Pressure sensor fault detection, analysis and processing system

Abstract

The invention discloses a pressure sensor fault detection analysis processing system, which comprises a real-time parameter sequence acquisition module and a local parameter sequence determination module, and is characterized in that a local parameter sequence is used for representing an ordered sequence of K adjacent real-time parameters of each real-time parameter; the to-be-detected vector construction module is used for constructing the plurality of local parameter sequences into to-be-detected vectors; the fault information output module is used for inputting the vector to be detected into a pre-trained fault detection model to obtain a fault label and a fault level of the pressure sensor; wherein the fault label is node extraction performed in forward propagation of a vector to be detected, and the fault level is mapping after fitting of a local parameter sequence; according to the method, the local parameter sequence is constructed by identifying the abrupt change parameters and the neighborhood of the abrupt change parameters, the local fault level is comprehensively quantified, and the evolution trend of potential faults in the operation process of the sensor can be meticulously represented.

Description

Technical Field

[0001] The present invention relates to a fault analysis and processing system, and specifically to a pressure sensor fault detection, analysis and processing system. Background Art

[0002] The pressure sensor fault detection technology usually relies on the modeling of the full parameter sequence, that is, directly based on all the parameter sequences continuously collected by the sensor during operation for fault identification. The patent document with the patent publication number CN116878728B discloses a pressure sensor fault detection, analysis and processing system, which ensures the transmission quality of the pressure sensor, extends the service life of the pressure sensor to a certain extent, and thus provides strong data support for the subsequent fault analysis of the pressure sensor.

[0003] However, since the amount of data generated by the sensor during the normal operation stage is huge and most of it is stable state data, if the parameters of the sensor during the normal operation stage are used as training samples in the prior art, it will lead to a low proportion of effective faults in the training samples, low training efficiency, and prone to false detection and missed detection problems. In addition, the existing methods usually judge faults based on the overall parameter abnormality degree, and generally lack the modeling of the local abnormality evolution process, and it is difficult to accurately identify potential evolving faults. Summary of the Invention

[0004] In view of the deficiencies of the prior art, the present invention provides a pressure sensor fault detection, analysis and processing system, which identifies mutation parameters and defines local parameter sequences based on the mutation parameters to calculate the fault level of the pressure sensor, so as to solve the technical problems proposed in the background art.

[0005] To achieve the above objectives, the present invention is realized through the following technical solutions: A real-time parameter sequence acquisition module, which is used to acquire N types of real-time parameter sequences of the pressure sensor within the current time period; A local parameter sequence determination module, which is used to determine several local parameter sequences for any real-time parameter of the real-time parameter sequence; Wherein, the local parameter sequence is used to represent an ordered sequence of K adjacent real-time parameters of each real-time parameter; A vector to be detected construction module, which is used to construct the several local parameter sequences into a vector to be detected; A fault information output module, which is used to input the vector to be detected into a pre-trained fault detection model to obtain the fault label and its fault level of the pressure sensor; Wherein, the fault label is the node extraction during the forward propagation of the vector to be detected, and the fault level is the mapping after the fitting of the local parameter sequence.

[0006] In some specific embodiments thereof, the modeling steps of the fault detection model include: A1. Obtain the mutation parameter sequence of the pressure sensor within the observation time period; A2. Extract the sequence macroscopic features from the mutation parameter sequence, and locate the local parameter sequence based on the sequence macroscopic features; A3. According to the local parameter sequence, construct a fault sample vector and calculate the local fault level of the local parameter sequence; A4. Pair the fault sample vector with the corresponding local fault level to form a fault detection sample; A5. Obtain multiple fault detection samples to form a fault detection training set for supervised training; A6. Input the fault detection training set into the supervised model, use the local fault level as the supervision target, and generate the fault detection model through supervised training.

[0007] In some specific embodiments thereof, obtaining the mutation parameter sequence of the pressure sensor within the observation time period includes: A1-1. Obtain N types of observation parameter sequences of the pressure sensor within the observation time period; Each observation parameter sequence represents the time sequence of the same type of parameter within the observation time period, including: pressure sequence, voltage sequence, output noise sequence, output delay sequence; A1-2. Extract the sequence macroscopic features of the same type of parameter from the N types of observation parameter sequences; A1-3. Obtain N types of reference parameter sequences of the pressure sensor within the reference historical time period; A1-4. Extract the reference parameter features of the same type of parameter from the N types of reference parameter sequences; A1-5. Calculate the parameter feature difference of the same type of parameter according to the reference parameter features and the sequence macroscopic features; A1-6. If the parameter feature difference exceeds the set threshold, determine that the observation parameter sequence of this type of parameter is a mutation parameter sequence.

