of determining a second set of sensor data corresponding to the inflation process based on the first data feature, the second data feature, and the at least one factor feature pair includes: For each of the aforementioned factor characteristic pairs Based on the factor feature pair, the first data feature, and the second data feature, determine the intermediate feature of the control factor corresponding to the factor feature pair; Based on the control factors corresponding to the factor feature pair, the weights of the factor feature pair, the first data feature, and the second data feature are determined, wherein the sum of the weights of the first data feature and the second data feature is less than the weight of the factor feature pair. Based on their respective weights, intermediate features, first data features, and second data features, a set of second sensor data corresponding to this inflation process is determined.

7. The method according to claim 6, characterized in that, The step of determining a set of second sensor data corresponding to the inflation process based on the respective weights, the intermediate features, the first data feature, and the second data feature includes: The first feature filtering network is invoked to process the first data feature to obtain the first filtering feature, process the intermediate feature to obtain the intermediate filtering feature, and process the second data feature to obtain the second filtering feature. The first feature filtering network includes at least two deep feature extraction layers, a pooling layer and an activation layer that are electrically connected in sequence. The first filtering feature and the intermediate filtering feature are upsampled to obtain their respective upsampled features. Based on their respective weights, the upsampled feature and the second filtering feature are weighted, and the weighted upsampled feature and the weighted second filtering feature are concatenated to obtain the concatenated feature. The second feature filtering network is invoked to process the spliced ​​features to obtain the processed spliced ​​features. The second feature filtering network includes a convolutional layer, a pooling layer, an activation layer and a convolutional layer that are electrically connected in sequence. Based on their respective weights, the intermediate feature, the first data feature, and the second data feature are weighted; The weighted intermediate features, the weighted first data features, the weighted second data features, and the processed spliced ​​features are fused together to obtain the second fused feature. Based on the second fusion feature, a second set of sensor data corresponding to this inflation process is determined.

8. An airtightness testing device, characterized in that, include: An inflation control module is used to control the inflation device to perform N inflation processes on the test chamber. The inflation amount of the first inflation process is determined based on the amount of gas that can make the air pressure of the test chamber reach the target air pressure. The test chamber is equipped with the test object and an air pressure sensor. N is an integer greater than 1. The data acquisition module is used to acquire N sets of sensor data collected by the air pressure sensor, which correspond one-to-one with the N inflation processes; The first processing module is used to adjust the N sets of sensor data based on the target air pressure and the sensor data corresponding to the first inflation process to obtain N sets of first sensor data. The second processing module is used to determine N sets of second sensor data based on the control information of each of the N inflation processes and the N sets of first sensor data. The detection result determination module is used to determine the airtightness detection result of the test object based at least on the inflation volume of each of the N inflation processes and the N sets of second sensor data.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method described in any one of claims 1-7.

10. A computer device, characterized in that, The method includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor, when executing the computer program, implements the method of any one of claims 1-7.

Images