[0008] In some specific embodiments thereof, the sequence macroscopic feature is the overall change feature of the observation parameter sequence within the observation time period; When the observation parameter sequence is the pressure sequence, its sequence macroscopic feature is the pressure change rate; When the observation parameter sequence is the voltage sequence, its sequence macroscopic feature is the voltage level difference; When the observation parameter sequence is the output noise sequence, its sequence macroscopic feature is the noise energy density; When the observation parameter sequence is the output delay sequence, its sequence macroscopic feature is the average delay time.

[0009] In some of the specific embodiments, the A1-1 includes: A1-1-1, anchoring the current observation time point; A1-1-2, backtracking a predefined duration from the current observation time point to determine the observation time period; A1-1-3, anchoring the type of parameter to be obtained; A1-1-4, obtaining the observed parameter scalars of the type of parameter to be obtained at each consecutive time point within the observation time period; A1-1-5, defining the time series of the observed parameter scalars at each consecutive time point as a sequence of the same type of observed parameters; A1-1-6, switching the type of parameter to be obtained and obtaining the sequence of the same type of observed parameters after switching until N types of sequences of observed parameters are obtained.

[0010] In some of the specific embodiments, extracting the sequence macroscopic features from the mutation parameter sequence and locating the local parameter sequence based on the sequence macroscopic features includes: A2-1, calculating the mutation index of each parameter of the same type in the mutation parameter sequence; A2-2, if the mutation index exceeds the set threshold, marking the corresponding parameter of the same type as a mutation parameter; A2-3, obtaining the K neighboring parameters before and after the mutation parameter in the corresponding observed parameter sequence; A2-4, arranging the mutation parameter and the K neighboring parameters before and after it in chronological order to form a local parameter sequence.

[0011] In some of the specific embodiments, calculating the mutation index of each parameter of the same type in the mutation parameter sequence includes: A2-1-1, within the observation time period, selecting several local subsequences in the mutation parameter sequence with a sliding window; A2-1-2, calculating the mean and standard deviation of each local subsequence; A2-1-3, for each parameter of the same type in each local subsequence, calculating its standard score according to the mean and standard deviation; The expression of the standard score is: ; Wherein, represents the t-th parameter of the same type, represents the mean of the parameters of the same type in the local subsequence, represents the standard deviation of the parameters of the same type in the local subsequence, represents the standard score of the parameter of the same type at the t-th time point within the observation time period; A2-1-4. Compare the standard score of each parameter of the same type with the set threshold; if the standard score is greater than the set threshold, determine that the parameter of the same type corresponding to the standard score is a mutant parameter.

[0012] In some specific embodiments, according to the local parameter sequence, a fault sample vector is constructed, including: A3-1. Characterize the parameters in the local parameter sequence to obtain K + 1 local parameter features; A3-2. Obtain the parameter type of the mutant parameter sequence and encode the parameter type to form the label feature of the local parameter; A3-3. Concatenate the K + 1 local parameter features and the label feature to construct a fault sample vector.

[0013] In some specific embodiments, the construction steps of the local fault level include: B1. Calculate the trend slope, local variance of the local parameter sequence, and the maximum offset of the mutant parameter; B2. Substitute the trend slope, local variance, and the maximum offset of the mutant parameter into the following formula to determine the local fault level of the local parameter sequence; ; where, represents the local fault level, represents the absolute value of the trend slope; V represents the local variance of the local parameter sequence; M represents the maximum offset of the mutant parameter in the local parameter sequence.

[0014] The present invention provides a pressure sensor fault detection, analysis and processing system, which has the following beneficial effects: By extracting the mutant parameter sequence and using the feature difference as the screening basis, the present invention only retains the parameter segments with significantly abnormal changes during the observation period as the source of training samples, effectively improving the density of fault information in the training samples, reducing the interference of irrelevant stable state data on the model training process, and significantly improving the training efficiency of the fault detection model.

[0015] Based on the mutant parameter and its neighborhood, the present invention constructs a local parameter sequence, and comprehensively quantifies the local fault level by calculating features such as local trend slope, local variance, and mutant offset, which can carefully characterize the evolution trend of potential faults during the operation of the sensor. Compared with the traditional overall anomaly determination method, this method shows higher accuracy in early fault detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 is a structural block diagram of a pressure sensor fault detection, analysis and processing system of the present invention; Figure 2Schematic diagram of the processing flow of a pressure sensor fault detection, analysis and processing system according to the present invention; Figure 3 Schematic diagram of the determination process of the mutation parameter sequence according to the present invention; Figure 4 Schematic diagram of the positioning process of the local parameter sequence according to the present invention. Specific implementation manners

[0017] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0018] Embodiment 1: Please refer to Figures 1 to 2 , the present invention provides a pressure sensor fault detection, analysis and processing system, and the processing system includes: A real-time parameter sequence acquisition module, configured to acquire N types of real-time parameter sequences of a pressure sensor within a current time period; A local parameter sequence determination module, configured to determine several local parameter sequences for any real-time parameter of the real-time parameter sequence; Wherein, the local parameter sequence is used to represent an ordered sequence of K adjacent real-time parameters of each real-time parameter; A vector to be detected construction module, configured to construct several local parameter sequences into a vector to be detected; A fault information output module, configured to input the vector to be detected into a pre-trained fault detection model to obtain a fault label and a fault level of the pressure sensor; Wherein, the fault label is node extraction during the forward propagation of the vector to be detected, and the fault level is mapping after fitting the local parameter sequence.

[0019] In this embodiment, N types of real-time parameter sequences of a target pressure sensor are acquired within a current time period, and each parameter sequence corresponds to a different physical quantity measurement dimension. For any real-time parameter in the real-time parameter sequence, several local parameter sequences on the time axis are determined to describe the parameter change characteristics in the local area. The above-mentioned several local parameter sequences are arranged according to a set rule to construct a vector to be detected in a unified format. Subsequently, the vector to be detected is input into a pre-trained fault detection model, and the corresponding fault label and fault level are output, so as to realize the fault detection and analysis of the current operating state of the pressure sensor.

[0020] Embodiment 2: Refer to Figures 3 to 4, the technical solution of this Embodiment 2 different from that of Embodiment 1 is that the modeling steps of the fault detection model described in Embodiment 1 are disclosed, and the modeling steps include: A1. Obtain the mutation parameter sequence of the pressure sensor within the observation time period; A2. Extract the sequence macroscopic features from the mutation parameter sequence, and locate the local parameter sequence based on the sequence macroscopic features; A3. According to the local parameter sequence, construct a fault sample vector and calculate the local fault level of the local parameter sequence; A4. Pair the fault sample vector with the corresponding local fault level to form a fault detection sample; A5. Obtain multiple fault detection samples to form a fault detection training set for supervised training; A6. Input the fault detection training set into the supervised model, use the local fault level as the supervision target, and generate the fault detection model through supervised training.

[0021] In this embodiment, by extracting the mutation parameter sequence of the target pressure sensor within the observation time period as the basic data for fault modeling. Based on the mutation parameter sequence, further extract the sequence macroscopic features to determine the local abnormal trend, and accordingly locate the local parameter sequence. For each located local parameter sequence, construct the corresponding fault sample vector and calculate the local fault level of each local parameter sequence. Pair the fault sample vector with the local fault level to form a fault detection sample. By accumulating multiple fault detection samples, construct a fault detection training set required for supervised training. Finally, input the fault detection training set into the supervised model, use the local fault level as the supervision target, and generate a fault detection model for the fault detection task through supervised training.

[0022] Exemplarily, the acquisition of the mutation parameter sequence can adopt the following steps, including: A1-1. Obtain N types of observation parameter sequences of the pressure sensor within the observation time period; Each observation parameter sequence represents the time sequence of the same type of parameter within the observation time period, including: pressure sequence, voltage sequence, output noise sequence, output delay sequence; A1-2. Extract the sequence macroscopic features of the same type of parameter from the N types of observation parameter sequences; A1-3. Obtain N types of reference parameter sequences of the pressure sensor within the reference historical time period; A1-4. Extract the reference parameter features of the same type of parameter from the N types of reference parameter sequences; A1-5. Calculate the parameter feature difference of the same type of parameter according to the reference parameter feature and the sequence macroscopic feature; A1-6. If the difference in parameter characteristics exceeds the set threshold, then the observed parameter sequence of this type of parameter is determined to be a mutation parameter sequence.

[0023] In this embodiment, within the observation time period, N types of observed parameter sequences are collected. Each observed parameter sequence is used to characterize the continuous change of a specific physical quantity in the time dimension, including indicators such as pressure, voltage, output noise, and output delay. For each type of observed parameter sequence, the macroscopic characteristics of the sequence within the observation time period are extracted to describe the parameter change characteristics under the current operating state. At the same time, the corresponding reference parameter sequences within the reference historical time period are collected, and historical characteristics are extracted to characterize the parameter behavior under normal conditions. By comparing the macroscopic characteristics of the sequence with the historical characteristics, the characteristic difference value is calculated, and based on a preset threshold, the observed parameter sequences with significant characteristic differences are screened out and identified as mutation parameter sequences. By only retaining the mutation parameter sequences as the source of subsequent training samples, it is possible to effectively focus on the data segments containing abnormal evolution trends and avoid the interference of redundant samples.

[0024] Exemplarily, the macroscopic characteristics of the sequence are the overall change characteristics of the observed parameter sequence within the observation time period, and are used to characterize the macroscopic change trend of the parameter sequence during operation; When the observed parameter sequence is a pressure sequence, its macroscopic characteristics of the sequence are the pressure change rate; When the observed parameter sequence is a voltage sequence, its macroscopic characteristics of the sequence are the voltage step; When the observed parameter sequence is an output noise sequence, its macroscopic characteristics of the sequence are the noise energy density; When the observed parameter sequence is an output delay sequence, its macroscopic characteristics of the sequence are the average delay time.

[0025] In this embodiment, for different types of observed parameter sequences, the macroscopic characteristics of the sequence that match their change characteristics are extracted respectively to accurately depict the macroscopic change trend during operation. For the pressure sequence, the pressure change rate is extracted as the macroscopic characteristics of the sequence, which is used to quantify the rising or falling speed of the pressure signal within the observation time period and reflect the continuous change behavior of the pressure system. For the voltage sequence, the voltage step is extracted as the macroscopic characteristics of the sequence. By measuring the amplitude between the maximum value and the minimum value, it characterizes the fluctuation range of the voltage supply. For the output noise sequence, the noise energy density is extracted as the macroscopic characteristics of the sequence, which is used to reflect the overall amplitude level of the noise signal in the frequency domain and identify possible abnormal vibrations or interference phenomena in the signal. For the output delay sequence, the average delay time is extracted as the macroscopic characteristics of the sequence, which quantifies the change trend of the system response delay and helps to judge the change of the response performance of the control system. By specifically extracting the characteristic indicators of different physical quantities, the normal operating state and abnormal change trends are effectively distinguished.

[0026] In this embodiment, the observation parameter sequence needs to define an observation time period to determine the observation parameter sequence. Therefore, step A1-1 includes: A1-1-1. Anchor the current observation time point; A1-1-2. Trace back a predefined duration from the current observation time point to determine the observation time period; A1-1-3. Anchor the parameter type to be obtained; A1-1-4. Obtain the observation parameter scalars of the parameter type to be obtained at each consecutive time point within the observation time period; A1-1-5. Define the time series of the observation parameter scalars at each consecutive time point as an observation parameter sequence of the same type; A1-1-6. Switch the parameter type to be obtained and obtain the observation parameter sequence of the same type after switching until N types of observation parameter sequences are obtained.

[0027] In this embodiment, by anchoring the current observation time point as the reference starting point for determining the observation data range. Subsequently, trace back a predefined time duration from this time point to determine the complete observation time period, so as to collect continuous time series data. For each parameter type to be obtained, sequentially extract the observation parameter scalars at each consecutive time point within the observation time period to form a corresponding time series data set. By arranging the consecutive observation parameter scalars in chronological order, they are constructed into an observation parameter sequence of the same type. After completing the sequence construction of one type of parameter, switch to the next parameter type to be obtained and repeat the above extraction and construction process until N predefined types of observation parameter sequences are obtained.

[0028] Exemplarily, the positioning of the local parameter sequence can be performed by the following steps: A2-1. Calculate the mutation index of each parameter of the same type in the mutation parameter sequence; A2-2. If the mutation index exceeds the set threshold, mark the corresponding parameter of the same type as a mutation parameter; A2-3. Obtain the first K and last K neighboring parameters of the mutation parameter in the corresponding observation parameter sequence; A2-4. Arrange the mutation parameter and the first K and last K neighboring parameters in chronological order to form a local parameter sequence.

[0029] In this embodiment, by calculating the mutation index of each parameter of the same type in the mutation parameter sequence, the abnormality degree of the parameter at each time point relative to the local feature is measured. Based on the set threshold condition, the mutation index is screened to identify the parameter points with significantly abnormal standard scores, which are marked as mutation parameters. For each mutation parameter, the neighborhood parameters at the K time points before and after it are further extracted from the corresponding observed parameter sequence to capture the local evolution characteristics around the mutation point. Finally, the mutation parameter and its neighborhood parameters are arranged in chronological order to form a local parameter sequence. The local parameter sequence centers around the mutation point and systematically organizes the parameter change information in the local area, providing a representative time series segment for fault trend modeling.

[0030] In this embodiment, step A2-1 is used to calculate the mutation index, and the calculation steps of the mutation index specifically include: A2-1-1. During the observation time period, several local subsequences in the mutation parameter sequence are selected by means of a sliding window; A2-1-2. Calculate the mean value and standard deviation of each local subsequence; A2-1-3. For the parameters of the same type in each local subsequence, calculate their standard scores according to the mean value and standard deviation; The expression of the standard score is: ; where represents the t-th parameter of the same type, represents the mean value of the parameters of the same type in the local subsequence, represents the standard deviation of the parameters of the same type in the local subsequence, represents the standard score of the parameter of the same type at the t-th time point during the observation time period; A2-1-4. Compare the standard score of each parameter of the same type with the set threshold; if the standard score is greater than the set threshold, it is determined that the parameter of the same type corresponding to the standard score is a mutation parameter.

[0031] In this embodiment, for the mutation parameter sequence, a sliding window method is adopted to select several local subsequences during the observation time period. Through localization processing, the change characteristics of the parameters on a short time scale are captured. For each local subsequence, its mean value and standard deviation are calculated respectively to describe the central tendency and fluctuation level of the parameter values in the local area. Further, for each parameter of the same type in the local subsequence, according to the corresponding local mean value and standard deviation, the standard score is calculated to quantify the deviation degree of the parameter relative to the local statistical characteristics. By comparing the standard score with the set threshold, the parameter points with abnormally large standard scores are screened out and determined as mutation parameters. This method can dynamically evaluate the abnormality degree of the observed parameters in the local area and improve the accuracy of mutation point identification.

[0032] Exemplarily, the construction of the fault sample vector can be carried out by the following steps: A3-1. Characterize the parameters in the local parameter sequence to obtain K + 1 local parameter features; A3-2. Obtain the parameter types of the mutation parameter sequence, and encode the parameter types to form the label features of the local parameters; A3-3. Concatenate the K + 1 local parameter features and the label features to construct the fault sample vector; Wherein, the local fault level is used to measure the trend mutation intensity of the mutation parameter relative to the neighborhood parameters, so as to quantify the potential fault trend of the local parameter sequence.

[0033] In this embodiment, for the local parameter sequence located by mutation detection, the feature extraction is sequentially performed on each parameter point in the sequence, and the extracted K + 1 local parameter features are used to describe the change trend and fluctuation characteristics in the local area. Further, the parameter type information corresponding to the mutation parameter is extracted and converted into label features through a predefined encoding method to identify the physical source category of the local sample. Subsequently, the K + 1 local parameter features and the label features are concatenated to construct a fault sample vector in a unified format, ensuring that the sample contains both local change information and type recognition information.

[0034] Further, the construction steps of the local fault level include: B1. Calculate the trend slope, local variance of the local parameter sequence, and the maximum offset of the mutation parameter; B2. Substitute the trend slope, local variance, and the maximum offset of the mutation parameter into the following formula to determine the local fault level of the local parameter sequence; ; Wherein, represents the local fault level, represents the absolute value of the trend slope, which is used to describe the overall change trend intensity of the local parameter sequence; V represents the local variance of the local parameter sequence, which is used to describe the overall faultiness of the local parameter sequence; M represents the maximum offset of the mutation parameter in the local parameter sequence, which is used to characterize the severity of the local mutation fault inside the local parameter sequence.

[0035] In this embodiment, the mutation parameter and its front and rear neighborhood parameters are extracted to form a local parameter sequence; then, based on this local parameter sequence, the following three types of local sequence features are calculated respectively: Trend slope (S): It is used to depict the strength of the overall change trend of the local parameter sequence, and reflects whether there is a continuous offset or drift phenomenon in the sensor output within the observed time range; Local variance (V): It is used to characterize the degree of fluctuation of the local parameter sequence, and reflects whether there are behaviors such as fault jitter, discreteness or decreased stability in the sensor output; Maximum offset (M): It is used to describe the instantaneous mutation amplitude of the mutation parameter relative to the mean value of the neighborhood parameters, and reflects the severity of sudden fault events (such as short circuit, impact, fracture, etc.).

[0036] After obtaining the local features of the three types of sequences, in order to comprehensively consider the contributions of each local feature to the local fault trend, the following fusion strategy is adopted: Add the absolute value of the trend slope to the local variance to form the overall fault basis quantity; then, using the maximum offset as the suppression factor, divide the overall basis quantity by 1 + M. The addition of 1 is used to avoid the division-by-zero fault. At the same time, the larger the maximum offset, the stronger the modulation effect on the overall fault, thus highlighting the dominant position of sudden faults; finally, perform a logarithmic transformation on the above ratio to compress the influence of extreme fault values, enhance the sensitivity to subtle change trends, and at the same time maintain the controllable range and numerical stability of the fault level.

[0037] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the processes or functions described in the embodiments of the present application are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center by wired (such as infrared, wireless, microwave, etc.) means.

[0038] The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or a data center that contains one or more collections of available media. The available media can be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., DVD ), or semiconductor media. The semiconductor media can be a solid-state drive.

[0039] In several embodiments provided by the present application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections between each other can be through some interfaces, and the indirect couplings or communication connections of devices or units can be in electrical, mechanical, or other forms.

[0040] As described above, this is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present application can easily think of changes or substitutions, which should all be covered within the protection scope of the present application.

Claims

1. A pressure sensor fault detection, analysis and processing system, characterized in that, Including: A real-time parameter sequence acquisition module, configured to acquire N types of real-time parameter sequences of a pressure sensor within a current time period; A local parameter sequence determination module, configured to determine several local parameter sequences for any real-time parameter of the real-time parameter sequences; Wherein, the local parameter sequence is used to represent an ordered sequence of K adjacent real-time parameters of each real-time parameter; A vector to be detected construction module, configured to construct a vector to be detected from several local parameter sequences; A fault information output module, configured to input the vector to be detected into a pre-trained fault detection model to obtain a fault label and a fault level of the pressure sensor; Wherein, the fault label is node extraction during the forward propagation of the vector to be detected, and the fault level is mapping after fitting the local parameter sequence.

2. The fault detection, analysis and processing system of a pressure sensor according to claim 1, characterized in that, The modeling steps of the fault detection model include: A1. Acquire a mutation parameter sequence of the pressure sensor within an observation time period; A2. Extract sequence macroscopic features from the mutation parameter sequence, and locate local parameter sequences based on the sequence macroscopic features; A3. According to the local parameter sequences, construct a fault sample vector and calculate the local fault level of the local parameter sequences; A4. Pair the fault sample vector with the corresponding local fault level to form a fault detection sample; A5. Acquire multiple fault detection samples to form a fault detection training set for supervised training; A6. Input the fault detection training set into a supervised model, use the local fault level as a supervision target, and generate the fault detection model through supervised training.

3. The pressure sensor fault detection, analysis and processing system according to claim 2, characterized in that Acquiring the mutation parameter sequence of the pressure sensor within the observation time period includes: A1-1. Acquire N types of observation parameter sequences of the pressure sensor within the observation time period; Each observation parameter sequence represents the time sequence of the same type of parameter within the observation time period, including: pressure sequence, voltage sequence, output noise sequence, output delay sequence; A1-2. Extract sequence macroscopic features of the same type of parameter from the N types of observation parameter sequences; A1-3. Acquire N types of reference parameter sequences of the pressure sensor within a reference historical time period; A1-4. Extract reference parameter features of the same type of parameter from the N types of reference parameter sequences; A1-5. Calculate the parameter feature difference of the same type of parameter according to the reference parameter features and the sequence macroscopic features; A1-6. If the parameter feature difference exceeds a set threshold, determine that the observation parameter sequence of the same type of parameter is a mutation parameter sequence.

4. The pressure sensor fault detection, analysis and processing system according to claim 3, characterized in that The sequence macroscopic feature is the overall change feature of the observation parameter sequence within the observation time period; When the observation parameter sequence is a pressure sequence, its sequence macroscopic feature is the pressure change rate; When the observation parameter sequence is a voltage sequence, its sequence macroscopic feature is the voltage level difference; When the observation parameter sequence is an output noise sequence, its sequence macroscopic feature is the noise energy density; When the observation parameter sequence is an output delay sequence, its sequence macroscopic feature is the average delay time.

5. The fault detection, analysis and processing system of a pressure sensor according to claim 4, wherein The A1-1 includes: A1-1-1. Anchor the current observation time point; A1-1-2. Trace back a predefined duration from the current observation time point to determine the observation time period; A1-1-3. Anchor the parameter type to be acquired; A1-1-4. Obtain the observed parameter scalars of the parameter type to be obtained at each continuous time point within the observation time period; A1-1-5. Define the time series of the observed parameter scalars at each continuous time point as the observed parameter sequence of the same type; A1-1-6. Switch the parameter type to be obtained, and obtain the observed parameter sequence of the same type after switching until the observed parameter sequences of N types are obtained.

6. The pressure sensor fault detection, analysis and processing system according to claim 2, characterized in that, Extract the sequence macroscopic features from the mutation parameter sequence, and locate the local parameter sequence based on the sequence macroscopic features, including: A2-1. Calculate the mutation index of each parameter of the same type in the mutation parameter sequence; A2-2. If the mutation index exceeds the set threshold, mark the corresponding parameter of the same type as a mutation parameter; A2-3. Obtain the first K and last K neighboring parameters of the mutation parameter in the corresponding observed parameter sequence; A2-4. Arrange the mutation parameter and the first K and last K neighboring parameters in chronological order to form a local parameter sequence.

7. A pressure sensor fault detection, analysis and processing system according to claim 6, characterized in that, Calculate the mutation index of each parameter of the same type in the mutation parameter sequence, including: A2-1-1. Within the observation time period, select several local subsequences in the mutation parameter sequence with a sliding window; A2-1-2. Calculate the mean and standard deviation of each local subsequence; A2-1-3. For each parameter of the same type in each local subsequence, calculate its standard score according to the mean and standard deviation; The expression of the standard score is: ； Among them, represents the t-th parameter of the same type, represents the mean of the parameters of the same type in the local subsequence, represents the standard deviation of the parameters of the same type in the local subsequence, and represents the standard score of the parameter of the same type at the t-th time point within the observation time period; A2-1-4. Compare the standard score of each parameter of the same type with the set threshold; if the standard score is greater than the set threshold, determine that the parameter of the same type corresponding to the standard score is a mutation parameter.

8. A pressure sensor fault detection, analysis and processing system according to claim 7, characterized in that, Construct a fault sample vector according to the local parameter sequence, including: A3-1. Characterize the parameters in the local parameter sequence to obtain K + 1 local parameter features; A3-2. Obtain the parameter types of the mutation parameter sequence, and encode the parameter types to form the label features of the local parameters; A3-3. Concatenate the K + 1 local parameter features and the label features to construct a fault sample vector.

9. A pressure sensor fault detection, analysis and processing system according to claim 8, characterized in that The construction steps of the local fault level include: B1. Calculate the trend slope, local variance of the local parameter sequence, and the maximum offset of the mutation parameter; B2. Substitute the trend slope, local variance, and the maximum offset of the mutation parameter into the following formula to determine the local fault level of the local parameter sequence; ； Among them, represents the local fault level, represents the absolute value of the trend slope; V represents the local variance of the local parameter sequence; M represents the maximum offset of the mutation parameter in the local parameter sequence.

Images